Mechanistic outlook on thermal degradation of 1,3-dialkyl imidazolium ionic liquids and organoclays

Abstract

The thermal degradation mechanisms of ionic liquids (ILs) 1-butyl-3-methylimidazolium chloride (BMImCl) and 1-butyl-3-methylimidazolium tetrafluoroborate (BMImBF₄) have been established using pyrolysis-GC-MS (Py-GC-MS) and B3LYP/6-311+G(d,p) level of density functional theory (DFT). BMImCl decompose through a bimolecular nucleophilic substitution (S_N2) while BMImBF₄ exhibit S_N2 along with a competitive E2 elimination pathway. Activation energy parameters obtained using Kissinger–Akahira–Sunose method and Ozawa–Flynn–Wall method is compared with the computed activation barriers. The montmorillonite based organoclay prepared using these ionic liquids exchange only the cation part ([BMIm]⁺) into the clay gallery leading to an expansion of d-spacing from 12.08 to 13.64 Å. The organoclay showed the maximum decomposition at 462 °C in the TGA experiment and the decomposition products were identified as methyl imidazole and 1-butene using Py-GC-MS. DFT studies employing a model compound Si(OH)₃O⁻ suggested a mechanism involving an imidazole-2-ylidine (carbene) intermediate for the decomposition of [BMIm]⁺ in the clay. Theoretical results were further supported by ¹³C NMR analysis of IL in presence of colloidal silica which showed a characteristic carbene NMR signal at 187.6 ppm.

---

1 Introduction

Ionic liquids (ILs) are a class of novel compounds typically composed of organic cations and inorganic or organic anions. ILs have large liquid ranges determined by their low melting point and high decomposition temperature. The high thermal stability of ILs attracted intense research interest owing to their applications in various fields such as energetic materials,^1,2 solvents for organic reactions,³ solvents for cellulose,⁴ thermal energy storage,⁵ clay modifiers to form thermally stable polymer composites,⁶ heat-transfer fluids,⁷ high-temperature lubricants,⁸ curing of ionogels⁹ and as a stationary phase in gas chromatography.¹⁰ It is well known that the physical and chemical properties of ILs vary with the choice of anion and cation while thermal stability is often tuned by the choice of anions.

For safe application of ILs at elevated temperatures, it is important to understand the degradation mechanism. The nucleophilic attack of the halide at primary alkyl substituents in imidazolium ring is known to be the major degradation pathway in N,N-dialkyl imidazolium halides.^11–16 Chowdhury et al.¹⁷ showed that imidazolium based ILs decompose at lower temperatures in the presence of nucleophiles. Baranyai et al.¹⁸ investigated the thermal decomposition mechanism of 1,3-dialkyl imidazolium bis(trifluoromethylsulfonyl)imide ILs, and suggested degradation of the anion as possible thermal decomposition pathway, but the anion degradation products were not detected. Lazzús¹⁹ introduced a group contribution method and Yan et al.²⁰ used quantitative structure property relationship method to predict the thermal decomposition temperature of ILs. Ohtani et al.²¹ studied decomposition of imidazolium ILs using Py-GC-MS and proposed mechanism based on the products formed. Quantum mechanical tools have also been extensively used to understand the degradation mechanism of compounds.^22–26

In spite of the several reports on the thermal stability of ILs,^27–30 systematic studies correlating the structural features as well as the degradation mechanism to understand the thermal stability of ILs are scarce. In the present study, the degradation reactions of 1-butyl-3-methylimidazolium chloride (BMImCl), 1-butyl-3-methylimidazolium tetrafluoroborate (BMImBF₄) and 1-butyl-3-methylimidazolium modified montmorillonite clay (Clay-BMIm) are analyzed using Py-GC-MS and formation of pyrolysis products are rationalized using density functional theory (DFT). We have proposed a new mechanism for decomposition of 1,3-dialkylimidazolium modified clays in general by taking into consideration of Clay-BMIm. To the best of our knowledge, this is the first report on thermal decomposition mechanism of 1,3-dialkylimidazolium incorporated montmorillonite clay.

