The seven stages of chemical technology from concept to industrialization (Issue 23/100) -- Exploration of separation solutions

Thread Content

This post was last edited by xiouxingzhe on 2026-6-23 at 15:28. Seven stages of chemical technology from concept to industrialization (Issue 23/100) —— Technology development: Exploration of separation methods. Dear friends: Hello everyone! In the previous issue, we discussed optimizing reaction conditions and identified the operating window. In this issue, we will discuss an aspect that is often underestimated but yet often determines the economic viability of a project: the exploration of separation options. I’ve seen many engineers devote 80% of their effort to reacting. The reaction went well; the conversion rate was good, and the selectivity was quite high. I felt that everything had been accomplished successfully. As for separation—“Isn’t it just steaming it, cooling it, and filtering it?” ”As a result, the separation scheme was not chosen properly; costs exceeded expectations, energy consumption was high, and the product quality failed to meet standards, rendering all the previous efforts in vain. In this issue, we’ll discuss the logic behind choosing separation solutions. I. Why separation is often more “expensive” than reaction From a thermodynamic perspective, the essence of separation is to “reverse” a spontaneous process. After the reaction is complete, the product, the unreacted reactants, the by-products, and the solvent mix together—a process of entropy increase that occurs spontaneously. To separate them again, you need to perform work to reduce entropy. Whether it’s distillation, crystallization, extraction, or adsorption, energy must be supplied. All of this input energy ultimately has to be converted into costs. In many fine chemical plants, the equipment and energy consumption of the separation units account for half or even more of the total investment and operating costs of the entire plant. A well-chosen separation scheme results in low energy consumption, simple equipment, and great operational flexibility ； If the choice is poor, all the money is spent on separation. Therefore, exploring separation methods is not something to be done after the reaction is complete as a form of make-up. It should be considered simultaneously with, or even earlier than, the optimization of reaction conditions. II. Starting from system analysis: The basis for the separation scheme is physical property data. Before conducting any separation experiments, first analyze the system clearly. What components are actually present in the reaction product? What are the boiling points of the product, the unreacted starting materials, the by-products, and the solvent? Are there any azeotropes? What is the relative volatility? Is there any significant difference in solubility in water? Which components are thermosensitive—do they decompose at higher temperatures? Which components tend to crystallize and precipitate? Some of these basic data can be found in literature and databases, while others require experimental measurement. Regardless of the source, first draw up a complete table of component properties. After the material system analysis is completed, a basis is available for selecting the separation principle. The boiling point difference is large enough—consider distillation. The solubility differences are significant—consider crystallization or extraction. There is an azeotrope – see if azeotropic distillation or extractive distillation can be used. Thermosensitive materials – consider vacuum distillation or molecular distillation. There is a basic priority logic: use physical separation whenever possible, rather than chemical separation. Physical separation does not require the introduction of a third substance, and it does not pose additional issues related to the recovery of separation agents. Where an existing energy gradient can be utilized, no additional high-quality energy needs to be supplied – for example, if circulating water can be used for cooling, then there is no need for refrigeration. III. Primary and Alternative Schemes: Based on system analysis, at least one primary separation scheme is determined. The main solution should be the most mature and cost-effective option given the current information. But there can’t be only the main plan – in case the main plan is disproven during the scaling phase, you need a fallback option. The alternative plan does not have to be as detailed as the main plan, but at least it is necessary to think through where to turn if the main plan fails. I’ve seen more than one project where the primary separation scheme worked well in the laboratory, but during the pilot-scale testing phase, there was a significant drop in mass transfer efficiency, or problems such as emulsification, foaming, and scaling occurred, issues that were not anticipated in the laboratory. Teams with backup plans only experience a few weeks of delay when changing direction ； Without a backup plan, starting the plan screening from scratch will result in a delay of several months. When evaluating the primary and alternative solutions, several factors must be taken into account: technical maturity – does this solution have any