During solvothermal fractionation of lignocellulose, lignin de- and repolymerization reactions occur simultaneously with complex transport phenomena. Modeling strategies able to capture intrinsic kinetics can provide key insights for better understanding these transformations. Accordingly, a reaction-diffusion model for simulating solvothermal delignification has been developed as part of this work. The biomass particles are represented as cylindrically-shaped, as this geometry allows for minimizing the complexity of the model, while maintaining its accuracy and general applicability, i.e., through specific descriptors such as porosity. The (effective) diffusivities of lignin species were calculated through the Stokes-Einstein equation, where lignin-solvent interactions were quantified according to the Flory theory. The reaction-diffusion model was applied for simulating hydrothermal pretreatment of birch using hot water and Organosolv pretreatment of poplar using ethanol-water solvent mixtures. Apart from the initial depolymerization, repolymerization reactions lead to the formation of more stable, or recalcitrant, 'lignin species' that can still 'depolymerize', albeit at a reduced rate compared to the initial one. This reactivity loss was found to obey a first-order decay with respect to the concentration of newly formed repolymerized species. A single set of kinetic descriptors, in particular, activation energies amounting to 93.4 and 52.5 kJ mol- 1 for de- and repolymerization reactions were obtained and molecular diffusivities of various species ranged from 1 & sdot;10- 10 to 1 & sdot;10- 8 m2 s-1. According to the Thiele modulus, minimizing diffusion limitations requires finely milled biomass, which is often impractical. To address this, we propose a robust reaction-diffusion framework applicable to different configurations for solvothermal delignification of real feedstocks.

The electrification of high-temperature reactors is a promising strategy for reducing CO2 emissions in energyintensive industries. ElectroThermal Fluidized Bed (ETFB) reactors are particularly attractive due to their ability to supply distributed heat directly through Joule heating. However, reactor-scale models that consistently link electrical design variables to temperature and conversion profiles remain limited. In this work, a coupled electrical-thermal modelling framework for ETFB reactors is developed to address this need. A field-resolved Laplacian Field Model (LFM) is formulated to compute electric potential, current paths, and spatially distributed Joule heating under prescribed electrical properties, enabling detailed analysis of geometry-dependent heating patterns. To support rapid parametric exploration, a simplified Current Partition Model (CPM) based on resistor analogies is also introduced, capturing dominant trends in power distribution, temperature, and conversion at substantially reduced computational cost, while exhibiting systematic deviations under certain geometric conditions. The framework is applied to the highly endothermic thermal decomposition of carbonyl sulfide (COS), illustrating how electrode configuration governs Joule heating distribution and indirectly influences thermal and reactive behaviour. In addition, the thermal formulation is assessed using available steady-state ETFB operating data from the literature. From the reported power and temperature data, the bed resistivity and effective heatloss parameters are estimated. With these values, the model reproduces the experimental electrode-tip temperatures within good accuracy across the range of immersion depths. Overall, the proposed methodology offers a structured reactor-scale approach to guide ETFB design and sizing, linking electrical configuration, heat demand, and conversion targets, while remaining compatible with future extensions.

Aldol reactions are important carbon‑carbon bond formation reactions for biomass upgrading and fine-chemical synthesis. Performing this reaction in aqueous media requires hydrothermally stable and highly active catalysts. To address this challenge, crosslinked copolymers containing polyethylene glycol methacrylate (PEGMA) and hydrophobic butyl (BMA) or lauryl methacrylate (LMA) monomers were developed as catalyst supports, aiming to create low-polarity microenvironments around the active sites and promote NH2/OH cooperativity. These copolymers were synthesized via bulk polymerization using poly(ethylene glycol) dimethacrylate as cross-linker, followed by chlorination and amination, and characterized using FT-IR, solid state 13C NMR, elemental analysis, and SEM. The catalytic performance was evaluated in the aldol reaction of 4-nitrobenzaldehyde with acetone in different solvents at 55 °C using batch and continuous flow reactors. The amine-functionalized PEGMA/BMA copolymer with 30 mol% BMA exhibited the highest activity in water (turnover frequency of 2.6 ± 0.17⨯10−3 s−1). Lower or higher BMA contents, or substitution with LMA, resulted in a reduced catalytic performance, attributed either to a lower hydrophobicity, unfavorable spatial arrangements, or steric hindrance. Also, the results showed that the activity of this catalyst decreased significantly in DMSO/n-hexane as solvent and increased gradually by water addition due to the essential role of water in limiting the formation of the site blocking species. Finally, continuous operation confirmed excellent stability with only 5% loss of activity after 16 h. Overall, this work highlights that precise engineering of the interfacial dielectric properties by tailoring the hydrophobic content of the catalyst is an effective strategy to achieve higher conversion in aqueous media.