2 Methodology

2.1 Materials

BMImCl and BMImBF₄ of purity > 99% were procured from M/s Otto Chemie Pvt. Ltd, Mumbai, India. Clay used was sodium montmorillonite (trade name: Cloisite-Na⁺, CAS no. 1318-93-0, Southern Clay Products, Inc., USA) for the ion exchange modification with ILs. Colloidal silica (30% suspension of SiO₂ in water, CAS no. 7631-86-9) was obtained from Sigma-Aldrich.

2.2 Functionalization of clay

Standard ion exchange procedures were employed for the functionalization of sodium montmorillonite clay (MMT) with ILs.³¹ Aqueous suspension of MMT clay (2%) was prepared by sonicating for 10 minutes using Hielscher-UIP1000hd probe sonicator at amplitude of 80%. To this, BMImCl (or BMImBF₄, both yield same product Clay-BMIm) was added and the exchange reactions were carried out by sonicating for 15 minutes. It was allowed to settle and filtered. The residue was repeatedly (5–8 times) washed with distilled water until no chloride traces were detected with silver nitrate solution. The residue (modified MMTs) was dried at room temperature for 4 h and then at 100 °C for 2 hour under vacuum. Characterisation of modified clay is included in the (ESI).†

2.3 Instrumental

Thermal stability of ILs and Clay-BMIm were studied using Perkin-Elmer, Pyris-1 TGA (Shelton, USA) in He atmosphere at a heating rate of 10 °C min⁻¹. Py-GC-MS studies were carried out using a system consists of a Thermo Electron Trace ultra GC directly coupled to a Thermo Electron PolarisQ (Quadrupole ion trap) mass spectrometer (Thermo Electron Corporation, Waltham, Massachusetts, USA) and SGE pyrolyser (Pyrojector II, SGE Analytical Science Pty Ltd, Ringwood, Victoria, Australia). The GC is equipped with 30 m, 0.25 mm ID capillary column (PDMS with 5% phenyl, SLB-5MS, Supelco). The Py-GC-MS conditions were set with ion source at 200 °C, inlet at 250 °C, transfer line at 280 °C, column temperature programme 40 °C to 250 °C with heating rate of 10 °C min⁻¹. The m/z range used was 20–600 amu and the pyrolysis was carried out at 350 °C and 600 °C. ¹³C NMR analysis was done using Bruker Avance-III 400 MHz spectrometer operated at 100.2 MHz with power 54 W, pulse width 8.9 μs and D₂O as locking solvent.

2.4 Computational method

Geometries of gaseous ion pairs of BMImCl, BMImBF₄, transition states and products were optimized at B3LYP³² level of DFT using 6-311+G(d,p) basis set as implemented in Gaussian09.³³ All the transition states were confirmed by the presence of single imaginary frequency. The single step conversion is further analysed using IRC calculations (ESI†). Molecular electrostatic potential (MESP) derived Merz–Singh–Kollman charge (MK charge)^34,35 was computed at the same level of theory. Basis set superposition error corrected binding energy³⁶ for the ion-pairs ([BMIm]⁺[X]⁻) was calculated using eqn (1).

[BMIm]⁺ + [X]⁻ ⇌ [BMIm]⁺[X]⁻ (1)

where [X]⁻ = Cl⁻ or BF₄⁻.

2.5 Kinetic studies

Kinetic parameters for the thermal decomposition of ILs were calculated using two methods, Kissinger–Akahira–Sunose (KAS) method^37–39 and Ozawa–Flynn–Wall (FWO) method.^40–42 KAS is a multiple heating rate method and we have selected heating rate of 2, 5, 10 and 15 °C min⁻¹. KAS method is based on the following eqn (2)where β is the heating rate, T_α is the temperature in Kelvin corresponding to a fixed degree of conversion α, A is the pre-exponential factor, R is the gas constant, E_a is the activation energy at a given degree of conversion and g(α) is the integral form of kinetic model function. E_a for a given degree of conversion is obtained from the slope of the linear fit of the plot versus 1/T_α. FWO method for determination of kinetic parameters is based on the assumption that the decomposition obeys first order kinetics using a point of constant conversion from a series of decomposition curves obtained at different heating rates. It helps in calculating the E_a for a desired conversion (eqn (3)).where b = 0.457 initially and it depends on the value of E/RT.³⁹ Value of b = 0.471, 0.453 and 0.456 respectively for BMImCl, BMImBF₄ and Clay-BMIm (ESI†).

3 Results and discussion

3.1 Electronic structure and properties of ion pairs

Optimized structures of BMImCl and BMImBF₄ at B3LYP/6-311+G(d,p) level of DFT are shown in Fig. 1. In BMImCl, the distances of the C2H⋯Cl (2.005 Å) are much longer than the covalent bond distance of H–Cl (1.31 Å) and shorter than the van der Waals distance of H⋯Cl (2.95 Å). Similarly in BMImBF₄, the shortest distance of C2H⋯F between the imidazolium cation and the anion (2.057 Å) are higher than covalent bond distance of H–F (1.07 Å) and smaller than the van der Waals distance of H–F (2.67 Å). The non-covalent binding energy (E_b) computed for ILs are very high, 373.4 kJ mol⁻¹ and 340.2 kJ mol⁻¹ respectively for BMImCl and BMImBF₄, indicating strong electrostatic interactions in ion pairs. The geometric features as well as the E_b values of these ion pairs agreed very well with those reported by Dong et al.⁴³ and Fumino et al. at B3LYP/6-31+G(d) level.⁴⁴ The MK charge can be considered as a measure of charge transfer from the anion to the cation. Higher charge transfer was observed for BMImCl (23%) than BMImBF₄ (12%) (Table 1).

Fig. 1 Optimized structures of BMImCl and BMImBF₄ at B3LYP level of DFT.

Table 1 MESP plots, MESP-derived charges, E_b and HOMO–LUMO gap of ion pairs

IL	MESP plot & MK charge (a.u)	Eb (kJ mol^−1)	HOMO–LUMO gap (eV)
BMImCl		−373.4	4.08
BMImBF4		−340.2	6.88

---

Generally the stability of a salt system is assessed by the large energy gap between the highest occupied molecular orbital and the lowest unoccupied molecular orbital (HOMO–LUMO gap).⁴⁵ Since, the charge transfer is more conducive in BMImCl than BMImBF₄, the latter with higher HOMO–LUMO gap (6.88 eV) can be considered as a more stable salt than BMImCl, which is further supported by TG analysis.

3.2 TG analysis

Fig. 2 shows the TG curves for BMImCl, BMImBF₄ and Clay-BMIm. The higher initial decomposition temperature observed for BMImBF₄ (260 °C) compared to BMImCl (180 °C), suggests the role of counter anion on thermal stability of ILs. The maximum decomposition temperatures (T_s) are 297, 427 and 462 °C respectively for BMImCl, BMImBF₄ and Clay-BMIm. The differences in the decomposition of selected ILs with different anions and Clay-BMIm, without an anion suggests the possibility of different mechanisms for their thermal decomposition reactions. In Clay-BMIm the weight loss observed between 500 °C and 600 °C was attributed to the dehydroxylation of structural hydroxyl groups.⁴⁶

Fig. 2 TG curves for BMImCl, BMImBF₄ and Clay-BMIm in He atmosphere at a heating rate of 10 °C min⁻¹.

E_a for thermal decomposition of BMImCl, BMImBF₄ and Clay-BMIm were calculated using KAS method and FWO method (Table 2). The calculated E_a values showed 6% deviation for BMImCl, 8% for BMImBF₄ and 9% for Clay-BMIm (ESI†).

Table 2 Kinetic studies using TG

β (°C min^−1)	KAS method					FWO method
	Ts (°C)				Ea (kJ mol^−1)	Ea (kJ mol^−1)
	2	5	10	15
BMImCl	257	285	297	300	101.5	107.4
BMImBF4	402	411	427	432	236.8	258.0
Clay-BMIm	437	442	462	468	229.3	208.5

---

3.3 Decomposition mechanism

3.3.1 BMImCl. Fig. 3 shows the pyrogram of BMImCl at two different temperatures, 350 and 600 °C. Four major decomposition products were identified, 1-chloromethane (m/z = 50), 1-chlorobutane (m/z = 92), 1-methylimidazole (m/z = 82) and 1-butylimidazole (m/z = 124). A decomposition mechanism based on the products obtained on pyrolysis is shown in Scheme 1.

Fig. 3 Pyrogram and decomposition products of BMImCl at 350 and 600 °C.

Scheme 1 Decomposition route of BMImCl.

The dealkylation reaction is initiated by the nucleophilic attack of chloride anion at the alkyl group of imidazolium cation which passes through the bimolecular nucleophilic (S_N2) transition state. Kroon et al.²² used B3LYP at 6-31G** level to predict the decomposition products of BMImCl. For pathway-I, E_a was 127 kJ mol⁻¹ while for pathway-II, a higher value 136 kJ mol⁻¹ was observed.

Semi quantitative estimation using standard solutions of 1-butylimidazole and 1-methylimidazole shows that 90% of the products are from pathway-1 (ESI†). We have calculated E_a of reaction pathways shown in Scheme 1 at a higher level, B3LYP/6-311+G(d,p) to confirm the product distribution as calculated from Py-GC-MS. Transition states identified and energy profile are shown in Fig. 4.

Fig. 4 Energy profile and transition states of BMImCl decomposition shown in Scheme 1. (a) Pathway-I and (b) pathway-II. Bond lengths in Å.

Demethylation reaction (pathway-I) proceed with an E_a of 122.7 kJ mol⁻¹ while debutylation reaction proceed with an E_a of 130.8 kJ mol⁻¹ (Fig. 4). These results match with Kroon et al.²² and shows the selectivity for pathway-I even though the E_a differ only by 8 kJ mol⁻¹ and the higher percentage of butylimidazole is accounted. Consequently the net E_a for the reaction was 123.5 kJ mol⁻¹ (Table 3) and experimental values for E_a were 101.5 kJ mol⁻¹ and 107.4 kJ mol⁻¹ calculated using KAS and FWO methods respectively. The results showed a deviation of 13% with FWO method and 18% with KAS method (Table 3). DFT studies also predict that both demethylation and debutylation reactions of BMImCl are endothermic by ∼10 kJ mol⁻¹ and DSC analysis shows endotherm for the same (ESI†).

Table 3 Comparison of computed and experimental E_a with extend of reaction (% R_n) for BMImCl and BMImBF₄

	Computational study					Experimental
	SN2(I)		SN2(II)		Avg. Ea (kJ mol^−1)	KAS method		FWO method
	Ea (kJ mol^−1)	Rn (%)	Ea (kJ mol^−1)	Rn (%)		Ea (kJ mol^−1)	Deviation (%)	Ea (kJ mol^−1)	Deviation (%)
BMImCl	122.7	90	130.8	10	123.5	101.5	17.8	107.4	13.0
E2
BMImBF4	188.5	51	237.0	49	212.1	236.8	10.4	258.0	17.8

---

3.3.2 BMImBF₄. Pyrogram of BMImBF₄ at 600 °C is shown in Fig. 5. Major peaks in the pyrogram are identified as 1-butene (m/z = 56), 1-methylimidazole (m/z = 82) and 1-butylimidazole (m/z = 124).

Fig. 5 Pyrogram of BMImBF₄ at 600 °C.

Owing to the steric bulkiness of BF₄⁻ compared to Cl⁻, the S_N2 reactions are likely to happen only at the N3–CH₃ site (Scheme 2) and it is unambiguous with the presence of butyl imidazole as the major product in the pyrogram.

Scheme 2 Decomposition routes of BMImBF₄.

Differing from BMImCl, the presence of 1-butene in the pyrogram suggests existence of competitive elimination reactions in the thermal decomposition of BMImBF₄. Three different elimination mechanism namely bimolecular E2 elimination, internal elimination and elimination via imidazole-2-ylidene (carbene route)^47,48 were analyzed by DFT (ESI†) and transition state was identified only for E2 elimination (pathway-II).

Though the intramolecular elimination reactions (Ei) are common in pyrolysis reactions, in the case of BMImBF₄ the basicity of the anion overrules and E2 elimination is found to be preferred over Ei route, as the basicity of BF₄⁻ anion is insufficient to form the imidazole-2-ylidene which is expected to undergo further abstraction of β-hydrogen to form the pyrolysis products viz. butene and methyl imidazole (ESI†).

Computed activation barrier for the formation of butyl imidazole, (Fig. 6) is 188.5 kJ mol⁻¹ (195 kJ mol⁻¹ by Kroon et al.²² at B3LYP/6-31** level of DFT) and is higher than the corresponding value for BMImCl (122.7 kJ mol⁻¹, Fig. 4, pathway-I). Transition state for E2 elimination was characterized by a four membered ring with breaking of C–N bond at C⋯N distance 2.537 Å and simultaneous β-hydrogen shift. The activation barrier for this reaction was 237 kJ mol⁻¹ (Fig. 6). The experimental values depicted in Table 2 using KAS method (236.8 kJ mol⁻¹), FWO method (258.0 kJ mol⁻¹) and a higher concentration of 1-butene and 1-methylimidazole in the pyrogram suggest the possibility of interplay of both the mechanisms in decomposition of BMImBF₄. The ratio of 1-methylimidazole to 1-butylimidazole was determined by a semi quantitative method as in the case of BMImCl and observed both reactions proceed equally (1 : 1) (ESI†). The average E_a predicted using DFT was 212.1 kJ mol⁻¹ (Table 3) and the deviation between computed and experimental E_a may be due to the lattice energy change in experiments where ILs in liquid phase are analysed unlike gas phase calculations in computational method. Both the mechanisms discussed are highly endothermic (172–212 kJ mol⁻¹) as shown in Fig. 6 and confirmed using DSC analysis (ESI†).

Fig. 6 Energy profile for (a) pathway-I (S_N2 mechanism) and (b) pathway-II (E2 elimination). Transition states are also shown with distance parameters in Å.

3.3.3 Clay-BMIm. In the present study organoclays were prepared using BMImCl and characterized (ESI†). Thermal stability of Clay-BMIm was studied using TGA (Fig. 2) and decomposition products were identified as butene and methyl imidazole using Py-GC-MS (Fig. 7).

Fig. 7 Pyrogram of Clay-BMIm at 600 °C.

Thermal decomposition of Clay-BMIm was analysed using DFT method. MMT clays own a pyrophyllite structure where the trivalent Al-cation in the octahedral layer is partially substituted by the divalent Mg-cation. Consequent negative charge generated in the clay layer is balanced by hydrated sodium ions adsorbed in the interlayer. The interlayer dimension is determined by the crystal structure of the silicate and the d-spacing is ∼1 nm for MMT clay. From the model of MMT clay given in Fig. 8, it is clear that the ions in the interlayer gallery face SiO₄ tetrahedral unit as their immediate neighbour.⁴⁹ In Clay-BMIm sodium ions are replaced by 1-butyl-3-methylimidazolium cation.

