Cationic polymerization of isobutylene catalysed by AlCl₃ with multiple nucleophilic reagents

Abstract

In this work, by exploiting the perfect performances of microflow reactors in mixing and residence time control, we systematically investigated the cationic polymerization of isobutylene (IB) catalysed by AlCl₃ with multiple nucleophilic reagents, isopropyl ether (iPr₂O) and ethyl benzoate (EB). Through properly introducing iPr₂O and EB, the polymerization of IB could produce PIBs with a narrow molecular weight distribution (PDI < 2.0), a relatively high molecular weight (>40 000 g mol⁻¹), and a high content of exo-olefin (w > 70%) at relatively high temperatures (−30 °C), during which most of the monomer conversion (>70%) could be fulfilled within 0.5 s and the chain scission mainly takes place after seconds. The expression [2[EB] + [iPr₂O]]/[AlCl₃] being equal to 1 is recognized as a quantitative criterion for achieving these outcomes, corresponding to H⁺iPr₂OAlCl₃(OH)⁻ initiating polymerization with inhibited chain transfer by introducing EB(AlCl₃)_n, but eliminating free AlCl₃. Increasing the flow capacity in a certain microflow system provides the potential to increase the molecular weight further and facilely tailor it for particular applications. This work verifies the various functions of multiple nucleophilic reagents and their ability to break the trade-off of conversion rate and propagation chain stability in cationic polymerization and develop new functional products of PIBs.

---

1. Introduction

Polyisobutylenes (PIBs), as typical products of cationic polymerization, have exhibited some distinguishing properties,^1,2 such as thermal stability, flexibility at ambient temperature, and impermeability to gases.^3–5 These properties allow PIBs to be used in a wide variety of applications, such as automobile tyres, medical bottle plugs, additives in fuels, and lubricants.^6–9 PIBs with a high content of exo-olefin can also be used to prepare PIB-supported catalysts,¹⁰ for which a high molecular weight with a low PDI can help to assure a stable catalytic performance and facilitate recycling.

AlCl₃, as a cheap and effective Lewis acid, is an important commercial catalyst for cationic polymerization. But the cationic polymerization of IB catalysed by AlCl₃ is usually an instantaneous, highly exothermic and uncontrolled reaction. It is accompanied by a series of side reactions, such as chain transfer, intramolecular shift, and chain termination. Until now, many attempts have been made to inhibit these side reactions and attain desired products, including adding nucleophilic reagents, exploiting a low-reactivity initiation system, decreasing the polymerization temperature, and tailoring solvent polarity.^11,12 Among these attempts, adding nucleophilic reagents is worth mentioning, due to its effectiveness, convenience, and diverse functions. For example, some types of ethers, such as iPr₂O, Bu₂O, and iBu₂O, were added to the initiation system to successfully synthesize highly reactive polyisobutylenes (HRPIBs), with a high content of exo-olefin and a narrow molecular weight distribution.^13–17 The ether added to the reaction system plays multiple roles, including promoting β-proton elimination of the growing chain ends to create exo-olefin end groups, decreasing or even suppressing the carbenium ion rearrangements to form the double bond isomers,^7,8,18–22 and promoting chain transfer to produce only low-molecular-weight products that are commonly available. On the other hand, some researchers focusing on medium-molecular-weight polyisobutylenes (MPIBs) or high-molecular-weight polyisobutylenes (HPIBs)^23–25 have found that the addition of ethyl benzoate (EB), a nucleophilic reagent different from ethers, is beneficial in attaining high-molecular-weight polymers, since EB could stabilize the carbocation to strongly inhibit the chain transfer,^26–29 and trap proton impurities to reduce some side reactions.^30–32 However, the monomer conversion could be seriously delayed due to the strong nucleophilicity of EB.^25,33

Herein, we envision that the preparation of PIBs with a low PDI, a high content of exo-olefin and a relatively high molecular weight may be accessed by carefully coordinating the effects of ethers like iPr₂O and EB. Correspondingly, the microflow system may be a good platform. Previous research by our group has indicated that the microflow system could provide good controllability of the cationic polymerization process with respect to both time and space.^17,23,34

In this work, with the addition of multiple nucleophilic reagents, iPr₂O and EB, we exploited a microflow reaction system to systematically investigate the polymerization of IB catalysed by AlCl₃ in a mixed solvent of CH₂Cl₂ and hexane (3 : 1, v/v). The main features include: (1) exploring the feasibility of synthesizing PIBs with a relatively low PDI, a relatively high molecular weight, and a high content of exo-olefin; (2) recognizing the quantitative criterion in the addition of iPr₂O and EB for achieving controlled polymerization; (3) investigating the evolution of conversion and molecular weight distribution to understand the kinetic characteristics of various chain reactions; (4) proposing the mechanisms of participation of iPr₂O and EB in the polymerization process; and (5) verifying the capability of the microflow reaction system in adjusting the molecular weight at a relatively high level. In overall, PIBs with a narrow molecular-weight distribution (PDI < 2.0), a relatively high molecular weight (>40 000 g mol⁻¹), and a high content of exo-olefin (w > 70%) were successfully prepared at −30 °C within 2 s, indicating that it is possible to break the trade-off between the conversion rate and propagation chain stability in cationic polymerization by introducing multiple nucleophilic reagents.

