Fairness in disclosure, I am the original author of the PeakFit software first published in 1990. Although I sold this more than a decade ago (2001), I have been working independently for the past two years to develop new software technology to better model and when necessary separate chromatographic peaks that elute as overlapping or with one or more hidden peaks.

I am also looking to test the new software technology with data where there is some possibility that the actual shape (as opposed to the quantity) of a given peak may have diagnostic or analytic value. An example would be a protein peak that is present at the same concentration, but where some form of degradation is expected that might be detectable in highly accurate modeled peak parameters.

The new technology is based on as yet unpublished generalizations of the HVL and Wade-Thomas NLC models that include instrumental effects and other deviations from non-ideality. I am posting here to solicit chromatographic data that is reasonably low-noise (allowing peak subtleties to be accurately modeled) but where the analysis has proven less than fully adequate with the instrumental software.

In return for sharing such data and helping me to test the new technology, I will send you the full analysis this new software presently generates (either privately or posted to this forum as you choose-I assume both are available?). I will do the analysis personally, and if there is any contribution to the chromatographic modeling science where your specific data highlights something important, I will invite you as a coauthor where you can describe your methodology and data in any paper that may be worthy of publication. No strings--nothing financial or commercial. From both sides, please, science only.

I will not use your data except to analyze it, but you should still remove all identifying information in any data you may furnish, and it should be exported to an ASCII format file. I still have AIA format import available, and that may work, but it is quite dated. If you wish to send such directly, you can use my personal e-mail, rebrown@hiwaay.net.

Thus far, with very clean IC data, I have been able to fit chromatographic peaks, both strongly fronted and tailed, to about 3-8 ppm unaccounted variance (r^2=0.999995). To put this in perspective, the original software (which is still on the market), might do at best about 3000 ppm on the more challenging peaks. This technology thus allows one to see what has been heretofore essentially invisible to modeling. Also, for those doing preparative chromatography, we have some indication we can separate the overload from the analytic component in the modeling, but this is nowhere as far along.