Application of online infrared spectroscopy to study the kinetics of precipitation polymerization of acrylic acid in supercritical carbon dioxide

Abstract

The kinetics of precipitation polymerization of acrylic acid (AA) in supercritical carbon dioxide (scCO₂) has been studied using in situ reaction monitoring of the AA/scCO₂ phase by Fourier transform infrared (FTIR) spectroscopy. A comparative in situ FTIR study has been performed in toluene. In both solvents, the polymerization kinetics display an induction period in which very little polymerization takes place followed by rapid onset of polymerization to reach a total conversion in about 70 min after the end of the induction period. The rate of polymerization displays an apparent 1.5-order dependence on the monomer concentration and a 0.5-order on the initiator concentration, which is consistent with a mechanism in which the main locus of polymerization is within the particles. The use of scCO₂ allows the production of higher molecular weight polyacrylic acid powder than in toluene without any solvent residues.

---

Introduction

Poly(acrylic acid) (PAA) is used in many applications, including as a superabsorbent polymer for hygiene products and agriculture, as a scale inhibitor for water treatment and as a dispersant for paints and the paper industry. It is produced industrially on large scale by free-radical polymerization.¹ Even though PAA may appear to be an ‘old’ and well-understood polymer, the free-radical polymerization of acrylic acid (AA) remains the subject of intense research activity with significant recent contributions on kinetics^2,3 and modeling⁴ of AA polymerization, reversible deactivation radical polymerization,⁵ process intensification,⁶ characterization⁷ and nanocomposites.⁸

The process for polymerizing AA can be either homogeneous (typically in aqueous solution) or heterogeneous. The heterogeneous polymerization of AA or its salts in organic media is a convenient way to prepare PAA nano- or microspheres at relatively high solid content with much better heat transfer and lower viscosity compared to solution polymerization. The reaction conditions (i.e. nature of initiator and continuous phase, degree of neutralization of AA, stabilizer) can be varied in order to meet the requirements for polymerizations in dispersion,⁹ inverse emulsion¹⁰ or inverse suspension.¹¹

Although widely used, these processes sometimes require tedious desorption of the surfactant from the particles after polymerization and removal of traces of volatile organic compounds (VOCs) like residual monomer and solvent. These drawbacks can be circumvented with precipitation polymerization,¹² which is a stabilizer-free process in which both initiator and monomer are soluble in the reaction medium while the polymer is insoluble and precipitates as it is formed. After filtration and washing of the polymer, this process usually allows the preparation of polymer powders of high purity for sensitive applications, e.g. cosmetics and food additives. The kinetics and mechanism of precipitation polymerization of AA in toluene have been studied in detail by several groups.^13–16 The polymerization is characterized by an induction period in which very little polymerization takes place, followed by an accelerating reaction leading to complete polymerization in a relatively short time compared to a solution polymerization at comparable overall monomer concentration.¹⁷ Several kinetic models have been developed to explain these features, differing on whether polymerization is assumed to occur exclusively in the dispersed phase^15,16 or in both the dispersed and continuous phases.^13,14

More recently, supercritical carbon dioxide (scCO₂) has shown great potential as a green solvent as its use can greatly reduce the production of toxic and environmentally damaging waste. Indeed, carbon dioxide is a readily available, inexpensive, nontoxic and nonflammable natural product with easily accessible critical coordinates (T_c = 31.1 °C and P_c = 7.4 MPa). In this context, heterogeneous controlled/living radical polymerizations in scCO₂ have been developed over the past decade.^18–20 Supercritical CO₂ is a good solvent for AA at relatively mild temperature and pressure, and is inert to radical reactions such as chain transfer, making it a solvent of choice for the precipitation polymerization of AA.^21–23 Its lack of toxicity and ease of removal by depressurizing the reactor are strong advantages over traditional solvents. During the 2000s, several groups studied the effect of various parameters such as the presence of a co-solvent²⁴ or a crosslinker,²² or the use of a continuous stirred tank reactor (CSTR)²¹ for the precipitation polymerization of AA in scCO₂. A kinetic and model study showed that polymerization in the CSTR occurred primarily in or on the polymer particles, with negligible contribution from polymerization in the continuous phase.²³

While the precipitation polymerization of AA in supercritical CO₂ has previously been investigated, a detailed study of its kinetics is still lacking. This prompted us to monitor the kinetics of the precipitation polymerization of AA in scCO₂ using in situ reaction monitoring of the AA/scCO₂ phase by Fourier transform infrared spectroscopy (in situ FTIR).