2. Experimental

2.1. Materials

Dichloromethane (CH₂Cl₂, 99.9+%, anhydrous), n-hexane (CH₃(CH₂)₄CH₃, 97.5+%, anhydrous), isopropyl ether (iPr₂O, 99.0+%), ethyl benzoate (C₉H₁₀O₂, 99.0+%), and aluminium chloride (AlCl₃, 99+%, anhydrous) were purchased from J&K Scientific (China). Isobutylene (IB, 99.9+%, anhydrous) was obtained from Korea Noble Gas (Korea). Ethanol (analytical reagent) was obtained from Sinopharm Chemical (China). Dichloromethane was dried with 5A molecular sieves overnight to decrease the content of water to about 7 ppm (determined with a coulometric Karl Fischer moisture meter (METTLER TOLEDO, Switzerland)); n-hexane was treated with a VAC solvent purification assembly (USA); isopropyl ether (iPr₂O) was distilled to remove the butylated hydroxytoluene (BHT) stabilizer, and then dried with 5A molecular sieves overnight; and ethyl benzoate (EB) was dried with 5A molecular sieves overnight. The catalytic complex solutions were prepared in a glovebox (MIKROUNA, China) shortly before experiments, in which the content of AlCl₃ was determined with a UV-Vis spectrophotometer (UV-2450, SHIMADZU).

2.2. Polymerization of IB

The preparation procedures for a catalytic complex solution include: (1) dripping iPr₂O on AlCl₃ powder; (2) pouring diluent for dissolving AlCl₃ for 2 h; (3) decanting the supernatant; and (4) adding EB or not. For entry 5, the EB is dripped on AlCl₃ powder and then diluent is added to dissolve it. The polymerization of IB was carried out in a microflow system composed of three T-shaped micromixers (M1 for the mixing of IB and diluent (the same with the solvent for preparing catalytic solution), M2 for the mixing of IB solution and catalytic complex solution, and M3 for the injection of the termination agent, ethanol), two precooling coiled tubes (C1 and C2, inner diameter 900 μm), and a microtube reactor (R1, inner diameter 900 μm), as shown in Fig. 1. The reaction time for IB polymerization could be adjusted by changing the length of R1. Four syringe pumps were used to deliver IB, diluent, catalytic solution, and termination agent. IB, transferred as vapour from the cylinder, was liquefied at −30 °C into the syringe and then pumped into the reaction system under a pressure of about 3 bars. As the content of water in each feed was determined experimentally, the content of water in the reaction system could be calculated.

Fig. 1 Schematic diagram of flow synthesis setup. M1, M2, and M3 are micromixers; C1 and C2 are curved tubes for achieving the pre-set temperature; R1 is the microtube reactor.

2.3. Characterization

The molecular weight and polydispersity index of the polymers were measured with a Waters gel permeation chromatography (GPC) system, composed of a Waters 2707 auto sampler, a 1515 Isocratic HPLC pump, a 2414 refractive index detector, and three GPC columns (the molecular weight could be detected from 500 to 4 × 10⁶) thermostated at 38 °C. Tetrahydrofuran (THF) was eluted at a flow rate of 1.0 mL min⁻¹. The instrument was calibrated with polystyrene standards. The data were processed with the Breeze 2 software from Waters. The content of exo-olefin was measured by proton nuclear magnetic resonance (¹H NMR) spectroscopy (JNM-ECA 600 MHz spectrometer, CDCl₃ as solvent). FTIR spectra were recorded in situ by using a Mettler Toledo ReactIR 15 instrument with a DiComp probe coupled to an MCT detector via AgX fibre. A spectrum was collected every 256 s by accumulating 256 scans with a wavenumber resolution of 4 cm⁻¹ over the spectral range 650 to 3000 cm⁻¹. The FTIR spectrum of the diluent (hexane : CH₂Cl₂ = 1 : 3 (v/v)) was chosen as the background for the spectra, which were recorded at 25 °C.

