Even Faster Analytics!

Matching the Speed of your Chemistry

For chemistry applications where data-density is important to monitor your processes, “normal” Analytics like online-GC can be too slow, requiring other techniques to collect the relevant data. We have validated multiple with the goal of collecting at least 1 datapoint per minute. This is necessary to follow fast chemistries or processes like deNOx, Methanol to Olefins, absorption breakthrough curves or feedback-loop controlled applications for iso-conversion like reforming. The easiest solution is taking online-MS (Mass-Spectroscopy), where a relatively simple mixture can be analyzed very fast, depending on the number of mass fragments that need to be measured, resulting in several datapoints per minute. 

As soon as the sample components have too many similar mass fragments the MS is no longer a suitable technique resulting in a lower accuracy and a lot of calibration work. However, if there are enough structural differences switching to FT-IR (Fourier-Transformation Infra-Red) is a suitable technique which can give very unique spectral absorbance for different bond types. Especially in the gas phase the wavelength bands can be very narrow and components can be identified even in the presence of water. In one of our applications, we have chosen a gas flow cell with a very narrow light path, resulting in a small sample volume needed to be flushed to get a fresh sample which is beneficial to get a real-time measurement of fast-changing concentrations.

The main advantage of the IR spectra is the ability to find a direct relation between concentration and area on a specific wavelength due to Beer’s Law, or create a calculation model with another analytical technique as a secondary input. See below the difference of sharp peaks in a deNOx example (ppm range) versus the broad spectrum changes in a hydrocarbon conversion (percentage range), wherein both cases we got a good correlation to use these as trending values for the applications.

Beer’s Law applied to sharp peaks versus PCR modelling to broad spectra.

Contact us