montmorillonite nanocomposites

montmorillonite nanocomposites

Polymer Degradation and Stability 94 (2009) 1548–1557 Contents lists available at ScienceDirect Polymer Degradation and Stability journal homepage: ...

726KB Sizes 73 Downloads 141 Views

Polymer Degradation and Stability 94 (2009) 1548–1557

Contents lists available at ScienceDirect

Polymer Degradation and Stability journal homepage: www.elsevier.com/locate/polydegstab

The thermal degradation behaviour of polydimethylsiloxane/montmorillonite nanocomposites James P. Lewicki a, *, John J. Liggat a, Mogon Patel b a b

Department of Pure and Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow G1 1XL, UK Atomic Weapons Establishment, Aldermaston, Reading RG7 4PR, UK

a r t i c l e i n f o

a b s t r a c t

Article history: Received 27 August 2007 Received in revised form 8 March 2009 Accepted 14 April 2009 Available online 8 May 2009

A series of novel polydimethylsiloxane/montmorillonite (PDMS/MMT) nanocomposites was prepared. The thermal degradation behaviour of these nanocomposites was studied by means of Thermal Volatilization Analysis (TVA) and Thermogravimetric Analysis (TGA). The major degradation products were identified as cyclic oligomeric siloxanes from D3 to D7, and higher oligomeric siloxane residues. Other minor degradation products include methane, bis-pentamethylcyclotrisiloxane, propene, propanal, benzene and dimethylsilanone. The results demonstrate that the nanoclay significantly alters the degradation behaviour of the PDMS network, modifying the profile of the thermal degradation and reducing the overall rate of volatiles evolution. The results also indicate that the nanoclay promotes the formation of dimethylsilanone and benzene by inducing low levels of radical chain scission. Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Poly(dimethylsiloxane) Nanocomposite Thermal degradation Montmorillonite

1. Introduction 1.1. Polysiloxane nanocomposites Polysiloxanes are arguably the most important of a class of noncarbon backboned polymers, having applications in many diverse areas from biomedical implant technology to their use as high temperature lubricants. Key features of polysiloxanes include a large degree of main-chain flexibility, high thermal stability and low electrical conductivity. The most commercially important polysiloxane today remains poly(dimethylsiloxane) (PDMS) which is the simplest of the polysiloxanes; having a Si–O backbone and two methyl substituents per silicon atom. Replacement of a methyl substituent on the repeating unit with a phenyl group results in poly(methylphenylsiloxane) (PMPS). PMPS has enhanced thermal and thermo-oxidative stability and is generally used in high temperature applications. Simple polysiloxanes are typically room temperature liquids and as such are often crosslinked by condensation or radical addition chemistry to form elastomers. Polysiloxane elastomers typically exhibit poor mechanical strength in their native states and are often reinforced by the addition of a heterogeneous filler material. Historically, polysiloxane elastomers have been reinforced with micron scale particles such as amorphous inorganic silica to * Corresponding author. Tel.: þ44 141 548 2269; fax: þ44 141 548 4822. E-mail addresses: [email protected] (J.P. Lewicki), [email protected] (J.J. Liggat), [email protected] (M. Patel). 0141-3910/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.polymdegradstab.2009.04.030

form polysiloxane microcomposites. However, with the continued growth of new fields such as soft nanolithography, flexible polymer electronics and biomedical implant technology, there is an ever increasing demand for polysiloxane materials with better defined, improved and novel physical, chemical and mechanical properties. In line with these trends, researchers have turned towards the development of polysiloxane nanocomposites; systems which incorporate a heterogeneous second phase on the nanometre scale. Over the last decade, there has been much interest in polymeric nanocomposite materials and the reader is directed towards the reviews by Alexandre and Dubois [1] or Joshi and Bhupendra [2] on the subject. Examples of the synthesis of polysiloxane nanocomposites reported in the literature include: work by Ma et al. [3] who modified montmorillonite with short segments of PDMS and blended this into a polymer melt/solution to yield examples of fully exfoliated or intercalated PDMS/clay nanocomposites. Pan, Mark et al. [4] synthesized well defined nano-fillers by reacting groups of four vinyl terminated POSS cages with a central siloxane core. These materials were subsequently chemically bonded into a PDMS network yielding a significant improvement in the mechanical properties of the polymer. 1.2. The thermal degradation of polysiloxanes Since the introduction of PDMS and its relatives as commercial polymeric materials some 50 years ago, there has been interest in

