Synthesis and bulk polymerization kinetics of monomer dehydroabietic acid-(2-acryloyloxy-ethoxy)-ethyl ester

A bulk polymerization monomer dehydroabietic acid-(2-acryloyloxy-ethoxy)-ethyl ester (DHADG-AC) was synthesized from dehydroabietic acid (DHA). The chemical structure of DHA-DG-AC was characterized by H NMR, C NMR, MS and FT-IR. The kinetics of the bulk polymerization of DHA-DG-AC was investigated by Differential Scanning Calorimeter (DSC). Two kinds of kinetic model (nth-order model and autocatalytic model) were used to investigate the polymerization process. The results showed that the experimental DSC curves were consistent with the computational data generated by the autocatalytic kinetic model, and the value of Ea was 95.73 kJ$mol .


Introduction
With the limitations in sourcing fossil fuels and the aggravation of environmental pollution, biomass resources such as rosin and oil are becoming more attractive [1] . Biomass resources are widely used to replace petrochemical resources, especially for polymers [1,2] . Hence, the design and synthesis of a naturally-derived polymerizable monomer is very important for sustainable development [3,4] .
Rosin is an abundant renewable resource which consists primarily of 90% rosin acids with characteristic hydrophenanthrene structures and 10% neutral compounds, and is widely used in many fields such as paper making, paint, adhesives and rubber [5] . The major component of rosin acid is abietic acid with a molecular formula C 20 H 30 O 2 . Abietic acid possesses a basic molecular structure of one carboxyl group and conjugated double bonds, which can react with different chemical substances to be transformed in polymerizable monomers [6] . Abietic acid reacts with maleic anhydride, acrylic acid, fumaric acid to be produced maleopimaric acid [7,8] , acrylopimaric acid [9,10] , and fumaropimaric acid [11,12] by the Diels-Alder reaction. The rosin based monomers can be used for further synthesis and to prepare polymers. Rosin based polymers, including epoxy resin [13,14] , polyester [15,16] and polyamide [17,18] , have been investigated as alternatives to petroleum based feedstock, as these exhibit better thermal stability and chemical resistance than petro-based polymers.
Vinyl, acrylic and allyl ester groups, which can undergo radical polymerization, have been grafted onto the rosin acid structure to afford rosin-derived vinyl polymeric monomers [19,20] . However, only a few applications of rosin-derived vinyl monomers have been reported, including use as crosslinking agent in UV-polymerization and thermal polymerization [21,22] . Due to characteristic hydrophenanthrene structures, rosin-derived vinyl monomers are used to synthesize pressure-sensitive adhesives with a high glass transition temperature and better adhesion performance by copolymerization with acrylic ester monomers [23] , and have great potential to be used in hydrophobically-modified water-soluble polymers. An appropriate kinetic model is one important prerequisite for understanding of polymerization. Such models are not only possible to predict the extent of polymerization, but they can also be used for process optimization in different systems [24] .
Differential scanning calorimetry (DSC) is the most widely used method to measure kinetic parameters, such as heat flow, extent of conversion and rate of conversion. A kinetic equation can then be used to compare with experimental data obtained by DSC. The study of cure kinetics of thermosetting polymers is useful for understanding the relationship between structure and properties [25] . Um et al. [26] proposed a procedure manipulating dynamic scanning data to determine cure kinetics, and then the cure kinetics were applied to study a three-component epoxy resin system. For free radical polymerization, several theories have been proposed to explain the kinetics and have achieved varying degrees of success in fitting the experimental data [27,28] . Zhang et al. [29] studied the kinetics of polymerization of 2,2-dinitropropyl acrylate, 2,2-dinitrobutyl acrylate and 2,2-dinitrobutyl methacrylate, which showed that the polymerization ability of three monomers decreased due to the presence of substituent methyl groups on the acrylyl double bond and 2,2dinitrobutyl on the ester group.
In this paper, a new radical polymerization monomer dehydroabietic acid-(2-acryloyloxy-ethoxy)-ethyl ester (DHA-DG-AC) was synthesized from dehydroabietic acid (DHA) which was separated from rosin. The kinetics were investigated by differential scanning calorimeter (DSC) in the non-isothermal mode and compared to the computational kinetic data. DHA-DG-AC has been shown to be a naturally-derived bulk polymerizable monomer.

Synthesis of DHA-DG-AC
The synthesized DHA-DG was used directly in the next step. DHA-DG (1.0 g, 0.0026 mol), triethylamine (0.26 g), hydroquinone (0.001 g) and dichloromethane were mixed in a flask. The combined solution of acrylylchloride (0.26 g, 0.0028 mol) and dichloromethane was added dropwise into the flask, and then the mixture was stirred at 25°C for 12 h. After the reaction, the solution was filtered and washed three times with Na 2 CO 3 aqueous solution. The organic solution was vacuum-distilled at 400 Pa and 45°C and dried in a vacuum oven, and a transparent viscous liquid, DHA-DG-AC (purity 96%), was obtained by silica gel column chromatography. DHA-DG-AC, 1 (Fig. S1).
2.4 Measurement of the progress of bulk polymerization of DHA-DG-AC by differential scanning calorimeter DHA-DG-AC and 2 wt% 2, 2'-azobis-isobutyronitrile were placed in a DSC cell. The kinetic data were obtained by DSC with nitrogen as flushing gas, heated from 25 to 200°C with different heating rates of 10, 15 and 20 K$min -1 . The heat flow was obtained, and then the kinetic data were processed further by DSC kinetics analysis software (Pyris Software, PerkinElmer, USA) to obtain the extent of conversion and the rate of polymerization.

Characterization
1 H NMR spectra were recorded with a Bruker 300 MHz spectrometer at room temperature with deuterated chloroform. FT-IR spectra were obtained using a Thermo Scientific Nicolet iS10 FT-IR spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) in the 400-4000 cm -1 wavenumber range. Mass spectrum was recorded on an Agilent-5973 spectrometer (ESI source; Agilent Technologies, Santa Clara, CA, USA).

Synthesis and characterization
The synthetic of DHA-DG-AC is shown in Fig. 1. The FT-IR spectra of DHA, DHA-DG and DHA-DG-AC are given in Fig. 2. In Fig. 2b, the characteristic DHA peak at 1687 cm -1 has disappeared after esterification, while peaks of ester carbonyl groups at 1725 cm -1 and hydroxyl group at 3435 cm -1 can be seen. In Fig. 2c, the characteristic peaks shown at 3435 cm -1 have disappeared and a peak at 1636 cm -1 indicates that the formation of the terminal C = C had occurred. The 1 H-NMR spectra of DHA-DG and DHA-DG-AC are shown in Fig. 3. The signals from 6.8 to 7.5 ppm were assigned to the protons on the aromatic ring. Compared to Fig. 3, the peaks from 5.5 to 6.6 ppm represented the protons on the unsaturated carbon of acrylic ester group. In Fig. 4, the carboxyl groups at 177.98 and 165.61 ppm, and C = C group at 130.57 and 127.69 ppm are present. These peaks confirm that DHA-DG-AC was synthesized.

Model examination and data correlation
The bulk polymerization kinetic parameters of DHA-DG-AC were measured by DSC. It is assumed that the measured heat flow (dH/dt) is proportional to the rate of polymerization (dα/dt). The rate of polymerization is defined as follows [31,32] : where DH is the enthalpy of the polymerization reaction. The rate of polymerization also can be described by  Eq. 2 [33] : where f(α) is a dependent kinetic model function, and K(T) is the temperature-dependent rate constant dependent on Arrhenius relationship [34] .
where E a is the apparent activation energy, A is the preexponential factor, R is the gas constant and T is the reaction temperature. The rate of polymerization is also obtained by Eqs. 2 and 3.
To develop a generalized model of the kinetics, two kinds of model (nth-order model and autocatalytic model) were investigated.

