CFD simulations of mixing conditions of isobutylene/maleic anhydride and their polymerization in continuous-flow synthesis

Abstract

Alternating copolymerization of isobutylene (IB) and maleic anhydride (MAh) affords valuable isobutylene–maleic anhydride (IBMA) materials. But conventional stirred-tank processes in ethyl acetate are hindered by heterogeneous gas–liquid–solid behavior, sluggish mass transfer, and product precipitation, while DMF, offering potential homogeneity, performs poorly under mild batch conditions. Here, we use CFD-guided reaction process design and a high-pressure continuous-flow strategy to convert the system into a single-phase, homogeneous operation in DMF by selecting an appropriate solvent/temperature/pressure window and enforcing rapid micromixing. This process-intensified approach delivers minute-scale residence times and an up to 85% yield at 100–130 °C and 3.3 MPa, with controllable M_n between 8 and 20 kg mol⁻¹ and narrow dispersity. By comparison, the ethyl acetate system required 4 h to reach an 84.9% yield, while batch in DMF at 70 °C and 0.5 MPa afforded only 2.66%. The method offers a route to otherwise intractable alternating copolymerization of gaseous monomers and precipitation products, with advantages in safety, productivity, and scalability.

---

1. Introduction

In the isobutylene–maleic anhydride (IBMA) copolymer obtained by alternating isobutylene and maleic anhydride, the reactive anhydride cycles in the polymer chains facilitate a range of functionalization and modifications. The primary applications of IBMA encompass ceramic additive components, water scale reduction, adhesives, battery packaging materials, etc.^1,2 Moreover, IBMAs have shown relatively new applications by altering the maleic anhydride portion to produce a variety of functional compounds, such as lithium battery electrolytes,³ crosslinking biological nanomaterials,⁴ and long-term cell tracers,⁵ to hydraulic lime modification.⁶

Alternating copolymers by an electron-deficient monomer, such as styrene, isobutylene, or butadiene, with maleic anhydride/maleimide could be obtained by highly exothermic polymerization. Thus, the highly active unstable center of the carbocation formed after a long initiation, combined with rapid and exothermic polymerization with increased viscosity, indicates the challenging control during the reaction.

Nowadays, continuous flow techniques are extensively explored from organic synthesis to various fields due to their high productivity, heat and mixing efficiency, safety, and reproducibility.^7–12 Additionally, the continuous flow method is emerging in polymer synthesis, offering accurate control over temperature and pressure compared to the traditional batch processes.^13–18 As one of the most extensively investigated types of maleic anhydride radical copolymerization with various monomers, the synthesis of styrene–maleic anhydride alternating copolymers (SMA) has been well studied.^19–23 Li and coworkers¹⁹ conducted the synthesis of SMA by reversible addition–fragmentation chain transfer (RAFT) polymerization in a continuous flow reactor based on a simple T-junction followed by a straight tube. At the same reaction temperature, the yield reached 69% in just 33 min, which was 2 h in the tank reactor.

In contrast, the synthesis of IBMA was rarely explored because of its specific properties. The boiling point of IB is as low as −6.90 °C at atmospheric pressure. What's more, the product is insoluble in commonly used weakly polar ester solvents.²⁴ Bacskai and coworkers conducted the synthesis of IBMA in THF or benzene solvent at 70 °C at 0.5 MPa with a solid product.²⁵ The information about the reaction characteristics of IBMA synthesis is still insufficient considering the inadequacy of existing works, and it requires further study.

For the continuous flow synthesis of IBMA, the obstacles must be effectively overcome. Firstly, it is necessary to increase the reaction pressure, ensuring that isobutylene is in the liquid state. Luo and his coworkers synthesized polyisobutylene at −20 °C to 50 °C, using liquid isobutylene at 0.3 MPa.²⁶ In the flow system, the miscible monomers could be well mixed and pumped into the reactors at high back pressure, allowing for reproducibility that is hardly reached in batch operation.

More importantly, the homogeneity of the reaction mixture must be considered in the flow process. The solid precipitation, which can pose a significant risk of clogging, as well as the hard mass transfer and heat transfer, inevitably leads to accidents and severely limits the applicability of the continuous-flow process. In the literature, solvent optimization approaches for solid products have been developed to improve the substance solubility.²⁷ Process intensification under continuous-flow conditions could enhance the performance of conventionally poor solvents. Mulks et al. used organomagnesium reagents in an optimized segmented/droplet microreactor, in which the green deep eutectic solvents were applied to dissolve lithium salts, avoiding clogging.²⁸ We also developed continuous in-line salt formation from tanshinone sulfonation in DCM solvent in a polyfluoroalkoxy (PFA) coil reactor, with a total yield of 90%.²⁹

In parallel, the computational fluid dynamics (CFD) study of mixing can provide insights into the potential state of the reaction and evaluate the impact on the reaction.^30–32 The mixing process was effectively described by the mixing index (MI) or mixing efficiency (ME). Luo and coworkers conducted experiments coupled with CFD simulation to predict the flow field of continuous synthesis of 1-ethoxy-2,3-difluoro-4-iodo-benzene,³³ or kilogram-scale synthesis of piperacillin³⁴ in a microreactor system. Garg et al. investigated the mixing conditions at the inlet of free radical polymerization of methyl methacrylate in tubular microreactors.³⁵ We explored the fluid mixing performance of a homemade designed 3D circular cyclone-type microreactor by CFD simulation.³⁶ The generated paired opposite vortices and local pressure enhanced the turbulence within laminar flow regimes.

