For years, we have advised that stopped-flow experiments which could produce rate constants which are orders of magnitude apart be done in two parts: one shot for the fast, one shot for the slow. After all, our clients have rapid-scanning spectrophotometers which collect 1,000 scans per second, so they don't need to worry about sample economy the way those with less powerful detection systems do.

However, what we have proved today is another extraordinary advance made possible by Matheson's implementation of Matrix Exponentiation. Just as Olis Global fitting software now supports 3D fits six species, five rates—which we would have argued to the ground as impossible just one year ago—Matheson's fits now prove that they can find two rates constants separated by five orders of magnitude.

A three species data set was synthesized with 2001 scans, 400 nm wide. Rate constants are 300 and 0.03. Noise of 0.001 AU was added, consistent with the level of noise which the Olis RSM obtains when scanning 1,000 scans/second.

We hand this data file (800,000 points total) to Matheson's Global fit. One second later, SVD returns eigenvectors suggesting three species. We select A → B → C as the chemical model. One second later, the fit to kinetics and spectra is complete.

In two seconds, with no user bias imposed on the fit, Matheson's Global fitting software found the three spectral shapes and found k1 and k2.

k1 = 329.9 ± 0.9 and k2 = 0.027 ± 0.003.
Exactly correct!!

Who needs log time when you have a resource like this available?

Logarithm time scales are unnecessary when you have the right hardware and software at your disposal. Just as our exquisitely fast, sensitive and accurate detection system requires no 'tricks' to protect the sample or get the right answer, so our software requires no 'tricks' to solve even the most difficult cases.