Ab initio study on thermal degradation reactions of polycarbonate

Journal of Molecular Structure (Theochem) 634 (2003) 305–310 www.elsevier.com/locate/theochem

Ab initio study on thermal degradation reactions of polycarbonate Jussi Katajisto, Tuula T. Pakkanen, Tapani A. Pakkanen*, Pipsa Hirva Department of Chemistry, University of Joensuu, P.O. Box 111, Joensuu 80101, Finland Received 2 May 2003; revised 13 May 2003; accepted 13 May 2003

Abstract An ab initio study on thermal degradation of polycarbonate, poly(bisphenol A carbonate), PC, was carried out using HartreeFock (HF) and density functional (B3LYP) methods. The polycarbonate chain was modelled by a dimer unit (PC-2), which represented a fragment of the polymer chain. The main emphasis was on the determination of reaction enthalpies and Gibbs energies for the different degradation reactions, which were calculated at a standard state (298.15 K) and at a higher temperature (673.15 K). On the basis of the calculated Gibbs energies, the various degradation paths were ranked according to their favourableness and spontaneity and found to agree well with experimental data. The CO2 elimination reaction and substitution reaction with water were found to be the most favourable reaction paths. q 2003 Elsevier B.V. All rights reserved. Keywords: Polycarbonate; Molecular modeling; Ab initio; Thermal degradation; Thermal isomerization

1. Introduction Polycarbonate, poly(bisphenol A carbonate), PC, is an important polymer with a variety of optical and technical applications. It is widely used in optical data storage devices, such as CD, CD – Rw and DVD disks [1]. Experimental and theoretical studies on the thermal degradation of PC are of particular interest, because they give information about the thermal stability, which can be applied in the tailoring of polycarbonate plastics for different purposes. The thermal degradation paths of polycarbonate have been studied by MALDI-TOF mass spectrometry [2] and found to include intramolecular exchange, * Corresponding author. Tel.: þ358-13-251-3345; fax: þ 358-13251-3344. E-mail address: [email protected] (T.A. Pakkanen).

substitution with water, CO2 elimination, disproportionation, and thermal isomerization. The reaction products are short-chain oligomers, containing 7 –12 monomer units. In this work, we study the thermodynamics of the various degradation paths of polycarbonate by applying ab initio methods. The selection of the reaction paths was made on the basis of mass spectrometry results [2 – 9]. The thermodynamics of the thermal degradation reaction paths of PC is not completely understood and the main aim of our study was to identify the energetically most favorable route. Some of the thermal degradation reaction paths of PC have recently been studied by B3LYP/6-311G** method and thermogravimetric analysis, but there the goal was to study the effect of different substituents in the polymer chain [10].

0166-1280/03/$ - see front matter q 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0166-1280(03)00385-3

306

J. Katajisto et al. / Journal of Molecular Structure (Theochem) 634 (2003) 305–310

2. Computational methods and models When there are several reaction paths, as in the case of thermal isomerization of PC, availability of the Gibbs energies of the different reaction steps allows estimation of the reaction path that is energetically the most favorable. All calculations were carried out by HF and B3LYP methods, as incorporated in the GAUSSIAN 98 program [11]. A theoretical study of the whole PC chain was not practical at the ab initio level. The size of the studied system was therefore limited to a dimer unit of polycarbonate (PC – 2.) The dimeric model represents the smallest polymeric unit based on the polymerization reaction between bis-phenol A and phosgene monomers. The chain ends (OH and Cl) represent the functional groups of the monomers (Fig. 1). At the first stage, the minimum energy conformations of the reactants and products were determined by a systematic conformational analysis with the Tripos force field [12]. At the second stage, the geometry of the minimum energy conformations was further optimized at the HF/3-21G, HF/6-31G* and B3LYP/6-31G* levels of theory. The results for the thermodynamic parameters were also calculated with a larger basis set (6-311G**). At the final stage we used HF/6-31G* method in the IR frequency calculations. These calculations were carried out at two different temperatures 273.15 and 673.15 K. The thermal enthalpy and Gibbs energy corrections of reactants and products were determined and the enthalpy of reaction was calculated according to Eq. (1). o

