Simulation of an experimental database of infrared spectra of complex gaseous mixtures for detecting specific substances. The case of drug precursors

SENSORS AND ACTUATORS B-CHEMICAL

卷:193

Article

Calderisi, Marco^1,2  Ulrici, Alessandro^1,2  Sinisalo, Sauli⁴  Uotila, Juho⁴  Seeber, Renato^2,3 

[1] Univ Modena & Reggio Emilia, Dipartimento Sci Vita, I-42122 Reggio Emilia, Italy

[2] Consorzio INSTM, I-50121 Florence, Italy

[3] Univ Modena & Reggio Emilia, Dipartimento Sci Chim & Geol, I-41125 Modena, Italy

[4] Gasera Ltd, FIN-20520 Turku, Finland

关键词: Spectra simulation;    Experimental design;    Fast wavelet transform;    FT-IR;    Laser photoacoustic spectroscopy;   

DOI  :  10.1016/j.snb.2013.12.035

来源: Elsevier

【 摘 要 】

This work is motivated by the need to develop suitable databases in absence of real experimental data, for instance when spectra measured with a newly developed instrumentation on real samples are not available yet. This notwithstanding, in fact, the realization of the physical project should be addressed by a starting database, also invaluable in order to test its effectiveness. In this article we face the issue of simulating gas mixtures spectra for the development of a new sensor for external cavity-quantum cascade laser photoacoustic spectroscopy (EC-QCLPAS) starting from literature FT-IR spectra of pure components: a dataset is realized suitable to realistically represent the ensemble of spectra of the gas mixtures of interest. The informative data deriving from the literature spectra were combined with the stochastic component extracted from a sample spectrum recorded with a prototype instrument, allowing us to build a matrix containing thousands of simulated spectra of gaseous mixtures, accounting for the presence of different components at different concentrations. Signal processing and experimental design techniques were used along the whole path leading to the dataset of simulated spectra. In particular, the goal of the construction of the database lies in the development of a final system to detect drug precursors in the vapor phase. The comparison of some EC-QCLPAS spectra with the corresponding simulated signals confirms the validity of the proposed approach. (C) 2013 Elsevier B.V. All rights reserved.

【 授权许可】

Free