The in situ FTIR approach allows the monomer concentration and the overall monomer conversion to be determined at any point during the polymerization. It is particularly suited to reactions in scCO₂ because it is otherwise difficult to sample the reaction mixture without changing the reaction conditions (pressure/concentration). Although this method has been applied to the online kinetic monitoring of homogeneous polymerizations in scCO₂,^25–29 it has never been reported to the best of our knowledge for a precipitation polymerization.

Thus, the aim of this study is to investigate the impact of initiator and monomer concentrations on the kinetics of polymerization and the molar mass, dispersity, and particle morphology of the resulting polymer. The in situ FTIR technique was validated by an initial study in toluene, then applied to the precipitation polymerization of acrylic acid in scCO₂.

Materials and methods

Materials

Acrylic acid (97%) containing 200 ppm of inhibitor (monomethyl ether of hydroquinone (MEHQ)), was provided by SEPPIC and used without prior purification. 2,2-Azobisisobutyronitrile (AIBN) was obtained from Sigma-Aldrich and recrystallized twice from methanol before use. Carbon dioxide with a purity of 99.95% was purchased from Air Liquide and used as received. Toluene (CHROMASOLV Plus, >99.9%) was obtained from Sigma Aldrich and used as received.

Infrared

Infrared absorption measurements were performed on a ThermoOptek Nicolet 6700 FTIR spectrometer equipped with a globar as the infrared source, a KBr/Ge beamsplitter and a DTGS (deuterated triglycine sulphate) detector in order to investigate the spectral range of 400–7500 cm⁻¹. Single beam spectra recorded with 2 cm⁻¹ resolution were obtained after the Fourier transformation of 32 accumulated interferograms.

Kinetic monitoring in toluene solution

The infrared kinetic monitoring of the polymerization of AA in toluene was performed using the ThermoOptek Nicolet 6700 FTIR spectrometer described above coupled with a variable temperature ATR (attenuated total reflectance) ‘Golden Gate’ accessory from Specac Ltd. This device has been modified by the addition of a stainless steel cell above the diamond crystal that allows measurements on liquids in a closed vessel (volume of 3.2 mL) that can be introduced through the upper port (see ESI† Fig. S1). The cell and the ‘Golden Gate’ accessory are sealed with a gasket in graphite. A magnetically driven stirrer in this cell constantly homogenizes the medium during the experiment. The device is equipped with cartridge heaters and a temperature controller to perform measurements from ambient temperature up to 300 °C. In a typical experiment, the monomer and AIBN were degassed under Ar for 10 min, then placed directly onto the ATR crystal. The cell was closed and an additional 5 min degassing was carried out by controlled injection of Ar via a capillary that was connected to the inlet port. The measuring cell was then heated to the desired temperature. IR spectra were recorded every 3 min in the spectral range 400–4000 cm⁻¹ until total disappearance of the band from AA at 1640 cm⁻¹ was observed.

Kinetic monitoring in scCO₂

Infrared kinetic monitoring of the polymerization of AA in scCO₂ was performed using an in-house built stainless steel cell³⁰ equipped with two cylindrical silicon windows and a variable path length which enabled the IR beam to traverse the supercritical phase and provide a quantitative analysis of the reaction medium in the frequency range of 1000–7500 cm⁻¹ (Fig. 1). The cell was configured with a path length of 1 mm and a volume of 4.7 mL. Flat Kapton® rings of 100 μm thickness were used to ensure the seal between the plug and the windows and between the plug and the cell body. The lower part of the reactor was fitted with a small tank containing a magnetically driven stirrer in order to constantly homogenize the medium during the experiment. The cell was heated using cartridge heaters disposed in the walls of the cell. Two thermocouples were used, one close to a cartridge heater for temperature regulation, and the other close to the sample to measure the sample temperature with a standard uncertainty u(T) = 1 K. The cell was connected via a stainless steel capillary tube to a manual pump (TOP Industrie) which allows the pressure to be raised to 50 MPa with a standard uncertainty u(P) = 0.1 MPa. This set-up allows online monitoring of the polymerization from room temperature up to 200 °C and up to 35 MPa of pressure.

