From copper-chromite to MXenes: A century of furfural hydrogenation – Experiments, DFT, machine learning, and path forward

Furfural hydrogenation is a key step in converting biomass-derived platform molecules into value-added chemicals and fuels. This review covers experimental and computational work from 1929 to 2025, starting with early copper and nickel catalysts and moving to today's bimetallic alloys, MXene-supported systems, and Pt/aluminosilicates. We examine how metal composition, support acidity, solvent choice, and hydrogen pressure steer selectivity toward furfural alcohol (FA), 2-methylfuran (2-MF), tetrahydrofurfural alcohol (THFA), or the two pentanediols. Density functional theory (DFT) calculations, including our recent work, reveal adsorption geometries, activation barriers, and branching pathways, while machine-learning (ML) models predict high FA selectivity for optimized bimetallic systems. A unified scheme is proposed that integrates initial carbonyl hydrogenation, side-chain hydrogenolysis, full ring saturation, and ring-opening routes, whether from THFA or directly from FA (our patented catalysts and recent Pt/aluminosilicate results). Key factors controlling selectivity are discussed in detail. Catalyst deactivation, noble-metal cost, and scale-up remain hurdles; non-precious metals, robust supports, and combined theory-experiment loops offer practical fixes. Techno-economic viability is highlighted, with minimum selling prices (MSP) of ∼$1300 t −1 for FA (single-step) and higher for diols due to complexity, alongside lifecycle benefits (near-zero or net-negative CO 2 with renewable H 2 for FA production). The discussion concludes with concrete directions for selective, stable, and industrially viable furfural upgrading. Mechanistic insights from XPS, STM, DFT, and ML guide catalyst design and performance optimization. • Furfural hydrogenation catalysts advance from 1920 copper-chromite to present Pt/aluminosilicates and MXenes. • Unified reaction mechanism links hydrogenation, hydrogenolysis, ring saturation and ring-opening via ML and DFT. • Selectivity is governed by metal-support acidity, solvent polarity, and hydrogen pressure. • Deactivation, noble-metal cost, and scale-up are met by non-precious metals, robust supports, process intensification. • Techno-economic viability and net-negative emissions enabled by renewable H 2 and circular biorefinery design.