Sulfided Homogeneous Iron Precatalyst for Partial Hydrogenation and Hydrodesulfurization of Polycyclic Aromatic Model Asphaltenes

Abstract

Here, we report the first study to use Fe 2 S 2 (CO) 6 as a well-defined petroleum-soluble precatalyst for partial hydrogenation of polycyclic aromatic and heteroaromatic compounds, including pyrene, phenanthrene, naphthalene, and benzothiophene. The in situ-generated heterogeneous catalyst was characterized using a combination of thermogravimetric analysis (TGA), combustion analysis (CHNS), Fourier transform infrared (FT-IR) spectroscopy, scanning transmission electron microscopy (STEM), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), and scanning electron microscopy-energy dispersive Xray (SEM-EDX). Catalytic performance of Fe 2 S 2 (CO) 6 was evaluated in a batch microreactor coupled with an agitator, varying reaction temperature, pressure, and precatalyst loading. The active phase, consisting of iron sulfide nanoparticles, was prepared from purified Fe 2 S 2 (CO) 6 in toluene; no sulfur additive is required. The results demonstrate that a "presulfided" iron catalyst leads to partial hydrogenation of polycondensed aromatics under moderate conditions, with naphthalene showing only low conversion. Activated carbon, γ-alumina, and activated silica are all effective dispersants for the Fe precursor. The reactivity shows self-consistent substrate dependence, varying with the resonance energy stabilization of the starting compounds and partially saturated intermediates coupled with the surface adsorption enthalpy of the aromatic ring system. The active catalyst derived from this precursor exhibits higher catalytic activity than commercial iron sulfide precatalysts. Importantly, the catalyst is active for hydrodesulfurization (HDS) of benzothiophene, albeit modestly so.

Active learning of chemical reaction networksviaprobabilistic graphical models and Boolean reaction circuits

“…7,8 While RNI is a common component of evaluating biomass reaction systems, 9,10 its application extends to petrochemical and pharmaceutical processes. 11,12 The research of Bhan and coworkers has especially highlighted the importance of identifying the reaction network to construct reliable kinetic models. [13][14][15][16][17] They effectively confirmed the occurring reactions by applying Wojciechowski's criteria to first-rank delplots 18,19 and conducting experiments with product, isotopic labeling, and probe molecule co-feeds.…”