Fig. 8 (a) Model structure of MMT clay, (b) Clay-BMIm, (c–h) optimized structures of species involved in decomposition of Clay-BMIm with transition state and energy profile.

A truncated model for the montmorillonite clay showing the interaction of Na⁺ with oxygen atoms that bridges two silicon centers is presented in Fig. 8(a). The [BMIm]⁺ could replace the Na⁺ cation owing to extended interaction of several CH bonds on to the layer dominated by Si–O bonds. Considering the negatively charged character of the clay and the possible CH⋯O–Si interactions from [BMIm]⁺ to the clay, we propose the use of the anionic moiety Si(OH)₃O⁻ to model the decomposition reaction of [BMIm]⁺ in the clay.

BMIm–Si(OH)₃O⁻ ion pair is found to be marginally higher in energy (9.1 kJ mol⁻¹) than the neutral complex of imidazole-2-ylidene–Si(OH)₄ (Fig. 8). Therefore the DFT results suggest that the intercalated imidazolium cations can easily form the imidazole-2-ylidene through a barrier less pathway. The singlet carbenes thus formed on heating can undergo elimination reactions involving beta hydrogen shift from the butyl substituent at N1 to the carbenic centre followed by bond breaking at the quaternary nitrogen. Such a transition state was located in the DFT analysis with single imaginary frequency corresponding to the bond shifting process which is shown in Fig. 8(g).

The activation energy for the decomposition was 195.6 kJ mol⁻¹ (Fig. 8(h)) and the reaction is exothermic by 39.9 kJ mol⁻¹ and DSC analysis shows an exotherm for the same (ESI†). The summarized decomposition mechanism is shown in Fig. 9. Anions with high basicity is capable of deprotonating imidazolium cation at C2 position, resulting in the formation of neutral carbon bases with nucleophilic singlet carbenes that are stabilized by the two neighboring nitrogen atoms at the carbenic centre. The formation of carbene from BMIm cation on heating was confirmed by the NMR analysis of BMImCl in D₂O solvent in presence of colloidal silica. Colloidal silica is 30% suspension of SiO₂ in water, which contains –SiOH species similar to our model compound Si(OH)₃O⁻ used in DFT study. ¹³C NMR spectra of BMImCl and BMImCl with colloidal silica heated at 90 °C for 5 minutes are shown in Fig. 10. The additional NMR signal observed at 187.64 ppm is attributed to singlet carbene carbon in imidazolium ring.

Fig. 9 Mechanism of Clay-BMIm decomposition through 1-butyl-3-methylimidazole-2-ylidene route.

Fig. 10 ¹³C NMR spectrum of (a) BMImCl in D₂O and (b) BMImCl with colloidal silica in D₂O after heating.

4 Conclusions

Thermal stability of BMImCl and BMImBF₄ were compared using TGA and decomposition products were identified using Py-GC-MS experiments. The inherent stability of BMImBF₄ is attributed to their very high HOMO–LUMO gap as well as the higher MK charge computed for ion-pair counterparts. Bimolecular nucleophilic substitution reactions are identified to be the lowest energy pathway for decomposition of both the ions pairs. However, BF₄⁻ anion facilitates competitive E2 elimination reactions in imidazolium based ILs. The extend of reactions were calculated from Py-GC-MS and DFT results. Demethylation reaction of BMImCl constitutes 90% of the reaction. Decomposition of BMImBF₄ proceeds with 50% demethylation and 50% elimination reaction. For the first time the thermal degradation of ionic liquid modified clay was elucidated using DFT and experimentally rationalized using Py-GC-MS and ¹³C-NMR experiments.