3. Results and discussion

3.1. Feasibility of synthesizing MPIBs with high reactivity

Using the microflow reaction system, we firstly conducted the polymerization of IB at −30 °C by directly mixing the monomer solution and a catalytic complex solution containing AlCl₃, iPr₂O, and EB, to test the feasibility of synthesizing PIBs with a relatively high molecular weight and high reactivity.

Entry 1 in Table 1 corresponds to introducing iPr₂O only. The ratio of iPr₂O to AlCl₃ is over 1. This is a prerequisite to guaranteeing a high content of exo-olefin end groups and a narrow molecular weight distribution, according to previous reports on HRPIB syntheses with AlCl₃ × iPr₂O co-initiating systems.^8,17,35 As seen, entry 1 shows a conversion as high as 89% within 2 s, and an M_n of 22 770, indicating the high activity of the AlCl₃ × iPr₂O co-initiating system and the tendency for chain elongation due to the decrease in temperature. However, the PDI is 2.87, which is very high compared with that of HRPIB (around 2.0 or usually less). For entries 2 to 4, EB is introduced to partly displace iPr₂O. When the addition of EB increases, the M_n increases persistently and the conversion decreases a little. In contrast, the variance of PDI is remarkable. The proper addition of EB can result in a PDI lower than 2.0, meeting the requirement for high reactivity. However, when using EB to replace iPr₂O in the catalytic complex solution preparation, for example entry 5, the conversion is too low to characterize. From the results in Table 1, PIBs with a high conversion and a narrow PDI could be attained by properly adding both iPr₂O and EB as nucleophilic reagents in the initiation system.

Table 1 The polymerization of IB catalysed by AlCl₃ with the addition of iPr₂O and EB^a

Entry	[iPr2O] (mmol L^−1)	[EB] (mmol L^−1)	conv.^b (%)	Mn (g mol^−1)	PDI	[PIB]^c (mmol L^−1)	Wexo (%)
a [IB] = 0.75 mol L^−1; F(IB) = 1 mL min^−1, F(diluent) = 7 mL min^−1, F(catalytic complex solution) = 8 mL min^−1; T = −30 °C; hexane/CH2Cl2 = 1 : 3 (v/v); t = 2 s; [H2O] = 0.82 mmol L^−1; [AlCl3] = 3.16 mmol L^−1.b Gravimetric conversion.c [PIB] = [IB] × 56 × conv./Mn.d [AlCl3] = 2.10 mmol L^−1.
1	3.54	0	89	22 770	2.87	1.64	30.5
2	2.97	0.25	76	26 660	1.95	1.19	60.4
3	2.50	0.49	75	36 560	1.87	0.86	65.9
4	1.79	0.84	72	47 540	1.82	0.63	76.0
5^d	0	1.10	2	—	—	—	—

---

Furthermore, we determined the ¹H NMR spectra of these products to determine whether the content of exo-olefin could meet the requirement for high reactivity. Fig. 2 shows a typical spectrum, corresponding to entry 4 of Table 1. As seen, the main resonance signals are located at δ = 1.1 (z), 1.41 (y), 0.99 (x), 4.85 (a1), 4.64 (a2), 5.17 (c1), 5.37 (c2), 2.83 (e), and 1.96 (h). Therein, the single peaks, z and y, are assigned to the –CH₃ and –CH₂ protons of the structural units along the main chain of PIB, respectively; the single peak, x is assigned to the protons in the head groups of –C(CH₃)₃ in PIB chains; the single peaks, a1 and a2, in the expansion of the olefin region, are assigned to the protons of exo-olefin end groups, –CH₂–C(CH₃)=CH₂; the quartet peaks, c1 and c2, are attributed to the –C(CH₃)=CH(CH₃) end groups in PIB chains; the multiple peak e is assigned to the protons in the –CH(CH₃)₂ end groups in the PIB chains; and the single peak h is as attributed to the protons in –CH₂C(CH₃)₂Cl terminal groups. From the calculation, structure A is predominant (w = 76%) in the PIB chains. The end-group distributions for the other entries are provided in Table S1.† In evidence, a fast synthesis of PIBs with medium molecular weight and high reactivity is available with the proper addition of EB in the co-initiation system of AlCl₃ × iPr₂O. A possible explanation is that the addition of EB barely affects the high reactivity of the co-initiation system of AlCl₃ × iPr₂O, but simultaneously inhibits the charge shift of carbenium as well as the β-proton elimination to achieve a controlled polymerization process.

Fig. 2 ¹H NMR spectrum of products from entry 4 of Table 1.