Fig. 1 HP cell for the online FTIR monitoring of the AA polymerization in scCO₂.

A typical polymerization is carried out according to the following experimental protocol: in a 5 mL flask, the AA monomer is degassed under Ar for 15 minutes. During this time the initiator is loaded into the lower part of the reactor. Then, a given quantity of monomer is introduced by syringe under a gentle stream of CO₂. An additional degassing with Ar is then carried out in the reactor under stirring for 5 minutes. The pressure is then increased to 6 MPa by introduction of CO₂. The temperature and pressure are then adjusted to the desired levels. A first spectrum was recorded immediately after the desired pressure was reached. Further spectra were recorded every 3 min until total disappearance of the band from AA at 1950 cm⁻¹. The temperature was then lowered to room temperature and the pressure decreased gradually to atmospheric pressure. The obtained polymer is a white powder that could be analyzed without further purification.

Size exclusion chromatography

Size exclusion chromatography (SEC) was carried out on a Agilent 1100 HPLC system including a vacuum degasser and an isocratic pump monitored by an eclipse 2 system (Wyatt technonology), a guard column Shodex SB807-G and two columns Shodex OHpak (SB-807 HQ (8 mm × 300 mm, 35 μm)) in series coupled with a Shodex RI-101 differential refractometer thermostated at 25 °C and a multi-angle laser light scattering detector (Dawn HELEOS® II, 18 angles, Wyatt technology). An aqueous solution of 0.1 M NaCl and 0.5 M of phosphate buffer (0.25 M NaH₂PO₄ + 0.25 M Na₂HPO₄) containing 100 ppm of NaN₃ was used as the eluent at a flow rate of 1 mL min⁻¹. Prior to injection, samples were diluted to 0.5 mg mL⁻¹, stirred for 24 h and filtered through 0.45 μm nylon filters. The incremental refractive index (dn/dc) of a high M_w PAA sample (M_w ∼ 10⁶ g mol⁻¹) was measured in the eluent using a PSS DnDc-2010 refractometer thermostated at 25 °C, at a wavelength of 620 nm. dn/dc (PAA) = 0.185 mL g⁻¹.

Scanning electron microscopy

The shape and surface morphology of the PAA particles was observed by scanning electron microscopy (SEM) with a MEB Quanta 250 FEG FEI microscope. The electron beam was produced with Schottky type field emission gun. The sample was analyzed under vacuum at 10⁻⁴ Pa. The electrons are emitted from a thermionic cathode. The information on PAA powder surfaces was obtained from secondary electron analysis. A SDD/Bruker 30 mm² detector was used to observe the morphology of PAA particles. Prior to scanning, the samples were coated with a 3 nm layer of platinum deposited with a vacuum metallizer.

Results and discussion

Online monitoring in toluene

We first performed online ATR-FTIR monitoring of the precipitation polymerization of AA in toluene at 62 °C with initial monomer concentrations [M]₀ of 0.58–1.71 M and a constant concentration of AIBN ([AIBN]₀) of 6.38 mM. The conditions of polymerization are compiled in Table S1 of the ESI.†

The concentration of acrylic acid at time t, [AA]_t, was calculated by following the evolution with time of the signal at 1639 cm⁻¹ that is associated to the ν(C=C) stretching mode of AA. According to the Beer–Lambert law, A₀ = ε₁₆₃₉·L·[AA]₀ and A_t = ε₁₆₃₉·L·[AA]_t where A₀ and A_t are the absorbances of the peak at 1639 cm⁻¹ at time 0 and time t, respectively, ε₁₆₃₉, the molar extinction coefficient at 1639 cm⁻¹ and L, the optical path length. Combining these two equations, the concentration of AA during the reaction is given by eqn (1):

[AA]_t = (A_t/A₀) × [AA]₀ (1)

The conversion, X, was calculated from eqn (2):

Fig. 2 shows the time dependence of the conversion of AA.