In this work, based on the exploration of the IBMA copolymerization conditions in a tank reactor, and CFD analysis of the mixing performance in the flow system, a continuous flow reaction was built, as shown in Fig. 1. The subject of this study is to provide new ideas and solutions in conducting IBMA reactions in an efficient and safe manner by exploiting the continuous-flow method.

Fig. 1 Schematic diagram of the continuous-flow copolymerization of isobutylene and maleic anhydride.

Results and discussion

1. Copolymerization of isobutylene and maleic anhydride in a tank reactor

For a continuous-flow reaction, it is essential that both starting materials and the product are soluble throughout the process. Thus, the reaction was conducted in a 50 mL stainless steel tank reactor to search for suitable conditions for the continuous flow. MAh was directly introduced from the cylinder to the tank in various solvents, and the reaction was carried out at 70 °C in N₂ at 0.5 MPa. As shown in Table 1, the yield of the copolymer obtained in traditional ethyl acetate (EA) solvent after 4 h reaction was 84.9% (entry 1) at a MAh concentration of 1.23 mol L⁻¹. It was found that the precipitate covered on the surface of the high viscosity reaction solution, with an M_n of 93.5 kg mol⁻¹. When the MAh concentration was increased to 2.47 mol L⁻¹, due to the severe clogging and high viscosity, the conversion was 70.5% and the M_n was up to 247 kg mol⁻¹. The product was hard to grind, suggesting the difficult handling in the tank reactor. The X-ray powder diffraction (XRD) pattern showed that the precipitate was amorphous. Thus, precise control over the initiation, propagation, and termination during polymerization is required in terms of pressure, temperature, concentration of the initiator, and so on. It is noticed that the rate of the copolymerization with 1.23 mol L⁻¹ MAh in EA solvent was relatively below 20% after the reaction for 2 h. For the slowly propagating monomers, the radical concentration needs to be rather high. Thus, the conversion drastically increased to 80% in the following 1 h, suggesting the rapid chain growth vs. slow initiation. Furthermore, when the reaction occurred within 1.5 h, almost no product was formed. The product formed at 1.5–2 h might be related to the precipitation acceleration effect during polymerization.^37,38

Table 1 Copolymerization of isobutylene–maleic anhydride in a tank reactor^a

Entry	MAh (mol L^−1)	Solvent	Temp. (°C)	Time (h)	Yield^d (%)	M n (kg mol^−1)	Đ
a 10 mL solvent, 8 mL IB. b Volume ratio. c Not detected. d Isolated yield.
1	1.23	EA	70	4	84.9	93.5	1.4
2	2.47	EA	70	4	70.5	247.0	1.9
3	2.47	EA : DMF = 1 : 1^b	70	4	20.1	31.5	1.6
4	2.47	EA : DMF = 1 : 3^b	70	4	4.87	24.0	1.7
5	2.47	DMF	70	4	2.66	—^c	—^c
6	2.47	DMF	80	6	37.4	36.5	2.2
7	2.47	DMF	80	8	42.8	20.1	1.4

---

Then, suitable conditions for flow chemistry were screened. In Table 1, at 70 °C, in the mixture of ethyl acetate and DMF at ratios of 1 : 1 (entry 3) and 1 : 3 (entry 4), the yields were 20.1% and 4.87%, respectively. But the whole solution was still turbid with small particles. In DMF, the yield was only 2.66% (entry 5). And at 80 °C, the yield was 4.8% in DMF after the reaction for 4 h, while it increased to 37.4% (entry 6) and 42.8% (entry 7) within 6 and 8 h, respectively. In addition, the viscosity of the system was gradually increased during the reaction. Then, the free radical polymerization showed an obvious self-acceleration period.³⁹ However, the subsequent rate reduction may be attributed to the decrease in the concentration of the active monomer in the system. In Fujimori's report, DMF was also tested as a solvent in this reaction.⁴⁰ It seemed that DMF was suitable for continuous-flow synthesis of IBMA, which might be benefited by the process intensification.

2. CFD simulations of the mixing of isobutylene in DMF

Before the synthesis of IBMA in the continuous-flow microreactor, CFD was carried out for the initial reaction screening. Due to the polar groups in the isobutylene–maleic anhydride copolymer, it is poorly soluble in hydrocarbon solvent, as well as the isobutylene. In DMF, the homogeneous polymerization of IB and MAh could avoid the lump blocking in the flow reactor. However, due to numerous interfaces within the system, the relationship of various types and flow states in solvent, the reaction performance is difficult to understand. Table S1 lists the physical properties of DMF and IB, and their mixture was simulated by CFD.