Dr H ¼

X products

ðE þ HCORR Þ 2

X

In Eq. (1) ðEÞ is the total energy of the molecule and ðHCORR Þ the thermal correction coefficient of the enthalpy. The Gibbs energy of a reaction was calculated by using thermal correction values of Gibbs energies ðGCORR Þ and total energies of molecules (E) according to Eq. (2) X X Dr Go ¼ ðE þGCORR Þ2 ðE þGCORR Þ ð2Þ products

reactants

HCORR ¼ Etot þkb T

ð3Þ

GCORR ¼ HCORR 2TStot

ð4Þ

Stot ¼ St þSr þSv þSe

ð5Þ

Etot ¼ Et þEr þEv þEe

ð6Þ

The terms HCORR and GCORR are thermal corrections to energy and they are defined in Eqs. (3) – (6). St ; Sr ; Sv and Se are translational, vibrational, rotational and electronic contributions to total molecular entropy. Et ; Er ; Ev and Ee are the corresponding energy terms. The total energies include zero-point corrections. The vibrational frequencies are calculated at the harmonic approximation level and are not scaled. The focus of this study was on the thermodynamics of the degradation reactions. The transition states and kinetics were not considered and therefore, there can be large energy barriers between reactants and products. However, the thermodynamics of the degradation reactions enable the identifying the energetically most favorable reaction paths. 3. Results and discussion 3.1. Geometry of the models

ðE þ HCORR Þ ð1Þ

reactants

Fig. 1. The dimer unit (PC-2) of polycarbonate.

The choice of the calculation method did not significantly affect the geometry of the PC-2 molecule. Therefore, the HF/3-21G method was found to be accurate enough for geometry optimizations. Instead, it failed in energy calculations. Fig. 2 presents all degradation reactions modeled with PC-2 as starting molecule. As for the starting molecule (PC –2) the choice of geometry optimization method did not significantly affect the geometries of the degradation products. The ˚ ) in the bond distance deviations were (0.001 – 0.03 A ˚ C– C bonds, (0.01 – 0.02 A) in the C –H bonds and ˚ ) in the C – O bonds and the deviations in (0.01 – 0.03 A

J. Katajisto et al. / Journal of Molecular Structure (Theochem) 634 (2003) 305–310

307

Fig. 2. The thermal degradation reactions of PC-2.

the bond angles were (1.0 – 7.0 8) at most. These results were also in agreement with experimental structures [13 – 14]. 3.2. Enthalpies and Gibbs energies for thermal degradation reactions of PC Table 1 shows the degradation reaction energies of the studied reaction routes. We note that HF-method

with basis set 6-31G* was sufficient enough in the determination of thermal enthalpy and Gibbs energy corrections (HCORR and GCORR). The choice of the calculation method did not significantly affect the thermal corrections. The predicted reaction enthalpies of the intramolecular exchange reaction (A) were positive in most cases. The most accurate calculation method (B3LYP/6-311G**) predicted that this reaction was

308

J. Katajisto et al. / Journal of Molecular Structure (Theochem) 634 (2003) 305–310

Table 1 Reaction energies for thermal degradation of PC Method

Basis set

Temperature (K)

Intramolecular exchange (kJ/mol)

Substitution with water (kJ/mol)

CO2 elimination (kJ/mol)

DrE8 DrE8 DrE8 DrE8

HF HF B3LYP B3LYP

3-21G 6-31G* 6-31G* 6-311G**

273.15 273.15 273.15 273.15

225.82 3.61 20.48 29.48

14.13 243.69 249.76 264.41

16.13 243.68 238.88 248.51

278.83 49.31 24.77 27.44

74.51 56.49 60.94 32.40

DrH8 DrH8 DrH8 DrH

HF B3LYP B3LYP B3LYP

6-31G* 6-31G* 6-311G** 6-311G**

273.15 273.15 273.15 673.15

27.73 9.14 18.14 14.44

248.19 254.26 268.91 269.65

250.68 245.88 255.51 257.57

23.41 21.14 1.53 26.51

44.41 48.87 20.32 21.57

DrG8 DrG8 DrG8 DrG

HF B3LYP B3LYP B3LYP

6-31G* 6-31G* 6-311G** 6-311G**

273.15 273.15 273.15 673.15

221.56 24.69 4.31 211.50

292.96 299.03 2113.68 2164.80

295.80 291.01 2100.63 2156.77

0.77 223.78 221.11 21.25

24.82 20.36 228.91 229.57

Dehydrogenation (kJ/mol)