J.P. Lewicki et al. / Polymer Degradation and Stability 94 (2009) 1548–1557

their thermal stability and the mechanisms by which PDMS and related siloxane polymers degrade at elevated temperatures. The most influential work on the subject of the thermal degradation of PDMS was carried out by Grassie et al. [5] who studied in detail the thermal and thermo-oxidative degradation of pure PDMS using a combination of Thermal Volatilization Analysis (TVA) and Thermogravimetric Analysis (TGA). Building upon earlier work by Pantode and Wilcock [6] and Thomas and Kendrick [7], Grassie and Macfarlane demonstrated that PDMS degrades via a depolymerisation reaction to yield cyclic oligomeric siloxanes. The depolymerisation proceeds from both free chain ends and as a result of intra-molecular backbiting reactions of continuous chain segments (see Fig. 1). Grassie and Macfarlane reported the major degradation products to be cyclic oligomeric siloxanes of ring sizes D3–D12 and higher oligomeric siloxane species. Lewis acids or bases were found to accelerate the depolymerisation reaction and lower the degradation temperature of the PDMS significantly. Methane was reported as a minor product when PDMS was degraded in the presence of a base such as KOH. The presence of oxygen was also reported to accelerate the degradation of PDMS and lead to significant crosslinking of the degradation products. Grassie and Macfarlane concluded that the degradation of pure PDMS is an equilibrium controlled depolymerisation reaction resulting in a high yield of monomer. More recently, Camino et al. [8] have made use of TGA in combination with kinetic analysis and computer simulation methods to study the thermal degradation of low crosslink density tri-methyl siloxy end-blocked PDMS. By studying the thermal degradation of PDMS at a range of heating rates, Camino et al. found that the rate of degradation is limited primarily by the diffusion of volatile cyclic oligomers from the system. Kinetic treatment of the TG data demonstrated that the thermal degradation of PDMS is a diffusion limited process which follows first order Arrhenius behaviour, having an activation energy of w27 kcal mol1. Molecular modelling studies of the primary degradation mechanism of PDMS (the intra-molecular backbiting reaction to form the D3 cyclic trimer) has shown that the cyclisation is energetically favoured due to the overlap of empty silicon d-orbitals with the orbitals of oxygen and carbon on

1549

a neighbouring chain segment. This interaction stabilizes the chain folded conformation necessary for the cyclisation reaction to take place. Today, the thermal degradation chemistry of comparatively simple siloxane polymer systems has been well studied and is relatively well understood. However, many commercial siloxane polymer systems are significantly more complex than the simple PDMS model systems generally studied. Commercial siloxane polymers are often heavily filled, crosslinked and chemically modified and as such their degradation behaviour is more complex and less well understood. This situation is most evident with siloxane nanocomposite materials which are currently being developed as the next generation of commercial polysiloxane systems. Currently, little is known about how the thermal degradation behaviour of these novel siloxane systems compares to current models of siloxane degradation. In this work the influence of a well known nano-filler; organically modified montmorillonite (MMT) clay on the thermal degradation behaviour of a crosslinked PDMS network is reported.

2. Experimental 2.1. Materials The following materials were purchased from the Aldrich Chemical Co. and used in this investigation: Two hydroxy-terminated polydimethylsiloxanes (PDMS), (Mn w550 g mol1 and w77000 g mol1 respectively), tetrapropylorthosilicate (TPOS) crosslinker and tin(II) 2-ethylhexanoate catalyst. Diphenylmethylsilanol (DPMS) chain terminator was obtained from the UK Atomic Weapons Establishment, Cloisite 6A, an organically modified montmorillonite clay was purchased from Southern Clay Products. In Cloisite 6A, a proportion of the native Naþ guest cations has been substituted for quaternary alkyl ammonium cations (of molecular formula [(CH3)2N(HT)2]þ, where HT is Hydrogenated Tallow ¼ 65% C18, 30% C16 and 5% C14) in order to increase the interlayer gallery spacing between platelets and hydrophobic nature of the clay. The modifier concentration of Cloisite 6A is 140 meq per 100 g of clay.

Fig. 1. Mechanisms of PDMS thermal depolymerisation.