For the slow reaction applied by continuous-flow techniques, it typically must be sufficiently fast to allow for reasonable reactor operation. The slow reactions are likely to be significantly accelerated without losing control by applying high temperatures in the microreactor.⁴¹ Due to the high heat and mixing control, a broader temperature window can be chosen than in classical batch operation. The critical temperature and pressure of isobutylene are 145 °C and 4.25 MPa, respectively. At 130 °C, the Aspen simulation of the mass ratio of the gas–liquid two-phase of IB and DMF are shown in Table S2. The liquid isobutylene can completely dissolve in DMF under simulated conditions above 2.8 MPa without obvious phase separation in the microreactor. Thus, the dilute reaction mixture of liquid isobutylene in DMF with maleic anhydride can be approximately regarded as a Newtonian fluid.

Fig. 2 shows the typical integrated system used in the CFD simulation. The mixture of isobutylene in DMF as the flow phases from two inlets was studied. The flow field in the microreactors and the delay loops are described as the incompressible model.^42,43 The governing equations include the conservation relationship of mass, momentum, and species.^44–46 For each phase in the flow, the flow constraint equation (eqn (1)) and volume fraction constraint equation (eqn (2)) must be satisfied:

Fig. 2 Diagram 3D model of the reaction system in CFD simulation.

In the equation, α_i and ρ_i are the volume fraction and density of the phase, respectively. In the continuous-flow reactor, the microfluid is at continuous flow state without slip boundary conditions, so the conventional macroscopic Navier–Stokes equation (eqn (3)) was used:

In the equation, P, g, μ, and F represent the pressure, gravity coefficient, viscosity, and the external force on the unit volume fluid, respectively. In the simulation process, the density and the viscosity were calculated by volume-weighted-mixing-law and mass-weighted-mixing-law, respectively. The flow in the microchannel is mainly laminar, so the free diffusion of molecules plays a dominant role in the mass transfer process. At a certain temperature, the diffusion caused by the thermal motion of molecules can be expressed by Fick's law (eqn (4)):

In the equation, c_i is the concentration of the component and D is the diffusion coefficient. In order to search for the diffusion coefficient, the Chapman–Enskog formula (eqn (5)) provided by Ansys was used for CFD simulation.

In this equation, D_ij is the diffusion coefficient. T is the environment temperature (K). M_w,i and M_w,j are the molecular weights (g mol⁻¹). P_abs is the absolute pressure (Pa). σ_ij and Ω_D are determined by the L–J characteristic length (Å) and L–J energy parameter (K), as shown in Table S1.

The convection between the phases should also be taken into account in the material transfer process when the flow velocity increases or the local turbulence is produced due to the change in the channel geometry. Therefore, the convection–diffusion equation can be obtained as shown in eqn (6):

With the increase of flow rate, the flow pattern changes from laminar flow to turbulent. As a result, the standard k–ε equations describe these situations, as shown in eqn (7) to (9):

In these equations, k is the turbulent kinetic energy (m² s⁻²). ε is the turbulent dissipation rate (m² s⁻³). μ_t is the turbulent (vortex) viscosity (Pa s). τ_ij is the turbulent stress tensor (Pa), and C_1ε = 1.44, C_2ε = 1.92, σ_k = 1.0, σ_ε = 1.3, and C_μ = 0.09.

The mixing index (M) is used to indicate the degree of mixing in a homogeneous system. In eqn (10), the value of M closer to 1 suggests the complete mixing:

N represents the number of sampling points in the statistical area. c_i is the component concentration on each sampling point, while c₀ is the component average concentration on each sampling point.⁴⁴

Mesh independence studies were conducted under the conditions of importing pure DMF with a flow rate of 0.00217 m s⁻¹ at inlet 1 and importing pure isobutylene with a flow rate of 0.00587 m s⁻¹ at inlet 2. The criterion for evaluation was the average concentration of isobutene at the same point, which remained basically unchanged after reaching 1000k cells, as shown in Table S3.

In the continuous-flow reaction, the flow rate influences the yield of the product in a certain range of reaction temperature, time, flow type, etc. In typical experiments, the flow is slow and limited in the laminar flow regime.^19,47 Here, the mixing simulation was carried out not only in the laminar flow region, but also at a high flow rate.

Table 2 presents the mixing index at various flow rates. The M increases from 0.82 to 0.88 as the total flow rate increases from 0.33 to 1.00 mL min⁻¹. Compared with the rapid mixing at a low flow rate, the mixing index decreases from 0.88 to 0.77 due to the short residence time at a high flow rate of 1.46 mL min⁻¹ (entry 3). The flow remains laminar in the flow rate range from 2.00 to 12.41 mL min⁻¹ (entries 1–5), during which the mixing indices are slightly changed. However, further increasing the flow rate to 18.16 mL min⁻¹, the intensified turbulence promotes the rapid mixing. After that, the higher speed slightly influences the mixing index. In Fig. 3, with the increased flow rate, the turbulence of the fluid becomes increasingly pronounced.

