ACS Engineering Au最新文献

To alleviate detrimental effects associated with anthropogenic emissions, the use of CO₂ and H₂ as feedstocks for their conversion to dimethyl ether (DME) with tandem catalysts is an attractive and sustainable route. First, we investigated the catalytic activity of bifunctional admixtures of Cu-ZnO-ZrO₂ (CZZ) and a silicoaluminophosphate, SAPO-34, for CO₂ hydrogenation to DME and optimized their reactivity with an emphasis on identifying optimum synthesis conditions for CZZ including Cu:Zn:Zr molar ratio and aging and calcination temperatures. The highest methanol (MeOH) productivity (10.8 mol kg_cat ^-1 h^-1) was observed for CZZ-611 aged at 40 °C and calcined at 500 °C. When coupled with SAPO-34, CZZ/SAPO-34 reached 20% CO₂ conversion and 56% DME selectivity at optimized conditions (260 °C, 500 psig, and 2000 mL g_CZZ ^-1 h^-1) and was stable for 50 h time-on-stream, with a slight reduction in activity. Next, we performed kinetic modeling to translate lab-scale findings to industrial packed-bed reactors followed by a techno-economic analysis (TEA) with cradle-to-gate environmental footprint evaluation to evaluate its industrial applicability. A TEA of a 20,000 tpy DME plant revealed raw material costs as the main operating cost drivers (H₂ cost comprises 47% of total cost). Considering green H₂ ($4/kg H₂) and captured CO₂ as feed, the minimum DME selling price (MDSP) was $3.21/kg, ∼2.7× higher than the market price ($1.2/kg). MDSP drops to $1.99/kg with gray H₂ ($1/kg H₂) and fluctuates ±$0.14 with changes in CAPEX (±30%) and other economic factors. The plant's carbon footprint was mainly affected by the H₂ source. Green and gray H₂ resulted in emissions of 0.21 and 4.4 kg CO₂ eq/kg DME, respectively. Importantly, a negative carbon footprint can be achieved by using green H₂ and CO₂ captured directly from air. Overall, our work shows tandem catalysis as a promising approach toward sustainable DME production and identifies the pathway toward making it cost-competitive with fossil fuels.

Knowledge of the equilibrium and kinetics of reactions is critical for optimizing industrial chemical processes. In this study, the equilibrium and kinetics of esterification reactions were systematically investigated for a series of carboxylic acids (acetic acid, propionic acid, formic acid, and levulinic acid) and alcohols (methanol, ethanol, n-propanol, and n-butanol), giving a total set of 16 esterification reactions at different temperatures. First, formation properties of reactants and products were utilized to calculate the reaction equilibrium constants K_eq of these reactions. These were compared with K_eq values obtained by one equilibrium experiment coupled to PC-SAFT predictions. The comparison yielded outstanding agreement between PC-SAFT-assisted K_eq values and the formation-property-derived K_eq values. The K_eq values were then used in activity-based kinetic expressions, and the predicted reaction kinetics were validated against experimental data to demonstrate the model’s accuracy. The deviations between PC-SAFT and experimental data were AAD% (K_eq) = 1.66% for the reaction equilibrium and AAD% (r) = 13.8% for the kinetic curves. The Arrhenius equation and van ’t Hoff equation were applied to depict the temperature dependence of reaction rate constants and of K_eq for each esterification reaction in a range of 303.15–423.15 K. Thus, activity-based thermodynamic standard properties are provided in this work, guiding the optimization of esterification reactions in a broad range of conditions.

In catalytic fixed bed reactors for highly exothermic reactions, the bed is often diluted with inert particles to prevent thermal runaway and to distribute the reaction more homogeneously along the reactor length. The partial oxidation of methanol to formaldehyde is an example with high industrial relevance, in which diluted fixed beds are applied. In this work, particle-resolved computational fluid dynamics (PRCFD) simulations are conducted for the hotspot region (0–0.5 m) of an industrial scale fixed bed for formaldehyde production to systematically investigate the impact of dilution on integral reactor performance and locally distributed quantities, such as the temperature and catalyst effectiveness factor. PRCFD is the most detailed modeling approach for the simulation of diluted fixed beds since the spatial resolution of the fixed bed geometry allows the inert particles to be considered directly without the implementation of averaged activity factors. Different catalyst distributions have a significant effect on integral conversion, hotspot formation, and catalyst overheating while increasing the inert thermal conductivity has only a minor impact on heat transport and hence reaction. The difference between the maximum catalyst temperature of two different catalyst arrangements can reach 34 K. Finally, the present study demonstrates that even highly diluted fixed beds with industrial particle and tube dimensions are not suited to perform intrinsic kinetic measurements for the partial oxidation of methanol because of catalyst overheating (ΔT = 23.12 K) and pore diffusion limitation (η_i,FA < 0.5).