industrial precedents in similar systems? Energy consumption — what is the approximate load for heating and cooling? Equipment investment – what type of equipment is needed, are there any special material requirements? Operational complexity – is it difficult to control, and are high requirements placed on the operators? Waste gases, waste liquids, and waste solids – does the separation process itself generate new waste gases, waste liquids, and waste solids? IV. Laboratory verification: Use actual product output, not simulated materials. Once the separation scheme is determined, it is verified in the laboratory. There is a common mistake that is easy to make, one I myself have made: using simulated materials instead of the actual reaction products for separation experiments. Simulated materials usually contain only the main components, lacking trace impurities. But it is precisely these trace impurities that sometimes have a particularly large impact on the separation efficiency. They may be surfactants – affecting the foam behavior and tray efficiency in distillation columns ； It might be a chelating agent – affecting extraction efficiency and phase separation speed ； It may be an inhibitor of crystal growth—affecting the crystallization yield and product purity. Therefore, for separation verification, it is essential to use the actual reaction output. If the reaction conditions have not yet stabilized and the composition of the output material is still changing, it is necessary to clearly indicate which batch of reaction output was used for each validation, so as not to mix together the separation data from different batches for analysis. The main aspects of verification include: whether the purity of the separated product meets the required standards, at which stage yield losses occur, and whether any phenomena unforeseen in the laboratory have arisen (emulsification, foaming, scaling, color changes). Make a rough estimate of the energy consumption for the separation process—how much is needed for heating, how much for cooling, and whether compression is required. If the cost of separation far exceeds that of the reaction itself, the economic viability of the entire process must be re-evaluated. V. Coupling of separation and reaction: Separation schemes are not isolated. There is a coupling relationship between it and the reaction conditions. Let’s take a simple example. If the reaction conversion rate increases by 5 percentage points, the load on subsequent separation may decrease by 20%. Conversely, if the separation scheme enables efficient recovery of the solvent, the reaction section can use a higher proportion of solvent to dilute the reactants, reduce viscosity, and improve mass and heat transfer. Therefore, during the exploration phase of separation schemes, it is worthwhile to take the time to conduct a coupling analysis between separation and reaction: to what extent do changes in reaction conditions affect the separation load? Does the separation scheme have a feedback effect on the reaction section? Is it possible to integrate reaction and separation—such as reactive distillation, reactive crystallization, membrane reactors—to carry out both reaction and separation simultaneously in a single device? This kind of coupled thinking often opens up some interesting technical directions. Of course, the complexity of integration is also higher, requiring more thorough verification. VI. Additional notes on alternative options: When exploring separation options, it often happens that while a primary option is chosen, the alternative options are merely considered casually without proper evaluation. The value of an alternative lies not in the fact that you must use it. It has two functions. First is risk hedging: if the main plan doesn’t work, the alternative plan serves as your escape route. The second is reverse verification: sometimes by comparing alternative options with the main solution, it is possible to identify certain issues within the main solution that were not noticed before—such as overly strict conditions for a particular unit operation, or high sensitivity to the quality of the raw materials. Therefore, alternative options deserve serious consideration. At the very least, this level should be achieved: it must be technically proven feasible, a rough economic estimate has been made, and the differences from the main plan have been clearly identified. Preview for the next issue: Issue 24 – Testing the setup of continuous reactors: The engineering transition from batch to continuous operation. The reaction conditions and separation methods have been identified, but all these tests have been carried out on small-scale batch reactors. True industrial production operates continuously 24 hours a day. Moving from batch to continuous operation is not as simple as adding a few more pumps; feeding, mixing, reaction, separation, and discharging – the entire process needs to be a closed loop. Next time, we’ll discuss how to set up a continuous small-scale testing setup, how to choose pumps, and how to determine pipe diameters. Table of Contents link: Seven-Stage Model of Chemical Technology from Idea to Industrialization (Issue 100) – Table of Contents (updated in real time)