Dear Bence and co,
Spec’s: Agilent 7500cs, PC, Windows 7, Igor Pro 6.37, and Iolite 2.41, 2.50 and 3.36.
I have been given U–Pb zircon data for a Cretaceous granitoid that was processed by a colleague using Iolite 2.41. I reprocessed the same data files using Iolite 2.50 and 3.36. The results I got from versions 2.50 and 3.36 are essentially identical. However, there are significant differences between my results and those of my colleague, as outlined below. I am hoping you might have a view of this that is bigger than mine.
1. Different 206Pb/238U dates for the same spot analyses. I am not too concerned about this because I assume that it is mainly due to different integration interval selections and exclusion of outliers >2SD from mean values by my colleague and 3SD outlier exclusion by me. We used the same number (34) of standards and the same spline for the standards (auto smooth), giving results for the standards that are <<1% different. I resolve 3 Cretaceous populations (~110, ~125 and ~135 Ma), whereas my colleague gets a single main peak at ~122 Ma that I think represents mixing of older and younger age zones within single grains (there is clear evidence for inheritance in CL images and age zoning in the raw signals).
2. Different 206Pb/238U percentage errors. The propagated uncertainties associated with 206Pb/238U ratios and dates obtained by my colleague (Iolite 2.41) are consistently ~50% smaller than I get with either Iolite version 2.50 or 3.36. This applies to both the standards and “unknowns”. I reprocessed the data using 2SD outlier exclusion to see what affect this had, but it can only account for an average ~10% reduction in uncertainties. Did Iolite 2.41 handle error calculations significantly differently to later versions? So far, I can’t think of anything else that we did differently when processing the data, other than what I’ve mentioned above, that might explain this difference.
Please let me know if you’d like any further information, the Iolite experiment file, or the data files. Any help/insights would be much appreciated! Thanks!
You must be logged in to reply to this topic.