Introduction

Cellulose derivatives are an important class of biopolymers and an attractive alternative to replace conventional petroleum-based plastics. The general characteristics of cellulose esters such as stiffness, high-moisture vapor transmission, resistance to grease and excellent transparency are interesting in plastic applications. However, cellulose esters present restricted applicability due to the limited mechanical properties, which are one of the most significant problems of their low dimensional stability under high humidity and temperature conditions. In the last two decades, cellulose esters have been used to develop biodegradable plastics, either alone or in blends with other polymers (Rosa et al. 2007).

Cellulose acetate (CA) is a polymer that can be easily molded, in the presence of plasticizers, to form films, membranes, fibers and microspheres. Cellulose acetate membranes, for instance, were successfully used in bioseparation systems for ultrafiltration, nanofiltration and reverse osmosis (Sivakumar et al. 1998; Samios et al. 1997). To expand the applications of cellulose acetate some strategies are being used: (1) formation of polymer blends (Sivakumar et al. 1998) (2) chemical modification of CA chains by the introduction of functional groups (Samios et al. 1997); (3) grafting of the CA polymer chain (Sivakumar et al. 1998; Samios et al. 1997; Narayan and Nie 1994a, b; Mansson and Westfelt 1981; Maldas and Bhattacharyya 1984; Biermann et al. 1987) and (4) preparation of hybrid organic–inorganic nanocomposites (Sivakumar et al. 1998; Samios et al. 1997; Park et al. 2004, 2006; Ruan et al. 2004; Wibowo et al. 2006; Romero et al. 2009; Zoppi and Gonçalves 2002; Silva et al. 2011; de Oliveira Junior et al. 2012).

Botaro and Gandini (1998) described a series of reactions for cellulose acetate modification in homogeneous medium using two structurally different isocyanates. The increase in the hydrophobic character of CA was confirmed when compared to non modified CA. Apart from this, a decrease in the tensile strength, which was associated with the decrease of the extent of hydrogen bonding between hydroxyl groups on the modified films, was observed.

CA composites prepared with silane coupling agents have been reported in literature (Silva et al. 2011; de Oliveira Junior et al. 2012). The preparation of CA composites using silane coupling agents can overcome some drawbacks, such as filler agglomeration and limited filler addition. In these composites, a dispersed inorganic phase was obtained by hydrolysis and condensation of reactive groups that may or not be linked to the organic matrix (de Oliveira Junior et al. 2012). The characteristics of these composites are determined by the composition and nature of the components (Bandyopadhyay et al. 2004; Amerio et al. 2005; Mackenzi and Bescher 2003).

Zoppi and Gonçalves (2002) prepared hybrids based on cellulose acetate and SiO2 by the hydrolysis of tetraethoxisilane using the sol–gel process. The authors demonstrated that the incorporation of a silica phase on a cellulose acetate film matrix resulted in increased stiffness, in decreased water permeation and in thermal stability similar to pure cellulose acetate film.

Silva et al. (2011) prepared nanocomposites derived from CA and a multifunctional alkoxysilane precursor using the sol–gel process. These nanocomposites showed improved dimensional stability in comparison to pure CA. However, the influence of the silane architecture on the composite morphology was not explored.

Organic–inorganic composites derived from CA and two types of silane coupling agents (methyltrimethoxysilane and phenyltriethoxysilane) were prepared by de Oliveira Junior et al. (2012). CA/siloxane composites were obtained using two different approaches: direct reaction between CA chains and a monomer; and reaction of CA with oligomers. The composites showed thermal stability similar to pure CA and reduced glass transition temperature. Silicon mapping images, obtained from transmission electron microscopy associated with electron loss spectroscopy technique, showed the presence of microdomains with higher polysilsesquioxane concentrations than the matrix and also the presence of silicon-rich nanodomains dispersed throughout the matrix.

On the other hand, the combination of cellulose acetate with elastomers has been used, especially to improve material mechanical properties, for example, material toughness. Some examples include cellulose acetate/poly(epichlorohydrin) blends (de Lima et al. 2012) and blends of cellulose acetate and castor oil-based polyurethane (PU) (Zhou et al. 2003).

Polysiloxanes are elastomers that may be considered for this purpose due to their interesting properties, such as low surface energy, excellent gas permeability, good heat stability, low temperature flexibility and biocompatibility (Mark et al. 1992).

Similar to any system of various components, the challenge when combining CA and polysiloxanes is the need for compatibilization between these components to reach good dispersion of the polymer phases as well as to ensure good adhesion between the phases.

The aim of the present work was to prepare cellulose acetate/polysiloxane composites using 3-isocyanatepropyltriethoxysilane, as a coupling agent. Initially, the CA was modified with the coupling agent. The CA/polysiloxane composites were then prepared using two different approaches: (1) hydrolysis followed by condensation reaction of modified CA in presence of water and (2) condensation reaction of modified CA in the presence of silanol-terminated poly(dimethylsiloxane). The composite structures, morphologies and thermal and mechanical properties were investigated by infrared spectroscopy, 29Si nuclear magnetic resonance, field emission scanning electron microscopy, transmission electron microscopy, thermogravimetric and dynamic-mechanical analyses. Due to its direct relation with the material properties, focus on the material morphology was given in this work.

Experimental

Materials

Cellulose acetate (CA) was purchased from Sigma-Aldrich, with 39.8 wt% acetyl content with a number average molar mass of the 30,000 g mol−1. 3-isocyanatepropyltriethoxysilane (IPTS) (95 % purity) was also purchased from Sigma-Aldrich. Dibutyltindilaurate (DBTDL) (DC 176 Catalyst) was obtained from Dow Corning and used as a catalyst. The silanol-terminated poly(dimethylsiloxane) (PDMS) with a number-average molar mass of 2,200 g mol−1 (DC Q1-3563) was also obtained from Dow Corning. Tetrahydrofuran (THF) was purchased from Merck and it was dried and distilled prior to use.

Preparation of CA-PSS and CA-PSS-PDMS composites

Cellulose acetate was dried under vacuum at 60 °C for 24 h before use. Then, 10.0 g of cellulose acetate was dissolved in 100.0 mL tetrahydrofuran (THF) and the solution was submitted to a nitrogen flux. Next, 1.0 g of IPTS (dissolved in 1.0 mL of THF) was added to the previous solution. The reaction was carried out for 6 h at room temperature (~26 °C), under mechanical stirring. The resulting CA-IPTS intermediary solution was divided into two fractions. In the first one, 0.22 g (0.012 mmol) of water (3:1) and 0.5 g of DBTDL catalyst were added. This procedure was carried out in order to promote the hydrolysis and condensation of Si-OEt reactive groups giving rise to the CA-PSS composite. In the second initial solution fraction, 1.0 g of PDMS (dissolved in 1.0 mL of THF) and 0.5 g of DBTDL catalyst were added. The mixture was then kept in a bottle and magnetically stirred for approximately 1 h at room temperature. This procedure was carried out to prepare cellulose acetate/PDMS composite (CA-PSS-PDMS) from the previously functionalized CA-IPTS intermediary solution.

The two obtained solutions were homogeneous and viscous and were separately cast onto Teflon™ Petri dishes, at room temperature under a nitrogen atmosphere. The samples were dried under vacuum at 60 °C for 24 h before analyses.

Characterization of the materials

Fourier transform infrared spectroscopy (FTIR) analysis was performed using a FTIR Bomem MB Series B102 spectrometer. Samples of the IPTS and PDMS were analyzed as thin films cast on NaCl windows. CA, CA-PSS and CA-PSS-PDMS samples were analyzed with the conventional KBr-supported film technique. Each spectrum was recorded with a 4 cm−1 resolution and a total of 16 scans in the region between 400 and 4,000 cm−1.

Cross polarization-magic angle spinning nuclear magnetic resonance (CP-MAS NMR) of 29Si was obtained at room temperature on a Bruker 400 spectrometer operating at 79.5 MHz for 29Si nucleus and at 400 MHz for 1H nucleus. The acquisition time was 40 ms, contact time 5 ms, recycling delay 1 s and spectral width 250 × 103 Hz.

Soxhlet extraction procedure was used to determine the insoluble fraction of the CA-PSS and CA-PSS-PDMS composites. This procedure was carried out using THF solvent, a good solvent for the three initial components. The extraction was performed for approximately 33 h. After Soxhlet extraction, the insoluble fractions were dried in a 50 °C air flow oven for 24 h. The residue of the extraction was called “insoluble fraction”.

Thermogravimetric analysis (TGA) was performed in a TA 2950 thermobalance, TA Instruments, in a 25–1,000 °C range, at 20 °C/min scanning rate under argon flux. The weight of the sample used was 5 mg in all cases.

Dynamic-mechanical analysis was carried out using a DMTA V Rheometric Scientific Instrument from −150 to 250 °C at 2 °C/min. Rectangular geometry samples (30 × 8.0 × 0.7 mm) were submitted to tension/compression deformation at a 1.0 Hz frequency and 0.02 % amplitude.

The localized mechanical properties of the films were characterized by nanoindentation using a Nano-indentation tester from CSM Instruments equipped with a three-sided pyramid diamond Berkovich tip. The samples for the nanoindentation test were previously embedded in epoxy resin. The surface was sandpapered and then polished with silica paste. The loading rate, unloading rate, maximum indentation force, maximum pause time, Poisson ratio, temperature, approach speed of the tip, slope contact delta and acquisition frequency were 4.00 mN min−1, 4.00 mN min−1, 2 mN, 30 s, 0.3, 19 °C, 3,000 nm min−1, 80 %, 10.0 Hz, respectively. The separation distance between each nanoindentation point was 0.04 mm.

The morphology of the composites was examined by emission scanning electronic microscopy (FESEM) using a JSM-6340 field emission scanning electron microscope from Jeol Ltd., operating at an accelerating voltage of 3.0 kV. The samples were prepared by fracturing the films in liquid nitrogen and mounting them on the sample holder with doublesided carbon adhesive tape. The samples were carbon and gold sputted coated in a Bal-Tec MD 020 instrument (Balzers).

The morphology of the composites was also investigated by transmission electron microscopy (TEM) using a Carl Zeiss CEM 902 transmission electron microscope (80 kV) equipped with a Castaing-Henry energy filter spectrometer within the column and a Proscan Slow Scan CCD camera. Ultrathin sections, approximately 45 nm thick, were cut perpendicular to the film plane at −100 °C, using a diamond knife (Diatome), in a Leica EM FC6 ultramicrotome. The thin sections were collected on uncoated 500 mesh copper grids (Ted Pella). Images were acquired using electrons with zero-loss energy and processed using AnalySis 3.0 software.

Contact angle measurements were carried out using the Krüss Easy drop DSA 20 equipment. A sessile drop of deionized water was deposited on the sample surfaces and the contact angle value was recorded. Measurements were performed at ambient temperature (24 ± 2 °C) and repeated 4 times on different fresh surfaces.

Results and discussion

The CA-IPTS intermediary was obtained by the reaction between CH2–OH groups present in CA chains and IPTS isocyanate groups, forming urethane linkages. The hydrolysis, followed by the self-condensation reaction of ethoxysilane groups from the CA-IPTS intermediary, promoted the formation of polysilsesquioxane domains, which can act as crosslinking centers in the CA-PSS composite formation. In order to prepare the CA-PSS-PDMS composite, the CA-IPTS intermediary was kept under an inert atmosphere, to react with the silanol groups present in the PDMS-end chains, in the presence of the Sn-catalyst. Figure 1 schematically illustrates the proposed structures of the CA-IPTS intermediary, CA-PSS and CA-PSS-PDMS.

Fig. 1
figure 1

Proposed reactions between a CA and IPTS, b hydrolysis and condensation in presence of water and c in presence of PDMS

Full size image

Figure 2 shows the infrared spectra of CA, PDMS, IPTS, CA-PSS and CA-PSS-PDMS. The infrared spectrum for IPTS shows C–H (propil group and CA units) stretching absorptions in the 2,960–2,850 cm−1 region, and an intense absorption of isocyanide groups at 2,273 cm−1 (Fig. 2b). This last absorption is not observed in both CA-PSS and CA-PSS-PDMS composites, which was an indication of the addition reaction between silane isocyanide groups and CH2OH groups present in the CA chains during the initial reaction, or its hydrolysis at the last step of the reaction. A broad absorption related to O–H (C–OH and Si–OH) stretching was observed in the 3,600–3,200 cm−1 region (Fig. 2d, e). The bands related to Si–O–Si are located in the 1,100–1,020 cm−1 region. According to Li et al. (2009), polysilsesquioxanes ladder structures show an intensive absorption related to Si–O–Si asymmetric vibration around 1,051 cm−1, while in cage structures this absorption occurs around 1,150 cm−1. The occurrence of a band at 802 cm−1, referring to C–H deformation and Si–C (Si–(CH3)2) stretching in CA-PSS-PDMS spectrum (Fig. 2e), can be observed, which is the characteristic band for the PDMS (Bellamy 1966). A characteristic C=O stretching band around 1,749 cm−1 and a O–H stretching band at 3,469 cm−1 for pure CA, were also observed. FTIR results of the materials suggest that the reaction between IPTS and cellulose acetate occurred, but this technique alone is not conclusive, being necessary to be confirmed by the use of other techniques.

Fig. 2
figure 2

Infrared spectra of a PDMS, b IPTS, c CA, d CA-PSS and e CA-PSS-PDMS

Full size image

Figure 3 shows the 29Si NMR spectra of the CA-PSS and CA-PSS-PDMS composites. The extent of condensation of Si sites can be evaluated by 29Si NMR peaks. The presence of different Si (T) sites indicates that there probably was incomplete condensation of the siloxane network. The CA-PSS composite show the T3 peak at around −65 ppm, referring to condensed alkoxides, T2 at around −57 ppm, corresponding to partially condensed Si–OH groups, T0 at around −47 ppm, corresponding to the hydrolyzed intermediary with no condensed Si–OH groups. This latter site is usually detected when Si is bonded to large voluminous substituents, which can be the case of the CA chains. In this work, the presence of T0 site suggests that CA promotes steric hindrance, or that there are associations by hydrogen bonds between Si–OH (from the siloxane network) and C–OH (from CA), which make the condensation reaction difficult (Abbehausen et al. 2010). The observation of three peaks for the T sites in the composites is an indication that the material obtained has structural heterogeneity, with more or less condensed domains, caused by the presence of the large voluminous group of CA chains.

Fig. 3
figure 3

29Si NMR spectra of CA-PSS and CA-PSS-PDMS

Full size image

The CA-PSS-PDMS composite showed the T3 and T2 peaks at around −67 ppm and −57 ppm, respectively. A peak at around −21 ppm was also observed, which corresponds to the D2 group, that is a characteristic peak of PDMS (Leite et al. 1995).

To confirm these previous results, the insoluble fraction of the CA-PSS and CA-PSS-PDMS composites was evaluated by Soxhlet extraction with THF, which is a good solvent for both CA and PDMS phases. The results of the solvent extraction were used to calculate the insoluble fraction percentage for the CA-PSS and CA-PSS-PDMS composites. In this case, the values of the residual dry mass in relation to the initial mass for CA-PSS and CA-PSS-PDMS composites were 44.4 % and 68.5 wt%, respectively. The high amount of non-soluble residue confirms that there was significant crosslinking network formation in both composites as a result of the occurrence of hydrolysis followed by condensation reaction of modified CA in presence of water as well as condensation reaction of modified CA in the presence of silanol-terminated PDMS.

Figure 4 shows the dynamic mechanical behavior of CA, CA-PSS and CA-PSS-PDMS films; E’ is the storage modulus, E″ is the loss modulus, and tan δ is the ratio between E″/E′. Table 1 presents the corresponding values for the glass transition temperatures obtained from the maximum values of the glass transition peaks in the loss modulus curve and in the tan δ curve. The CA loss modulus curves (E″) show a 169 °C glass transition temperature (Tg). There is also a secondary transition related to the β relaxation peak at approximately −23 °C. It is a well known fact that the presence of acetyl and hydroxyl groups promotes intra- and intermolecular interactions in the CA chains by hydrogen bonding. These hydrogen bond bridges between cellulose structural segment neighbors, making CA particularly rigid and result in the high value of the glass transition temperature of cellulose acetate films. Usually three to five peaks appear in the dynamic mechanical spectrum of CA (Backman and Lindberg 2001) and are associated to the relaxation of a part of the main chain, of the glucose ring and also to groups associated to water (Silva et al. 2011; Backman and Lindberg 2001) and/or adsorbed water (Vidéki et al. 2007). Only two transitions were observed in this work: Tg and β relaxation.

Fig. 4
figure 4

Dynamic mechanical behavior of CA, CA-PSS and CA-PSS-PDMS films: a storage modulus (E′), b loss modulus (E″) and c tan δ curves

Full size image
Table 1 Glass transition temperature (Tg) and indentation hardness (HIT) of CA, CA-PSS and CA-PSS-PDMS films
Full size table

A decrease of the CA-PSS and CA-PSS-PDMS Tg values (156 and 139 °C, respectively), in relation to that of CA (169 °C), was observed. This decrease could be interpreted by the contribution of three factors: (1) increase of free volume due to the insertion of silane side groups, (2) high mobility of the substituent groups, due to the presence of Si–O–Si groups, and (3) reduction of hydrogen bonds due to the substitution of the hydroxyl groups by silanes, which contribute as an extra factor to the increased mobility of the functionalized polymer (de Oliveira Junior et al. 2012). On the other hand, the isolated influence of the crosslinking network formation, as experimentally observed, is expected to result in an increase of the Tg value. However, this effect was not pronounced when compared to those above mentioned. Regarding CA-PSS-PDMS composite, the insertion of PDMS flexible units promoted an additional reduction of the Tg value compared to the CA matrix, due to the low rotational energy barrier which is characteristic of the Si–O bond (Westman and Marthy 1996).

