Investigation and Optimisation of THz Simulants for Explosives and Illegal Drugs: An Investigation of Simulant Improvements for TNT and MDMA by Constrained Optimisation of Powders in the THz Spectral Range
Translated title
Investigation and Optimisation of THz Simulants for Explosives and Illegal Drugs
Author
Hansen, Carsten Pagaard
Term
4. term (FYS10)
Education
Publication year
2026
Submitted on
2026-05-29
Abstract
This master’s thesis investigates how to identify and combine materials that can mimic the measurable properties (in particular infrared absorbance spectra) of the explosive TNT and the illegal drug MDMA. The goal is to find so‑called simulant materials – harmless mixtures that behave like TNT and MDMA in measurement instruments – and to optimise the mixing ratios of these simulant blends. A range of different ingredient types is identified and combined mathematically using linear combinations of their absorbance spectra. In the same way, linear combinations are constructed to reproduce the spectra of TNT, the explosive PETN, MDMA, and the illegal drug cocaine using different sets of ingredients. Two main approaches are used to find the best simulant mixtures: (1) an unconstrained optimisation, where the difference between the target and the obtained spectrum is minimised, and (2) a constrained objective function developed in this thesis, which also incorporates practical constraints on the mixtures. The method used to minimise this objective function is described, and the results of both approaches are presented and compared. The thesis identifies optimised simulants for TNT and MDMA and concludes that certain pharmaceutical compounds are better suited as ingredients than, for example, sugars, artificial sweeteners, and amino acids. To assess how closely the simulants resemble the real substances, the Pearson correlation coefficient and a weighted cross‑correlation coefficient are used as measures of spectral similarity. Overall, simulant mixtures that include pharmaceuticals are found to show a higher degree of spectral similarity to TNT and MDMA. The theoretically derived mixing ratios for TNT and MDMA simulants are tested experimentally. Taking all relevant uncertainties into account, the experimental measurements are found to agree well with the theoretical predictions.
Dette speciale undersøger, hvordan man kan finde og sammensætte materialer, der efterligner de målbare egenskaber (særligt infrarøde absorbansspektre) for sprængstoffet TNT og det illegale rusmiddel MDMA. Formålet er at finde såkaldte simulantmaterialer – altså uskadelige blandinger, der opfører sig som TNT og MDMA i måleudstyr – samt at optimere blandingsforholdene i disse simulantblandinger. Der identificeres en række forskellige typer ingredienser, som kombineres matematisk ved hjælp af lineære kombinationer af deres absorbansspektre. På samme måde laves lineære kombinationer, der efterligner TNT, sprængstoffet PETN, MDMA og det illegale rusmiddel kokain ved brug af forskellige ingredienser. I arbejdet anvendes to tilgange til at finde de bedste simulantblandinger: (1) en ukonstraint optimering, hvor forskellen mellem det ønskede og det opnåede spektrum minimeres, og (2) en konstrueret, begrænset målfunktion, som også tager praktiske hensyn i blandingerne med i beregningen. Metoden til at minimere denne målfunktion beskrives, og resultaterne af de to tilgange sammenlignes. Specialet finder optimerede simulantblandinger for både TNT og MDMA og konkluderer, at visse lægemiddelstoffer er bedre egnede som ingredienser end fx sukkerarter, kunstige sødemidler og aminosyrer. For at vurdere, hvor godt simulantblandingerne ligner de rigtige stoffer, anvendes Pearson-korrelationen og en vægtet krydskorrelationskoefficient som mål for spektral lighed. Generelt viser resultaterne, at simulantblandinger, der inkluderer lægemidler, har en højere grad af spektral lighed med TNT og MDMA. De teoretisk udledte blandingsforhold for TNT- og MDMA-simulantblandingerne testes i laboratorieforsøg. Inden for de vurderede usikkerheder stemmer de eksperimentelle målinger godt overens med de teoretiske forudsigelser.
[This abstract has been rewritten with the help of AI based on the project's original abstract]