**1. Introduction**

Acetonitrile is an organic solvent often used to purify butadiene and fatty acids. Given its high chemical activity, acetonitrile is also used as a raw material and synthetic intermediate for pharmaceutical chemicals [1–3]. A considerable amount of acetonitrile–water mixture is produced during the chemical production process. Acetonitrile is toxic and expensive, so its recovery from chemical production waste liquid is beneficial from environmental and economic points of view. However, owing to the azeotropic phenomenon, acetonitrile cannot be separated from its aqueous solution by a conventional rectification method. Thus, techniques, such as extractive distillation (ED) and pressure swing distillation (PSD), are often used in the industry [4–6].

ED involves the addition of an extractant to the azeotrope to increase the relative volatility of light and heavy components and achieve the purpose of separation [7–9]. Rodriguez-Donis et al. [10] used butyl acetate as extractant to separate acetonitrile and water mixture and analyzed the feasibility of seven configurations. Raeva et al. [11] proposed that the selection of extractants should not be limited to the relative volatility, but selective analysis is also an effective method. Sazonova et al. [12] found that when ED was used to separate acetonitrile and water mixture, the energy consumption of glycerol as an extractant was lower than that of dimethylsulfoxide and 1,2-ethandiol. You et al. [13] used a multi-objective genetic algorithm to simulate and optimize the separation process of acetonitrile and water mixture by ED with ethylene glycol as an entrainer. For multi-component mixtures, many scholars adopt the technology of the dividing wall column to reduce energy consumption and equipment investment [14,15].

The advantage of PSD over ED is that PSD does not introduce a third component and can efficiently separate pressure-sensitive azeotropes [16–18]. The separation of acetonitrile and water mixture by PSD has already been studied by some scholars [19,20]. Repke et al. [21] studied the separation of

acetonitrile and water mixture by PSD and carried out dynamic control. It was found that when the feed concentration fluctuated greatly, it could also maintain stability. Kim and Huang [22,23] optimized the process variables by taking the reboiler heat duty and total energy consumption as objective functions, respectively. Huang and Matsuda [24,25] studied the rectifying/stripping section heat integration to reduce the energy consumption in PSD, i.e., heat transfer between the rectifying section of the high pressure column (HPC) and the stripping section of the low pressure column (LPC), with a remarkable energy-saving effect. However, there is another way of heat integration, that is, heat transfer between the condenser of the HPC and the reboiler of the LPC.

For increasingly complex industrial technologies, Gao and Dai [26–28] proposed the fault detection and diagnosis (FDD) and data-driven approaches in the modeling, control, and optimization of complex industrial systems. Many design variables are involved in the chemical distillation industry, so commercial software, such as Aspen Plus, can better design, optimize, and control the distillation process [29,30]. Aspen Plus (Aspen Tech, Bedford, MA, USA) is a large-scale general process simulation system for plant design, steady-state design, and dynamic simulation. It is recognized as a large scale process simulation software in the chemical industry, which has been widely used in industrial design and academic research [31,32]. In this study, the conceptual design and optimization of PSD for the separation of acetonitrile and water mixture was carried out, so Aspen Plus was selected for steady-state simulation and optimization design.

This work aims to compare PSD with full heat integration and traditional PSD for the separation of acetonitrile and water mixture. At present, there is no comparison between the two methods in the literature. Therefore, in this work, Aspen Plus software was used to simulate and optimize the PSD process. The entire optimization process uses the total annual cost (TAC) as the objective function to find the optimal values of the design variables through the sequential iteration method. At the same time, full-heat integration process design for PSD was optimized and compared with PSD without heat integration from an economic point of view. The results from this work can provide some technical support for such azeotrope separation designs.

#### **2. Design of Separation Scheme for PSD**

#### *2.1. Feasibility Analysis*

The calculation accuracy of the process simulation depends on the selection of the thermodynamic equation. Since the azeotropic concentration under different pressures predicted by the Wilson equation agrees well with the experimental data published in Azeotropic Data [33], the Wilson equation was used in the simulation process, and the binary interaction parameters involved in Wilson equation can be obtained in Aspen Plus software. Table 1 lists the variation of acetonitrile–water azeotrope with the pressure. The acetonitrile–water azeotrope composition is sensitive to the pressure changes. As the pressure increases from 1 to 4 atm, the mole concentration of acetonitrile decreases from 0.6971 to 0.5886, indicating that PSD is feasible for separating acetonitrile–water azeotrope.


**Table 1.** Effect of the pressure on acetonitrile–water azeotrope.