1550

J.P. Lewicki et al. / Polymer Degradation and Stability 94 (2009) 1548–1557

2.2. Thermal Volatilization Analysis

2.4. Ultrasonic processing of PDMS/MMT pre-polymer blends

All TVA analysis was carried out using a TVA line which was built in-house, based upon the apparatus and techniques described by McNeill et al. [9] The apparatus consisted of a sample chamber (heated by a programmable tube furnace) connected in series to a primary liquid nitrogen cooled subambient trap and a set of four secondary liquid nitrogen cooled cold traps. The whole system was continuously pumped to a vacuum of 1  104 torr by means of a two stage rotary pump and oil diffusion pumping system. Volatile condensable products could be initially trapped at two stages: The water jacket cooled ‘cold-ring’ (T w 12  C) immediately above the heated area of sample tube (this condensed high boiling point materials) and the primary liquid nitrogen cooled sub-ambient trap (T w 196  C) which collected all the lower boiling point species. Two linear response Pirani gauges were positioned at the entrance and exit of the primary sub-ambient trap to monitor the evolution of both condensable and non-condensable volatiles as a function of pressure vs. temperature/time from the sample. The use of linear response Pirani gauges allows valid pressure peak integrations to be carried out; where peak area corresponds to the quantity of evolved volatiles. Trapped, lowboiling species could be distilled into separate secondary cold traps by slowly heating the primary sub-ambient trap to ambient temperatures. These separated fractions could be subsequently removed into gas-phase cells for FTIR and GC–MS analysis. A series of non-linear Pirani gauges were placed at the entrance and exits of all secondary fraction traps to monitor the pressure changes as volatile species were distilled into separate traps and gas cells. All TVA runs were conducted under vacuum using 25 mg samples of each model system. The heating rate was 20  C min1 to a temperature 550  C. A 1–200 amu VG single quadrapole Q300D mass spectrometer sampled a continuous product stream during both the degradation and the differential distillation runs. Subambient differential distillation of collected volatiles was carried out by heating the primary sub-ambient trap at a rate of 4  C min1 from 196 to 40  C. Volatiles were separated into four major fractions for subsequent IR and GC–MS analysis. A significant coldring fraction was also collected for each sample. All FTIR analysis of the collected TVA products was carried out using a Mattson 5000 FTIR Spectrometer used in transmission mode. High boiling ‘cold-ring’ fractions were cast from chloroform solution onto NaCl disks for analysis. Low-boiling volatiles were analyzed in the gas phase using gas phase cells with NaCl windows. All GC–MS analysis of the collected TVA products was carried out using a Finnigan ThermoQuest capillary column trace GC and Finnigan Polaris Quadrapole Mass Spectrometer. Suitable fractions were dissolved in chloroform and subsequently analyzed.

Dispersion and exfoliation of the montmorillonite clay platelets within the pre-polymer resin were achieved by means of low frequency ultrasonic processing (sonication). The apparatus used was a Cole-Palmer Ultrasonic Processor equipped with a 4 mm titanium probe. Ultrasonic processing of montmorillonite clay in a pre-polymer resin has been shown to be an effective method of achieving exfoliation [10] of the system. Ultrasonic processing effectively disrupts the initial tactiod structure of the clay, overcoming the electrostatic interactions which hold individual platelets in a regular stacked formation. This allows the individual platelets to be dispersed throughout the pre-polymer matrix forming a nanometre scale dispersion. Confirmation of the formation of a nanodispersion of Cloisite particles within these PDMS matrices, obtained through X-ray diffraction and transmission electron microscopy is reported in our related paper on the thermal ageing of PDMS nanocomposites [11]. Experimentally it has been found that the dispersion achieved falls between an intercalated and a fully exfoliated state.

2.3. Thermogravimetric analysis All TGA analysis was carried out using a Shimadzu TGA-51 Thermogravimetric analyzer, with 25 mg samples under a purge flow of 20 ml min1 nitrogen at a heating rate of 10  C min1 from 25 to 800  C. In order to better define onset temperatures for nonoxidative degradation, the first derivative of the sample mass was calculated as a function of temperature. A derivative mass value of 0.005 mg/ C was defined as a standard value to indicate the onset of significant mass loss/degradation. Individual onset degradation temperatures were then defined from the intercept of a theoretical y-axis line at this value with each derivative plot. This method yielded a repeatable, comparative method of determining onset degradation temperatures.