Table 2 Effect of flow rates on the mixing index at a constant flow rate ratio in the continuous-flow reactor

Entry	MAh flow rate (mL min^−1)	Total flow rate (mL min^−1)	Flow regime	Mixing index
1	0.24	0.33	Laminar	0.82
2	0.73	1.00	Laminar	0.88
3	1.46	2.00	Laminar	0.77
4	3.65	5.00	Laminar	0.80
5	9.08	12.41	Laminar	0.79
6	18.16	24.83	Turbulent	0.86
7	45.39	62.07	Turbulent	0.91
8	90.78	124.15	Turbulent	0.91

---

Fig. 3 CFD results on mixing and segregation of isobutylene concentration in DMF at total flow rates of a) 0.33, b) 1.00, c) 2.00, d) 5.00, e) 12.41, f) 24.83, g) 62.07, and h) 124.15 mL min⁻¹.

Generally, the mixing between the two fluids mainly occurs in the injection, where the fluid streams collide with the obvious disturbance between the liquid phases. Fig. 4(A) shows the variation in concentration distribution of the two inlets at different flow rates with various injection angles. When the total flow rate is increased from 1.00 mL min⁻¹ to 12.41 or 124.15 mL min⁻¹ with an isobutylene flow ratio of 27%, the laminar flow transitions into turbulence, resulting in the reversal of concentration distribution at the injection angles of 60° or 120°. The mixing is enhanced at smaller injection angles in the laminar flow rate range, while the mixing efficiency is higher at increased flow rates. As a result, similar reaction yields at 90° or 180° injection angle were observed at low flow rates.

Fig. 4 (A) Distributions of isobutylene concentration at different flow rates with feed angles of a) 60°, b) 90°, c) 120°, and d) 180°; (B) flow diagrams at different flow rates with feed angles of a) 60°, b) 90°, c) 120°, and d) 180°; (C) velocity vector diagrams within 90° at different flow rates of 1.00, 12.41, and 124.15 mL min⁻¹.

In Fig. 4(B), due to the proximity of the DMF phase fluid to IB at feed angles of 60° and 120°, the flow space of IB is constricted, manifesting a sudden drop in mixing and an inversion of the concentration distribution. With the total flow rate increasing from 12.41 to 124.15 mL min⁻¹, the angle of 90° or 180° achieves a higher mixing index. A T-type micromixer with 90° injection angle is optimal for the mixing at low flow rates in terms of pressure drop. For high flow rates within the simulated transition zones, a 180° T-micromixer might be suitable. Fig. 4(C) illustrates the transition from laminar to turbulent flow as the flow rate increased from 1.00 to 124.15 mL min⁻¹ at the 90° injection angle. The velocity vector lines tend to form vortexes, which directly improves the mass transfer efficiency and results in the mixing index fluctuation. Further increasing the flow velocity, the vortex becomes apparent in the vector diagram with intensified disturbance, and the mixing index increases rapidly.

Therefore, according to the simulation results about the flow velocity, flow direction, and flow ratio, we chose a T-type mixer with an injection angle of 90° in the following flow chemistry experiments to prove the effect of the mixing index on the yield.

3. Continuous-flow copolymerization of isobutylene and maleic anhydride

Based on the above results, the copolymerization of isobutylene and maleic anhydride was tested in the continuous-flow reactor, as shown in Fig. S1. Maleic anhydride and the initiator in DMF were pumped into the reaction system by an injection pump, while isobutylene was directly pumped into the reaction system in the liquid form by a double-plunger micropump. According to the gas–liquid phase simulation calculation of the isobutylene–DMF binary system and their respective physical properties (Table S2), the isobutylene–DMF liquid system is in a homogeneous phase at 3.0 MPa at 130 °C. The preliminary experiments were successfully conducted between 1.3 and 4.3 MPa. The reaction was stable under 3.3 MPa, which was selected as the reaction pressure.

Firstly, two common initiators, such as benzoyl peroxide (BPO) and 2,2′-azodiisobutyronitrile (AIBN), were compared. It is known that the rate of polymerization is proportional to the square root of the rate of initiation in the radical polymerization.⁴⁸ Due to the good heat transfer, the reaction could be carried out at high temperature. In Table S4, the initiation efficiency of the BPO initiator was 34.2% at 110 °C (entry 2), which was less than that of AIBN with 64.2% (entry 1). Thus, AIBN was selected as the initiator in the continuous flow reactor. In Table 3, it was obvious that increasing the initiator content from 1 mol% to 3 mol% could effectively improve the yield from 51.5% to 85.3%, related to the reaction product with fluctuating M_n and polydispersity.

Table 3 Effect of AIBN concentration on isobutylene–maleic anhydride copolymerization in the continuous-flow reactor^a

Entry	AIBN (mol%)	Yield^b (%)	M n (kg mol^−1)	Đ
a Temperature: 110 °C, reaction time: 10 min, pressure: 3.3 MPa; total volume of reactor: 5 mL; MAh concentration: 2 mol L^−1; ratio of IB and MAh: 2 : 1. b Isolated yield.
1	1	51.5	7.40	2.2
2	2	73.2	19.7	1.7
3	3	85.3	7.41	2.7

---

As shown in Table 4, the yield of the copolymer increased as the concentration of MAh and the ratio of IB to MAh were raised. The yield of the copolymer reached 39.42% with 3 mol L⁻¹ MAh (entry 3), and 48.57% at a molar ratio of IB and MAh of 2 : 1 (entry 5). However, the copolymers at high MAh concentration might prefer to crystallize, leading to clogging in the coil. Thus, the maleic anhydride concentration was set at 2 mol L⁻¹ in the investigation.