Nth-order kinetic model
For nth-order kinetic, f(α) is described as follows.
where n is the reaction order.
The rate of polymerization defined as follows is used.
The nth-order kinetic parameters are obtained by the Kissinger method [34] and Crane method [35] . At the DSC peak exotherm (T p ), it is assumed that the extent of polymerization reaction is constant and not dependent on the heating rates (β). The Kissinger method and the Crane method are shown in Eqs. 7 and 8. Based on the slope and the intercept, the apparent activation energy (E a ) of 82.77 kJ$mol -1 was obtained and the pre-exponential factor A was 1.58 Â 10 11 . A linear relationship of ln(β) versus 1/T p was obtained in Fig. 9 and the linear equation could be described by Eq. 10.
From the slope and E a , the reaction order n was calculated as 0.9047. Based on the kinetic parameters obtained, the rate of bulk polymerization could be expressed as follows. The rate of polymerization equation was used to compute dα/dt versus T curves and comparisons with the experimentally obtained curves are shown in Fig. 10. The calculated value of dα/dt based on Eq. 11 was much higher than the experimental data. Evidently, the nth-order reaction is not able to accurately describe the bulk polymerization of DHA-DG-AC.

Autocatalytic kinetic model
The kinetic parameters of the bulk polymerization of DHA-DG-AC were obtained by the method of Mαlek on kinetic analysis. The apparent activation energy E a was obtained by the isoconversional method and the logarithmic form of the kinetic Eq. 4.
The apparent activation energy E a can be calculated by    plotting of ln(dα/dt) versus 1/T p at different conversions as shows in Fig. 11.
As shown in Fig. 12, E a increased with the conversion due to the decreased mobility of the reactive group [25] . E a obtained a value of 95.73 kJ$mol -1 .
There are two special functions needed, y(α) and z(α), to find an appropriate kinetic model [36] . y(α) and z(α) are described as follows: zðαÞ ¼ πðuÞ The values of y(α) and z(α) were normalized within the range 0 to 1. From Figs. 13 and 14, α M and α p 1 are obtained and correspond to the max values of y(α) and z(α), respectively. Based on Fig. 15 and Table 1, it is evident that the value of α M was lower than the value of α p 1 , while α p 1 was less than 0.632. So the model of Šestαk-Berggren, Eq. 17 was the most appropriate.
where m, n refer to the the reaction order. The kinetic parameters are calculated by Eqs. 18 and 19. Fig. 10 Comparison of computational data and experimental data. da/dt, the rate of polymerization; T, the reaction temperature. Fig. 11 Plots of ln(dα/dt) versus 1/T p at different conversions. da/dt, the rate of polymerization; T p , the DSC peak exotherm. Fig. 12 The apparent activation energy E a at different conversions a Fig. 13 Plots of y(α) versus α. y(α), special function; α, the conversion.
The plots of ln[(dα/dt)exp(u)] versus ln[α p (1-α)] are shown in Fig. 16. The values of m, n and A were obtained from the slope, the intercept and Eq. 19 listed in Table 2. So the kinetic equation of bulk polymerization is described in Eq. 20.        Fig. 17 Comparison of computational data and experimental data. da/dt, the rate of polymerization; T, the reaction temperature.

Conclusions
DHA-DG-AC was successfully synthesized and characterized by FT-IR spectra, MS, 13 C NMR and 1 H NMR spectra. The kinetics of the bulk polymerization of DHA-DG-AC were studied by DSC in the non-isothermal mode. Two kinds of kinetic model: nth-order and autocatalytic were investigated. It was established that the autocatalytic kinetic model was the most suitable model for description of the bulk polymerization of DHA-DG-AC, and the value of E a was 95.73 kJ$mol -1 . DHA-DG-AC is a naturallyderived polymerizable monomer.