Loading field map
Reaction engineering is the branch of chemical engineering concerned with the design, analysis, and operation of systems in which chemical transformations occur. Its central task is to answer a deceptively simple question: given a desired chemical product, what reactor should be built, and how should it be run? The answer requires combining knowledge of chemical kinetics—how fast reactions proceed and under what conditions—with the physics of flow, mixing, and heat transfer inside the reactor vessel. The field exists because a chemical reaction does not occur in a vacuum; it occurs in a specific physical environment that shapes how fast the reaction proceeds, how much product is formed, and whether the process is safe and economical.
At its heart, reaction engineering addresses the interaction between two distinct timescales: the timescale of the chemical reaction itself and the timescale of physical processes that bring reactants together, remove products, and manage heat. A reaction may be intrinsically fast, but if reactants are not mixed quickly enough, the observed rate will be limited by mixing rather than by chemistry. Conversely, a slow reaction may require a large reactor volume to achieve meaningful conversion, but that volume introduces temperature gradients that can alter the reaction pathway or deactivate the catalyst.
The field's foundational conceptual tool is the ideal reactor model. These are idealized descriptions of reactor behavior that assume perfect mixing or perfect plug flow, allowing the engineer to isolate the effects of kinetics from the complications of real flow patterns. The two canonical ideal reactors are the continuous stirred-tank reactor (CSTR) and the plug flow reactor (PFR). In a CSTR, the contents are assumed to be perfectly mixed, so the composition and temperature are uniform throughout, and the outlet stream has the same composition as the reactor contents. In a PFR, fluid moves through a tube with no axial mixing; each infinitesimal slice of fluid behaves like a tiny batch reactor, and composition changes along the length. A third ideal model, the batch reactor, is simply a closed vessel where the reaction proceeds over time without inflow or outflow.
These ideal models are not accurate descriptions of most real reactors, but they serve as bounding cases. A real reactor will behave somewhere between a CSTR and a PFR, and the engineer's job is to determine where it falls and whether that matters. The residence time distribution (RTD) is the experimental tool for this purpose: by injecting a tracer pulse and measuring its concentration in the outlet over time, one can characterize how long different fluid elements spend in the reactor. The RTD does not, by itself, predict conversion for arbitrary kinetics, but it provides a diagnostic of non-ideal flow and a basis for more detailed models.
Reaction engineering emerged as a distinct discipline in the mid-twentieth century, though its roots lie in earlier industrial practice. Chemical manufacturers had long operated reactors—acid digesters, polymerization vessels, catalytic converters—but design was largely empirical, based on trial and error and scale-up from pilot plants. The field crystallized when engineers began to formalize the relationship between kinetics and reactor design, treating the reactor as a system that could be modeled mathematically rather than merely built and tested.
The key intellectual move was the systematic use of the design equation, which expresses the reactor volume required to achieve a given conversion as an integral or algebraic function of the reaction rate. For a PFR, the volume is the integral of the inverse rate over the desired conversion range; for a CSTR, it is simply the product of the feed rate and the inverse rate evaluated at the outlet conditions. These equations made explicit what had been implicit: that reactor size depends not only on how fast the reaction is but on the flow pattern that determines how long each fluid element stays in contact with the catalyst or other reactants.
A second major development was the treatment of catalysis as an integral part of reactor design. Heterogeneous catalysis—where the reaction occurs on the surface of a solid catalyst—introduced a new set of complications. The reactant must diffuse from the bulk fluid to the catalyst surface, adsorb onto the surface, react, and then the product must desorb and diffuse back. Each step has its own rate, and the slowest step controls the overall rate. Reaction engineering developed formalisms for describing these steps, including the Langmuir–Hinshelwood and Eley–Rideal mechanisms for surface reactions, and for quantifying the effects of pore diffusion inside catalyst particles. This led to the concept of the effectiveness factor, which measures how much the observed reaction rate is reduced by diffusion limitations inside a porous catalyst pellet.
A third strand was the treatment of non-isothermal operation. Many industrial reactions are strongly exothermic or endothermic, and the heat released or absorbed can dramatically alter the reaction rate and selectivity. The field developed energy balances alongside mass balances, leading to the analysis of thermal runaway in exothermic reactors, the design of heat-exchange reactors, and the concept of adiabatic temperature rise. This thermal perspective also gave rise to the study of multiple steady states—situations where a reactor can operate stably at more than one temperature and conversion for the same feed conditions, depending on its history.
Within reaction engineering, several distinct approaches coexist, each addressing a different aspect of the overall problem. They are not rival schools that displaced one another but complementary tools that are often used together.
