Synthesis of adipic acid (AA) through the oxidation of cyclohexanol and cyclohexanone (K/A oil) with nitric acid was conducted in a capillary microreactor system. Effects of the temperature, the nitric acid concentration, the volumetric flow rate ratio of nitric acid to K/A oil, and the capillary length on the selectivity and the product yield were investigated systematically to achieve optimal reaction conditions. Notably, a high yield of adipic acid (i.e., 90%) was achieved just in 6 seconds at 85℃ with the use of 55 wt% nitric acid. Gas components produced in this oxidation process and its total volumetric flow rate were determined under various operating conditions, which was beneficial for reaction mechanism characterization and process optimization. Finally, a kinetic model was established, which was of crucial theoretical significance and practical value for optimizing the reactor design and better understanding such fast and highly exothermic multiphase processes with abundant gas production.
For the ionic liquid (IL)-solute systems of broad interest, a deep neural network based recommender system (RS) for predicting the infinite dilution activity coefficient (γ∞) is proposed and applied for a large extension of the UNIFAC model. In the RS, neural network entity embeddings are employed for mapping each IL and solute and neural collaborative filtering is utilized to handle the nonlinearities of IL-solute interactions. A comprehensive experimental γ∞ database covering 215 ILs and 112 solutes (totally 41,553 data points) is established for training the RS, where the cross-validation and test are performed based on a rigorous dataset split by IL-solute combinations. The obtained RS shows superior performance than the state-of-the-art γ∞ models and is thus taken to complete the solute-in-IL γ∞ matrix. Based on the completed γ∞ database, a large extension of the UNIFAC-IL model is finally presented, filling all the parameters between involved groups.
We introduce a straightforward method for the preparation of novel starch-based ultramicroporous carbons (SCs) that demonstrate high CH4 uptake and excellent CH4/N2 selectivity. These SCs are derived from a combination of starch and 1-6 wt. % of acrylic acid, and the resulting materials are amenable to surface cation exchangeability as demonstrated by the formation of highly dispersed K+ in carbon precursors. Following activation, these SCs contain ultramicropores with narrow pore-size distributions of <0.7 nm, leading to porous carbon-rich materials that exhibit CH4 uptake values as high as 1.86 mmol/g at 100 kPa and 298 K, the highest uptake value for CH4 to date, with the IAST-predicted CH4/N2 selectivity up to 5.7. Both the potential mechanism for the formation of narrow pores and the origin of the favorable CH4 adsorption properties are discussed and examined. This work may potentially guide future designs for carbon-rich materials with excellent gas adsorption properties.
The manuscript describes a computational study that provides molecular-level insight into shale gas adsorption and transport in shale rocks, which are composed of organic and inorganic matter. Atomistic simulations were used to generate realistic models of the organic matter structures with both micro- and mesoporosity, and correspond to mature and overmature type-II kerogens. These porous material models are unique to most other previous kerogen models since they contain other components (asphaltene/resin, hydrocarbons and carbon dioxide/water fractions) that are typically not modeled. The inclusion of these additional components significantly influences the resulting porous structure characteristics. The adsorption and diffusion behavior of methane (as a shale gas proxy) and methane/carbon dioxide mixtures were simulated in the model structures. Several key industrial-relevant findings are described in the manuscript.
We present a new family of fast and robust methods for the calculation of the vapor-liquid equilibrium at isobaric-isothermal (PT-flash), isochoric-isothermal (VT-flash), isenthalpic-isobaric (HP-flash), and isoenergetic-isochoric (UV-flash) conditions. The framework is provided by formulating phase-equilibrium conditions for multi-component mixtures in an effectively reduced space based on the molar specific value of the recently introduced volume function derived from the Helmholtz free energy. The proposed algorithmic implementation can fully exploit the optimum quadratic convergence of a Newton method with the analytical Jacobian matrix. This paper provides all required exact analytic expressions for the general cubic equation of state. Computational results demonstrate the effectivity and efficiency of the new methods. Compared to conventional methods, the proposed reduced-space iteration leads to a considerable speed-up as well as to improved robustness and better convergence behavior near the spinodal and coexistence curves of multi-component mixtures, where the preconditioning by the reduction method is most effective.
The low concentration methyl iodides (CH3I) adsorption process on reduced silver-functionalized silica aerogel (Ag0-Aerogel) was studied. The kinetic data were acquired using a continuous flow adsorption system. Because the corresponding physical process was observed, the shrinking core model (SCM) was modified and applied. An average CH3I pore diffusivity was calculated, the CH3I-Ag0-Aerogel reaction was identified as a 1.37 order reaction instead of first order reaction, and the nth order reaction rate constant was determined. This modified SCM significantly increases the accuracy of adsorption behavior prediction at low adsorbate concentration. Modeling results indicate that the overall adsorption process is controlled by the pore diffusion. However, at low adsorbate concentration (<100 ppbv), the CH3I adsorption is limited to the surface reaction due to the low uptake rate in a predictable time period.