Table 4 Copolymerization of isobutylene–maleic anhydride with various molar ratios of isobutylene to maleic anhydride in the continuous-flow reactor^a

Entry	MAh (mol L^−1)	Ratio of IB/MAh	Yield^b (%)
a Temperature: 120 °C; reaction time: 5 min; pressure: 3.3 MPa; AIBN: 1 mol% of MAh; total volume of the reactor: 5 mL. b Isolated yield.
1	1	1	15.0
2	2	1	22.9
3	3	1	39.4
4	1	2	35.3
5	2	2	48.6

---

Then, the effects of residence time (T_R), temperature (Temp.), and flow rate (F_R) on the continuous-flow polymerization were investigated. In Table 5, the reaction was carried out from 100 to 130 °C respectively to initiate AIBN (entries 1–4). It was noted that the solution was yellow after the reaction in the continuous reactor, while it became light pink in the tank reactor after a long time, suggesting the side reaction between MAh and DMF. The flow rate of MAh of 0.73 mL min⁻¹ corresponds to a residence time of 5 min in a 5 mL coil. The yields rapidly increased from 7.7% (entry 1) to 40.8% (entry 4). When the residence time increased to 10 or 15 min, corresponding to a flow rate of MAh of 0.36 or 0.24 mL min⁻¹, the yields were raised. The longer residence time suggested the better reaction of the reactants, obtaining a yield of 29.7% at 100 °C (entry 5). However, the yield was 69.5% at 130 °C (entry 16), which was lower than that at 120 °C with 81.8% (entry 15). Besides, the M_n and Đ obtained after 15 min were higher and narrower than those in the residence time of 5 min. Furthermore, when the residence time was increased from 15 to 20 min, the yield of the copolymer slightly increased from 73% to 78%. At 110 °C, with a flow rate of MAh of 0.73 mL min⁻¹ and with a residence time of 15 min, the yield of the copolymer was slightly increased from 64.1% (entry 10) to 83.3% (entry 14), while the M_n and Đ were decreased from 20.0 to 18.7 kg mol⁻¹ and 1.4 to 1.2, respectively, suggesting that the increased total flow rate might benefit the copolymerization. As shown in Table 2, the mixing index increased from 0.82 to 0.88 as the total flow rate increased from 0.33 to 1.00 mL min⁻¹, which is consistent with the yields from 64.1% to 83.3%.

Table 5 Copolymerization of isobutylene–maleic anhydride under different conditions in the continuous-flow reactor^a

Entry	Temp. (°C)	T R (min)	F R (mL min^−1)	Yield^d (%)	M n (kg mol^−1)	Đ
a Pressure: 3.3 MPa; AIBN: 1 mol% of MAh; MAh concentration: 2 mol L^−1; ratio of IB and MAh: 2 : 1. b Residence time. c MAh flow rate. d Isolated yield.
1	100	5	0.73	7.7	14.5	1.2
2	110	5	0.73	23.5	4.9	2.1
3	120	5	0.73	36.1	7.8	1.8
4	130	5	0.73	40.8	16.6	1.2
5	100	10	0.36	29.7	20.3	1.3
6	110	10	0.36	51.5	7.4	2.2
7	120	10	0.36	61.1	15.2	1.8
8	130	10	0.36	64.5	16.8	1.2
9	100	15	0.24	42.8	18.8	1.3
10	110	15	0.24	64.1	20.0	1.4
11	120	15	0.24	73.3	19.6	1.4
12	130	15	0.24	65.5	17.7	1.3
13	100	15	0.73	71.6	15.7	1.3
14	110	15	0.73	83.3	18.7	1.2
15	120	15	0.73	81.8	18.8	1.0
16	130	15	0.73	69.5	15.6	1.2

---

In Table S5 and Fig. 5, with the flow ratio of isobutylene to DMF gradually increasing from 9 : 91 to about 27 : 73 under the experimental conditions, the corresponding reaction yield increased from 15% to 48%, while the mixing index of DMF to isobutylene changed from 0.77 to 0.88 according to eqn (10). The fluid in the tube exhibits laminar flow under the experimental conditions in Table 5. The concentration distribution is almost parallel to the longitudinal plane. As a result, the higher the flow rate of isobutylene, the stronger the mass transfer efficiency and the higher the mixing efficiency. The concentration of 2 mol L⁻¹ maleic anhydride and the monomer molar ratio of 2 : 1 presented the best performance.

Fig. 5 Effect of flow rate ratios of isobutylene to DMF at the total flow rate of 1.00 mL min⁻¹ on the yield and distributions of isobutylene mass fractions.

Here, we summarized the various test conditions to obtain the optimized conditions in Table S6. The highest yield of the copolymer could be achieved at 85.8% (entry 1), in which the reaction conditions were the same as those in entry 14 in Table 5 except that the concentration of the initiator was 2%. When the temperature was raised to 120 °C, the yield of the product decreased from 85.8% to 81.7% (entry 2). Furthermore, the product yield was 83.4% (entry 6), suggesting that the initiator concentration had a slight effect on the yield.

