ABSTRACT This work presents a dynamic, model‐based kinetic measurement method that couples vapor‐liquid‐equilibrium with reaction kinetics for the bulk‐phase polymerization of propylene in a bench‐scale reactor. This model couples Peng‐Robinson‑based flash calculations with a simplified kinetic scheme (activation, propagation, deactivation). Reactor pressure‐time trajectories from polymerizations (at 20°C–30°C) are used as model input. The novelty of this approach lies in explicitly embedding vapor‐liquid equilibrium into the dynamic pressure‑based calibration of polymerization kinetics under bulk pre‑polymerization conditions. Through the process of parameter tuning, the simulated and measured reactor pressure drop caused by conversion of liquid propylene to polypropylene in a multi‐component mixture are aligned, yielding the simulated polymerization rate and the elementary reaction rate constants. The model reproduces the characteristic pressure decrease and end‐point yields within acceptable deviations and demonstrates instantaneous catalyst activity profiles, the activity decay over reaction time, and the temperature response of the catalyst. This approach enables access to detailed catalytic polymerization kinetics of propylene at mild temperatures which are typically used under industrial pre‐polymerization reactors.
ABSTRACT Atom transfer radical polymerization (ATRP) has historically required an oxygen‐free environment to prevent oxidation of both the polymeric radical and the active copper catalyst, limiting its deployment in industrial‐scale processes. In this study, ATRP of styrene was successfully carried out in the presence of oxygen by using triethylborane (TEB) complexed with an aryl amine to form air‐stable complexes, which served as an essential component of the reaction mixture. Polymerizations conducted in DMSO at 80°C reached high yields within 10 h while maintaining relatively low dispersity (Ð ≈ 1.11), although the molecular weights exceeded theoretical values. When the reaction was allowed to proceed for 24 h, dispersity increased and a low‐molecular‐weight shoulder became apparent on GPC. When the complexed triethylborane was omitted, no polymerization was observed to occur when carried out in the presence of oxygen. The brominated polystyrene precursor formed with oxygen present was successfully chain extended, indicating chain end fidelity and consistent with the ATRP mechanism.
ABSTRACT Polyvinylidene fluoride (PVDF) is a semicrystalline fluoropolymer mostly produced by free‐radical suspension or emulsion polymerization. It has the second largest production capacity among all fluoropolymers, following only polytetrafluoroethylene. This great industrial attention is due to its excellent chemical and physical properties, among which its inertness and resistance to chemicals and heat, its high melting point and high tensile strength. As a result, it is one of the first materials of choice in the field of membrane separations, especially for micro and ultrafiltration. In addition, due to its piezoelectric, pyroelectric, and optical properties, PVDF is finding more and more extensive application in batteries and sensors. Given the important role played by PVDF in the polymer market, this review is aimed at reporting the latest advancements in its manufacturing, with particular reference to the innovations introduced following the new legislation on environmental protection from perfluoroalkyl substances. The main focus is on the emulsion polymerization process, where experimental conditions and models from the literature are presented and discussed. After having clarified the polymer reaction engineering aspects for this material, the latest advancements in the application of PVDF are reviewed, with the aim of highlighting its strengths as well as points for improvement.
ABSTRACT The free‐radical polymerization behavior of methyl methacrylate (MMA) is investigated in the presence and absence of trisilanol isobutyl polyhedral oligomeric silsesquioxane (TSI‐POSS) at 70°C and 80°C using a calorimetric approach based on continuous temperature monitoring. The reaction temperature is experimentally recorded, and the corresponding heat flow is estimated from an energy balance, enabling the calculation of monomer conversion and polymerization rate ( R p ) as functions of time. The experimental data are further analyzed to estimate the k t /k p 2 ratio (ratio of the termination to the squared propagation rate coefficients), which serves as a comparative indicator of changes in the balance between termination and propagation. The results reveal a non‐monotonic evolution of the k t /k p 2 ratio, characterized by an initial decrease followed by an increase and a subsequent decline, suggesting behavior consistent with diffusion‐controlled polymerization regimes. The presence of TSI‐POSS is associated with a decrease in the polymerization rate and an increase in the k t /k p 2 ratio, suggesting the influence of diffusion‐related limitations on the polymerization process. These observations are qualitatively consistent with diffusion‐related effects, although independent characterization would be required to confirm the underlying mechanisms.
