EPR spin labelling studies of molecular dynamics in elastomer-silica composites
Descripción
Res. Chem. Intermed., Vol. 28, No. 2,3, pp. 191– 204 (2002) Ó
VSP 2002. Also available online - www.vsppub.com
EPR spin labelling studies of molecular dynamics in elastomer-silica composites A. BUTTAFAVA 1 , G. M. GHISONI 1 , A. FAUCITANO 1;¤ , G. NEGRONI 2 , A. PRIOLA 3 , F. PEDITTO 3 , A. TURTURRO 4 and M. CASTELLANO 4 1 General Chemistry Department, University of Pavia, V. le Taramelli 12, 27100 Pavia, Italy 2 Pirelli Research Centre, Milan, Italy 3 Department of Material Science, Polytechnic of Turin, Turin, Italy 4 Department of Chemistry and Industrial Chemistry, University of Genova, Genova, Italy
Received 28 October 2001; accepted 20 December 2001 Abstract—Novel methods of nitroxyl spin labelling suitable for molecular dynamics studies within the interface regions of SBR elastomer/ silica composites have been developed and used together with the nitroxyl spin probe technique. Fast and slow motional components have been identi ed within the interface regions and the corresponding rotational diffusion tensors have been measured as a function of the temperature and the SiO2 concentration. The fast rotational frequency is found to be orders of magnitude slower than that measured in the absence of SiO2 . This difference is suggested to arise from a closer packing of the macromolecules near the silica surface caused by the van der Waals bonding interactions. Increase of the SiO 2 concentration results in a decrease of the molecular mobility. This effect has been imputed to the overlapping of the bonding interaction regions. Spin probe measurements in the SBR-SiO 2 matrices using TEMPO, strongly suggest that the hindrance to chain segmental motion induced by the SiO2 -SBR interactions propagates beyond the interface regions thus involving the bulk polymer matrix. It is suggested that the information on the segmental chain dynamics obtained through the spin labelling and spin probe measurements can be developed as a method for investigating the polymer/ ller interactions within the reinforcing mechanism by the ller. Keywords: EPR; spin label; rubber- ller composites; molecular dynamics.
INTRODUCTION
The mechanism of reinforcement of polymeric matrices by inorganic llers is a matter of active investigation by theoretical modelling and experiments. The major contributions to the viscoelastic behaviour of the elastomer- ller composites are believed to arise from the ller particle network, the ller-elastomer interactions, ¤
To whom correspondence should be addressed.
192
A. Buttafava et al.
hydrodynamic effects and the elastomer network; however, the relative importance of such mechanisms is at present unknown [1 – 4]. The aim of this work is to exploit the possibility of using molecular dynamics parameters as probes for detecting bonding interactions between SiO 2 particles and styrene – butadiene (SBR) copolymer chains and to investigate the ability of such interactions to extend their effects beyond the elastomer / ller interface. The molecular dynamics study is based on the EPR spectral analysis of nitroxyl type radicals either used as free molecules distributed throughout the whole matrix (spin probe technique) or bound to the silica particles (spin label technique). In the latter case, the method is suitable for exploring selectively the molecular dynamics at the ller /polymer interface.
EXPERIMENTAL
Sample preparation and materials Styrene – butadiene rubber (23.5% styrene) prepared by emulsion polymerisation and SiO2 (total surface area 160 m2 /g), obtained by precipitation, were used. The composites SBR /SiO2 were prepared by precipitation with methanol from tetrahydrofuran solutions containing the rubber and the labelled SiO 2 (15– 30% by weight, with respect to SBR). After precipitation, the samples were kept for several hours under high vacuum to eliminate the solvent. Spin labelling of SiO2 The spin labelling method was based on the following reaction sequence: (a) Functionalisation of the silica particles surface through the reaction of the SiO2 hydroxyl groups with bis-triethoxysilyl propyltetrasulphane (TESPT). At the end of this reaction, reactive ethoxysilyl groups are bound to the SiO 2 surface. The reaction is performed at 120 ± C and takes place with elimination of ethanol. (b) Reaction of the SiO2 bound ethoxysilyl groups with 4-hydroxy-2,2,6,6-tetramethyl piperidine-N-oxyl (TEMPOLO). The reaction is performed at 120 ± C under vacuum with the reaction vessel connected to a cold nger kept at ¡78 ± C in order to favour the elimination of the ethanol resulting from the condensation. The labelling reaction is described in the scheme shown in Fig. 1. SBR and SBR-SiO2 matrices doped with the nitroxyl TEMPO Spin probe molecular dynamics measurements within the SBR and SBR-SiO2 (37% SiO2 ) matrices were made by using the commercially available (Sigma) 2,2,6,6-tetramethyl piperidine-N-oxyl radical (TEMPO) as motional probe. The
EPR spin labelling studies of molecular dynamics
193
Figure 1. Spin labelling of SiO2 / ller particles.