3.2. Effects of EB on the polymerization of IB, co-initiated by AlCl₃ × iPr₂O

To quantitatively understand the effects of EB on the AlCl₃ × iPr₂O co-initiation system, we carried out a series of polymerization experiments with constant contents of AlCl₃ and iPr₂O in the catalytic complex solution and varied contents of EB. Table 2 shows the results. For entry 6, without adding EB, the PDI is 3.66, reflecting an uncontrollable cationic polymerization co-initiated by AlCl₃ due to the deficiency of iPr₂O, compared with AlCl₃. When adding EB gradually, a general tendency is that M_n increases and the PDI decreases. A transition takes place at entry 13, corresponding simultaneously to the maximum M_n and the minimum PDI. Afterwards, increasing EB further results in M_n decreasing and PDI increasing. For entry 13, PDI is only 1.62. This implies that the polymerization co-initiated by AlCl₃ may be totally suppressed by the quantitative interaction between AlCl₃ and EB.

Table 2 Effects of EB on IB polymerization co-initiated by AlCl₃ × iPr₂O^a

Entry	[EB] (mmol L^−1)	conv.^b (%)	Mn (g mol^−1)	PDI	[PIB]^c (mmol L^−1)	Wexo (%)
a [IB] = 0.75 mol L^−1; F(IB) = 1 mL min^−1, F(diluent) = 7 mL min^−1, F(catalytic complex solution) = 8 mL min^−1; T = −30 °C; hexane/CH2Cl2 = 1 : 3 (v/v); t = 12 s. [H2O] = 0.85 mmol L^−1; [AlCl3] = 3.84 mmol L^−1; [iPr2O] = 1.79 mmol L^−1.b Gravimetric conversion.c [PIB] = [IB] × 56 × conv./Mn.
6	0	86	19 610	3.66	1.84	28.5
7	0.19	85	17 610	3.45	2.02	29.2
8	0.40	80	20 130	3.27	1.66	30.5
9	0.51	74	18 640	3.72	1.66	31.4
10	0.59	76	25 110	2.88	1.27	30.9
11	0.70	82	27 520	3.03	1.25	32.2
12	0.80	81	34 450	2.39	0.98	46.7
13	1.01	58	46 060	1.62	0.53	70.0
14	1.21	21	14 360	2.20	0.61	53.8

---

Our previous work¹⁷ has indicated that AlCl₃ and iPr₂O generate a complex at a molar ratio of 1 : 1. As EB is added into the AlCl₃ × iPr₂O co-initiation system, it may also interact with the AlCl₃ not participating in generating iPr₂O·AlCl₃ at a specific stoichiometric ratio of 1 : n (EB : AlCl₃). Thus, the transition from uncontrolled polymerization to controlled polymerization will occur as [AlCl₃] becomes equal to [iPr₂O] plus n[EB]. Fig. 3 shows the variances of M_n and PDI with ([iPr₂O] + 2[EB])/[AlCl₃]. The declining tendency of PDI changes at ([iPr₂O] + 2[EB])/[AlCl₃] = 1, indicating that n may be around 2. In other words, [iPr₂O] + 2[EB] being greater than or equal to [AlCl₃] is a criterion to achieve controlled polymerization in the co-initiation system of AlCl₃ × iPr₂O with the addition of EB. Entries 2 to 4 in Table 1 also meet the criterion well, since they both have ([iPr₂O] + 2[EB])/[AlCl₃] around 1.1.

Fig. 3 The dependence of the molecular weight distribution on the composition of the catalytic complex solution.

It is also worth noticing that entry 14 shows a low conversion and a low M_n compared with entries 2 to 4, although ([iPr₂O] + 2[EB])/[AlCl₃] is also around 1.1. Their differences in terms of polymerization conditions lie in that entry 14 exploits the low molar ratio of iPr₂O to AlCl₃ and the long reaction time. A low ratio of iPr₂O to AlCl₃ could reduce the polymerization rate to decrease the conversion, while a long reaction time leads to a low M_n.