Ozone-assisted catalytic oxidation (OzCO) has emerged as a promising alternative to conventional catalytic oxidation (CCO) for the elimination of volatile organic compounds (VOCs), offering efficient conversion at lower temperatures. In this work, a series of monometallic (Cu/HAP, Mn/HAP) and bimetallic (CuzMn/HAP, z = 0.5, 1, 2) catalysts were synthesized by wet impregnation using hydroxyapatite (HAP) as support and systematically evaluated for propane (C3H8) OzCO. Comprehensive physicochemical characterization demonstrated that Cu-Mn interactions improved metal dispersion, reducibility, oxygen mobility, and the concentration of surface-adsorbed oxygen species. Mn/HAP exhibited the highest C3H8 conversion, whereas Cu/HAP showed the greatest CO2 selectivity. The bimetallic catalysts displayed clear synergistic effects, with Cu1Mn/HAP achieving the optimal balance between C3H8 conversion and CO2 selectivity. Mechanistically, MnOx species facilitated O(3 )decomposition to generate active oxygen species, while CuOx species rather enhanced the further oxidation of CO into CO2. The Cu1Mn/HAP catalyst exhibited optimal performance at intermediate space times and temperatures, high O3 partial pressures, and low C3H8 partial pressures. Water vapor inhibited C3H8 conversion at 80 degrees C but enhanced CO2 selectivity, with the effect being reversible and less pronounced at higher temperatures. Overall, Cu1Mn/HAP demonstrated robust performance, highlighting its potential as an efficient catalyst for low-temperature C3H8 OzCO.

This study investigates how the structure of hydrolysis lignins (HLs) impacts both their solubility and behavior in mild reductive catalytic depolymerization (RCD). Thorough characterization of 8 different HLs shows that solubility and depolymerization outcomes depend on intrinsic lignin properties as well as carbohydrate content. A 70/30 vol% ethanol/water mixture is generally the most effective solvent mixture, although the absolute solubility varies substantially among the HLs, with a high carbohydrate content and molecular weight typically lowering the solubility. The β-O-4 bond content is identified as the primary factor dictating both the monomer yield and number of para-substituted side chains formed during RCD, with a linear relationship between β-O-4 bond cleavage and monomer/side chain formation. However, the syringyl/guaiacyl ratio further steers monomer yields, with syringyl-rich lignins generating more monomers per cleaved β-O-4 bond. Despite structural diversity, para-substituted side chain selectivity in the monomer fraction remains consistent for all HLs, with approximately 85% propanol-substituted monomers. Similar trends were observed for other technical lignins, underscoring the broad applicability of the RCD protocol, although minor variations in side-chain selectivity were noted. Carbohydrate content also affects the evolution of aliphatic and carboxylic acid hydroxyl groups during RCD, with higher carbohydrate contents leading to sharp initial decreases in the former and increases in the latter. In all HLs formyl native end groups are reduced to methyl groups, and ethyl ester formation via esterification of carboxylic acid groups is confirmed. Lastly, high-molecular-weight HLs (> 9000 g/mol in this study) may experience initial diffusion limitations, delaying efficient depolymerization.