Experimental results on pressure drop and flow patterns for gas-liquid flow through packed beds obtained in the International Space Station with two types of packing are presented and analyzed. It is found that the pressure drop depends on the packing wettability in the viscous-capillary (V-C) regime and this dependence is compared with previously published results developed using short duration low-gravity aircraft tests. Within the V-C regime, the capillary contribution is the dominant force contributing to the pressure drop for the wetting case (glass) versus the viscous contribution dominating for the non-wetting case (Teflon). Outside of the V-C regime, it is also found that hysteresis effects that are often strong in normal gravity gas-liquid flows are greatly diminished in microgravity and pressure drop is nearly independent of packing wettability. A flow pattern transition map from bubble to pulse flow is also compared with the earlier aircraft data.
Exploring highly active and stable electrocatalyst for oxygen evolution reaction is important for the development of water splitting and rechargeable metal-air batteries. Herein, a hybrid electrocatalyst of CoFe alloy and CoxN heterojunction encapsulated and embedded in N-doped carbon support (CoFe-CoxN@NC) was in situ coupling via a pyrolysis process of a novel coordination polymer from lignin biomacromolecule. CoFe-CoxN@NC exhibited an excellent OER activity with a low overpotential of 270 mV at 10 mA•cm−2 and stability with increment of 20 mV, comparable to commercial Ir/C. DFT calculations revealed that CoxN and N-doped grapheme encapsulation can reduce the binding strength between *O and CoFe alloy, prevent metals leaching and agglomeration, and improve electron transfer efficiency, thereby, remarkably enhancing the OER activity and stability. In situ coupling strategy of alloy and nitride heterojunction on N-doped lignin-derived carbon provided a promising and universal catalyst design for the development of renewable energy conversion technologies.
Separation of higher hydrocarbons from methane is an important and energy-intensive operation in natural gas processing. We present a detailed investigation of thin and oriented MFI zeolite membranes fabricated from 2D MFI nanosheets on inexpensive α-alumina hollow fiber supports, particularly for separation of n-butane, propane, and ethane (“natural gas liquids”) from methane. The present MFI membranes display high permeances and selectivities for C2-C4 hydrocarbons over methane, driven primarily by stronger adsorption of C2-C4 hydrocarbons. We study the separation characteristics under unary, binary, ternary and quaternary mixture conditions, including the pressure dependence. The membranes are highly effective in quaternary mixture separation at elevated feed pressures, for example allowing n-butane/methane separation factors of 170–280 and n-butane permeances of 710–2700 GPU in the 1-9 bar feed pressure range. Furthermore, we parametrize and apply multicomponent Maxwell-Stefan transport equations to predict the main trends in separation behavior over a range of operating conditions.
Molecular simulation has emerged as an important sub-field of chemical engineering, due in no small part to the leadership of Keith Gubbins. A characteristic of the chemical engineering molecular simulation community is the commitment to freely share simulation codes and other key software components required to perform a molecular simulation under open-source licenses and distribution on public repositories such as GitHub. Here we provide an overview of open-source molecular modeling software in Chemical Engineering, with focus on the Molecular Simulation Design Framework (MoSDeF). MoSDeF is an open-source Python software stack that enables facile use of multiple open-source molecular simulation engines, while at the same time ensuring maximum reproducibility.
In this letter, we investigate the rebound dynamics of two equally sized droplets simultaneously impacting a superhydrophobic surface via lattice Boltzmann method (LBM) simulations. We discover three rebound regimes depending on the droplet distance: a complete-coalescence-rebound (CCR) regime, a partial-coalescence-rebound (PCR) regime, and a no-coalescence-rebound (NCR) regime. We demonstrate that the rebound regime is closely associated with dynamic behaviors of the formed liquid ridge or bridge between two droplets. We also present the contact time in the three regimes. Intriguingly, although partial coalescence takes places, the contact time is still dramatically shortened in the PCR regime, which is even smaller than that of a single droplet impact. These findings provide new insights into the contact time of multiple droplets impact, and thereby offering useful guidance for some application such as anti-icing, self-cleaning, and so forth.
In this investigation, CO2 capture performance of zeolite 13X monoliths with 600 and 800 cpsi in presence of SO2/NO impurities under dry and humid conditions were evaluated and compared with that of 13X beads. Dynamic breakthrough tests demonstrated a drastic reduction in CO2 capture capacity and deterioration of kinetics under dry-clean conditions, whereas, upon switching the feed from a clean gas to contaminated gas which contained SO2 and NO, different adsorption performance was observed. Specifically, in dry-contaminated mode, the adsorbents retained their capture capacities with comparable kinetics to that of dry-clean feed conditions, however, in humid-contaminated mode, the adsorbents experienced improved CO2 uptake and CO2/N2 selectivity, albeit at the expense of deteriorated kinetics. These findings indicate that the presence of SO2 and NO contaminants, especially SO2 contaminants, lead to dramatic changes in the adsorption performance of zeolite 13X monoliths, indicating the importance of evaluating adsorbent materials under realistic conditions.