2.5. Preparation of siloxane–montmorillonite nanocomposites A series of five model elastomer systems were prepared incorporating 0, 1, 2, 4 and 8% (on total resin mass) of Cloisite 6A. Each variant was prepared in the following manner: 14 g of PDMS (Mn w 77,000 g mol1) was mixed with 6 g of PDMS (Mn w 550 g mol1) by an IKA-Werke overhead stirrer. To this, the appropriate mass of Cloisite was added and mixed mechanically for twenty minutes. This blend was then sonicated under a 9 s pulsed regime at an intensity of 35% for twenty minutes in order to exfoliate the clay platelets. The resin blend was degassed and cooled to 2  C. 5% (on resin mass) of DPMS and a stoichiometric quantity of TPOS (1.748 g) was mechanically mixed into the blend. Finally, 5% of tin(II) 2-ethylhexanoate catalyst was mixed into the blend and the mixture was transferred into a 10 cm2 mould. This was cured in an oven at 65  C for twenty minutes. The elastomer that was formed was removed from the mould and post-cured for a further fifteen hours at 65  C. 3. Results and discussion 3.1. TGA of PDMS/Cloisite model systems TGA was employed in this study to determine if the inclusion of a nano-scale dispersion of Cloisite into a model PDMS elastomer matrix affected the overall non-oxidative thermal stability of the elastomer matrix. Fig. 2 is a composite plot of the TGA mass loss curves obtained from the 0, 1, 2, 4 and 8% Cloisite filled systems. It can be clearly observed from Fig. 2 that the introduction of Cloisite into the elastomer lowers the non-oxidative degradation temperature of the model systems significantly. Analysis of the derivative of each mass loss curve allowed onset degradation temperatures to be defined (see Fig. 3). The inclusion of as little as 1% Cloisite 6A into the matrix was found to lower the non-oxidative degradation temperature by w17  C and as Fig. 3 illustrates, increasing the mass fraction of Cloisite in the matrix leads to further decreases in onset degradation temperature. Interestingly however, it can also be observed from Fig. 2 that at loadings above 1%, Cloisite appears to reducing the rate of mass loss in the degradation region between 400 and 500  C, suggesting that the rate of volatiles evolution is being altered by the nanoclay.

J.P. Lewicki et al. / Polymer Degradation and Stability 94 (2009) 1548–1557

Fig. 2. TGA mass loss curves for PDMS/Cloisite model elastomer systems. Squares, circles, up-triangles, down-triangles and diamonds represent 0, 1, 2, 4 and 8% filled systems respectively. (Note – symbols are not representative of total number of data points).

Overall, the TGA data has shown that the inclusion of a Cloisite nano-dispersion into the PDMS matrix has a negative impact on its non-oxidative stability; however it is also reducing the rate of volatiles evolution once thermal degradation has commenced. 3.2. TVA analysis of PDMS/Cloisite model systems The thermal degradation of the model elastomer systems under high vacuum in each case produced a significant quantity of condensable volatile species, a small quantity of a non-condensable gas (which was detected at temperatures above 400  C and identified by online mass spectrometry as methane) and a significant cold-ring fraction. The TVA plots showing the rate of volatiles evolution vs. temperature for each model system are presented in Fig. 4. Table 1 summarises the onset degradation temperatures and evolution rate peak maxima for the model systems. The relative quantities of condensable volatiles have also been calculated by integrating the total peak areas in Fig. 4. This data is also given in Table 1. From Fig. 4 and Table 1 it can be observed that the unfilled system degrades in a single step process with the rate of volatiles evolution reaching a maximum at 418  C. Significant onset of volatiles evolution commences at w380  C which corresponds

1551

Fig. 4. TVA plots of the degradation of the model elastomer systems showing the rate of volatiles evolution as a function of pressure vs. the furnace temperature. Solid line represents the unfilled elastomer, dashed line; the 1% system, dotted line; the 2% system, square symbols þ line; the 4% system and circles þ line; the 8% system. Heavy black line represents the furnace temperature vs. time.

closely with the onset degradation temperature derived from the TGA curves. The TVA plot of the unfilled system showing a single stage degradation process is in agreement with what is reported in the literature for PDMS polymers [5]. The Cloisite filled systems exhibit markedly altered degradation behaviour. As with the TGA results, the TVA data clearly shows that the onset degradation temperature of the elastomer is lowered by the presence of Cloisite. The TVA data in Fig. 4 also shows that the maximum rate of volatiles evolution occurs at lower temperatures in the Cloisite filled systems. For example; there is a 9  C decrease in the temperature at which the maximum rate of volatiles evolution occurs in the 1% Cloisite filled elastomer. Just as significantly however, the overall shape of the degradation profiles are altered appreciably in the Cloisite filled systems. The 1 and 2% filled systems evolve a greater overall quantity of volatiles (see Table 1), however the 4% system shows a significantly different rate profile and 8% system shows a reduced overall level and change in the rate of volatiles evolution throughout the degradation. From Fig. 4 it can be observed that as the level of Cloisite is increased, the degradation changes from a single stage process into a more complex event. Low levels of Cloisite effectively shift the degradation to lower temperatures and increase the production of volatiles; suggesting that at low levels Cloisite makes the PDMS more labile to thermal depolymerisation. However, as the level of Cloisite is increased there is an overall decrease in volatiles evolution, a reduction in the primary degradation peak intensity and the introduction of a broad tail to this peak, extending to above 500  C.