ABSTRACT The interrelationship between chain transfer agent (CTA) mass transfer and reaction in emulsion polymerization is examined. Three dimensionless parameters describing these phenomena are proposed. The Damkohler Number ( Da cta ) is defined as the maximum rate of CTA consumption divided by the maximum rate of CTA transfer from the monomer droplets to the polymerizing polymer particles. The probability of mass transfer‐ rather than reaction‐limitation for each CTA‐monomer pair is assessed. S D is the maximum instantaneous rate of consumption of CTA by reaction, divided by the desired instantaneous rate of consumption of CTA, assuming no mass‐transfer limitation on CTA. S D is defined so that it has a value of one when the CTA is consumed at the correct rate in order to control molecular weight and minimize branching and crosslinking over the entire polymerization. It is demonstrated that the desired value of S D in emulsion polymerization is different from that in homogeneous polymerization. S is defined as the maximum rate of CTA mass transfer divided by the desired rate of CTA consumption and is useful if CTA mass transfer is used to “meter” the CTA into the polymerizing polymer particles. Proper selection of CTA, and polymerizer operation are discussed in light of these three dimensionless parameters.
ABSTRACT The free‐radical polymerization of ethylene under high‐pressure conditions involves highly reactive radicals and short reaction times. Especially in the vicinity of the initiator feed location, mixing can influence the radical distribution and, consequently, the polymer properties. This work aims to link mixing, modeled using computational fluid dynamics (CFD), with polymer property information such as molecular weight distribution (MWD) and branching density through a Monte Carlo approach. The presented modeling framework uses the concept of compartmentalization to transfer mixing information from the CFD model to the Monte Carlo model. Comparison with polymer samples from a mini‐plant reactor shows good agreement between the simulated and measured MWDs. In contrast to most existing models based on ideal reactor concepts, the approach presented here does not require parameter adjustment, for example, for initiator efficiency. Furthermore, the developed compartmentalization model can be used as a deterministic model in place of a CFD simulation, significantly reducing simulation time.
A fault detection system based on a recurrent autoencoder is proposed, suitable for continuous processes with complex nonlinear dynamics. The system consists of three modules: 1) a recurrent autoencoder; 2) a fault state detection module; and 3) an interpretation module. The training strategy only requires process data obtained under normal operating conditions. The fault detection module analyzes the total reconstruction error of the measured variables obtained by the recurrent autoencoder. The interpretability module examines the reconstruction errors of the individual variables and highlights variables potentially responsible for the detected anomaly. The developed system is applied to a computationally simulated Styrene-Butadiene rubber latex production process. The latex is synthesized in a train of continuous stirred tank reactors that may be affected by various types of faults. Simulation results show a satisfactory performance of the proposed fault detection system. Failures whose data patterns differ significantly from those of normal operation are effectively detected. In contrast, some subtle faults only cause minor deviations in the data patterns and are therefore difficult to detect rapidly, leading to delays in detection. In most cases, the interpretability module can correctly identify the variable more strongly associated with the detected fault.
A comprehensive model of free radical entry into particles during emulsion polymerization, recently reported in this journal, was modified to include the calculation of radical-particle interactions based on the DLVO theory as well as the simplified calculation of secondary nucleation. Electrostatic repulsion was found to play a crucial role, causing changes in the size of the predominant entering oligomer radical, whether due to variations in ionic strength or surfactant coverage (theta). As theta increases to similar to 0.1, the energy barrier for the entry of water-soluble radicals raises sharply, causing their entry frequency (rho) to decrease drastically and be controlled by diffusion. For theta>similar to 0.1, rho becomes constant, and the predominant entering species are primary particles, whose formation is the rate-determining step. It was also found that the onset of secondary nucleation occurs at a critical theta value rather than a critical particle concentration. Based on the results obtained and their agreement with experimental data, key mechanistic aspects of radical entry into the system under study are proposed, which contradict several postulates of one of the most accepted theories of free radical entry. These findings also suggest that nucleation mechanisms above and below the surfactant's CMC should be revisited.
The primary synthesis of liquid rubber via anionic polymerization requires substantial alkali metal initiators, resulting in high costs. This study introduces a metal-free chain transfer agent (CTA), 9,10-dihydroanthracene (DHA), into the anionic polymerization of isoprene (Ip). A series of liquid polyisoprenes (PIp) with number-average molecular weights (M n) ranging from 2100 to 6800 g/mol were synthesized. Furthermore, the correlation between DHA, THF dosage, and reaction temperature and the resulting polymer structures was established. The results indicate that at 20 degrees C, DHA induces chain transfer during the anionic polymerization of Ip. Moreover, higher dosages of both DHA and THF led to lower M n and polydispersity index (PDI) of the resulting PIp, corresponding to a lower degree of polymerization (). Concurrently, the initiator efficiency (E I) and chain transfer constant (C M) increased. Conversely, -25 degrees C effectively suppressed chain transfer, while 60 degrees C did not significantly promote the process. To validate its practical potential, a 100-fold scale-up experiment was performed. This yielded a series of liquid PIp with M n below 2500 g/mol, PDI below 1.21, E I greater than 20.0, and C M greater than 2.8 & times; 10- 2. It provides crucial theoretical and experimental foundations for advancing the preparation technology of liquid rubber.