In virtue of the process intensification in the continuous-flow microreactor, the copolymerization of IB and MAh could be carried out in homogeneous mode in poor solvent DMF. And the yield was 85% at 110 °C in 15 min. However, in the tank reactor, 4 h was needed to achieve this conversion with the solid product covered over EA at 70 °C, accompanying a long initiation period and fast acceleration in the propagation period. The improved control of polymerization creates a significantly safer advance.

4. Characterization of copolymers

The NMR results of IBMA obtained by the continuous-flow method are shown in Fig. S2. The double peaks in the vicinity of 1.1 ppm indicated that isobutene and maleic anhydride existed alternately in the copolymer.

The morphology of products (Table S7) was investigated by SEM and TEM. The particles obtained in the batch reaction in EA (Fig. S3(a)–(c)) were almost spherical due to the good agglomeration during the 4 h reaction. In contrast, the particles obtained in the batch reaction in DMF (Fig. S3(e) and (h)) displayed irregular shapes. For the products obtained by continuous-flow polymerization in DMF, the particles formed quickly during washing in ethanol, resulting in the irregular products. With the lower concentration of the product (Fig. S3(d), (f), and (g)) (yields of 73.3%, 61.1% and 36.1%, respectively), the particle size decreased. As shown in the TEM images in Fig. S4, the sample obtained by the tank reactor was spherical after a long reaction time. For the one by the continuous-flow reactor, besides the spherical form, the particles were generally irregular.

An FTIR spectroscopy trace of this prepared polymer is shown in Fig. S5, which shows the characteristic C=O stretching of the carboxylic acid present in maleic acid at 1770 cm⁻¹ as well as 1850 cm⁻¹. The peaks at 1080 and 925 cm⁻¹ are indicative of C–O–C stretching vibrations. The C–H stretching vibrations of (CH₃ and CH₂) alkane groups in isobutylene present at 2970 cm⁻¹ indicate the presence of C–H alkane groups in the polymer.

Conclusion

We established and validated a CFD-guided, high-pressure continuous-flow implementation of the alternating copolymerization of isobutylene and maleic anhydride in DMF. Thus, a traditional stirred-tank process involving precipitation with limited mass-transfer was transformed into a homogeneous, single-phase operation. The yield of the copolymer can be effectively promoted in minutes owing to the enhancement of mass and heat transfer by the increased flow rate, reaction temperature and pressure in DMF solvent. At a molar ratio of isobutylene to maleic anhydride of 2 : 1, the highest yield of 85% was achieved at a maleic anhydride concentration of 2 mol L⁻¹, within 15 min at 110 °C and 3.3 MPa. This work provides clear novelty in reaction implementation (homogeneous, pressurized continuous flow) and reaction process design (CFD-guided tubular system, phase-behavior and micromixing control), and offers safe, scalable alternating copolymerization of gaseous monomers and precipitation products.

Conflicts of interest

There are no conflicts to declare.

Data availability

Supplementary information is available. See DOI: https://doi.org/10.1039/D5RE00245A.

All relevant data are within the manuscript and its additional files.

Acknowledgements

This work was partially supported by the National Natural Science Foundation of China (No. 22278087), the Science and Technology Projects in Guangzhou (2023A04J1371), and the Guangdong Basic and Applied Basic Research Foundation (2023A1515110391).