Table 1 Onset degradation temperatures, maximum volatile evolution rate peak temperatures and overall level of evolved volatiles – determined from peak integrals. Data taken from Fig. 1: TVA plots of the degradation of the model elastomer systems.

Fig. 3. Onset degradation temperature vs. mass fraction of Cloisite present in the elastomer matrix. (As determined from 1st derivative of individual mass loss curves).

Model elastomer

Onset degradation temperature/ C

Max. evolution rate: Peak temperature/ C

! P(T)/ torrs

Unfilled elastomer 1% Cloisite 2% Cloisite 4% Cloisite 8% Cloisite

380 362 360 354 367

418 409 400 382 373

1.31 1.65 1.86 2.02 1.22

1552

J.P. Lewicki et al. / Polymer Degradation and Stability 94 (2009) 1548–1557

3.3. TVA characterisation of the extent of cure It is clear from the TGA and TVA data that the inclusion of Cloisite into the PDMS elastomer matrix significantly alters the non-oxidative degradation behaviour of these systems. What is not clear from the data presented so far is however, if the Cloisite is acting chemically within the system as a pro-degradant or whether the inclusion of Cloisite has inhibited curing of the system - leading to a less densely crosslinked network with residual free silanol chain ends. In order to determine if Cloisite does indeed inhibit the crosslinking reaction, two further resin blend samples were prepared: one remained unfilled and the other contained a 4% loading of Cloisite. These blends were prepared in an identical manner to the main sample group with the exception that 500 mg samples of each were transferred to sealed TVA sample tubes immediately after the inclusion of the catalyst. These sealed samples were allowed to cure in the sealed environment of the sample tubes for 20 minutes at 65  C. Each sample tube was then connected to the TVA system and the volatile contents collected under vacuum for 90 minutes with the sample held at 65 C. After this period of collection, the condensed volatiles were characterised by differential distillation. The levels of propanol present from each sample were characterised directly using the online mass spectrometer, operating in a multiple ion monitoring (MIM) mode. The levels of propanol evolved from each sample are shown in Fig. 5 From Fig. 5 it can be observed that the 4% Cloisite filled sample does not evolve a decreased level of propanol when compared to the unfilled elastomer. In fact, under the conditions of the test, the 4% Cloisite filled sample evolved a slightly increased level of propanol over the unfilled system. The analysis of effective cure as a function of propanol evolution suggests strongly that Cloisite does not inhibit the crosslinking reaction. The reduction in thermal stability of the Cloisite filled systems cannot therefore be reasonably attributed to an accompanying decrease in crosslink density and increase in residual silanol groups These data behaviour would seem to suggest that chemically, Cloisite can act as a pro-degradant within the PDMS matrix. It is possible that the acid sites on the clay platelets or a breakdown product of the quaternary ammonium modifier are promoting the scission of Si–O bonds, making the depolymerisation reaction more labile. It is also clear however, that this trend is reversed at higher

Fig. 5. Comparative MIM-MS plots of propanol evolution from an unfilled and a 4% filled PDMS elastomer systems during cure. It can be observed that the inclusion of 4% Cloisite 6A does not reduce the overall quantity of propanol evolved - a marker of the curing reaction. Solid black and dashed lines represent the unfilled and filled systems respectively.

Fig. 6. Composite differential distillation plots of 0, 1, 2, 4, and 8% Cloisite filled systems showing minor products over the range of 180 to 50  C: Solid line represents the unfilled elastomer, dashed line; the 1% system, dotted line; the 2% system, square symbols þ line; the 4% system and circles þ line; the 8% system.

clay loadings: It is reasonable to assume that this reversal is due to the physical barrier properties of the clay platelets becoming significant. PDMS degradation in the bulk is a diffusion controlled process and the action of Cloisite here is likely to be inhibiting the evolution of volatiles, thereby reducing the rate of volatiles evolution and allowing time for volatile depolymerisation products to undergo some degree of condensed phase action.

3.4. Differential distillation and characterisation of collected degradation products Collected volatiles for all model elastomer systems were separated by sub-ambient distillation. The differential distillation plots for each system are shown in Figs. 6 and 7. The data has been divided into two separate plots so that both the major and the minor product distributions can be evaluated separately. Individual peaks represent discrete components of the total volume of collected volatile species.

Fig. 7. Composite differential distillation plots of 0, 1, 2, 4, and 8% Cloisite filled systems over the range of 180 to 40  C, scaled to show major products: Solid line represents the unfilled elastomer, dashed line; the 1% system, dotted line; the 2% system, square symbols þ line; the 4% system and circles þ line; the 8% system.

J.P. Lewicki et al. / Polymer Degradation and Stability 94 (2009) 1548–1557

1553

Fig. 8. GC-MS Chromatogram of product peak 8 from the 2% Cloisite filled system showing D3, D4 and D5 cyclic oligomeric siloxanes in decreasing abundance.

In Figs. 6 and 7 the complete range of collected volatile species are shown. Each separated component peak has been numbered. Peaks 1–7 in Fig. 6 are the minor fraction of degradation products. These minor products were identified by online mass spectrometry as: (1) – propene and CO2, (2) – butene, (3) – the enolate form of dimethylsilanone, (4) – an unidentified silicone compound, (5) – propanal, (6) – benzene and (7) – an unknown (possibly a carboxylic acid). Peaks 8 and 9 in Fig. 7 represent the major volatile products of degradation. Identification of all products has been achieved through a combination of FTIR spectroscopy, online mass spectrometry and GC–MS. Product peak 8 has been identified as consisting, primarily of D3, D4 and D5 cyclic oligomeric siloxanes (with the D3 cyclic being w3 times more abundant than the D4 cyclic in all of the Cloisite filled systems). Bis-pentamethylcyclotrisiloxane (BPS) has also been identified as a minor component of

peak 8 and bis(pentamethylcyclotrisiloxy)tetramethyldisiloxane (BPSTS) has been identified in peak 9. The GC–MS trace of peak 8 for the 2% Cloisite filled system is shown in Fig. 8. As peak 8 is the largest product peak, The D3 cyclic, hexamethylcyclotrisiloxane is therefore the major volatile product of the degradation. This is significant as it indicates that the fully crosslinked, filled siloxane systems still degrade in a manner that is similar to that of simple PDMS, yielding the hexamethylcyclotrisiloxane monomer as the major thermodynamic depolymerisation product. Peak 9 was determined to consist primarily of D6 and D7 cyclics, silanoate esters and aromatic di-phenyl residues. The GC–MS chromatogram of peak 9 is shown in Fig. 9. GC–MS chromatograms of peaks 8 and 9 were obtained for all model systems however the product distributions were found to be nearly identical in each case,

Fig. 9. GC-MS Chromatogram of product peak 9 from the 2% Cloisite filled system. All significant identified products are labelled accordingly.

1554

J.P. Lewicki et al. / Polymer Degradation and Stability 94 (2009) 1548–1557

Fig. 10. Mechanism of BPSTS formation. Two molecules of methane are yielded for every BPSTS molecule.

suggesting that Cloisite is not having a significant impact on mechanisms of the major degradation reactions. BPS was observed to be a minor degradation product in peak 8 for all model systems. BPS was first reported as a degradation product of PDMS by Grassie and co-workers [5]. Grassie reported BPS as a degradation product in PDMS systems that contained a Lewis base such as KOH and evolved methane as a minor degradation product. On the basis of this, Grassie proposed a mechanism to account for both methane and BPS production in the presence of KOH and it is likely that a similar reaction is taking place in the model elastomer systems to form both BPS and BPSTS. This reaction is summarised in Fig. 10 and shows a likely route to BPSTS formation. It is proposed that tin hydroxide, present in the elastomers as a catalyst residue is providing the OH ions in this reaction: Tin oxide is the final hydrolysis product of the tin(II) 2-ethylhexanoate catalyst used in the synthesis of these systems and is has been established previously by Patel et al. [12] that tin(IV) in the form of SnO2 is present in such room temperature vulcanised (RTV) PDMS elastomers. It is thought that tin oxide within the elastomer system reacts with residual water present in the matrix or water from the surroundings to form tin hydroxide; which acts as source of hydroxyl ions.

Fig. 11. Thin film IR spectra of collected cold-ring fractions. Lines numbered 1 to 5 correspond to the cold-rings from 0, 1, 2, 4 and 8% Cloisite filled elastomers respectively.

J.P. Lewicki et al. / Polymer Degradation and Stability 94 (2009) 1548–1557

1555

Fig. 12. Benzene formation through thermally induced radical cleavage and hydrogen abstraction.

Fig. 13. Mass spectrum of the enolate ion form of di-methylsilanone having primary peaks at 73, 45 and 43 amu corresponding to the parent ion, the Si–O and the SiCH3 fragments.

2-Ethylhexanoic acid, another by-product from the hydrolysis of tin(II) 2-ethylhexanoate catalyst is also known to remain within RTV cured systems such as these after formulation [11,13]. It is thought that this free acid within the model systems becomes reactive during thermal degradation; as evidenced by the detection of alkyl silanoate esters within peak 9 in all systems (see Fig. 9). It is known from the literature [11,14,15] that 2-ethylhexanoic acid acts as a pro-degradant within PDMS elastomer systems at moderate temperatures (~150–250  C); the acid attacks the relatively polar Si–O bond making inter- or intra-molecular backbiting reactions more favourable. 2-Ethylhexanoic acid is therefore likely to be contributing to the degradation of the model systems in the initial stages of the thermal degradation process - however at high temperatures >300  C, The carboxylic acid will its-self decompose. While acid and base catalysed scission reactions cannot occur simultaneously, it is possible that acid catalysed degradation occurs at moderate temperatures (<300  C) and tin oxide acts as a Lewis base to catalyse the formation of BPS and BPSTS in the temperature region between 400 and 550  C. In both peaks 8 and 9 there are also smaller, yet significant quantities of more complex, branched siloxane species. These have

Fig. 14. Elimination of dimethylsilanone from polydimethylsiloxane.

1556

J.P. Lewicki et al. / Polymer Degradation and Stability 94 (2009) 1548–1557

proved difficult to identify explicitly but suggest that a complex series of side reactions are occurring in addition to the main-chain depolymerisation process. These more complex branched species are likely to be the result of the degradation of crosslinked portions of the matrix, which result in the formation of branched siloxane species. The cold-ring fractions collected from all model systems were analyzed by IR spectroscopy and size exclusion chromatography (SEC). It was determined by SEC that the cold-ring fractions consisted of a relatively broad molecular weight range of higher oligomeric siloxanes with an Mn of w6500 g mol1 and a polydispersity index, Mw/Mn ¼ 1.88. Fig. 11 shows the IR spectra of the cold-ring fractions for each model system. From the IR spectra in Fig. 11 it can clearly be observed that the cold-ring fractions consist primarily of siloxane oligomers having characteristic peaks at w2970, 1265, 1090 and 1033 cm1. Again, higher siloxane oligomers have been reported as major products of the degradation of simple polysiloxanes. There is no observable silanol content from the IR data suggesting that the oligomeric fraction does not simply consist of un-reacted starting materials. There is also a clear indication of di-phenyl silanol in the cold-ring (aromatic peaks at 3071 and 3052 cm1 in Fig. 11) suggesting that a proportion of the end-cap has detached from the network relatively intact or has in fact failed to react completely during cure. Analysis of volatile peaks 8, 9 and the cold-ring fractions of the degraded Cloisite model elastomer systems have demonstrated that the major high boiling components of the degradation of these nanocomposite elastomers are very similar to that of simple PDMS systems - suggesting that the primary degradation of the model PDMS elastomers proceeds through the same thermal de-polymerisation process that linear PDMS polymers undergo. There is however, also a clear indication from Fig. 6 that the distributions of the minor, low boiling products are influenced by the Cloisite nanoclay. Propene and propanal (peaks 1 and 5) are thought to be oxidation products of residual propanol in the elastomer matrix from the initial cure reaction. Interestingly, as the level of Cloisite is increased in the matrix, the levels of propene and propanal also increase. It is thought that this is due to Cloisite filled elastomers retaining more propanol than the unfilled systems. Once again due to the physical barrier effects of the clay. With increasing Cloisite levels, the quantity of benzene (peak 6) was observed to increase significantly. Benzene can only have originated from the thermal degradation of the di-phenylmethylsilanol end cap. A general mechanism for the formation of benzene is outlined in Fig. 12: Thermal cleavage of the Ar–Si bond at high temperatures (400–550  C) results in the formation of a benzene radical and an aryl silanol radical fragment. The benzene radical will readily abstract hydrogen from a convenient source such as the methyl pendants on PDMS; forming benzene. The remaining radicals can either re-combine with other radical species present or propagate further chain scission reactions. These data suggest that the presence of Cloisite promotes the formation of benzene and raises the question of why Cloisite promotes such a radical scission process. The transient species; dimethylsilanone was not detected as a degradation product of the unfilled elastomer however, its presence was inferred in all Cloisite filled systems from the detection of what is thought to be the enolate form of dimethylsilanone by online mass spectrometry. The levels of dimethylsilanone/enolate are observed to increase with increasing loadings of Cloisite. Shown in Fig. 13 is the mass spectrum of product peak 3 which has been assigned as the enolate form of dimethylsilanone, obtained during the thermal degradation of the 1% Cloisite filled elastomer. Dimethylsilanone formation in siloxane systems has been reported in the literature: Barton et al. [16] reported that

