https://www.euro-online.org/enog/inoc2007/Papers/mac-slots.html https://www.euro-online.org/enog/inoc2007/Papers/m https://www.euro-online.org/enog/inoc2007/Papers/mac-slots.html

Miniature NIR device develops a taste for ethanol

By customising a commercially-available miniature NIR spectrometer, Austrian scientists have produced a portable device for determining the concentration of ethanol in gasoline.

Known as the MicroNIR 2200, the miniature NIR spectrometer was developed by JDS Uniphase, a US optical technology company. It is a hand-held device that employs two tungsten light bulbs as the light source, a linear variable filter to select specific NIR wavelengths and a diode-array detector (DAD), and can be powered by simply hooking it up to a laptop with a USB cable. JDS Uniphase specifically offers the MicroNIR 2200 for customisation for various different applications, including pharmaceutical testing, food analysis and forensic investigations.

So when a team of chemists from the University of Innsbruck, led by Christian Huck, were looking to produce a portable device that could determine the concentration of ethanol in gasoline, they decided to try customising the MicroNIR 2200. Ethanol is now regularly added to gasoline in many countries around the world, in order to replace fossil fuels with renewable biofuels, but because ethanol can potentially damage engines and pipelines its concentration in gasoline usually needs to be kept below around 15%.

To turn the MicroNIR 2200 into a device that can determine ethanol concentrations in gasoline, Huck and his team developed a sampling cell that fits onto the front of the MicroNIR 2220. This sampling cells consists of a spherical gold mirror and a thermoelectric cooler, both of which were designed to enhance the sensitivity of the detector by improving the signal-to-noise ratio.

Using this customised MicroNIR 2200 device, Huck and his team analysed 36 samples of gasoline spiked with ethanol at concentrations from 0% to 10%, and then converted this data into a predictive model using partial least squares regression. As they report in Analytica Chimica Acta, this model could accurately predict the concentration of ethanol in commercial gasoline samples from NIR data, with the measured concentrations varying between 3.0% and 5.6%.

This model was also much more accurate than a model produced when the 36 samples were analysed by a non-customised MicroNIR 2220. The customised device and associated model could even accurately determine ethanol concentrations in gasoline samples with varying chemical compositions, which other teams looking to determine ethanol concentrations with NIR spectroscopy have struggled to achieve.

Blog tags: 
Techniques: