Abstract

AbstractRaman spectroscopy provides an efficient method for non‐contact determination of liquid water temperature with high spatial resolution. It can be also used for remote in situ determination of subsurface water temperature. The method is based on temperature‐dependent changes of the molecular OH stretching band in the Raman spectra of liquid water. These in turn are attributed to a decrease in intermolecular hydrogen‐bonding interactions with increase in temperature. Here, the results of an experimental study employing three different approaches in the determination of temperature from recorded OH stretching band in the Raman spectra of liquid and supercooled water are presented and discussed. The first two methods are based on deconvolution of the spectral band into Gaussian components whose intensities and associated specific spectral markers are temperature dependent, and the third approach is based on Raman difference spectroscopy (RDS). The presented measurements were conducted on distilled and deionized supercooled and liquid water in the temperature range between −12.5 and +32.5 °C. The results are compared in terms of linearity of response, sensitivity and accuracy. It is shown that the method based on RDS even in the supercooled temperature range provides better accuracy (the standard deviation from the true temperature is ±0.4 K) and linearity in temperature determination than more complicated methods based on Gaussian deconvolution of the OH stretching band. Copyright © 2005 John Wiley & Sons, Ltd.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call