3.3. Evolution of the polymerization of IB co-initiated by AlCl₃ × iPr₂O with EB addition

To understand the polymerization procedure, two groups of experiments were carried out to monitor the variance of the monomer conversion and the molecular weight of the product with time. All the experimental conditions in each group were identical, with the exception of the reaction time. Between the two groups, the contents of AlCl₃, iPr₂O, and EB are different, but ([iPr₂O] + 2[EB])/[AlCl₃] always remains at around 1.1. The results are listed in Table 3. For the first group (entries 15 to 17), we found that 72% of IB was converted to MPIBs with a PDI less than 2.0 after 2 s. Consequently, as the residence time increased, the conversion and the PDI showed minimal change, but both the M_n and the content of exo-olefin clearly decreased. At 2 s, these were 42 470 g mol⁻¹ and 73.0%, respectively; at 12 s, they were 29 380 g mol⁻¹ and 60.3%, respectively. The only way to explain these phenomena is that chain scission occurs as the residence time is prolonged, especially for longer chains. For the second group (entries 18 to 20), the time profiles of the conversion, M_n, PDI, and the content of exo-olefin show similar tendencies; for example, 70% conversion could be reached after 0.5 s. We can confirm that the conversion from IB to MPIBs with high reactivity could proceed sufficiently on a sub-second scale, and the chain scission becomes marked after a few seconds. Thus, decreasing the time for achieving sufficient mixing to far less than one second and controlling the residence time at the second scale or less are prerequisites for obtaining PIBs with relatively high M_n and reactivity. The microflow system is an appropriate platform to meet these requirements. Based on these judgements, PIBs with a higher M_n and content of exo-olefin are available by using a reaction time of less than 0.5 s, as long the conversion is acceptable; the M_n could be facilely adjusted by changing the reaction time, as long the reactivity is acceptable.

Table 3 Evolution of IB polymerization co-initiated by AlCl₃ × iPr₂O with the addition of EB at −30 °C^a

Entry	t (s)	conv.^d (%)	Mn (g mol^−1)	PDI	[PIB]^e (mmol L^−1)	Wexo (%)
a [IB] = 0.75 mol L^−1; F(IB) = 1 mL min^−1, F(diluent) = 7 mL min^−1, F(catalytic complex solution) = 8 mL min^−1; hexane/CH2Cl2 = 1 : 3 (v/v).b [H2O] = 0.93 mmol L^−1; [AlCl3] = 3.47 mmol L^−1; [iPr2O] = 1.79 mmol L^−1; [EB] = 1.01 mmol L^−1.c [H2O] = 0.96 mmol L^−1; [AlCl3] = 4.59 mmol L^−1; [iPr2O] = 3.57 mmol L^−1; [EB] = 0.80 mmol L^−1.d Gravimetric conversion.e [PIB] = [IB] × 56 × conv./Mn.
15^b	2	72	42 470	1.88	0.71	68.2
16^b	6	73	37 180	1.84	0.82	70.2
17^b	12	73	29 380	1.79	1.04	60.3
18^c	0.5	71	41 400	1.84	0.72	72.6
19^c	6	77	31 820	1.90	1.01	68.5
20^c	12	86	27 870	1.97	1.29	59.2

---

3.4. Roles of iPr₂O and EB in the process control for the polymerization of IB

As mentioned above, the nucleophilic reagents, iPr₂O and EB, have multiple effects on the polymerization, endowing the PIBs with a relatively high molecular weight as well as high reactivity. Fig. 4 shows a schematic illustration of the proposed evolution routes of species in the catalytic complex solution. According to the procedures for the catalytic complex solution preparation, some of the AlCl₃ initially combines with iPr₂O to generate the complex iPr₂O·AlCl₃. Then, both AlCl₃ and iPr₂O·AlCl₃ have the opportunity and enough time to be combined with adventitious H₂O from the diluent during the dissolution step. This generates two types of initiating species, H⁺AlCl₃OH⁻ and H⁺iPr₂OAlCl₃OH⁻, simultaneously. Thus, the added EB can be combined with residual free AlCl₃ to generate EB(AlCl₃)_n, but the corresponding initiating species cannot be generated, since most of the H₂O has participated in forming H⁺iPr₂OAlCl₃OH⁻ and H⁺AlCl₃OH⁻. Besides, during the polymerization experiments, the adventitious H₂O from the diluent feed may have little effect on the distributions of these existing species due to the short contact time. H⁺iPr₂OAlCl₃(OH)⁻ is expected to be formed uniquely because the insertion of iPr₂O can enhance the stability of the carbenium ion to inhibit intramolecular charge shifts.

Fig. 4 Evolution routes of species in catalytic complex solution.