Experimental data from different sources present challenges due to variability and noise from various experimental conditions, apparatuses, and environmental factors. In this work, we propose a general method to address these challenges to build a consistent data set. As a case study, we analyze experimental data sets of methane’s thermal conductivity across the liquid, vapor, and supercritical phases. The method is based on machine learning (ML) techniques, which consistently integrate data from various experimental sources. It feeds raw data compiled by the National Institute of Standards and Technology (NIST) database to different ML algorithms to achieve this purpose. Our findings indicate that ML models yield predictions closer to the NIST’s processed data than to the original raw experimental data used to train the models. This demonstrates the models’ generalization from heterogeneous, noisy, and untreated data sets. While our approach does not eliminate preprocessing, it suggests that ML can autonomously handle noisy data, providing a faster and cost-effective alternative to traditional pre- and postprocessing methods. By guiding the refinement of labor-intensive methods, ML proves adaptable for real-time data, enabling immediate adjustments and revolutionizing industrial and scientific optimizations. Therefore, the proposed ML approach is general and efficient in handling complex and heterogeneous data to deliver reliable predictions without extensive preprocessing.

Traditional fixed-bed reactors use pellets that only allow species transport through diffusion, not convection. However, using highly porous pellets, such as open-cell foams, allows bulk gas to move through them. Such a fixed bed results in a lower pressure drop and intimate contact between the gas and solid phases, which is desirable for catalytic reactions and adsorption processes. A common strategy for modeling fixed beds is to use the porous medium assumption, where the gas and solid phases are represented as an effective porous medium. This approach necessitates several effective properties calculated using analytical relations for simple geometries and empirical correlations for complex geometries. However, such a representation for a fixed bed of highly porous pellets is unavailable. This study addresses this problem by developing a mathematical framework for a fixed bed reactor of open-cell foam pellets as a porous medium. To this end, the volume averaging and asymptotic averaging techniques are employed. The governing equations for the porous medium (continuum) model are developed based on the volume averaging technique, and the effective properties are calculated using the unit cell simulations. The developed mathematical framework is assessed against three-dimensional particle-resolved simulations for linear and nonlinear catalytic kinetics and CO₂ adsorption. For all the test cases, the developed framework can reproduce the pressure drop and species concentration predicted by the particle-resolved simulations with orders of magnitude reduction in the simulation time.

We envision periodic open cellular structures (POCS) with streamlined elliptical struts as potential intensified structured catalytic supports. Streamlined elliptical struts aligned to the flow direction substitute conventional cylindrical ones, aiming at reducing the pressure drop while increasing the surface area for catalyst deposition. Reactive computational fluid dynamics simulations are employed for the fundamental investigation of mass transfer coefficients and friction factors. The effects of the design parameters (i.e., porosity ε, angle between the struts’ axis and the streamwise direction α and elliptical strut elongation R) are evaluated. The POCS transport properties are significantly affected by increasing ellipse elongation R and decreasing the angle α. For low R, the same Sherwood number and friction factor are obtained as those for the regular diamond lattice with circular struts. For high elongation, the geometry approaches a honeycomb-like shape, and the properties of the honeycomb are recovered as asymptotic conditions. Decreasing α results in a streamlined structure with a reduced friction factor and a reduced transport coefficient, consistent with previous observations for POCS with circular struts. The effects of α and R on the transport coefficient and friction factor cannot be decoupled from individual contributions. To address this complexity, a machine learning-aided approach was proposed for the prediction of the mass transfer coefficients and friction factors of the POCS as a function of the design parameters. POCS with intensified properties are characterized by a 2-fold larger trade-off index between transport coefficient and pressure drop than the state-of-the-art honeycomb. These advantages are manifested across various operating conditions and design parameters of the POCS, showcasing its high flexibility in manufacturing.

Ammonia production is one of the most important industrial chemical processes, but the synthesis reaction is strongly limited by chemical equilibrium. This is commonly compensated by applying high pressures, but large recycle ratios and purging losses are still unavoidable. Equilibrium limitations can alternatively be evaded by sorption enhancement, where NH₃ is selectively removed from the reaction mixture by a solid sorbent material. One material class commonly applied in this approach are metal halides like MgCl₂, as they typically show high NH₃ capacity even at elevated temperatures. In this study, a thermodynamic equilibrium model based on Gibbs energy minimization is established that is able to predict the simultaneous NH₃ synthesis and sorption equilibrium. After parametrization for metal chloride-based sorbents, the model is used to estimate the potential effect of sorption enhancement on the NH₃ synthesis in equilibrium. For kinetic studies under realistic operating conditions, a reactor model was established using kinetics for both iron and ruthenium-based catalysts. Simulations reveal that near-full conversion is possible in sorption-enhanced NH₃ synthesis under a wide range of realistic operating conditions. At thermodynamically unfavorable conditions, the process benefits from overstoichiometric amounts of sorbent as this keeps the sorbent saturation low and thus increases the sorption driving force. The integration of a sorbent material into the NH₃ synthesis reaction was shown to result in increased conversion, but at the same time also allows for a higher NH₃ formation rate. An increase in H₂ conversion by up to 550% was found at 350 °C, 100 bar, 15,000 h^–1 for twice the stoichiometrically required sorbent. While it has been demonstrated experimentally before, these findings quantify and emphasize the vast potential of sorption-enhanced NH₃ synthesis under a wide range of conditions.