Fig. 2 Conversion profiles for precipitation polymerization of acrylic acid in toluene showing effect of changing initial [AA] ([AIBN]₀ = 6.38 mM). Dashed lines represent best fit to Barrett and Thomas' model.¹⁷

Whatever the initial concentration of AA, we observe an induction period where there is very little conversion of AA followed by rapid onset of polymerization. Total conversion is reached in about 70 min after the end of the induction period.

These characteristics are typical of precipitation and dispersion polymerizations, in which, although the monomer and initiator are soluble in the continuous phase, most polymerization takes place in the dispersed phase.^14–17 The induction period corresponds to an initial period of slow polymerization in the continuous phase and particle nucleation. When a critical volume of particles has been generated, they become swollen with monomer and the polymerization accelerates.

In the later stages of the reaction, the rate of polymerization decreases as the monomer is depleted. Longer induction periods are observed at low initial [AA] because under these conditions it takes longer to produce the critical volume of particles required for the onset of autoacceleration. The maximum rate of change in conversion (dX/dt)_max shows a slight dependence on [AA]₀. In previous studies of precipitation polymerization, the observed dependence of dX/dt on [AA]₀ has ranged from zero-order³¹ to 0.7-order.¹⁴ Similar dependence on [AA]₀ has been observed in solution polymerizations, which has been ascribed to monomer-enhanced decomposition of the initiator.³ Additionally the rate constant of propagation of AA in water varies with [AA]₀, and is significantly lower at 50% AA than at 10% AA.²

Online monitoring in scCO₂

We then used in situ FTIR to investigate the influence of the initial AIBN and monomer concentrations on the kinetics of the precipitation polymerization of AA in scCO₂. An FTIR spectrum recorded at T = 62 °C, P = 160 bar, [M]₀ = 0.128 M and [AIBN]₀ = 6.38 mM is shown in Fig. 3. Using a path length of 1 mm to ensure a good homogenization of the mixture, we observe that the main peaks associated with scCO₂ at about 3600 cm⁻¹ and 2300 cm⁻¹ are saturated. While some peaks associated with AA are also saturated (for example the ν(C=O) stretching vibration at 1720 cm⁻¹), we observed AA peaks at 1620, 1639, 1950 and 2664 cm⁻¹ with a good signal to noise ratio that we can use to quantify the monomer concentration during the polymerization (see ESI† 2). At [AA] > 0.128 M, the peaks at 1620, 1639 and 2664 cm⁻¹ also became saturated. Therefore, we monitored the disappearance of the peak at 1950 cm⁻¹ (see Fig. S3 of the ESI†).

Fig. 3 Initial FTIR spectrum of AA in scCO₂ at T = 62 °C, P = 160 bar [M]₀ = 0.128 M and [AIBN]₀ = 6.38 mM.

The peak at 1950 cm⁻¹ was used to check the initial concentration of AA in order to limit systematic errors due to the loss of a small amount of AA during the degassing of the cell with CO₂. The extinction coefficient, ε₁₉₅₀ was determined from an average of four FTIR spectra measured with the same cell at [AA] of 0.73 M. The concentration of the AA monomer during the reaction and the conversion were calculated from eqn (1) and (2), as for the polymerizations in toluene. The conditions of polymerization are compiled in Table S1 of the ESI.†

Fig. 4 shows the effect of changing the initial [AA] on the polymerization kinetics. As in toluene, the polymerization kinetics are characterized by an induction period during which very little polymerization occurs, followed by rapid onset of polymerization. Lower initial [AA] is associated with longer induction periods and a slightly slower rate of change of conversion (dX/dt).

Fig. 4 Selected conversion profiles for precipitation polymerization of acrylic acid in scCO₂ showing effect of changing initial [AA] (initial [AIBN] = 6.4 mM). Dashed lines represent best fit to Barrett and Thomas' model.¹⁷

In Fig. 5, the effect of changing initiator concentration is shown. As for [AA]₀, reducing the initiator concentration results in longer induction periods and slower polymerizations.