Although the addition of EB does not change the contents of H⁺iPr₂OAlCl₃OH⁻ and H⁺AlCl₃OH⁻ in the catalytic complex solution, the other species around these initiating species may change to have various effects on their initiation activities. The block diagrams (a)–(d) in Fig. 4 show the variation in the main species resulting from adding EB. In detail, for the EB-free case (Fig. 4a), free AlCl₃ and iPr₂OAlCl₃ exist without the initiating species; in Fig. 4b, the addition of EB can decrease the free AlCl₃ and generate EB(AlCl₃)_n; in Fig. 4c, the addition of EB is allows the conversion of all the free AlCl₃ to EB(AlCl₃)_n; in Fig. 4d, the EB is added in excess and free EB exists. If only H⁺AlCl₃(OH)⁻ and H⁺iPr₂OAlCl₃(OH)⁻ exist in the catalytic complex solution, the former has higher initiation activity due to the strong acidity of AlCl₃ and the absence of iPr₂O as electron donor. For the H⁺AlCl₃(OH), the free AlCl₃ can stimulate the initiation activity, and EB(AlCl₃)_n is an inhibitor for the initiation activity. For H⁺iPr₂OAlCl₃OH⁻, EB(AlCl₃)_n has little effect on the initiation activity, but could delay the chain transfer. Thus, when the content of AlCl₃ is considerable, H⁺AlCl₃(OH)⁻ plays a decisive role in initiating the polymerization compared with H⁺iPr₂OAlCl₃OH⁻, resulting in uncontrolled polymerization; when the content of EB(AlCl₃)_n is considerable, with the content of free AlCl₃ being minimal, the initiation activity of H⁺iPr₂OAlCl₃OH⁻ will prevail over that of H⁺AlCl₃OH⁻, resulting in controlled polymerization. Meanwhile, the free EB could weaken the stabilization effect of EB(AlCl₃)_n on propagation chains, resulting in a decrease in M_n.

In our previous work,¹⁷ we found that free iPr₂O could promote chain transfer. Thus, a general and rational inference is that for the nucleophilic reagents, the free one can promote the chain transfer and the one coordinated with AlCl₃ can delay the chain transfer, while both these effects for EB are much stronger than those for iPr₂O. Additionally, H⁺iPr₂OAlCl₃OH⁻, as the expected and effective initiating species in this work, is generated by H⁺iPr₂OAlCl₃OH⁻, combining water under the competitive effect of AlCl₃, so the content of H⁺iPr₂OAlCl₃OH⁻ is also dependent on the original ratio of iPr₂O to AlCl₃. By exploiting all these parameters, with the exception of the iPr₂O content, a high ratio of iPr₂O to AlCl₃ corresponds to a high content of H⁺iPr₂OAlCl₃OH⁻, as well as a high initiation efficiency and a low molecular weight, as shown in Table 1.

The above mechanism, speculated from the results of polymerization experiments, could also be confirmed by the ATR-FTIR spectra. Fig. 5a shows the spectra of iPr₂O or AlCl₃ in diluent. The signal bands at 1012, 1109, 1128, and 1169 cm⁻¹ are assigned to C–O–C stretching in iPr₂O, and no signal appears in the range from 800 to 1200 cm⁻¹ for AlCl₃. When adding iPr₂O into AlCl₃, all the FTIR bands assigned to the C–O–C stretch disappear, and two new peaks are generated at 904 and 1099 cm⁻¹. This indicates the formation of a complex between iPr₂O and AlCl₃, which has been recognized as an effective catalyst for IB polymerization. Fig. 5b shows the spectra of EB in diluent with and without the addition of AlCl₃. The signal bands at 1029, 1074, 1111, 1178, 1266, 1280, 1316, 1370, and 1719 cm⁻¹ belong to the characteristic peaks of EB. When EB makes contact with solid AlCl₃ before adding diluent, most of these peaks disappear and many new peaks emerge, such as at 994, 1191, 1342, 1389, 1424, 1500, 1577, and 1611 cm⁻¹. This indicates that the interaction between AlCl₃ and EB is complex and strong, which may seriously inhibit the activity of AlCl₃ as catalyst and result in the low conversion shown in entry 5. However, when EB makes contact with the AlCl₃ solution, we can find that almost all of the main peaks for EB in diluent are preserved, and only a small new peak emerges at 1342 cm⁻¹. In evidence, the interaction between AlCl₃ and EB becomes much weaker in this case due to the prior existence of the diluent. The species EB(AlCl₃)_n corresponds to this weak interaction. Fig. 5c shows the effects of the addition of EB on the spectra of the complex of iPr₂O and AlCl₃ in the diluent. The characteristic bands of the complex (904 and 1099 cm⁻¹) seem to be always present with the addition of EB. This implies that the complex could still play a role as an effective catalyst, and EB interacts weakly with AlCl₃ weakly to inhibit uncontrolled polymerization, as well as chain transfer.

Fig. 5 ATR-FTIR spectra for understanding the species in the catalytic complex solution. (a) iPr₂O or AlCl₃ in diluent; (b) EB in diluent with and without AlCl₃ addition; (c) iPr₂O and AlCl₃ in diluent with various EB additions. The diluent is hexane/CH₂Cl₂ = 1 : 3 (v/v). The note ‘+’ represents adding the reagent on the right into the reagent on the left during sample preparation. After adding diluent, the solution stood for 2 h and the supernatant was separated for further preparation or characterization.