The modeling of the radical polymerization of n-butyl acrylate (nBA) was investigated by a deterministic approach in Predici and a stochastic model with Monte Carlo (mcPolymer). The same kinetic network and coefficients were used, but chain-length-dependent termination, the beta-scission, and the macromonomer formation and incorporation had to be implemented differently. Thereby, the modeling approaches gave consistent results, and the benchmark of the two modeling approaches was successful. Moreover, the strong effect of temperature on polymeric microstructure, such as branching densities and macromonomer formation, was investigated with both models.
Well defined polystyrene in varied sizes with polydispersity indices < 1.2 were obtained in reduced time using Microwave Assisted Activator Re-Generated by Electron Transfer Atom Transfer Radical Polymerization Reactions. The use of pre-synthetized Cu(II)Br/Me6TREN catalyst allows for rapid set-up of the reaction and addition of an excess of stannous octoate allows for oxygen tolerant conditions during the reaction set-up. The reaction mixture is stable for several days at room temperature, allowing for a series of polymerizations to be set up using an automatized polymerization reaction process. The reaction conditions were optimized in regards of the temperature (80 degrees C-100 degrees C), catalyst loading (0.15 equiv) and reaction times at 85 degrees C (<= 7 h). Well-controlled polymerizations are achieved for sizes up to 15 kDa, while maintaining polydispersity index values below 1.05. Kinetic comparisons between microwave irradiation and conventional heating methods show a higher apparent propagation rate constant while obtaining well-defined polymers. A norbornene-functionalized initiator allows us to obtain alpha, omega-end-chain functionalized polystyrene, providing potential expansion of applications in fields such as the synthesis of nanoparticles and polymer design.
Polymer reaction kinetics link chemical mechanisms to the evolving polymer microstructure, but describing structural statistics efficiently remains challenging: population-balance models may become intractable, while chain-based Monte Carlo (MC) simulations can be computationally demanding. We review and formalize reaction-related observables (RROs), implemented as cumulative counters-massless, non-chemical entities that record reaction events such as monomer incorporation, branching, scission, and network-forming steps without affecting the mass balances. RROs extend deterministic rate-equation models by embedding event statistics directly into the kinetic framework, providing compact, interpretable observables and enabling reduced descriptions that remain easy to extend across model variants. We illustrate how counter-based evaluations can be used to construct derived structural indicators and heuristic structure-property relations, including examples for crosslinking networks.
In this work, we present a new class of silane chain-transfer agents (CTAs) for the molecular weight and the molecular weight distribution control during the production of low-density polyethylene (LDPE). These silanes exhibit a wide range of chain-transfer propensity as measured by their C s values (the ratio of the rate coefficient for hydrogen abstraction and ethylene propagation, k h/k p) with the best performers being tris(trimethylsilyl)silane (C s = 16.0) and trimethylsilylsilane (C s = 15.7). Density functional theory (DFT) calculations were used to estimate the C s values. The quantum chemical calculations were validated by pilot-plant experiments for triethylsilane (C s = 0.22) and tris(trimethylsilyl)silane (C s = 36.5). Given the expected variability in the pilot-plant data, an excellent agreement between theory and experiment was achieved. Furthermore, a model utilizing the DFT calculated Mulliken charge on the silicon atom to predict the Cs value was developed to facilitate silane screening in the future.
Fluid simulations are cost-effective and zero-waste alternatives for research and development of polymer reactors. However, many polymer-specific simulation software packages assume a homogeneous reactant mixture, overly simplifying the physics. Computational Fluid Dynamics (CFD) simulations provide more insight into this process but are difficult to utilize with detailed chemistry mechanisms. This work seeks to verify the implementation of free radical polymerization chemistry into a plant-scale CFD reactor model for Low-Density Polyethylene (LDPE) to investigate spatial gradients and their impact on the system. The process resulted in the most accurate CFD polymer reactor simulation to date, known to the authors, and sets a precedent for the verification and validation of other reactor models. It was found that the axial flow profiles formed distinct regions within the reactor wherein there were unmixed polymer properties. Significantly more variability was also found in the upstream-most reactor zone compared to a downstream zone, exemplified by a coefficient of variation of 447% in the former and 0.372% in the latter for the polydispersity index on a central plane of the reactor. Additionally, noticeable differences in properties were found between the inside and outside of the mixing shaft, which were only fractions of a percent different.
