The bi-atom catalysts (BACs) have attracted increasing attention in important electrocatalytic reactions such as oxygen reduction reaction (ORR). Here, by means of density functional theory simulations coupled with machine-learning technology,...

We have studied the batch and continuous extractive distillation of minimum-and maximum-boiling azeotropic mixtures with a heavy entrainer. These systems exhibit class 1.0-1a and 1.0-2 ternary diagrams, each with two subcases depending on the location of the univolatility line. The feasible product and feasible ranges of the operating parameters reflux ratio (R) and entrainer/feed flow rate ratio for continuous (F E /F) and batch (F E /V) operation were assessed. Class 1.0-1a processes allow the recovery of only one product because of the location of the univolatility line above a minimum value of the entrainer/feed flow rate ratio for both batch and continuous processes. A minimum reflux ratio R also exists. For an identical target purity, the minimum feed ratio is higher for the continuous process than for the batch process, for the continuous process where stricter feasible conditions arise because the composition profile of the stripping section must intersect that of the extractive section. Class 1.0-2 mixtures allow either A or B to be obtained as a product, depending on the feed location. Then, the univolatility line location sets limiting values for either the maximum or minimum of the feed ratio F E /F. Again, the feasible range of operating parameters for the continuous process is smaller than that for the batch process. Entrainer comparison in terms of minimum reflux ratio and minimum entrainer/feed ratio is enabled by the proposed methodology.

A deep learning approach coupling the Tree-LSTM network and back-propagation neural network for predicting the octanol–water partition coefficient.

Extractive distillation is one of the most attractive approaches for separating azeotropic mixtures. Few contributions have been reported to design an extractive distillation for separating maximum-boiling azeotropes and no systematic approaches for entrainer screening have been presented. A systematic approach to design of two-column extractive distillation for separating azeotropes with heavy entrainers has been proposed. A thermodynamic feasibility analysis for azeotropes with potential heavy entrainers was first conducted. Then, five important properties are selected for entrainer evaluation. Fuzzy logic and develop membership functions to calculate attribute values of selected properties have been used. An overall indicator for entrainer evaluation is proposed and a ranking list is generated. Finally, the top five entrainers from the ranking list have been selected and use process optimization techniques to further evaluate selected entrainers and generate an optimal design. The capability of the proposed method is illustrated using the separation of acetone-chloroform azeotropes with five potential entrainers.

Deep learning rapidly promotes many fields with successful stories in natural language processing. An architecture of deep neural network (DNN) combining tree‐structured long short‐term memory (Tree‐LSTM) network and back‐propagation neural network (BPNN) is developed for predicting physical properties. Inspired by the natural language processing in artificial intelligence, we first developed a strategy for data preparation including encoding molecules with canonical molecular signatures and vectorizing bond‐substrings by an embedding algorithm. Then, the dynamic neural network named Tree‐LSTM is employed to depict molecular tree data‐structures while the BPNN is used to correlate properties. To evaluate the performance of proposed DNN, the critical properties of nearly 1,800 compounds are employed for training and testing the DNN models. As compared with classical group contribution methods, it can be demonstrated that the learned DNN models are able to provide more accurate prediction and cover more diverse molecular structures without considering frequencies of substructures.

The separation of ternary nonideal systems with multi‐azeotrope is very important because they are often found in the waste of chemical and pharmaceutical industries, which is much more difficult due to the formation of multi‐azeotrope and distillation boundary. We propose a systematic procedure for design and control of a triple‐column pressure‐swing distillation for separating ternary systems with three binary minimum azeotropes. This procedure involves thermodynamic insights, a two‐step optimization method, and effective control strategy. The separation of tetrahydrofuran (THF)/ethanol/water is used to illustrate the capability of the proposed procedure. It is found that the pressure limits in columns can be determined through the analysis of residue curve maps, distillation boundary, and isovolatility curves. The optimal triple‐column pressure‐swing distillation is generated with the minimum total annual cost (TAC) of $2.181 × 106 in sequence A. The operating conditions are well controlled approaching their desired specifications in an acceptable time when disturbances occur. © 2019 American Institute of Chemical Engineers AIChE J, 65: 1281–1293, 2019

Extractive dividing-wall column (EDWC) was proved to be a promising energy-saving technique for the separation of multiple azeotropes or close-boiling mixtures; however, its controllability is very challenging due to its intensified structure with smaller physical space and strong interactions. Most studies on the EDWC control focused on evaluating the performance of temperature or composition control using proportional-integral (PI) control; nevertheless, significant steady-state offsets and overshoots are present in the control of two product purities of EDWC under different disturbances. In this paper, the performance of single temperature control (TC), temperature difference control (TDC) and double temperature control (DTDC) schemes for PI control of EDWC was first examined for the separation of toluene and 2-methoxyethanol using dimethyl sulfoxide as the entrainer. The results show that the steady-state offsets in product purities of TDC scheme are much smaller than those of TC and DTDC schemes. Subsequently, the offset-free model predictive control (MPC) based on temperature differences was proposed to improve the operation of EDWC. The results indicate that this MPC scheme can achieve much better control performance than PI control in terms of maximum transient deviation, amplitude of oscillations and setting time.

No abstract