The transformation of the chemical industry to renewable energy and feedstock supply requires new paradigms for the design of flexible plants, (bio-)catalysts, and functional materials. Recent breakthroughs in machine learning (ML) provide unique opportunities, but only joint interdisciplinary research between the ML and chemical engineering (CE) communities will unfold the full potential. We identify six challenges that will open new methods for CE and formulate new types of problems for ML: (1) optimal decision making, (2) introducing and enforcing physics in ML, (3) information and knowledge representation, (4) heterogeneity of data, (5) safety and trust in ML applications, and (6) creativity. Under the umbrella of these challenges, we discuss perspectives for future interdisciplinary research that will enable the transformation of CE.

A promising application of demand-side management is the chlor-alkali electrolysis. However, storing the produced chlorine for flexibility should be avoided whenever possible. If PVC is produced from chlorine, storing the intermediate 1,2-dichloroethane resulting from direct chlorination of ethene is a better alternative as it is less toxic than chlorine and can be easily stored. Currently, no dynamic process models to study the process behavior or to develop optimal trajectories for the 1,2-dichloroethane production under different demand response scenarios are available. Hence, we formulate and solve a dynamic, pressure-driven model of the synthesis of 1,2-dichloroethane and validate it with real process data in this contribution. As part of this dynamic model, differentiable formulations for weeping and the flow over a weir of a distillation tray are presented, which are also valid whenever certain trays run dry.

The implementation of the hydroformylation reaction for the conversion of long-chain alkenes into aldehydes still remains challenging on an industrial scale. One possible approach to overcoming this challenge is to apply tunable systems employing surfactants. Therefore, a novel process concept for the hydroformylation of long-chain alkenes to aldehydes in microemulsions is being investigated and developed at Technische Universität Berlin, Germany. To test the applicability of this concept for the hydroformylation in microemulsions on a larger scale, a miniplant has been constructed and operated. This contribution presents the proof of concept for hydroformylation in microemulsions carried out during a 200 h miniplant operation. Throughout the operation a stable aldehyde yield of 21% and a catalyst loss in the product phase below 0.1 ppm were achieved, which confirms previous lab scale findings. Additionally, solution strategies for a stable continuous operation to overcome challenges such as foaming, phase separation issues, and coalescence dynamics are discussed herein.

We investigate aqueous multiphase systems for catalytic gas/liquid reactions, namely, the rhodium-catalyzed hydroformylation of the long-chain olefin 1-dodecene. The multiphase system was formulated from 1-dodecene, water, and a nonionic surfactant, which increases the solubility between the two nonmiscible liquid phases. On the basis of these systems, we present in this paper a transfer of lab experiments (semibatch) to a successful operation of a miniplant in continuous mode. Under optimized conditions, the reaction showed turnover frequencies of ∼200 h–1 and high selectivity of 98:2 to the desired linear aldehyde. The miniplant was operated continuously for a total of 130 h. The control of the phase separation and catalyst recycling for product isolation for a long time period appeared to be challenging. Nevertheless, the separation was kept stable for over 24 h. The organic components in the product phase amounted to desired values between 95 and 99 wt %. The desired 99.99% of the catalyst remained in the aqueous catalyst phase.

No abstract