Abbreviations

IL	Ionic liquid
[BMIm]^+	1-Butyl-3-methylimidazolium cation
BMImCl	1-Butyl-3-methylimidazolium chloride
BMImBF4	1-Butyl-3-methylimidazolium tetrafluoroborate
MMT	Montmorillonite clay
Clay-BMIm	Clay modified with ILs
DFT	Density functional theory
[BF4]^−	Tetrafluoroborate anion
Ea	Activation energy
Eb	Binding energy
Ts	Maximum decomposition temperature
HOMO	Highest occupied molecular orbital
LUMO	Lowest unoccupied molecular orbital
ΔE*	Energy gap between the HOMO and LUMO
MK charge	Merz–Singh–Kollman charge
KAS	Kissinger–Akahira–Sunose method
FWO	Osawa–Flynn–Wall method

Acknowledgements

The authors thank Director, Vikram Sarabhai Space centre for granting permission to publish this work. One of the authors (ET) thanks Indian Space Research Organization for providing research fellowship. The authors also thank Dr Suresh C. H., CSIR-National Institute for Interdisciplinary Science and Technology, Trivandrum for his valuable comments.

References

1. E. Sebastiao, C. Cook, A. Hu and M. Murugesu, J. Mater. Chem. A, 2014, 2, 8153–8173.
2. E. Thomas, K. P. Vijayalakshmi and B. K. George, RSC Adv., 2015, 5, 71896–71902.
3. R. H. Wang, C. M. Jin, B. Twamley and J. M. Shreeve, Inorg. Chem., 2006, 45, 6396–6403.
4. F. Wendler, L. N. Todi and F. Meister, Thermochim. Acta, 2012, 528, 76–84.
5. L. Bai, X. Li, J. Zhu and B. Chen, Energy Fuels, 2011, 25, 1811–1816.
6. J. Gilman, W. H. Awad, R. D. Davis, J. Shields, R. H. Harris, C. Davis, A. B. Morgan, T. E. Sutto, J. Callahan, P. C. Trulove and H. C. DeLong, Chem. Mater., 2002, 14, 3776–3785.
7. E. B. Fox, A. E. Visser, N. J. Bridges and J. W. Amoroso, Energy Fuels, 2013, 27, 3385–3393.
8. Z. Zeng, B. S. Phillips, J. C. Xiao and J. M. Shreeve, Chem. Mater., 2008, 20, 2719–2726.
9. J. L. Bideau, L. Viau and A. Vioux, Chem. Soc. Rev., 2011, 40, 907–925.
10. Z. Breitbach and D. Armstrong, Anal. Bioanal. Chem., 2008, 390, 1605–1617.
11. H. L. Ngo, K. LeCompte, L. Hargens and A. B. McEwen, Thermochim. Acta, 2000, 357–358, 97–102.
12. T. Erdmenger, J. Vitz, F. Wiesbrock and U. S. Schubert, J. Mater. Chem., 2008, 18, 5267–5273.
13. A. G. Glenn and P. B. Jones, Tetrahedron Lett., 2004, 45, 6967–6969.
14. L. Bai, X. Li, J. Zhu and B. Chen, Energy Fuels, 2011, 25, 1811–1816.
15. C. Maton, N. D. Vos and C. V. Stevens, Chem. Soc. Rev., 2013, 42, 5963–5977.
16. K. R. J. Lovelock, J. P. Armstrong, P. Licence and R. G. Jones, Phys. Chem. Chem. Phys., 2014, 16, 1339–1353.
17. A. Chowdhury and S. T. Thynell, Thermochim. Acta, 2006, 443, 159–172.
18. K. J. Baranyai, G. B. Deacon, D. R. MacFarlane, J. M. Pringle and J. L. Scott, Aust. J. Chem., 2004, 57, 145–147.
19. J. A. Lazzús, J. Mol. Liq., 2012, 168, 87–93.
