EUVIMEDEUVIMED entdecken
Crossref

Complementary application of high-speed gas chromatography and Fourier transform infrared spectroscopy for enhanced partial least squares modeling and prediction of the physical properties of kerosene-based rocket and jet fuels

Wenjing Ma, Valencia M Parker, Robert E Synovec

Journal of Chromatography A · 2026

DOI: 10.1016/j.chroma.2026.467308 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Chromatography A, 2026. Verfasst von Wenjing Ma und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Chemistry, Box 351700, University of Washington, Seattle, WA, 98195, USA. Hauptthemen laut MEDLINE: Kerosene. Die Arbeit nennt 51 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

It is well established that partial least squares (PLS) regression is a powerful chemometric tool for linking complex fuel composition data to physicochemical properties when coupled with either gas chromatography (GC) or Fourier transform infrared spectroscopy (FT-IR). In this report, we investigate combining the analysis of high-speed GC with flame ionization detection (GC-FID) data with FT-IR data via data augmentation strategies with PLS modeling to optimize the prediction of viscosity, hydrogen content, heat of combustion, and density across a diverse set of 50 kerosene-based fuels. High-speed GC-FID separations were performed at 2-min and 5-fold slower at 10-min separation times, while comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry (GC × GC-TOFMS) served as a high-resolution method to benchmark the PLS modeling performance. PLS models constructed from GC × GC-TOFMS achieved normalized root mean square error of prediction (NRMSEP) values of 2.9% (viscosity), 6.7% (hydrogen content), 10.6% (heat of combustion), and 3.4% (density). GC-FID models with 10-min separations produced similar performance with NRMSEPs of 3.1%, 7.0%, 9.8%, and 3.2%, respectively, while the high-speed 2-min separations yielded 2.7%, 7.4%, 8.1%, and 5.7%. FT-IR models provided competitive performance, achieving NRMSEPs of 5.1% (viscosity), 3.4% (hydrogen content), 9.9% (heat of combustion), and 2.0% (density). Data augmentation of 2-min GC-FID separations and FT-IR spectra further improved the PLS models, yielding NRMSEPs of 2.7% (viscosity), 4.5% (hydrogen content), 6.9% (heat of combustion), and 3.6% (density). Overall, the results demonstrate that high-speed GC-FID and FT-IR spectroscopy provide synergistic analytical modalities with data augmentation for rapid and robust fuel property prediction.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

KeroseneLeast-Squares AnalysisSpectroscopy, Fourier Transform InfraredChromatography, GasViscosity

Stichwörter der Autor:innen

Comprehensive two-dimensional gas chromatography-mass spectrometryFourier transform infrared spectroscopyFuel property modelingHigh speed gas chromatographyPartial least squares regression

Substanzen

Kerosene

Autor:innen und Institutionen

  • Department of Chemistry, Box 351700, University of Washington, Seattle, WA, 98195, USA.
  • Department of Chemistry, Box 351700, University of Washington, Seattle, WA, 98195, USA. Electronic address: synovec@chem.washington.edu.

Zitieren

APA

Ma, W., Parker, V. M., & Synovec, R. E. (2026). Complementary application of high-speed gas chromatography and Fourier transform infrared spectroscopy for enhanced partial least squares modeling and prediction of the physical properties of kerosene-based rocket and jet fuels. Journal of Chromatography A. https://doi.org/10.1016/j.chroma.2026.467308

Vancouver

Ma W, Parker VM, Synovec RE. Complementary application of high-speed gas chromatography and Fourier transform infrared spectroscopy for enhanced partial least squares modeling and prediction of the physical properties of kerosene-based rocket and jet fuels. Journal of Chromatography A. 2026. doi:10.1016/j.chroma.2026.467308

BibTeX
@article{Ma2026,
  author = {Ma, Wenjing and Parker, Valencia M and Synovec, Robert E},
  title = {Complementary application of high-speed gas chromatography and Fourier transform infrared spectroscopy for enhanced partial least squares modeling and prediction of the physical properties of kerosene-based rocket and jet fuels},
  journal = {Journal of Chromatography A},
  year = {2026},
  doi = {10.1016/j.chroma.2026.467308}
}
Kostenlos starten →