3.5. Tailoring molecular weight by flow capacity

As is well known, the mixing of species participating in a fast reaction may considerably affect the course of the reaction, and the residence time is an important parameter for controlling a reaction with sequential side reactions. It has been confirmed that the polymerization of IB co-initiated by AlCl₃ × iPr₂O with EB can be fast, and the products experience distinct chain scission after seconds of residence time. So, a platform with precise controllability of both the mixing performance and the residence time is favourable for obtaining PIB products with relatively high molecular weight and low PDI, and the microflow system is a good choice. For a certain microflow system, the change in the flow capacity can change both the residence time and the mixing performance, providing a convenient and effective way to tailor the products.

Herein, fixing the volume of the delay tube on 0.27 mL, we prepared and characterized PIBs at various flow capacities. From the results shown in Table 4, we could see that a higher flow capacity, corresponding to better mixing and a shorter residence time, is beneficial in obtaining PIBs with a higher M_n, a lower PDI, and a higher content of exo-olefin, with an acceptable decrease in conversion. The reasons are that better mixing could guarantee a sufficient monomer supply and a uniform distribution of nucleophilic reagents around the living polymer chains to promote chain propagation and inhibit chain transfer and termination, and a short residence time can reduce the possibility of chain scission. Overall, introducing iPr₂O and EB into the polymerization of IB catalysed by AlCl₃ endows the feasibility to prepare PIBs with a relatively high molecular weight, a low PDI and high reactivity, and exploiting the microflow system to carry out this process allows facile tailoring of the molecular weight of PIBs by flow capacity in an extended scope.

Table 4 Effects of mixing on IB polymerization co-initiated by AlCl₃ × iPr₂O with the addition of EB at −30 °C^a

Entry	Ft^b (mL min^−1)	conv.^c (%)	Mn (g mol^−1)	PDI	[PIB]^d (mmol L^−1)	Wexo (%)
a Hexane/CH2Cl2 = 1 : 3 (v/v); [IB] = 0.75 mmol L^−1; [EB] = 0.86 mol L^−1; [H2O] = 0.79 mmol L^−1; [AlCl3] = 3.27 mmol L^−1; [iPr2O] = 1.79 mmol L^−1.b F(IB) : F(diluent) : F(catalytic complex solution) = 1 : 7 : 8.c Gravimetric conversion.d [PIB] = [IB] × 56 × conv./Mn.
21	8	70	43 120	2.00	0.68	67.0
22	16	63	50 030	1.91	0.53	73.0
23	32	60	59 050	1.73	0.43	79.0

---

4. Conclusions

PIBs with different molecular weights have various applications, and the characteristics of a low PDI and a high content of exo-olefin are quite valuable to make PIBs easy to functionalize. In this work, we firstly attempted to use two kinds of nucleophilic reagents to co-ordinately control the polymerization process. Through exploiting the microflow reactor and properly introducing iPr₂O and EB, PIBs with a high conversion (>70%), a relatively high molecular weight (>40 000 g mol⁻¹), a low PDI (<2.0), and a high content of exo-olefin (>70%) could be attained at a conventional temperature (−30 °C) within 2 s. Most monomers could participate in the generation of PIBs within 0.5 s. Chain scission, aggravated by the higher molecular weight, mainly takes place in the following seconds. Mechanistic studies suggest that two initiating species, H⁺AlCl₃OH⁻ and H⁺iPr₂OAlCl₃OH⁻, exist simultaneously in the polymerization of IB catalysed by AlCl₃ with the addition of iPr₂O and EB. When the content of EB(AlCl₃)_n is considerable, and the content of free AlCl₃ is minimal, the initiation activity of H⁺iPr₂OAlCl₃OH⁻ will prevail over that of H⁺AlCl₃OH⁻ to achieve controlled polymerization. In addition, increasing the flow capacity in a certain microflow system could intensify the mixing and reduce the residence time simultaneously, providing the potential to further increase the molecular weight of the PIBs and facilely tailor it in an extended scope. Above all, this work shows a new horizon for controlled cationic polymerizations and functional PIB products. However, the functions of various nucleophilic reagents, dependent on their amounts and method of addition, still need further investigation to reveal the essential principles and guidelines for more extensive applications.

Acknowledgements

The authors are grateful for the support of the National Natural Science Foundation of China (21176136, 21422603) and the National Science and Technology Support Program of China (2011BAC06B01) in this work.