Via mild reductive catalytic depolymerization of lignin, the most abundant aromatic biopolymer on Earth, various functionalized aromatic monomers and oligomers can be produced. Noble metal catalysts, either monometallic or bimetallic, are commonly used for this purpose. In this study, the catalytic activity and selectivity of noble monometallic (Pd, Pt, Ru) and bimetallic noble-base (combination with Cu, Co, Fe) metal catalysts are investigated for the mild reductive catalytic depolymerization of technical hydrolysis lignin. Calcined Pd-and Pt-based materials are reduced during the reaction and exhibit comparable catalytic activity, both in terms of molecular weight reduction and monomer yield after depolymerization. In contrast, calcined Ru-based catalysts must be prereduced to become active, form larger nanoparticles after calcination, and exhibit lower catalytic activity. Additionally, Pd-based catalysts are uniquely able to modify various interunit linkages present in lignin, such as the (3-(3 bond and (3-5 bond. For the selectivity towards (3-O-4 bond derived side chains, Pd favors para-propanol side chains, Pt both para-propanol and para-propyl, while Ru predominantly forms para-propenyl side chains. The incorporation of a base metal, particularly Cu, leads to altered catalytic performance, resulting among others in changes in side chain selectivity. Lastly, the results indicate that formyl native end groups are fully converted to para-methyl side chains with the Pd-and Pt-based catalysts, whereas many of these groups remain intact with the Ru-based catalysts.

When aiming at designing performant catalysts for existing or novel (sustainable) reactions, a wide variety of methodologies can be adopted, turning catalyst design into a science on its own. However, fundamental knowledge and clear guidelines for the application of these methodologies for catalyst innovation are lacking, impeding truly efficient and adequate catalyst design, with serendipity still playing too important a role. This review fills this gap by systematically categorizing and reviewing the most relevant methodologies and providing their fundamentals. Despite the strong benefit from systematically collected data incoming from digital laboratories, the main common drawback remains how to design catalysts truly "out of the box"? While various methodologies can serve more humble purposes (e.g., fine-tuning catalyst compositions), truly innovative catalyst design can only be achieved via a knowledge-driven approach, strongly supported by kinetic models in combination with suitable catalysis informatics tools.

Developing stable, hydrophilic heterogeneous catalysts with high active site densities remains a major challenge for the scalability of the green, sustainable, and aqueous-phase production of complex chemicals. Herein, a new highly stable 1-(2-aminoethyl) piperazine (AEP)-functionalized poly[ethylene glycol] methacrylate (PEGMA) catalyst is synthesized and investigated for the aqueous-phase aldol reaction between acetone and 4-nitrobenzaldehyde. Density-functional theory (DFT) was used to elucidate the stepwise PEGMA-to-AEP functionalization pathway and to rationalize the preferential reactivity of the primary amine. Spectroscopic and elemental analyses confirmed successful covalent anchoring of AEP onto the PEGMA backbone. The catalyst exhibits a high active site concentration (0.66 +/- 0.02 mmol g(-)(1)), substantially enhancing the catalytic performance. Increasing the catalyst active-site concentration from 0.100 to 0.397 mmol g(-)(1) led to an increase in aldol product formation from 0.008 to 0.024 mol L--(1) after 4 h, indicating a proportional catalytic activity. Comparative kinetics evaluation using zero-, first-order, and Langmuir-Hinshelwood (L-H) models indicates that the L-H best describes the reaction kinetics, with the lowest residual value (2.19 & times; 10(-5)). Under continuous-flow operation, AEP-PEGMA maintains similar to 30% conversion and similar to 99% aldol selectivity for over 40 h at room temperature. Overall, this work demonstrates that AEP-PEGMA is a potential heterogeneous catalyst for scalable and sustainable chemical processes, effectively addressing issues related to low active site availability, synergistic precursor-support interaction, and catalyst deactivation.

Process development and optimization of multiphase reactions is often challenging due to the complex interplay between reaction kinetics and mass transfer. This work analyzes an industrial-scale pharmaceutical debenzylation of an amine-containing precursor catalyzed by a Pd/C catalyst. Using historical lab-scale data, a kinetic model was constructed that reveals competitive adsorption between hydrogen and the precursor, resulting in a reduced debenzylation rate at higher dissolved hydrogen concentrations. When accounting for gas-liquid mass transfer at the pilot scale, lower pressures and stirring speeds limit hydrogen availability due to mass transfer, while higher pressures and stirring speeds increase the dissolved hydrogen concentration, inducing more pronounced adsorption competition and, hence, reducing the reaction rate. These findings indicate the existence of an optimal intermediate operational window that minimizes conversion times by balancing hydrogen availability against catalyst inhibition. Methodologically, this work highlights how multiscale modeling based on historical lab and pilot data can improve industrial-scale predictions and supports more efficient, better-understood pharmaceutical manufacturing.