A new transport model is proposed for paraffin wax deposition onto a cold finger from flowing wax-containing oils. The model solves transient energy and mass balances simultaneously for a reversible first-order kinetic rate for precipitation of pseudo-single-component wax, and the effects of yield stress using a critical solid wax concentration to withstand flow-induced stress at the deposit-fluid interface, Cpi. The model can predict the time evolution of the deposit thickness, and the spatial and temporal evolution of temperature and wax concentration and was validated using experiments involving a cylindrical cold finger. We found that for oils with Cpi close to zero, the deposit thickness growth is dominated by heat transfer. However, mass transfer cannot be neglected as diffusion of wax into the deposit continues to take place even after the deposit has stopped growing. For oils with non-zero Cpi, the deposit growth is slow and accompanied by occasional sloughing.
Diverse engineering fields request flash calculations like isothermal flash, isenthalpic flash, and isentropic flash. They can be cast as minimization of a thermodynamic state-function and solved by Michelsen’s Q-function approach. Flash calculations for open systems, i.e. systems where chemical potentials are specified instead of the mole numbers for some components, also belong to this scope. By analyzing the construction of Q-functions through Legendre transforms, we extend this approach to the flash for open systems in the absence or presence of chemical reactions, resulting in general formulations for various specifications. For systems without reactions, the classical framework using mole numbers as independent variables is employed; for those with reactions, the modified-RAND framework is employed. We present examples for open systems at constant temperature and pressure. Using the Q-function minimization, we can solve multicomponent non-reactive or reactive systems at a specified chemical potential with quadratic convergence over a wide range of conditions.
Error-in-variables model (EVM) methods are used for parameter estimation when independent variables are uncertain. During EVM parameter estimation, output measurement variances are required as weighting factors in the objective function. These variances can be estimated based on data from replicate experiments. However, conducting replicates is complicated when independent variables are uncertain. Instead, pseudo-replicate runs may be performed where the target values of inputs for repeated runs are the same, but the true input values may be different. Here, we propose a method to estimate output-measurement variances for use in multivariate EVM estimation problems, based on pseudo-replicate data. We also propose a bootstrap technique for quantifying uncertainties in resulting parameter estimates and model predictions. The methods are illustrated using a case study involving n-hexane hydroisomerization in a well-mixed reactor. Case-study results reveal that assumptions about input uncertainties can have important influences on parameter estimates, model predictions and their confidence intervals.
While decomposition techniques in mathematical programming are usually designed for numerical efficiency, coordination problems within enterprise-wide optimization are often limited by organizational rather than numerical considerations. We propose a ‘data-driven’ coordination framework which manages to recover the same optimum as the equivalent centralized formulation while allowing coordinating agents to retain autonomy, privacy, and flexibility over their own objectives, constraints, and variables. This approach updates the coordinated, or shared, variables based on derivative-free optimization (DFO) using only coordinated variables to agent-level optimal subproblem evaluation ‘data’. We compare the performance of our framework using different DFO solvers (CUATRO, Py-BOBYQA, DIRECT-L, GPyOpt) against conventional distributed optimization (ADMM) on three case studies: collaborative learning, facility location, and multi-objective blending. We show that in low-dimensional and nonconvex subproblems, the exploration-exploitation trade-offs of DFO solvers can be leveraged to converge faster and to a better solution than in distributed optimization
This work considers a seeded fesoterodine fumarate (FF) cooling crystallization and presents the methodology and implementation of a real-time machine learning modeling-based predictive controller to handle batch-to-batch (B2B) parametric drift. Specifically, an autoencoder recurrent neural network-based model predictive controller (AERNN-MPC) is developed to optimize product yield, crystal size, and energy consumption while accounting for the physical constraints on cooling jacket temperature. Deviations in the kinetic parameters are considered in the closed-loop simulations to account for the B2B parametric drift, and two error-triggered online update mechanisms are proposed to address issues pertaining to the availability of real-time crystal property measurements and are incorporated into the AERNN-MPC to improve the model prediction accuracy. Closed-loop simulation results demonstrate that the proposed AERNN-MPC with online update, irrespective of the accessibility to real-time crystal property data, achieves a desired closed-loop performance in terms of maximizing product yield and minimizing energy consumption.
The performance of advanced controllers depends on the selection of several tuning parameters that can affect the closed-loop control performance and constraint satisfaction in highly nonlinear and nonconvex ways. There has been a significant interest in auto-tuning of complex control structures using Bayesian optimization (BO). However, an open challenge is how to deal with uncertainties in the closed-loop system that cannot be attributed to a lumped, small-scale noise term. This paper develops an adversarially robust BO (ARBO) method that is suited to auto-tuning problems with significant time-invariant uncertainties in a plant simulator. ARBO uses a Gaussian process model that jointly describes the effect of the tuning parameters and uncertainties on the closed-loop performance. ARBO uses an alternating confidence-bound procedure to simultaneously select the next candidate tuning and uncertainty realizations, implying only one expensive closed-loop simulation is needed at each iteration. The advantages of ARBO are demonstrated on two case studies.