probe was allowed to diffuse through the rubber matrices from a 10¡4 M ethanol solution, where the polymer is insoluble. The doped samples have then been kept for several hours under vacuum at room temperature in order to eliminate the solvent. EPR measurements and spectral analysis The samples were sealed under N2 atmosphere in quartz tubes to avoid the oxygen Heisenberg broadening effect. The EPR spectra were recorded in the
A. Buttafava et al.
194
temperature range from 77 K to 393 K using a Bruker EMX /12 spectrometer with data acquisition system and temperature control. The EPR analysis for obtaining the molecular dynamics parameters was performed using the stochastic Liouville approach as implemented in the code developed by J. H. Freed [5 – 7]. The core of the method is embodied in the equation: I.! ¡ ! ± / D .1=¼ /hVj[.0 ¡ iL / C i.! ¡ ! ± /I ]¡1 jVi; where I.! ¡ ! ± / is the line shape function, L is the Liouville super-operator associated to the time dependent Hamiltonian, 0 is the symmetrised diffusion superoperator used for modelling a reorentational motion which is the result of several contributions: 0 D 0iso C 0ex C 0u C 0av C 0dj ; where 0iso D free rotational diffusion; 0ex D spin exchange; 0u D rotational diffusion controlled by a potential; 0av D effects from viscosity anisotropy; 0dj D jump between equivalent sites. I is the identity operator, the vector jVi includes the spin operator for the allowed transitions and the equilibrium distribution of molecular orientations, .0 ¡ iL / is the Liouville stochastic super-operator. The magnetic and rotational diffusion tensors principal axis systems are related to the molecular geometry of the spin label as shown in the scheme below. The hf and g tensors are assumed to share the same principal axis system with the major hf principal value lying along the symmetry axis of the half lled orbital. The rotational diffusion tensor D is assumed to be axially symmetric with the fast rotation direction making a tilt angle µ with respect to the z magnetic axis. The simulation procedure was developed along the following steps: (a) Simulation of the rigid limit powder EPR spectrum of the spin label for obtaining the principal values of the g and hf tensors shown below. (b) Choice of the tilt angle. (c) Choice of the motional model. (d) Computer simulation using the rotational frequency as tting parameter: if O.K. ! extraction of the motion parameters (correlation time, rotational frequency, motional model). If unsuccessful ! back to point b starting with different parameters. The motional models implemented in the code are the Brownian diffusion, the free diffusion and the strong jump diffusion [8]. Brownian diffusion was used through all the simulations since it proved to be the best suited model for matching the experimental data. A second simpli ed method of analysis was also used based on the measurements of the outer peaks separation in the nitroxyl EPR spectrum: this parameter
EPR spin labelling studies of molecular dynamics
195
Figure 2. Computed dependence of the motion parameter 2A0zz (N) as a function of the rotational frequency. Isotropic Brownian diffusion.
corresponds to 2Azz (N) in the rigid limit spectrum and decreases with increasing the frequency [5 –7, 9, 10] (Fig. 2).
RESULTS
Spin probe measurements in the pure SBR matrix The spectral changes taking place as a function of the temperature in the rubber matrix between 393 K and 313 K are reported in Fig. 3, whilst Fig. 4 shows the temperature dependence of the molecular mobility as monitored through changes of the effective 2A0zz splitting. The maximum slope in the 2A0zz vs. temperature curve is observed at 228 K corresponding to a calculated frequency of 7 £ 107 s¡1 . This temperature can be considered as an EPR de ned Tg . The Tg of the SBR rubber as detected from thermal or speci c volume measurements is indeed ca. 20 ± C lower. The probe approaches the fast motion limit of 109 s¡1 at ca. 320 K. The calculated intermediate motional range spectra show a mismatch in the line intensity ratio with respect to the experimental patterns which cannot be accounted for in terms of diffusion tensor anisotropy, tilt angles or motional model. Only the assumption of a restoring potential, as in liquid crystals, allowed a signi cant improvement of the matching with the experimental spectra to be obtained. This hypothesis however seems unreasonable in the case of relaxed elastomer samples. Spin label measurements in SBR elastomer — SiO2 composites The nitroxyl spin labels bound to the silica particles will undergo a rotational diffusion motion spanning a maximum depth of 15 Å within the SiO2 / rubber interface. The characteristics of such motions are expected to be under the in uence of the molecular packing of the elastomer chain in the interface regions.
196
A. Buttafava et al.
Figure 3. EPR spectra of the spin probe TEMPO in the pure SBR rubber matrix recorded as a function of temperature. The rotational diffusion tensors were determined by computer simulation within the Brownian rotational diffusion model.
A sample sequence of the EPR spectra recorded with a sample of SBR rubber containing 30% of labelled SiO2 as a function of the temperature is shown in Fig. 5. Up to 200 K, satisfactory computer simulations could be obtained (Fig. 6) within the Brownian rotational diffusion model by assuming a D== /D? anisotropy D 10
EPR spin labelling studies of molecular dynamics
197
Figure 4. Spin probe molecular dynamics in the pure SBR rubber matrix. The diagram shows the decrease of the 2A0zz (N) parameter with increasing the temperature. The EPR glass transition is at ca. 318 K.
and a tilt angle of 50± . The tilt angle is close to that calculated for the rotation of the label about the C O bond. Above 200 K, a satisfactory matching could be achieved only by assuming the contribution by two different motional components (Figs 7, 8). The fast component frequency is 8:5£107 s¡1 at 363 K and increases to 1£108 s¡1 at 393 K. Within the same temperature range, the slow component is calculated to undergo a rotational diffusion which increases from 1:2 £ 107 s¡1 to 2:5 £ 107 s¡1 . This large difference of mobility strongly suggests that part of the labels may be under the constraint arising from dipolar bonding with hydroxyl groups at the surface of the SiO2 particles. This would indeed be the situation for those labels bound through one of the two ethoxy groups still available near the SiO2 surface in the TESPT structure after the rst step (Fig. 1). The mobile fraction of the spin labels instead is expected to reproduce the macromolecular segmental chain mobility within the SiO2 -rubber interface region up to a depth of 15 Å. This mobility, when it is compared with that determined within the pure SBR rubber matrix, is seen to be signi cantly slower. In fact, in the pure rubber matrix, the fast rotational diffusion limit of 109 s¡1 is reached at 313 K whilst within the rubber/ SiO2 interface region the rotational frequency remains one order of magnitude slower at a much higher temperature (1 £ 108 s¡1 at 393 K) (Figs 3, 5). It may be suggested that the observed decrease of the label molecular mobility with respect to the pure rubber matrix may be a consequence of the enhancement of the chain packing density near the silica surface due to the van der Waals elatomer / SiO2 bonding interactions. The decrease from 30% to 15% of the SiO2 concentration in the SBR matrix has a signi cant effect on the label molecular mobility, which consists in an increase of
198
A. Buttafava et al.
Figure 5. Spin label molecular dynamics in SBR-SiO 2 matrices (30% of labelled SiO2 ). EPR spectra recorded as a function of temperature. The rotational diffusion tensors were determined by computer simulation using the Brownian rotational diffusion model. The D values shown a side the spectra recorded above 203 K pertain to the fast and slow motional components.
the rotational frequency from 8:5 £ 107 s¡1 to 1 £ 108 s¡1 at 363 K. This mobility enhancement is also evident from the changes of the 2A0zz (N) parameter measured as a function of the temperature (Fig. 7). The effect of the increase of the ller concentration on the segmental chain mobility can be attributed to the superimposition of the interface regions with consequent cumulation of the effects stemming from the SiO2 -rubber interactions, such as the decrease of the free volume and the enhancement of the chain packing density.
EPR spin labelling studies of molecular dynamics
199
Figure 6. Spin label molecular dynamics in SBR-SiO 2 matrices (30% labeled SiO2 ). Brownian rotational diffusion model. Dashed lines: computed; continous line: experimental.
Spin probes measurements in the SiO2 / SBR rubber composites The capability of the SiO2 -SBR van der Waals interactions to extend their effect beyond the interface regions to the bulk of the elastomer matrix has been tested using the spin probe (TEMPO) technique. The diffusion process is expected to lead
200
A. Buttafava et al.
Figure 7. Molecular dynamics in SBR-SiO 2 matrices (30% of labeled SiO2 ). Deconvolution by computer simulation of the fast and slow components in the EPR spectrum recorded at 363 K. Brownian rotational diffusion model. Dashed lines: computed; continous lines: experimental.
to a distribution of the spin probes throughout the whole matrix thus including sites at the SiO2 -interfaces, on the silica surfaces and within the bulk polymer matrix. By using adequate concentrations of the nitroxyls (diffusion up to a stationary stage from 10¡4 M solutions of the probe in ethanol), the relative abundance of the spin probe population in the bulk matrix is made to be prevalent. The EPR spectra thus obtained result from the superimposition of a slow and a fast motional component the latter being the dominant one.
EPR spin labelling studies of molecular dynamics
201
Figure 8. Molecular dynamics in SBR-SiO 2 matrices (15% of labelled SiO2 ). Deconvolution by computer simulation of the fast and slow components in the EPR spectrum recorded at 363 K. Brownian rotational diffusion model (continuous lines: experimental; dashed lines: computer simulations).
In Table 1, the fast rotational frequencies as measured in the temperature range from 263 K to 303 K in the pure SBR and SiO2 -SBR matrices are reported. The presence of SiO2 has the effect of inducing a limited but signi cant decrease of the average bulk segmental chain mobility. This effect is seen to be further enhanced by crosslinking via a vulcanisation process. These results strongly suggest that the effects induced by the ller on the chain dynamics start at the SiO2 /SBR
A. Buttafava et al.
202
Figure 9. Spin label measurements in SBR/SiO2 (labelled SiO2 ) matrices. The diagrams show the effect of the SiO2 concentration on the outer peaks separation (2A0zz (N)) as a function of the temperature. Squares: 15% SiO2 ; triangles: 30% SiO2 . Table 1. Principal values of the spin label g and hf tensors obtained by computer simulation of the rigid limit spectrum at 77 K gxx D 2:0090 Axx D 8:0 G
gyy D 2:0063 Ayy D 4:0 G
gzz D 2:0021 Azz D 35:6 G
giso D 2:0058 Aiso D 15:9 G
Table 2. Effect of the SiO2 and crosslinking on the rotational diffusion tensor of the spin probe TEMPO trapped in SBR and SBR / SiO2 (37% SiO2 ) matrices. Crosslinked samples were obtained through a conventional sulphur vulcanisation procedure at 423 K T (K)
SBR D (s¡1 )
SBR-SiO 2 1D (%)
SBR-SiO 2 after vulcanisation 1D (%)
263 268 293 298 303
7:5 £ 107 8:1 £ 107 2:0 £ 108 3:0 £ 108 4:1 £ 108
¡4:0 ¡7:4 ¡5:0 ¡6:6 ¡14:6
¡17:3 ¡9:8 ¡30:0 ¡26:0 ¡29:0
interfaces but are capable of propagating beyond these regions into the bulk polymer matrix. CONCLUSIONS
A novel spin labelling technique has been developed for exploring the chain segmental mobility within the interface regions in SBR rubber matrices containing SiO2 as ller. The method is based on the chemical linking of a 2,2,6,6-tetramethyl
EPR spin labelling studies of molecular dynamics
203
piperidine-N-oxyl probe to the surface of the SiO2 particles, via a two-stage reaction schematically represented by the sequence:
The major results of the molecular dynamics measurements within the SiO2 -elastomer interface regions can be summarised as follows: (a) Fast and slow motional components have been identi ed and their rotational diffusion tensors have been measured as a function of the temperature and the SiO2 concentration. (b) The fast rotational frequency is found to be orders of magnitude slower than that measured in the absence of SiO 2 . The hindrance to the molecular mobility in the
A. Buttafava et al.
204
interface region is suggested to arise from closer packing of the macromolecules near the silica surface caused by the van der Waals bonding interactions. (c) The increase of the SiO2 concentration results in a decrease of the molecular mobility. This effect has been imputed to the overlapping of the bonding interaction regions. (d) Spin probe measurements in SBR-SiO2 matrices using TEMPO strongly suggest that hindrance to segmental motion induced by the SiO2 -SBR interactions propagate beyond the interface regions thus involving the bulk polymer matrix. The results of this investigation demonstrate that spin labelling and spin probe EPR measurements are suitable methods for obtaining detailed information regarding the intensity and the extension of polymer/ ller bonding interactions which are among the factors determining the reinforcing mechanism by the ller.
REFERENCES 1. 2. 3. 4. 5. 6.
7.
8. 9. 10.
G. Raos and G. Allegra, J. Chem. Phys. 113, 7554 (2000). A. I. Medalia, Rubber Chem. and Technol. 51, 437 (1978). M. J. Wang, Rubber Chem. and Technol. 71, 520 (1998). S. Wlff and M. J. Wang, in: Carbon Black Science and Technology, 2nd edn., J. B. Bonnet, R. C. Bansal and M. J. Wang (Eds). Dekker, NY, New York (1993). J. H. Freed, Theory of slow tumbling ESR spectra for nitroxides, in: Spin Labeling: Theory and applications, L. J. Berliner (Ed.), p. 53. Academic Press (1976). D. J. Schneider and J. H. Freed, Calculating slow motional magnetic resonance spectra, in: Spin Labeling: Theory and applications, L. J. Berliner and J. Reuben (Eds), p. 1. Plenum Press, NY, New York (1989). P. Tormala, G. Weber and J. J. Lindberg, Spin label and probe studies of relaxations and phase transitions in polymeric solids and melts, in: Molecular Motion in Polymers by ESR, R. Boyer and S. E. Keinath (Eds), p. 81. Harwood Academic Press, London (1978). S. A. Goldman, G. V. Bruno, C. F. Polnaszek and J. H. Freed, J. Chem. Phys. 56, 716 (1972). P. A. Egelstaff, J. Chem. Phys. 53, 2590 (1970). G. G. Cameron, Pure Appl. Chem. 54, 483 (1982).
Lihat lebih banyak...
Comentarios