References

1. L. Ma, Y. Shan, X. Sun, M. Gao, X. Han and J. Xu, Fabrication of highly transparent AlON ceramics by fast formed direct aqueous slip casting green body, Mod. Phys. Lett. B, 2022, 36, 2242027.
2. R. Del Olmo, J. Harwood, D. d. M. Zanata, F. Santino, J. L. Olmedo-Martinez, P. Howlett and I. Villaluenga, Enabling room temperature solid-state lithium batteries by blends of copolymers and ionic liquid electrolytes, J. Power Sources, 2024, 621, 235233.
3. S.-T. Oh, Y.-W. Jeong, S.-S. Kim and S.-W. Ryu, Synthesis of poly(isobutylene-alt-maleic anhydride)-based water-soluble binders and their electrochemical properties, Ionics, 2022, 28, 4303–4310.
4. S. Cao, S. Peeters, S. Michel-Souzy, N. Hamelmann, J. M. J. Paulusse, L. L. Yang and J. Cornelissen, Construction of viral protein-based hybrid nanomaterials mediated by a macromolecular glue, J. Mater. Chem. B, 2023, 33, 7933–7941.
5. X. Guan, Y. Ding, S. Lai, X. Yang, J. Wei, J. Zhang, L. Zhang, K. Wang, J. Tong and C. Li, Nonconjugated fluorescent polymer nanoparticles by self-assembly of PIMA-g-beta-CD for live-cell long-term tracking, Carbohydr. Polym., 2022, 291, 119633.
6. X. Wei, Y. Li, J. Hui, W. Wang, B. Zhang, L. Chang, Y. Jiao, Z. Sang, H. Luo and X. Wang, Effects of an isobutylene-maleic anhydride copolymer on the rheological behavior and early hydration of natural hydraulic lime, Polymers, 2022, 14, 4104.
7. T. Tsubogo, H. Oyamada and S. Kobayashi, Multistep continuous-flow synthesis of (R)- and (S)-rolipram using heterogeneous catalysts, Nature, 2015, 520, 329–332.
8. T. Yang, Z. Chen, Y. Wang and J. Xu, Fundamental and development of microreaction technology for the application of fine chemicals synthesis, J. Chem. Eng. Jpn., 2023, 56, 2197011.
9. A. Saikia, R. Newar, S. Das, A. Singh, D. J. Deuri and A. Baruah, Scopes and challenges of microfluidic technology for nanoparticle synthesis, photocatalysis and sensor applications: A comprehensive review, Chem. Eng. Res. Des., 2023, 193, 516–539.
10. V. V. K. Doddapaneni, A. Chang, H.-Y. Jun, C.-H. Choi, Z. Feng and C.-H. Chang, Recent advances in microfluidics-enabled controlled reaction, assembly and exfoliation of inorganic nanomaterials, CrystEngComm, 2023, 25, 5606–5628.
11. S. Khizar, N. Zine, A. Errachid, N. Jaffrezic-Renault and A. Elaissari, Microfluidic-based nanoparticle synthesis and their potential applications, Electrophoresis, 2022, 43, 819–838.
12. Z. Jin, H. Wang, X. Hu, Y. Liu, Y. Hu, S. Zhao, N. Zhu, Z. Fang and K. Guo, Anionic polymerizations in a microreactor, React. Chem. Eng., 2022, 7, 1026–1036.
13. M. Rubens, J. H. Vrijsen, J. Laun and T. Junkers, Precise polymer synthesis by autonomous self-optimizing flow reactors, Angew. Chem., Int. Ed., 2019, 58, 3183–3187.
14. M. H. Reis, F. A. Leibfarth and L. M. Pitet, Polymerizations in continuous flow: recent advances in the synthesis of diverse polymeric materials, ACS Macro Lett., 2020, 9, 123–133.
15. H. Wang, Z. Jin, X. Hu, Q. Jin, S. Tan, A. Reza Mahdavian, N. Zhu and K. Guo, Continuous flow cationic polymerizations, Chem. Eng. J., 2022, 43, 132791.
16. L. Xiang, M. Qiu, M. Shang and Y. Su, Continuous synthesis of star polymers with RAFT polymerization in cascade microreactor systems, Polymer, 2021, 222, 123669.
17. L. Xiang, Y. Song, M. Qiu and Y. Su, Synthesis of branched poly(butyl acrylate) using the Strathclyde method in continuous-flow microreactors, Ind. Eng. Chem. Res., 2019, 58, 21312–21322.
18. Y. Song, M. Shang, G. Li, Z.-H. Luo and Y. Su, Influence of mixing performance on polymerization of acrylamide in capillary microreactors, AIChE J., 2018, 64, 1828–1840.
19. W. Liu, Q. Li, Y. Zhang, T. Liu, L. Wang, H. Li and Y. Hu, Continuous-flow RAFT copolymerization of styrene and maleic anhydride: Acceleration of reaction and effect of polymerization conditions on reaction kinetics, J. Flow Chem., 2021, 11, 867–875.
20. S. Bag, S. Ghosh, S. Paul, M. E. H. Khan and P. De, Styrene-maleimide/maleic anhydride alternating copolymers: Recent advances and future perspectives, Macromol. Rapid Commun., 2021, 42(23), e2100501.
21. D. J. Keddie, A guide to the synthesis of block copolymers using reversible-addition fragmentation chain transfer (RAFT) polymerization, Chem. Soc. Rev., 2014, 43, 496–505.
22. M. J. Derry, L. A. Fielding and S. P. Armes, Polymerization-induced self-assembly of block copolymer nanoparticles via RAFT non-aqueous dispersion polymerization, Prog. Polym. Sci., 2016, 52, 1–18.
23. M. D. Nothling, Q. Fu, A. Reyhani, S. Allison-Logan, K. Jung, J. Zhu, M. Kamigaito, C. Boyer and G. G. Qiao, Progress and perspectives beyond traditional RAFT polymerization, Adv. Sci., 2020, 7(20), 2001656.
24. K. Liu, H. Teng, X. Qu and X. Jin, Study on the synthesis of maleic anhydride-isobutylene, Ind. Water Treat., 2013, 33, 56–59.