C –H transfer (kJ/mol)

DrE8 ¼ Eproducts 2 Ereactants.

not favorable at standard state (273.15 K) but favorable at higher temperature (673.15 K). Since cyclic oligomers have been observed experimentally at higher temperatures, this result is in agreement with experimental findings [15 –22]. The CO2 elimination (C1) and substitution (B) reactions were both exothermal and favorable. The CO2 elimination and the substitution with water were the most exothermic degradation reactions and thus produced the thermodynamically most favored products. The observation is supported by mass spectrometry studies [2 – 3], in which high CO2 concentrations have been detected. The CO2 elimination produces an intermediate product, which can undergo dehydrogenation reaction. The DH8 values were close to zero and the calculated Gibbs energies, DG8, indicated that reaction step (C2) was also favorable. Disproportionation (C – H transfer) (D) reaction produced two molecular fragments as products (Fig. 2). All methods with different basis sets indicated that this reaction is thermodynamically favorable and endothermic. Only exception was B3LYP/6-311G** method, which indicated that this reaction was slightly exothermal at higher temperature. The HF/3-21G method, also used in this study, gave opposite results to the other more accurate methods and it was considered insufficient for energy considerations.

3.3. Enthalpies and Gibbs energies for thermal isomerization of PC Thermal isomerization was different from the other degradation reactions because it can proceed along two main paths. The final product is an oligomer containing xanthone or fluoronone unit. Fig. 3 shows reaction steps and reaction products. The calculated reaction enthalpies and Gibbs energies are listed in Table 2. The first main degradation path (a), which produces only an oligomer with xanthone unit, comprises two steps. The more accurate methods B3LYP/6-31G* and B3LYP/6-311G** indicated that the first step (I) and second step (II) are exothermal and thermodynamically favored. In the second main reaction path (b) the first two reaction steps (III and IV) are endothermic and not favorable. All calculation methods gave similar results excluding B3LYP/6-311G**, which suggested that the reaction step (IV) was exothermal and favorable. The final reaction step can proceed in two ways. Cleavage of water (V) produces a xanthone unit, whereas reaction with hydrogen and simultaneous cleavage of water (VI) produces a fluoronone unit. Both steps were exothermal and favorable regardless of the calculation method and all calculation methods predicted that the formation of the fluoronone unit was more favorable

J. Katajisto et al. / Journal of Molecular Structure (Theochem) 634 (2003) 305–310

309

Fig. 3. Thermal isomerization of PC-2.

than the formation of the xanthone unit. The nonfavorable reaction step (III) explained the failure to detect the fluoronone unit under 500 8C in experimental data [23 –24]. At higher temperature the effect of entropy in degradation reactions increased. The calculated Gibbs

energies at the high temperature varied from 1.17 kJ/mol (thermal isomerization, reaction step (I)) to 69.34 kJ/mol (thermal isomerization, reaction step (IV)) compared to the Gibbs energies at standard state. The largest Gibbs energy deviations from the standard state were observed in degradation reactions, in which

Table 2 Reaction energies for thermal isomerization of PC Method

Basis set

Temperature (K)

I (kJ/mol)

DrE8 DrE8 DrE8

HF B3LYP B3LYP

6-31G* 6-31G* 6-311G**

273.15 273.15 273.15

7.39 22.44 22.76

DrH8 DrH8 DrH8 DrH

HF B3LYP B3LYP B3LYP

6-31G* 6-31G* 6-311G** 6-311G**

273.15 273.15 273.15 673.15

DrG8 DrG8 DrG8 DrG

HF B3LYP B3LYP B3LYP

6-31G* 6-31G* 6-311G** 6-311G**

273.15 273.15 273.15 673.15

DrE8 ¼ Eproducts 2 Ereactants.

