Selection of Operating Wavelengths for THz Temperature Tomography

Paul Wright, Krikor B. Ozanyan

School of Electrical and Electronic Engineering, The University of Manchester, Manchester M13 9PL, United Kingdom

ABSTRACT

The temperature distribution within subjects containing water vapour can be inferred from two absorption images, obtained at photon energies appropriate to the rovibrational spectrum of the water molecule. Two spatially overlapping beams are required, with wavelengths suitable for each molecular transition being targeted. Because of the resonant character of the molecular interaction with the probing light beams, Temperature Tomography necessitates careful selection and precise control of these wavelengths, if reliable measurement across the required temperature range is to be achieved. Here, we suggest and discuss rules for the choice of a pair of wavelengths in the terahertz range, where the rovibrational fingerprint of water vapour is remarkably rich. This region is also particularly suitable for imaging of larger molecules, including many organic materials of importance in the chemical and biological sciences, and their derived technologies.

We report results in the spectral interval from 1.0 to 3.0 THz, determined by the availability of a narrow line (~5 GHz) terahertz source and the information in the spectral databases. The main adopted criteria are i) the individual line­strengths, ii) their robustness against neighbour interference and overlap, under conditions typical for the imaged subject, and iii) the separation of lower state energy E’’ required to achieve temperature sensitivity. Secondary criteria, such as usefulness under ambient pressure and/or temperature conditions and freedom from interference by molecules other than water are also applied. All known terahertz water absorption lines in the interval of interest are simulated from the HITRAN database. A figure of merit, defined as the ratio between linestrength and summed overlap from other lines, assuming 10 GHz Lorentzian broadening at T=1800 K and T=296 K, is used to shortlist highly­independent lines from more than 400 candidates within the spectral region of interest. A more detailed examination of the properties of the shortlisted lines is then undertaken, with particular emphasis on the availability of a pair of lines having appropriate separation in their lower state energies. We also demonstrate the importance of candidate lines being checked for excessive reduction of line­strength at higher temperatures. 1800 K spectra confirm that sufficient air broadening takes place to ensure the general form of the spectrum remains reasonably stable across an acceptably wide range of water concentration.

In our case study, the applied criteria result in the choice of Line 209 (1.869 THz, high figure of merit at both 296 K and 1800 K) and any of the two lines 111/131 (2.430 THz/2.319 THz, close to the maximum of the terahertz source power envelope, yielding corresponding lower state energy separation of ?E’’ = 756 cm­1 / 837 cm­1).

Keywords Temperature Tomography, Terahertz Tomography, spectral database, tunable source