20. F. Yan, S. Xia, Q. Wang and P. Ma, J. Chem. Eng. Data, 2012, 57, 805–810.
21. H. Ohtani, S. Ishimura and M. Kumai, Anal. Sci., 2008, 24, 1335–1340.
22. M. C. Kroon, W. Buijs, C. J. Peters and G. J. Witkamp, Thermochim. Acta, 2007, 465, 40–47.
23. S. Reshmi, K. P. Vijayalakshmi, D. Thomas, B. K. George and C. P. Reghunadhan Nair, J. Anal. Appl. Pyrolysis, 2013, 104, 603–608.
24. Z. Xue, Y. Zhang, X. Zhou, Y. Cao and T. Mu, Thermochim. Acta, 2014, 578, 59–67.
25. P. Ballone and R. C. Huerto, Faraday Discuss., 2012, 154, 373–389.
26. Y. Hao, J. Peng, S. Hu, J. Li and M. Zhai, Thermochim. Acta, 2010, 501, 78–83.
27. Y. Cao and T. Mu, Ind. Eng. Chem. Res., 2014, 53(20), 8651–8664.
28. W. H. Awad, J. W. Gilman, M. Nyden, R. H. Harris, T. E. Sutto, J. Callahan, P. C. Trulove, H. C. DeLong and D. M. Fox, Thermochim. Acta, 2004, 409, 3–11.
29. C. P. Fredlake, J. M. Crosthwaite, D. G. Hert, S. N. V. K. Aki and J. F. Brennecke, J. Chem. Eng. Data, 2004, 49, 954–964.
30. M. E. Van Valkenburg, R. L. Vaughn, M. Williams and J. S. Wilkes, Thermochim. Acta, 2005, 425, 181–188.
31. L. B. de Paiva, A. R. Morales and F. R. V. Diaz, Appl. Clay Sci., 2008, 42, 8–24.
32. A. D. Becke, Phys. Rev. A, 1988, 38, 3098–3100.
33. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, Revision A.1, Gaussian Inc., Wallingford, CT, 2009.
34. U. C. Singh and P. A. Kollman, J. Comput. Chem., 1984, 5, 129–145.
35. B. H. Besler, K. M. Merz Jr. and P. A. Kollman, J. Comput. Chem., 1990, 11, 431–439.
36. S. F. Boys and F. Bernardi, Mol. Phys., 1970, 19, 553–566.
37. H. Tanaka, Thermochim. Acta., 1995, 267, 29–44.
38. S. Vyazovkin, K. Chrissafis, M. L. Di Lorenzo, N. Koga, M. Pijolat and B. Roduit, Thermochim. Acta, 2014, 590, 1–23.
39. B. V. L'Vov, Spectrochim. Acta, Part B, 2011, 66(7), 557–564.
40. ASTM, Standard Test Method for Decomposition Kinetics by Thermogravimetry, E-1641–1699.
41. T. Ozawa, Bull. Chem. Soc. Jpn., 1965, 88, 1881–1886.
42. J. H. Flynn and L. A. wall, Polym. Lett., 1966, 4, 323–328.
43. K. Dong, S. Zhang, D. Wang and X. Yao, J. Phys. Chem. A, 2006, 110, 9775–9782.
44. K. Fumino, A. Wulf and R. Ludwig, Phys. Chem. Chem. Phys., 2009, 11, 8790–8794.
45. G. Zhao and M. Lu, J. Mol. Model., 2012, 16, 2443–2451.
46. P. Singla, R. Mehta and S. N. Upadhyay, Green Sustainable Chem., 2012, 2, 21–25.
47. O. Holloczki, D. Gerhard, K. Massone, L. Szarvas, B. Nemeth, T. Veszpremi and L. Nyulaszi, New J. Chem., 2010, 34, 3004–3009.
48. E. A. Turner, C. C. Pye and R. D. Singer, J. Phys. Chem. A, 2003, 107, 2277–2288.
49. S. Cadars, R. Guegan, M. N. Garaga, X. Bourrat, L. L. Forestier, F. Fayon, T. V. Huynh, T. Allier, Z. Nour and D. Massiot, Chem. Mater., 2012, 24, 4376–4389.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra24907d

---