The effect of the presence of polysilsesquioxane domains and PDMS on the dynamic mechanical behavior can be analyzed from the storage modulus, loss modulus and tan δ curves. The CA-PSS and CA-PSS-PDMS composite curves (Fig. 4a) show a sharper drop with the increase in temperature, in the temperature range between −150 and 0 °C, when compared to pure CA, which can be an indication of the presence of siloxane. In addition, the width of the glass transition in the E’ curves can also be used as a criterion for the identification of structural heterogeneities in the materials. It is observed that the width of the CA-PSS and CA-PSS-PDMS composite glass transitions are broader compared to that of pure CA. This behavior is also associated to the siloxane domains, which caused heterogeneities in the materials.

Nanoindentation technique can provide an opportunity to understand the response of composites to a load in a nano-scale. Average indentation hardness values (HIT) for pure CA, CA-PSS and CA-PSS-PDMS composites are presented in Table 1. The addition of the silane coupling agent to CA did not significantly change the average hardness value, while the addition of the silane coupling agent together with PDMS decreased the average hardness values for CA-PSS-PDMS by 27 % when compared to pure CA. The decrease in the average value for the CA-PSS-PDMS composite is probably due to the combination effect of the crosslinking network and the presence of the low modulus PDMS. Hardness values were almost constant in pure CA in different nanoindentation points analysed, however the variations in hardness became high in composites, especially in the CA-PSS-PDMS composite. These variations in hardness are associated to the non-homogeneity of the surface composition and reflect different localized mechanical properties.

The morphological characteristics of pure CA, CA-PSS and CA-PSS-PDMS bulk fractures were investigated by FESEM and the micrographs of the fractured surfaces are presented in Fig. 5. Both the CA (Fig. 5a) and CA-PSS films (Fig. 5b) presented optical transparency and their fracture surface showed a rough and uniform morphology. On the other hand, the CA-PSS-PDMS film was translucent, which is an indication of phase separation. The CA-PSS-PDMS film phase separation was confirmed by the presence of domains with 9.5 ± 3.2 μm average diameter, which showed different surface roughness as observed in the FESEM images (Fig. 5c). These samples presented smooth PDMS and rougher CA rich domains, with PDMS domains as the dispersed phase. In addition to this, the dispersed domains were elliptically deformed as a result of the contraction generated during evaporation of the solvent due to the films being prepared by the solvent casting method. The film morphology obtained by the solvent casting technique depends on a number of factors such as the composition of the film, the rate of the solvent evaporation, and the solution viscosity (Kim et al. 2000). The image in Fig. 5c also shows that, besides the relatively large domains, a good adhesion between PDMS phase and the cellulose acetate matrix is observed, since the interfaces between phases showed no cracks or voids.

Fig. 5
figure 5

FESEM of bulk fractures: a CA, b CA-PSS and c CA-PDMS films

Full size image

Figure 6 shows the TEM images of pure CA, CA-PSS and CA-PSS-PDMS composites. The TEM analysis was used in order to observe the fine details of the matrix morphology, not possible to be investigated by FESEM. The CA-PSS composite presented silicon-rich nanodomains dispersed in the matrix, which appear as dark regions in the TEM images. These nanodomains were not detected by the scanning electron microscopy analyses (Fig. 5b), but only visualized by TEM (Fig. 6b). On the other hand, in addition to the microdomains previously observed by FESEM, the CA-PSS-PDMS composite matrix also showed nanodomains dispersed in the CA matrix (Fig. 6c). These nanodomains are associated to high concentration of polysilsesquioxane condensation products. The TEM images also showed the integrity of the interfaces between the polysiloxane nanodomains and the matrix, which also indicates good interfacial adhesion, as expected from a covalent CA-polysilsesquioxane interface.