A promising route for the valorization of acid gas (CO2 and H2S) components involves their simultaneous transformation into carbon monoxide and sulfur, through carbonyl sulfide (COS) intermediate. In this work, we systematically explore the catalytic performance of 13X and 4 A for COS formation under varying conditions of temperature, acid gas partial pressure, and zeolite hydration state. H2S, CO2, and COS breakthrough experiments at 45 degrees C reveal that the capacities of all three molecules are higher for 13X than for 4 A. Thermal gravimetry on hydrated zeolites specifies water contents of 13.1 and 12.0 mmol/g for 13X and 4 A, respectively. COS yield is highest at 100 degrees C, showing temperature dependence in the case of 13X; in contrast, 4 A retains more than 70 % of its maximum activity over an extended range of temperatures. An increase in acid gas partial pressure from 0.2 to 0.8 bar gradually increases the total COS in 13X, whereas the activity of 4 A remains constant. Likewise, COS formation increases with decreasing zeolite hydration; threshold-dependent in 13X but progressive and relatively less pronounced in 4 A. Both zeolites, independent of conditions, undergo a decline in COS formation over time due to the water-induced inhibition of active sites, attributed to poisoning. While activity in 4 A decays rapidly, 13X exhibits a more gradual decay, corresponding to the inhibitory effect of the produced water being less pronounced in 13X than in 4 A. This reduces competitive adsorption on active sites and mitigates site blockage in 13X, which in turn preserves catalytic performance over time, indicating that 13X is more sensitive to changing conditions than 4 A. An optimum operating window identified for the two materials can help reduce the energy required for the industrial conversion of acid gas and subsequent catalyst regeneration. This corresponds to reaction at 120 degrees C and 250 degrees C and the regeneration at 250 degrees C and 300 degrees C for 13X and 4 A, respectively.

The Single-Event MicroKinetic modelling methodology has been successfully applied to "steady-state" methanol-to-olefins data acquired over a highly acidic H-ZSM-5 catalyst with Si/Al ratio of 40 at temperatures between 375 degrees C and 450 degrees C with space times ranging from 0.74 to 15.5 kg(cat) s mol(MeOH)(-1) at 40 kPa MeOH partial pressure. Next to the alkene homologation cycle, also the aromatic hydrocarbon pool cycle is implemented in the model in terms of elementary steps. 14 of the 26 model parameters (mainly so-called kinetic descriptors) are established literature values, without further adjustment, owing the fundamental nature of the model, while the remaining ones (mainly the catalyst descriptors) were statistically estimated with physicochemical relevance. The relative importance of the alkene homologation cycle and the aromatic hydrocarbon pool was probed through newly introduced model-calculated descriptors, such as the 'methanol consumption ratio' and the 'ratio of aromatic methylation to alkene reactions'. These descriptors revealed that at higher temperatures and space times, the alkene homologation cycle dominates due to reduced methanol availability and faster alkene reactions. Furthermore, increasing temperature or conversion shifts the importance towards cracking and alkylation reactions within the cycles. The model-based descriptors and, more particularly, how their trends as a function of temperature and conversion aligned (or deviated) from the experimentally observed ethylene/(isobutane+isobutene) ratio, provide valuable insights into how operating conditions influence catalytic cycles, aiding in optimizing product selectivity in MTO processes.

Increasing temperature shifts COS pyrolysis from bimolecular consumption toward pressure-dependent unimolecular decomposition, requiring reversible bimolecular and sulfur-mediated pathways for reactor-scale prediction.

Machine learning is increasingly used to guide catalyst discovery, but it remains unclear how much reported performance reflects genuine predictive ability. In this work, the evaluation protocol alone is shown to manufacture or erase apparent success. Using the largest experimental dataset in heterogeneous catalysis, 12,708 measurements across 59 catalysts for the oxidative coupling of methane, we find a stark performance hierarchy: (i) random splitting looks excellent (MAE ≈0.75 %, R2 ≈0.90), (ii) holding out process conditions is harder, and (iii) holding out entire catalyst compositions, the only test that reflects real catalyst discovery, collapses accuracy nearly three-fold (MAE ≈2.1 %, R2 ≈0.32). To probe why, we generate and openly release a noisefree, condition-matched library of 70,400 datapoints across 326 virtual catalysts from a validated microkinetic model. The same gap appears without any measurement noise, so the barrier is intrinsic: predicting across a discrete compositional space sampled by only dozens of catalysts. Two findings point the way forward. Data quality beats quantity, a small curated set of active catalysts rivals a library five to six times larger. Additionally, feature attribution reveals that models lean almost entirely on process conditions while ignoring composition, exposing compositional encoding as the central unsolved problem. Honest evaluation changes the verdict on what machine learning can, and cannot, yet do for catalyst discovery.