II (kJ/mol)

III (kJ/mol)

IV (kJ/mol)

V (kJ/mol)

VI (kJ/mol)

8.18 11.27 5.52

46.75 32.02 31.95

21.81 0.82 26.68

252.99 224.02 222.50

298.98 234.53 250.57

8.81 21.02 21.33 20.23

20.93 2.15 23.59 28.40

46.27 31.54 31.47 33.47

22.54 1.54 25.96 24.60

260.92 231.95 230.43 237.49

289.44 224.99 241.02 255.19

9.46 20.37 20.69 20.48

234.55 231.46 237.21 282.75

49.77 35.04 34.97 38.27

25.06 4.07 23.43 20.98

299.92 270.95 269.43 2120.51

2141.37 276.92 292.95 2162.29

310

J. Katajisto et al. / Journal of Molecular Structure (Theochem) 634 (2003) 305–310

the entropy changes were largest. These reactions were substitution with water, CO2 elimination and reaction steps II, V and VI in thermal isomerization. In addition, the intramolecular exchange reaction became favorable at higher temperature.

4. Conclusions In the present work, the CO2 elimination and substitution with water were thermodynamically the most favorable reaction paths in the thermal degradation of PC, but intramolecular exchange was also possible. In the thermal isomerization reaction of PC the calculations indicated that the formation of fluoronone unit is more favorable than the formation of xanthone unit but the formation of xanthone unit is still significant due to the spontaneity of the first main reaction path (a).

References [1] G. Ka¨mpf, D. Freitag, G. Fengler, Kunststoffe 82 (1992) 385. [2] C. Puglisi, F. Samperi, S. Carroccio, G. Montaudo, Macromol. 32 (1999) 8821. [3] C. Puglisi, L. Sturiale, G. Montaudo, Macromol. 32 (1999) 2194. [4] L. Przybilla, H.J. Ra¨der, K. Mu¨llen, Eur. Mass. Spectrom. 5 (1999) 133. [5] X.G. Li, M.R. Huang, Polym. Int. 48 (1999) 387. [6] G. Montaudo, C. Puglisi, F. Samperi, Polym. Deg. and Stab. 26 (1989) 285. [7] H. Tagaya, K. Katoh, J. Kadokawa, K. Chiba, Polym. Deg. and Stab. 64 (1999) 289. [8] J. Lub, G.H.W. Buning, Polym. 31 (1990) 1009.

[9] D.J. Brunelle, T.L. Evan, T.G. Shannon, E.P. Boden, K.R. Steward, L.P. Fontana, D.K. Bonauto, J. Am. Chem. Soc. 30 (1989) 569. [10] V. Van Speybroeck, Y. Martele, M. Waroquier, E. Schacht, J. Am. Chem. Soc. 123 (2001) 10650. [11] Gaussian 98, Revision A.7. M.J. Frisch, G.W. Trucks, H.B. Schlegel. G.E. Scuseria, M.A. Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery. Jr. R.E. Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, A.G. Baboul, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, J.L. Andres, C. Gozalez, M. Head-Gordon, E.S. Replogle, J.A. Pople, Gaussian. Inc., Pittsburgh PA, 1998. [12] Sybyl þ , version 6.2, Tripos, Inc., USA. [13] B. Montanari, P. Ballone, R.O. Jones, Macromol. 32 (1999) 3396. [14] B. Montanari, P. Ballone, R.O. Jones, O. Hahn, J. Phys. Chem. A. 103 (1999) 5387. [15] P. Ballone, B. Montanari, R.O. Jones, J. Phys. Chem. A. 104 (2000) 2793. [16] P. Ballone, R.O. Jones, J. Phys. Chem. A. 105 (2001) 3008. [17] J. Akola, P. Ballone, R.O. Jones, Macromol. 35 (2002) 2327. [18] P. Ballone, R.O. Jones, J. Chem. Phys. 117 (2002) 6841. [19] P. Ballone, R.O. Jones, J. Chem. Phys. 116 (2002) 7724. [20] P. Ballone, R.O. Jones, J. Chem. Phys. 115 (2001) 3895. [21] A.G. Shaikh, S. Sivaram, C. Puglisi, F. Samperi, G. Montaudo, Polym. Bullet. 32 (1994) 427. [22] E.C. Aquino, Kinetics and mechanism of cyclic oligomeric carbonate formation, Academic Dissertation, The university of Akron, 2000, p. 3. [23] K. Oba, H. Ohtani, S. Tsuge, Polym. Deg. and Stab. 74 (2001) 172. [24] C. Puglisi, L. Sturiale, G. Montaudo, Macromol. 32 (1999) 2198.