Size-exclusion chromatography (SEC) analysis of polymers with nontrivial architectures presents specific challenges, seeing its separation of macromolecules by variations in hydrodynamic volume rather than true molar mass. Care should thus be taken with respect to calibration and detection strategies, with a coupling to (i) other higher scale experimental tools (e.g. rheology), and (ii) modeling techniques worthwhile in view of a detailed molecular characterization at the distributed level. We here formulate a first set of guidelines for improving SEC-based characterization, putting forward the relevance of detailed experimental data recording and error analysis as well as in silico characterization support. We consider both linear homopolymers, copolymers, branched and crosslinked polymers, and highlight that within the polymer community we should further work on the even better embedding of multi-detector approaches and advanced modeling techniques such as kinetic Monte Carlo. Overall this work supports the development of robust workflows for the analysis of advanced polymer materials, bridging the worlds of experimentalists and modelers.
An exploratory study on the possibility of developing bimodal PVC microsuspensions via the in situ generation of small particles in the presence of a large microsuspension seed is presented in order to reduce the number of process steps. The in situ nucleation step was investigated using 3 different redox initiators, and the final products were compared to the reference product made by mixing 2 seeds and copolymerizing them to increase the solid content. It was found that while none of the alternatives considered allowed to reproduce the exact PSD of the reference system, a redox system based on a combination of ammonium persulfate and Iron (II) sulfate came the closest to the reference product. It enabled us to generate well-defined small particles in the right size range. The other systems that did not generate enough negatively charged free radicals led to poorly defined small particle populations.
Kinetic models are essential tools for providing a fundamental understanding of polymerization processes. However, they are constrained by the current level of understanding of the system under study and often rely on numerous unknown parameters that are difficult to measure experimentally or estimate reliably. To address these challenges, polymer chemistry-informed neural networks (PCINNs) have been developed in which the domain knowledge embedded in these kinetic models is combined with data-driven machine learning tools. Herein, we investigate the robustness and limitations of this methodology through an in-silico investigation in which an imperfect kinetic model is combined with small datasets. As a case study, high-temperature solution polymerization of methyl methacrylate was considered. It is demonstrated that PCINNs are able to utilize imperfect kinetic models in order to reach high levels of predictive performance, with reliable extrapolation at reaction temperatures significantly beyond the range of the original dataset. Integrating domain knowledge was found to substantially reduce variance in the outcomes of the final model when using a relatively small dataset. Finally, it is shown that PCINNs can be used to provide insights into the causes of poor predictive performance of kinetic models, offering the ability to use data-driven learning to inform first-principles approaches.
Starved-feed semi-batch radical polymerization of isobornyl acrylate (iBoA) at high temperatures is used to produce a high macromonomer content polymer without a mediating agent. The p(iBoA) macromonomer solution serves as an addition-fragmentation agent to synthesize acrylate blocky copolymers via a single-pot sequential feeding process, as has been demonstrated using n-butyl acrylate (BA) as the second monomer. While previous studies confirmed macromonomer incorporation, the copolymer yield was not quantified. In this work, post-polymerization crosslinking is used to quantify macromonomer incorporation and the impact of reaction temperature and monomer feed ratio. Reaction temperature has little effect on the fraction of copolymer product (70 wt.% at 140 degrees C and 160 degrees C), but strongly influences the relative amount of block versus comb copolymer structure. Maintaining a 50/50 feed ratio of the two monomers is, however, necessary to achieve copolymer levels greater than 60 wt.% in the final product. The ability to synthesize a structured copolymer product by radical polymerization without a mediating agent and control the amounts of block versus comb structure offers the potential to efficiently produce polymeric dispersants for industrial applications.
In the commercial production of acrylic and vinyl polymers, trace amounts of oxygen can significantly disrupt free-radical polymerization by generating toxic volatile organic compounds (VOCs), retarding polymerization, and altering polymer properties. To establish a mechanistic understanding of these effects across industrially relevant monomers, this work examines the influence of oxygen on the solution homopolymerizations of butyl acrylate (BA), vinyl acetate (VAc), and methyl methacrylate (MMA). Systematic experiments conducted under nitrogen and air reveal pronounced monomer-dependent behavior: BA and VAc exhibit severe retardation and form only oligomeric species, while MMA continues to polymerize but shows substantial reductions in molar mass. In all three systems, the presence of oxygen promotes the formation of toxic VOCs, with the extent of oxidation increasing sharply as monomer concentration decreases. To complement the experimental findings, a mechanistic kinetic model is developed for MMA, enabling prediction of monomer conversion, VOC formation, and molecular-weight evolution under oxygen-containing conditions. Together, this work provides a unified framework for understanding oxygen-induced defects in industrial free-radical polymerization processes involving acrylic and vinyl monomers.