Fig. 6
figure 6

TEM images of a CA, b CA-PSS and c CA-PSS-PDMS films

Full size image

The TEM images in Fig. 6 also show the differences in relation to the density of the polysilsesquioxane nanodomains, which are greater for CA-PSS-PDMS. This greater density can be due to the presence of PDMS in this composite distributed not only in microdomains, visualized by FESEM images, but also in the polysilsesquioxane nanodomains observed by TEM.

A schematic representation of the CA-PSS-PDMS composite is proposed based on the TEM observations, as can be seen in Fig. 7. This representation is coherent with the result that the organosilanes were incorporated in the matrix in the form of dispersed nanodomains of polysilsesquioxane and also in PDMS microdomains.

Fig. 7
figure 7

Schematic representation of CA-PSS-PDMS composite

Full size image

The thermal behavior of the CA-PSS and CA-PSS-PDMS composites was also evaluated and the results were compared with the pure cellulose acetate. Figure 8 shows the thermogravimetric curves for the CA and CA-PSS and CA-PSS-PDMS films, carried in inert atmosphere and under vacuum. Three steps of degradation were observed for cellulose acetate. The first step from 25 to ~320 °C corresponds to the volatilization of the volatile matter, and/or the evaporation of residual absorbed water. The second step from ~320 to ~420 °C, can be associated to the main thermal degradation of cellulose acetate chains. The carbonization of the products to ash can occur above 420 °C. These three steps may correspond to the three steps which were suggested by Chatterjee and Conrad (1968) and Hanna et al. (1999), as representing the thermal degradation of the cellulosic materials.

Fig. 8
figure 8

TGA curves of CA, CA-PSS and CA-PSS-PDMS films

Full size image

In addition to CA degradation steps, in the case of CA-PSS, a desamination and decarbamation at around 190 °C are expected, followed by degradation of cellulose acetate and the propyl groups up to 400 °C. Mineralization is also expected to start at higher temperatures. At 950 °C, a 4 wt% residue amount was produced, probably consisting of SiCxOy, which is the pyrolysis product of the polysilsesquioxane component (Abbehausen et al. 2010).

The thermal degradation of the PDMS, present in the CA-PSS-PDMS composite, was characterized by a chain depolymerization, which is expected to occur between approximately 350 and 550 °C to produce cyclic oligomers. The hexamethylcyclotrisiloxane trimer is the most abundant oligomer, presenting small quantities of tetramers, pentamers, hexamers and higher oligomers (Camino et al. 2001, 2002). Works in literature have shown that depolymerization occurs preferably at the free ends of the chains as a result of the intramolecular rearrangement of continuous segments of the chain involving siloxane linkages (Si–O) and Si–OH groups at the end of chains (not involved in the formation of the polymer network) (Camino et al. 2001, 2002; Lewickn et al. 2009). Intra- or intermolecular rearrangements involving only siloxane bonds also occur in the degradation of PDMS. At 950 °C, the SiCxOy residue content is around 10 wt%.

Water contact angle measurements were used to evaluate the effect of the polysilsesquioxane network and PDMS on the composite surface properties. The mean contact angles for the CA, CA-PSS and CA-PSS-PDMS composite are 70.1 ± 0.2, 98.3 ± 0.2 and 110.6 ± 0.4, respectively. An increase of the contact angle in composites is related to the decrease of the hydrophilic character. Therefore, the modification of the CA confers a hydrophobic character to the free surface, reducing its wettability by water. In addition, a higher increase of the contact angle was observed for the CA-PSS-PDMS composite compared to the pure CA and CA-PSS composite. Besides the presence of polysilsesquioxane network, an additional effect can exists. In this case, it is possible that free chains may have migrated to the film surface increasing its concentration in the surface due to the low surface tension of PDMS (Tang et al. 1999).

Conclusions

Composites of cellulose acetate and poly(dimethylsiloxane) using 3-isocyanatepropyl triethoxysilane as a coupling agent were prepared. The proposed reactions of CA and silane chains were indicated by infrared and 29Si NMR spectra and confirmed by insoluble fraction formation. The CA-PSS-PDMS composite presented both polysilsesquioxane nanodomains and polysiloxane microdomains. In addition, good adhesion between the PDMS phase and CA matrix was observed. The CA-PSS-PDMS composite showed thermal stability similar to pure CA, increased flexibility, confirmed by the lower Tg values, decreased average value of hardness and increased hydrophobic surface character. This demonstrated the usefulness of this methodology for the preparation of cellulose acetate/polysiloxane composites with different properties which can be useful for extending the utility of cellulose acetate in plastic applications, in the development of coatings, packing, laminates, semi-permeable membranes and fibers.