Fig. 5 Conversion profiles for precipitation polymerization of acrylic acid in scCO₂ showing effect of changing initial [AIBN] (initial [AA] = 0.71 M). Dashed lines represent best fit to Barrett and Thomas' model.¹⁷

Model fitting

Conversion vs. time data were fitted to the model developed by Barrett and Thomas,¹⁷ in which all polymerization is assumed to occur within the growing particles. This leads to the following expression for the rate of change of conversion with time (eqn (3)):

In eqn (3), the coefficient k is a composite of numerous rate constants and reagent concentrations (eqn (4)):

In eqn (4), k_p, k_t and k_i represent the rate constants of propagation, termination and initiation, respectively; [M]₀ and [I] represent the initial concentration of monomer and the concentration of initiator (assumed to be constant throughout the reaction); V_p is the molar volume of the polymer; and α is the partition coefficient of the monomer between the continuous and dispersed phases.

Solution of eqn (3) leads to the following expression for conversion as a function of time (eqn (5)):

This equation results in a sigmoidal conversion vs. time curve, displaying the induction period, t₀, and auto-acceleration that is characteristic of precipitation and dispersion polymerizations.

The maximum rate of polymerization, R_p,max, occurs at a conversion of 1/3, and is given by eqn (6):

As the coefficient k is proportional to [M]^0.5 and [I]^0.5 (eqn (4)), R_p,_max should be proportional to [M]^1.5 and [I]^0.5.

The coefficient k was evaluated for each polymerization individually by non-linear least squares (NLLS) fitting of eqn (5) to the experimental conversion data (Fig. 2, 4 and 5), and used to determine R_p,max. In general, the model successfully reproduced the major features of the polymerization, e.g. the induction period and the autoacceleration. During the induction period, however, low levels of conversion were observed experimentally which were not predicted by the model. This likely corresponds to polymerization in the continuous phase leading to the nucleation of new particles.

As expected, the maximum rates of polymerization in both toluene and scCO₂ are proportional to [AA]^1.5 (Fig. 6). The rates of polymerization in the two solvents are similar.

Fig. 6 Maximum rate of polymerization for precipitation polymerization of acrylic acid in scCO₂ (circles) or toluene (squares), showing 3/2 order dependence on [AA]. Solid line shows model prediction based on combined fit to all scCO₂ data points, dashed lines show ±1 standard error.

An average value of k = k′([M]₀[I])^0.5 was obtained by combining all the scCO₂ data, and NLLS fitting to eqn (7). In this model, the induction period t_i,0 was allowed to vary between experiments, while k′ was constant across all experiments.

This procedure resulted in a value of = 0.96 ± 0.11 M⁻¹ min⁻¹ for polymerizations of AA in scCO₂. A similar procedure for the polymerizations in toluene resulted in a value of = 0.84 ± 0.14 M⁻¹ min⁻¹. The error reported is a conservative estimate, calculated by assuming that the errors in each experiment are highly correlated, and that each additional experiment adds only two degrees of freedom to the total. Good agreement was obtained between the experimental and modelled data, as shown in Fig. 7 (see also Fig. S4 and S5 of the ESI†).

Fig. 7 Agreement between modelled and experimental conversion data for precipitation polymerizations of acrylic acid in scCO₂ (left) or toluene (right).

Modelled induction periods (t_i,0 in eqn (7)) for toluene and scCO₂ are shown in Fig. 8. In general, longer induction periods were observed in toluene than in scCO₂, although the maximum rate of polymerization was similar in either solvent. The difference in induction period is presumably related to slower particle nucleation in toluene than in scCO₂, either due to lower inherent solubility of poly(acrylic acid) in scCO₂, or due to the lower molecular weight of the polymer formed in toluene. Differences in the initiator efficiency (f = 0.88 in scCO₂, 0.55 in benzene, which should be similar to toluene) may also play a role, although this could be offset by slower decomposition of AIBN in scCO₂.³²

Fig. 8 Induction time as a function of monomer concentration (left, circles: scCO₂; squares: toluene) or initiator concentration (right, circles: scCO₂) for precipitation polymerizations of acrylic acid.