Lignin depolymerization is an important yet very challenging step in its valorisation into chemicals (or fuels) due to the feedstock's recalcitrance and structural variability. Advanced process modelling and, more particularly, kinetic modelling accounting for the feedstock composition, could enable rapid tailoring of operating conditions to achieve the desired product quality and, thus, enhance the flexibility of biorefineries in processing various streams. Inspired by the continuous theory of lumping as applied to hydrocarbon reactions, our present work introduces a modelling approach to simulate lignin depolymerization kinetics in the presence of a catalyst. The model extrapolates the reactivity of typical lignin moieties such as ether and carbon-carbon bonds towards (macro)molecules of various sizes, allowing to simulate the evolution of the molecular weight distribution (MWD) in terms of parent and newly formed lignin species. The yield distribution function (YDF), a key element for describing the quantities of the various species formed, brings an elementary understanding of the cracking mechanism and, more particularly, the role of different functionalities therein, within reach. A proof-of-concept application for an Organosolv lignin depolymerization in the presence of Ru/C catalyst at 325 degrees C demonstrates the robustness of the model. The proposed modelling methodology is deemed to be potentially applicable to any lignin type as it is described in terms of depolymerization kinetics based on the transformation of typical functional groups, c.q., reactive moieties.

The mounting crisis of plastic waste, coupled with the shortcomings of mechanical recycling, underscores the urgent need for scalable chemical recycling solutions. Among these, pyrolysis followed by steam cracking emerges as a promising pathway to transform polyolefin-rich waste into virgin-grade olefins. Yet, the complex nature of pyrolysis oils has hindered their industrial use. In this study, hydrotreatment was investigated as an upgrading strategy for distilled pyrolysis oils derived from real mixed plastic waste. Two diesel-range fractions of pyrolysis oil, raw and hydrotreated, were characterized using two-dimensional gas chromatography (GC × GC) and evaluated in a bench-scale steam cracker. Hydrotreatment provided significant compositional refinement, lowering nitrogen from 1546 mg·kg-1 to <10 mg·kg-1, reducing olefins from 51 wt% to 8 wt%, and eliminating detectable oxygen (1776 mg·kg-1 initially). These changes brought the product within industrial feedstock specifications. These improvements translated into enhanced cracking performance. The hydrotreated pyrolysis diesel blend achieved ethylene yields of up to 32 wt% at 880 °C, surpassing fossil naphtha while suppressing CO and aromatic by-products such as benzene, toluene, and xylenes (BTX). The results establish hydrotreatment as a critical enabler for converting waste-derived pyrolysis oils into cracker-ready feedstocks, supporting their integration into petrochemical infrastructure and advancing a circular economy for plastics.

Heterogeneously catalyzed hydroformylation over a commercial 5% rhodium nanoparticle catalyst has been performed within the intrinsic kinetics regime in a high-throughput kinetics setup. Owing to the use of a paraffinic solvent, the reaction was carried out either in the gas or the liquid phase. An ethylene conversion of around 4% mol/mol was obtained in the gas phase, whereas liquid-phase operation allowed achieving around 9% mol/mol conversion under comparable reaction conditions. The presence of the paraffinic solvent is supposed to better tune the reactant concentration to which the catalyst is exposed. Propanal and ethane were the main products observed, the highest propanal selectivity, 75% mol/mol, being obtained at the lowest temperature, 120 degrees C. Apparent activation energies for both hydroformylation (59 kJ/mol (l) / 68 kJ/mol (g)) and hydrogenation (87 kJ/mol (l) / 94 kJ/mol (g)) were found to be lower at liquid compared to gas-phase conditions, suggesting a lower overall surface coverage at liquid-phase conditions. Ethylene and hydrogen were found to exhibit a positive impact on gas-phase hydroformylation and hydrogenation, resulting in higher ethylene conversion when increasing their molar reactant ratios. Based on the positive impact of ethylene observed in the gas-phase operation, it is likely that in the liquid phase, where ethylene solubility is higher, the altered molar ratio distribution at the catalyst surface further enhances yields toward propanal.