25. R. Bacskai, L. P. Lindeman and D. L. Rabenstein, Stereochemistry of alternating isobutene-maleic anhydride, isobutene-dimethyl fumarate, and isobutene-dimethyl maleate copolymers, J. Polym. Sci., 1972, 10, 1297–1310.
26. S. Zhu, Y. C. Lu, K. Wang and G. Luo, Fast flow synthesis of highly reactive polyisobutylene co-initiated by an AlCl₃/isopropyl ether complex, RSC Adv., 2016, 6, 9827–9834.
27. H. L. D. Hayes and C. J. Mallia, Continuous flow chemistry with solids: A review, Org. Process Res. Dev., 2024, 28, 1327–1354.
28. F. F. Mulks, B. Pinho, A. W. J. Platten, M. R. Andalibi, A. J. Expósito, K. J. Edler, E. Hevia and L. Torrente-Murciano, Continuous, stable, and safe organometallic reactions in flow at room temperature assisted by deep eutectic solvents, Chem, 2022, 8, 3382–3394.
29. M. Sun, C. Liang, L. Cao, Y. Wang, J. Yang, S. Hou, W. Yu, Y. Ma, R. Cheng and J. Ye, Two-step continuous flow process of sodium tanshinone IIA sulfonate using a 3D circular cyclone-type microreactor, Chin. Chem. Lett., 2024, 35, 108738.
30. K. Ekambara, M. T. Dhotre and J. B. Joshi, CFD simulation of homogeneous reactions in turbulent pipe flows-tubular non-catalytic reactors, Chem. Eng. J., 2006, 117, 23–29.
31. H. Zhong, J. Chen, F. Gao, J. Zhang, Y. Zhu and B. Niu, 3D virtual full-loop CFD simulation of industrial two-stage FCC reaction-regeneration system, Int. J. Chem. React. Eng., 2022, 20, 1179–1191.
32. N. Padoin, T. Matiazzo, H. G. Riella and C. Soares, A perspective on the past, the present, and the future of computational fluid dynamics (CFD) in flow chemistry, J. Flow Chem., 2024, 14, 239–256.
33. Q. Chen, S. Xia, Y. Wang, G. Luo, H. Shang and K. Wang, Continuous synthesis of 1-ethoxy-2,3 difluoro-4-iodo-benzene in a microreactor system and the Gaussian and computational fluid dynamics simulations, AIChE J., 2021, 67, e17217.
34. Y. Xie, Q. Chen, G. Huang, Y. Wang, W. Hu, Z. Yan, X. Wang, J. Huang, M. Gao, W. Fei and G. Luo, Scaling up microreactors for kilogram-scale synthesis of piperacillin: Experiments and computational fluid dynamics simulations, AIChE J., 2021, 67, e17231.
35. D. K. Garg, C. A. Serra, Y. Hoarau, D. Parida, M. Bouquey and R. Muller, Numerical investigations of perfectly mixed condition at the inlet of free radical polymerization tubular microreactors of different geometries, Macromol. Theory Simul., 2020, 29, e2000030.
36. S. Hou, M. Sun, L. Cao, C. Liang, J. Yang, X. Zhou, J. Ye and R. Cheng, Computational fluid dynamics simulation and experimental study on mixing performance of a three-dimensional circular cyclone-type microreactor, Chin. Chem. Lett., 2024, 35, 108761.
37. G. T. Caneba, Z. Xu and Y. L. Dar, Free-radical retrograde-precipitation copolymerization of vinyl acetate and acrylic acid, J. Appl. Polym. Sci., 2009, 113, 3872–3882.
38. N. Y. Abu-Thabit, Chemical oxidative polymerization of polyaniline: A practical approach for preparation of smart conductive textiles, J. Chem. Educ., 2016, 93, 1606–1611.
39. G. A. O'Neil, M. B. Wisnudel and J. M. Torkelson, A critical experimental examination of the gel effect in free radical polymerization: Do entanglements cause autoacceleration?, Macromolecules, 1996, 29, 7477–7490.
40. F. Lauterbach, M. Rubens, V. Abetz and T. Junkers, Ultrafast photoRAFT block copolymerization of isoprene and styrene facilitated through continuous-flow operation, Angew. Chem., Int. Ed., 2018, 57, 14260–14264.
41. N.-T. H. Ha and K. Fujimori, Theoretical study of the copolymerization of styrene and maleic anhydride prepared in carbon tetrachloride and in N,N-dimethylformamide, Acta Polym., 1998, 49, 404–410.
42. A. A. Yagodnitsyna, A. V. Kovalev and A. V. Bilsky, Flow patterns of immiscible liquid-liquid flow in a rectangular microchannel with T-junction, Chem. Eng. J., 2016, 303, 547–554.
43. M. Kashid and L. Kiwi-Minsker, Quantitative prediction of flow patterns in liquid–liquid flow in micro-capillaries, Chem. Eng. Process., 2011, 50, 972–978.
44. H. Lv, Z. Yang, J. Zhang, G. Qian, X. Duan, Z. Shu and X. Zhou, Liquid flow and mass transfer behaviors in a butterfly-shaped microreactor, Micromachines, 2021, 12, 883.
45. A. Alam, A. Afzal and K.-Y. Kim, Mixing performance of a planar micromixerwith circularobstructionsin a curved microchannel, Chem. Eng. Res. Des., 2014, 92, 423–434.
46. S. Zhu, K. Wang and Y. Lü, Effects on the mixing process of a coiled tube after a T-junction: Simulation and correlation, Chin. J. Chem. Eng., 2018, 26, 2441–2447.
47. T. Iwasaki and J. Yoshida, Free radical polymerization in microreactors. Significant improvement in molecular weight distribution control, Macromolecules, 2005, 38, 1159–1163.
48. H. R. Allcock, F. W. Lampe and J. E. Mark, Contemporary Polymer Chemistry, Pearson Education, Inc., 2003, p. 832.

---