Characteristics of obtained poly(acrylic acid)s

We studied the influence of process parameters on weight-average molar mass of the PAAs obtained at the end of the polymerization in scCO₂ and toluene.

The effects of initiator and monomer concentration were studied at 62 °C and 160 bar. The residual AA content at the end of the polymerization ranged from 2–7%. The results, reported in Fig. 9, show the dependence of M_w on [AA]₀ and [AIBN]₀, respectively. All the conversion and SEC data are collected in Table S1 in ESI.† When the polymerization was carried out in scCO₂, M_w were high and reached 1.3 × 10⁶ g mol⁻¹ for the lowest [AIBN]₀ and the highest [AA]₀. We found a square-root dependence of M_w on [AIBN] and a first-order dependence on [AA], in agreement with the theoretical expression for kinetic chain length in free-radical polymerization. Dispersities were 2.1–2.6, and M_n values followed a similar trend to M_w. Sample recovery values (i.e. the weight fraction of PAA sample which was eluted in the SEC line and detected by the refractometer) were inversely proportional to M_w and ranged from 48 to 82%. The low observed recovery is due to chain transfer to polymer, which results in the formation of a fraction of insoluble, crosslinked polymer that is lost during filtration of the SEC sample. Polymerizations in toluene led to PAAs with much lower M_w of 2–4 × 10⁵ g mol⁻¹ (Fig. 9 and Table S1†), which can be explained by pronounced chain transfer to solvent.³³

Fig. 9 Weight-average molecular weight (M_w) as a function of monomer concentration (left, circles: scCO₂; squares: toluene) or initiator concentration (right, circles: scCO₂) for precipitation polymerizations of acrylic acid.

The size and morphology of the formed PAA particles were characterized by scanning electron microscopy (SEM). All the PAAs reported in this study formed a coagulum of fine particles of 50–200 nm in diameter (Fig. 10). This can be explained by the precipitation of high molecular weight chains which are in a solid, glassy state at a polymerization temperature well below the T_g of PAA.²¹

Fig. 10 SEM of PAA particles obtained by precipitation polymerization in scCO₂. Magnification of 17 000 × (left) and 3000 × (right). T = 62 °C, P = 160 bar, [AIBN]₀ = 6.38 mM, [AA]₀ = 0.71 M.

Conclusions

This study shows that the kinetics of polymerization in toluene and scCO₂ are consistent with a mechanism in which the main locus of polymerization is within the particles. This leads to characteristic polymerization kinetics showing an induction period in which very little polymerization takes place and an apparent dependence of the rate of polymerization on [AA]₀^1.5 and [AIBN]^0.5. The polymer molecular weight, meanwhile, is proportional to [AA]₀ and [AIBN]^−0.5, in line with expectation for a radical polymerization. The use of scCO₂ allows the production of high molecular weight polyacrylic acid powder with minimal contamination by solvent or unreacted monomer.

Acknowledgements

SEPPIC is gratefully acknowledged for financial support of this work.