Kinetic modeling is the foundation. It seeks to express the reaction rate as a function of concentration, temperature, and catalyst state. The simplest form is the power-law rate expression, where the rate is proportional to the product of reactant concentrations raised to some exponents. More sophisticated approaches include Langmuir–Hinshelwood kinetics for catalytic reactions, which account for adsorption equilibria, and microkinetic models that attempt to describe every elementary step on a catalyst surface. The limitation of kinetic modeling is that it requires experimental data and often involves significant uncertainty; different rate expressions can fit the same data equally well, and extrapolation beyond the tested range is risky.
Reactor modeling takes the kinetic expression and embeds it in a physical description of the reactor. Ideal reactor models are the simplest, but real reactors often require more detailed treatment. The dispersion model adds an axial mixing term to the PFR equations, characterized by a dispersion coefficient that quantifies the degree of back-mixing. Compartment models divide the reactor into a network of ideal zones—some well-mixed, some plug-flow—connected by flows, which can capture complex behavior while remaining computationally tractable. Computational fluid dynamics (CFD) represents the most detailed approach, solving the full Navier–Stokes equations coupled with species transport and reaction. CFD can capture fine-scale phenomena like recirculation zones, dead volumes, and local hot spots, but it is computationally expensive and requires careful validation.
Stability and control analysis addresses the dynamic behavior of reactors. A reactor is not a static device; it responds to disturbances in feed composition, temperature, and flow rate. The field studies whether a reactor will return to its original operating point after a perturbation, whether it will oscillate, or whether it will drift to a different steady state. This analysis is particularly important for exothermic reactions, where the feedback between temperature and reaction rate can lead to instability. The stability criterion for a CSTR, which relates the heat generation and heat removal rates, is a classic result that remains central to reactor safety.
Scale-up and design methodology is the practical bridge between laboratory kinetics and industrial reactors. The central problem is that a reaction that behaves one way in a small flask may behave very differently in a large vessel, because mixing times, heat transfer areas, and flow patterns do not scale proportionally with volume. The field has developed heuristics and dimensionless numbers—such as the Damköhler number, which compares the reaction rate to the transport rate—to guide scale-up. The modern trend is toward model-based scale-up, where a detailed reactor model is used to predict performance at industrial scale, rather than relying on empirical rules of thumb.
A substantial portion of reaction engineering is concerned with reaction networks—the fact that most industrial reactions do not produce a single product but a mixture of desired and undesired products. The engineer must not only maximize conversion but also selectivity, the fraction of reacted feed that goes to the desired product. This introduces the concept of parallel and series reactions: a reactant may form product A and waste B simultaneously (parallel), or it may form A which then further reacts to waste C (series). The choice of reactor type can dramatically affect selectivity. For example, a PFR is generally better than a CSTR for a series reaction where the desired product is an intermediate, because the PFR exposes the product to less further reaction.
Catalyst deactivation is another central concern. Catalysts lose activity over time through mechanisms such as coking (carbon deposition), sintering (loss of surface area), and poisoning (strong adsorption of impurities). Deactivation affects reactor design because the reactor must be sized to maintain acceptable production over the catalyst lifetime, and it affects operation because the temperature may need to be raised to compensate for declining activity. The field has developed models for different deactivation kinetics and design strategies such as moving-bed reactors that continuously remove and regenerate catalyst.
Contemporary reaction engineering is shaped by several developments. Computational power has made CFD and detailed kinetic modeling routine, allowing engineers to simulate reactors with far greater fidelity than was possible even a few decades ago. Machine learning is increasingly used to accelerate kinetic parameter estimation and to build surrogate models that approximate expensive CFD simulations. However, these tools do not replace the fundamental conceptual framework; they extend it.
Sustainability has become a major driver. Reaction engineering is central to the design of processes for carbon capture and utilization, biomass conversion, and the production of renewable fuels and chemicals. These applications often involve complex feedstocks, multiphase systems, and novel catalysts, pushing the field toward more integrated approaches that combine reaction engineering with separations and process systems engineering.
Multiphase reactors—where gas, liquid, and solid phases coexist—remain an active area. Bubble columns, fluidized beds, trickle beds, and slurry reactors are all used industrially, and their design requires understanding of hydrodynamics, mass transfer between phases, and reaction kinetics simultaneously. The interaction between these phenomena is often the limiting factor in reactor performance, and it remains a challenging area where empirical correlations and detailed simulation are both used.
The field's enduring contribution is not any single model or equation but a way of thinking: every chemical transformation is embedded in a physical environment, and the engineer's task is to understand how that environment shapes the transformation and to design the environment to achieve the desired outcome. This perspective—the systematic coupling of chemistry with transport phenomena—remains the defining feature of reaction engineering and the source of its practical value.