Lignin holds enormous potential for biofuel production through thermal and thermochemical processes. However, its high oxygen content in the form of phenolics leads to a low heating value, low stability, and high viscosity. Catalytic hydrodeoxygenation (HDO) with metal-containing catalysts with appropriate oxygen vacancies and active sites represents an efficient route for removing oxygen and producing high-quality liquid hydrocarbons. In this context, our present work focuses on synthesising Ni-doped mixed oxide (Ti1-xZrxO2) supports for the HDO of lignin model compounds (anisole, phenol, cresol). Notably, 3 wt% Ni doped Ti0.50Zr0.50 (Ni/Ti0.50Zr0.50O2) exhibits superior activity and selectivity, achieving a 94% yield of cyclohexane as the primary deoxygenated product, with complete conversion of anisole at 230 degrees C. In contrast, the mono metal oxide-based support catalysts Ni/ZrO2 and Ni/TiO2 yield 29% and 49% of cyclohexane, respectively. The higher activity and selectivity towards deoxygenated products can be attributed to high oxygen vacancies, Lewis-acid strength and strong metal-support interaction compared to Ni/TiO2 and Ni/ZrO2, as confirmed by the O2-temperature programmed desorption, X-ray photoelectron spectroscopy, NH3-diffuse reflectance infrared Fourier transform spectroscopy and H2-temperature programmed reduction studies. The use of a low percentage (3 wt%) of non-precious metal doped mixed oxide (Ti1-xZrxO2) supports with enhanced oxygen vacancies opens a new window for exploring effective heterogeneous catalysts for the transformation of lignin-derived bio-oil into fuel-grade hydrocarbons.

The photocatalytic degradation of ethanol via oxidation reaction on a TiO2 photocatalyst is studied experimentally and numerically for UV light intensities up to 400 W m-2 on the photocatalyst surface. The effect of increasing light intensity on the reaction rates for ethanol, acetaldehyde, and acetic acid oxidation is quantified, as is the transition point from a proportional dependency to one determined by a fractional power dependence (alpha), Ri proportional to I alpha. The kinetic regime limited by the fractional power dependence reflects a condition that favors electron-hole recombination during the photocatalytic process. Reactive 3D CFD simulations using a discrete ordinate model are carried out. The extended kinetic model that incorporates the effect of UV incident radiation on the oxidation rates is included. As computational costs are significant, a surrogate model is developed. A 1D plug flow reactor model is constructed by including dimensional reductions in both the applied hydrodynamic model and radiation model. The 1D surrogate model reduces computational time and data storage requirements and provides quantitatively reliable ethanol photocatalytic oxidation results with a mean absolute error of 5 % compared to 3D CFD calculated values.

ZSM-22 zeolites with different Si/Al ratios (38, 50, 80) were prepared via a hydrothermal synthesis method, investigated for the catalytic dehydration of 1,3-butanediol (1,3-BDO) to butadiene (BD) at 300 °C. The catalytic performance of the synthesized materials was related to their properties and compared to a commercial ZSM-22 zeolite (Si/Al = 30). ZSM-22 (50) exhibited a quick decline in conversion, a lower BD selectivity, and higher propylene selectivity compared to the other materials, which could be attributed to the presence of strong Lewis acid sites and silanol nests. The Lewis sites favor the cracking of the intermediate 3-buten-1-ol (3B1OL) into propylene, while the silanol nests interact with the free hydroxyl group of 3B1OL, potentially inhibiting further dehydration towards BD. The highest initial BD yield of 74% was observed over ZSM-22 (80), while the highest initial BD productivity of 2.7 gBD·g−1cata·h−1 was achieved over ZSM-22 (38). After 22 h time on stream (TOS), c-ZSM-22 and ZSM-22 (38) outperformed previously reported catalysts from the literature, with productivities amounting to 1.3 gBD·g−1cata·h−1 and 1.2 gBD·g−1cata·h−1, respectively, at a site time of 6.6 molH+·s·mol−11,3-BDO.