Notes and references

1. F. L. Buchholz, in Ullman's Encyclopedia of Industrial Chemistry, ed. S. Hawkins and G. Schulz, VCH, Weinheim, 1992, vol. A21, p. 143.
2. J. Barth, W. Meiser and M. Buback, Macromolecules, 2012, 45, 1339–1345.
3. N. Lorber, B. Pavageau and E. Mignard, Macromolecules, 2010, 43, 5524–5529.
4. N. F. G. Wittenberg, C. Preusser, H. Kattner, M. Stach, I. Lacík, R. A. Hutchinson and M. Buback, Macromol. React. Eng., 2016, 10, 95–107.
5. I. Chaduc, A. Crepet, O. Boyron, B. Charleux, F. D'Agosto and M. Lansalot, Macromolecules, 2013, 46, 6013–6023.
6. N. Micic, A. Young, J. Rosselgong and C. Hornung, Processes, 2014, 2, 58.
7. I. Lacík, M. Stach, P. Kasák, V. Semak, L. Uhelská, A. Chovancová, G. Reinhold, P. Kilz, G. Delaittre, B. Charleux, I. Chaduc, F. D'Agosto, M. Lansalot, M. Gaborieau, P. Castignolles, R. G. Gilbert, Z. Szablan, C. Barner-Kowollik, P. Hesse and M. Buback, Macromol. Chem. Phys., 2015, 216, 23–37.
8. J. M. Silva-Jara, R. Manríquez-González, F. A. López-Dellamary, J. E. Puig and S. M. Nuño-Donlucas, J. Macromol. Sci., Part A: Pure Appl. Chem., 2015, 52, 732–744.
9. H. Minami, A. Kimura, K. Kinoshita and M. Okubo, Langmuir, 2010, 26, 6303–6307.
10. J. F. Ge, F. J. Lu and W. Ding, Chin. Chem. Lett., 2002, 13, 993–996.
11. B. Renard and T. F. McKenna, Macromol. Symp., 2000, 150, 251–257.
12. D. Zhang and X. Yang, in Encyclopedia of Polymeric Nanomaterials, ed. S. Kobayashi and K. Müllen, Springer Verlag, Berlin Heidelberg, 2014, pp. 1–10.
13. C. Bunyakan, L. Armanet and D. Hunkeler, Polymer, 1999, 40, 6225–6234.
14. C. Bunyakan and D. Hunkeler, Polymer, 1999, 40, 6213–6224.
15. H.-G. Poersch-Panke, A. Avela and K.-H. Reichert, Angew. Makromol. Chem., 1993, 206, 157–169.
16. A. Avela, H.-G. Poersch and K.-H. Reichert, Angew. Makromol. Chem., 1990, 175, 107–116.
17. K. E. J. Barrett and H. R. Thomas, J. Polym. Sci., Part A-1: Polym. Chem., 1969, 7, 2621–2650.
18. H. Lee, E. Terry, M. Zong, N. Arrowsmith, S. Perrier, K. J. Thurecht and S. M. Howdle, J. Am. Chem. Soc., 2008, 130, 12242–12243.
19. C. Boyère, C. Jérôme and A. Debuigne, Eur. Polym. J., 2014, 61, 45–63.
20. G. Hawkins, P. B. Zetterlund and F. Aldabbagh, J. Polym. Sci., Part A: Polym. Chem., 2015, 53, 2351–2356.
21. T. Liu, J. M. DeSimone and G. W. Roberts, J. Polym. Sci., Part A: Polym. Chem., 2005, 43, 2546–2555.
22. T. Liu, J. M. DeSimone and G. W. Roberts, Polymer, 2006, 47, 4276–4281.
23. T. Liu, J. M. DeSimone and G. W. Roberts, Chem. Eng. Sci., 2006, 61, 3129–3139.
24. Q. Xu, B. Han and H. Yan, Polymer, 2001, 42, 1369–1373.
25. M. Buback, Angew. Chem., Int. Ed. Engl., 1991, 30, 641–653.
26. S. Beuermann, M. Buback, C. Isemer and A. Wahl, Macromol. Rapid Commun., 1999, 20, 26–32.
27. S. Beuermann, M. Buback, C. Schmaltz and F.-D. Kuchta, Macromol. Chem. Phys., 1998, 199, 1209.
28. P. Chambon, E. Cloutet, H. Cramail, T. Tassaing and M. Besnard, Polymer, 2005, 46, 1057–1066.
29. S. Beuermann, M. Buback and M. Jürgens, Ind. Eng. Chem. Res., 2003, 42, 6338–6342.
30. V. Martinez, S. Mecking, T. Tassaing, M. Besnard, S. Moisan, F. Cansell and C. Aymonier, Macromolecules, 2006, 39, 3978–3979.
31. M. K. Mishra and S. N. Bhadani, J. Macromol. Sci., Chem., 1985, 22, 235–242.
32. Z. Guan, J. R. Combes, Y. Z. Menceloglu and J. M. DeSimone, Macromolecules, 1993, 26, 2663–2669.
33. T. J. Romack, E. E. Maury and J. M. DeSimone, Macromolecules, 1995, 28, 912–915.

---

Footnote

† Electronic supplementary information (ESI) available: ATR-IR device for the online monitoring of polymerization, characteristic absorption bands of AA, combined fits to polymerization in toluene and scCO₂, SEC results. See DOI: 10.1039/c6re00022c

---