Insight into the alkali metal poisoning sensitivity of V2O5-WO3/TiO2 catalysts for NOx abatement via machine learning and in situ Raman spectroscopy

Jiang Deng, Shiqi Guo, Yuejin Li, Penglu Wang, Li Chen, Jun Liu, Xingchi Li, Guang Yan, Sha Wang, Zaisheng Jin, Ming Xie, Dengsong Zhang

Research output: Contribution to journalArticlepeer-review

2   Link opens in a new tab Citations (SciVal)
5 Downloads (Pure)

Abstract

V2O5–WO3/TiO2 (VWTi) catalysts for NH3–SCR suffer severe poisoning by alkali metals, especially K, yet the site-specific poisoning mechanism remains unclear. Herein, we elucidate the poison mechanism based on a comprehensive investigation consisting of experimental work, theory calculation, and machine learning, conducted by controlling the VOx density and K/V ratio. Using a variety of characterization techniques, we found that the SCR activity of a VWTi catalyst was governed by its redox ability and the Lewis acidity dominated by V4+. The terminal V=O group is a Lewis acid and can adsorb NH3, while the bridging V–O–V group serves as a redox center, capable of activating NO/O2. K poisons a VWTi catalyst by attacking the strong Brønsted acids first and then the strong Lewis sites, resulting in a nonlinear progression of activity decline, which is slow initially but accelerates with increasing K accumulation. This phenomenon is especially evident for high–V loading catalysts dominated by the polymeric VOx species. Density functional theory calculations reveal that K poisons VWTi catalysts by binds K to the terminal V=O sites, forming the chemically inactive KVO3 compound and weakening the NH3 adsorption on the neighboring VOx. This work offers a comprehensive understanding of the site-specific sensitivity of VOx species to alkali metal poisoning and provides important insights to the deactivation process, which could be used to design practical VWTi catalysts for commercial applications.

Original languageEnglish
Article number122956
JournalChemical Engineering Science
Volume321
Issue numberPart C
Early online date9 Nov 2025
DOIs
Publication statusPublished - 1 Feb 2026

Data Availability Statement

Data will be made available on request.

Keywords

  • In situ Raman Technology
  • K Poisoning
  • Machine Learning
  • Selective Catalytic Reduction of NH
  • V–based Catalysts

ASJC Scopus subject areas

  • General Chemistry
  • General Chemical Engineering
  • Industrial and Manufacturing Engineering

Fingerprint

Dive into the research topics of 'Insight into the alkali metal poisoning sensitivity of V2O5-WO3/TiO2 catalysts for NOx abatement via machine learning and in situ Raman spectroscopy'. Together they form a unique fingerprint.

Cite this