References

1. K. Kunal, M. Paluch, M. Roland, J. E. Puskas and A. P. Sokolov, J. Polym. Sci., Part B: Polym. Phys., 2008, 46, 1390–1399.
2. J. E. Puskas and G. Kaszas, in Encyclopedia of Polymer Science and Technology, Wiley-InterScience, New York, 2003, vol. 5, pp. 382–418.
3. J. P. Kennedy and B. Ivan, Designed Polymers by Carbocationic Macromolecular Engineering: Theory and Practice, Hanser, Munich, 1991, pp. 173–177.
4. J. P. Kennedy and E. Marechal, Carbocationic Polymerization, 1982, pp. 475–488.
5. S. F. Rach and F. E. Kühn, Sustainability, 2009, 1, 35–42.
6. I. V. Vasilenko, D. I Shiman and S. V. Kostjuk, Polym. Chem., 2014, 5, 3855–3866.
7. R. Kumar, P. Dimitrov, K. J. Bartelson, J. Emert and R. Faust, Macromolecules, 2012, 45, 8598–8603.
8. Q. Liu, Y. X. Wu, P. F. Yan, Y. Zhang and R. W. Xu, Macromolecules, 2011, 44, 1866–1875.
9. P. Dimitrov, J. Emert and R. Faust, Macromolecules, 2012, 45, 3318–3325.
10. D. E. Bergbreiter, H. L. Su, H. Koizumi and J. Tian, J. Organomet. Chem., 2011, 696, 1272–1279.
11. R. F. Storey, B. J. Chisholm and L. B. Brister, Macromolecules, 1995, 28, 4055–4061.
12. I. V. Vasilenko, D. I. Shiman and S. V. Kostjuk, J. Polym. Sci., Part A: Polym. Chem., 2012, 50, 750–758.
13. I. V. Vasilenko, A. N. Frolov and S. V. Kostjuk, Macromolecules, 2010, 43, 5503–5507.
14. Q. Liu, Y. X. Wu, Y. Zhang, P. F. Yan and R. W. Xu, Polymer, 2010, 51, 5960–5969.
15. S. V. Kostjuk, H. Y. Yeong and B. Voit, J. Polym. Sci., Part A: Polym. Chem., 2013, 51, 471–486.
16. S. V. Kostjuk, RSC Adv., 2015, 5, 13125–13144.
17. S. Zhu, Y. C. Lu, K. Wang and G. S. Luo, RSC Adv., 2016, 6, 9827–9834.
18. D. I. Shiman, I. V. Vasilenko and S. V. Kostjuk, Polymer, 2016, 99, 633–641.
19. S. V. Kostjuk, I. V. Vasilenko, D. I. Shiman, A. N. Frolov and L. V. Gaponik, Macromol. Symp., 2015, 349, 94–103.
20. D. I. Shiman, I. V. Vasilenko and S. V. Kostjuk, Polymer, 2013, 54, 2235–2242.
21. P. Dimitrov, J. Emert and R. Faust, Macromolecules, 2012, 45, 3318–3325.
22. K. J. Bartelson, P. De, R. Kumar, J. Emert and R. Faust, Polymer, 2013, 54, 4858–4863.
23. S. Zhu, Y. C. Lu, K. Wang and G. S. Luo, Eur. Polym. J., 2016, 80, 219–226.
24. V. Seidl, C. N. Duecker and A. Wiedemann, et al., US Pat., 7 662 900, 2010.
25. Y. Li, Y. X. Wu, L. H. Liang and G. Y. Wu, Chin. J. Polym. Sci., 2010, 28, 55–62.
26. G. Kaszas, J. E. Puskas, C. C. Chen and J. P. Kennedy, Macromolecules, 1990, 23, 3909–3915.
27. J. P. Kennedy, I. Majoros and A. Nagy, Adv. Polym. Sci., 1994, 112, 1.
28. K. Matyjaszewski and P. Sigwalt, Polym. Int., 1994, 35, 1–26.
29. M. Szwarc, Makromol. Chem., Rapid Commun., 1992, 13, 141–145.
30. M. Fodor, Y. C. Bea and R. Faust, Macromolecules, 1998, 31, 4439–4446.
31. S. Hadjikyriacou, M. Acar and R. Faust, Macromolecules, 2004, 37, 7543–7547.
32. R. F. Storey, C. L. Curry and L. K. Hendry, Macromolecules, 2001, 34, 5416–5432.
33. Y. Li, Y. X. Wu, X. Xu, L. H. Liang and G. Y. Wu, J. Polym. Sci., Part A: Polym. Chem., 2007, 45, 3053–3061.
34. Y. C. Lu, S. Zhu, K. Wang and G. S. Luo, Ind. Eng. Chem. Res., 2016, 55, 1215–1220.
35. Y. X. Wu, Y. X. Tan and G. Y. Wu, Macromolecules, 2002, 35, 3801–3805.

---

Footnote

† Electronic supplementary information (ESI) available: Table with the end group distribution for all the experimental entries in the paper. See DOI: 10.1039/c6ra21983g

---