dimethylsilanone can be formed as an intermediate through the thermal decomposition of alkoxy-silanes and work by Khabashesku et al. [17] demonstrated that under certain conditions, dimethylsilanone can be formed from the vacuum pyrolysis of octamethylcy-clotetrasiloxane. The most likely source of dimethylsilanone in the model elastomers is from the radical scission of the siloxane main-chain at high temperatures (see Fig. 14): Radical scission of a section of the PDMS chain leads to the elimination of dimethylsilanone and the formation of two new radical chain ends which can either recombine or propagate further scission processes. It is postulated that the dimethylsilanone is in equilibrium with the enolate form. Like the formation of benzene, dimethylsilanone formation is promoted by the presence of Cloisite within the elastomer network. The mechanism by which Cloisite promotes the radical degradation of diphenylmethylsilanol and dimethylsilanone elimination is not clear, however it is speculated that at high temperatures (400–550  C) the nanoclay is acting to catalyse these thermal cleavage processes. What must also be made clear however, is that these are minor reactions that occur only at low levels during the thermal degradation of the model elastomer systems. Nonetheless, they are significant and point towards a previously unreported action of montmorillonite nanoclay within PDMS. 4. Conclusions This work has demonstrated clearly that the novel PDMS/MMT nanocomposite elastomers studied, thermally degrade primarily, though a thermal depolymerisation processes that closely correlates with the accepted model of the thermal degradation of linear PDMS: TVA analysis has demonstrated that the major products of the thermal degradation of the model elastomer systems are D3–D7 cyclic oligomeric siloxanes (with the D3 cyclic being the most abundant) and a series of higher oligomeric siloxane residues. The presence of Cloisite within the network does not affect the distributions of these major products. It is clear from the TGA and TVA data that Cloisite has a negative impact on the thermal stability of these PDMS elastomer systems. However the TVA data has also demonstrated that at clay loadings above 2% the physical barrier effects of the nanoclay in the degrading systems become significant and serve to reduce the rate of volatiles evolution and the quantity of volatile species formed. An analysis of the minor, low boiling degradation products has indicated that there are low levels of radical degradation chemistry occurring within the PDMS networks. Significantly these thermally induced radical scission processes are promoted by the presence of Cloisite within the networks and it is hypothesised that at high temperatures (>400  C) Cloisite may be actively catalysing these radical cleavage reactions. If Cloisite is indeed promoting the formation of benzene and dimethylsilanone through radical cleavages of the network, then the associated increase in the level of side radicals may be contributing to the reduction in thermal stability observed in the nanocomposite systems. References [1] Alexandre M, Dubois P. Materials Science and Engineering R-Reports 2000;28:1. [2] Joshi M, Bhupendra SB. Journal of Macromolecular Science Part C: Polymer Reviews 2004;44:389. [3] Ma J, Xu J, Ren J, Yu Z, Mai Y. Polymer 2003;44:4619. [4] Pan GR, Mark JE, Schaefer DW. Journal of Polymer Science Part B: Polymer Physics 2003;24:3314. [5] Grassie N, MacFarlane IG. European Polymer Journal 1978;14:875. [6] Patnode W, Wilcock DF. Journal of the American Chemical Society 1946;68:358.

J.P. Lewicki et al. / Polymer Degradation and Stability 94 (2009) 1548–1557 [7] Thomas TH, Kendrick J. Journal of Polymer Science Part A: Polymer Chemistry 1969;7:537. [8] Camino G, Lomakin SM, Lazzari M. Polymer 2001;42:2395. [9] McNeill IC, Ackerman L, Gupta SN, Zulfiquar M, Zulfiquar S. Journal of Polymer Science Part A: Polymer Chemistry 1977;15:2381. [10] McIntyre S, Kaltzakorta I, Liggat JJ, Pethrick RA, Rhoney I. Industrial Engineering Chemistry Research 2005;44:8573. [11] Lewicki JP, Liggat JJ, Patel M, Pethrick RA, Rhoney I. Polymer Degradation and Stability 2007;93:158.

1557

[12] Patel M, Skinner AR, Chaudhry A, Billingham NC, Mahieu B. Polymer Degradation and Stability 2004;83:157. [13] Patel M, Skinner AR, Maxwell RS. Polymer Testing 2005;24:663. [14] Patel M, Skinner AR. Polymer Degradation and Stability 2001;73:399. [15] Patel M, Soames M, Skinner AR, Thomas S. Polymer Degradation and Stability 2004;83:111. [16] Barton TJ, Bain S. Organometallics 1988;7:528. [17] Khabashesku VN, Kerzina ZA, Maltsev AK, Nefedov OM. Journal of Organometallic Chemistry 1989;364:301.