Influence of cure conditions on properties of resol/layered silicate nanocomposites.
Polymer-layered nanocomposites using nanosized clay particles as reinforcing agents have attracted a great interest because of their superior properties such as mechanical strength, heat resistance, gas permeability, and flammability compared with neat polymers, especially when an exfoliated state is achieved (1-15). Clays, as montmorillonite (MMT), are inexpensive, chemically and thermally stable, and have good mechanical properties. The enhanced properties that produce the presence of MMT in composites are presumably a result of nanometer size, large aspect ratio, and large surface area of the silicate layers (7), (8). To increase the organophility of these naturally hydrophilic phylosilicates, the cations in the galleries of the clay have to be exchanged by cationic modifiers (e.g., quaternary ammonium salts) (2), (6), (8). The modified clay (or organoclay), whose surface energy is decreased, tends to be more compatible with polymers. Moreover, the modified clay can react or interact with the monomer or the polymer (1), (2), (4), (5), (7-11) thus improving the interfacial strength between clay nanolayers and the polymer matrix (3), (6). The state of dispersion of clays in a polymer matrix can result in the formation of different kind of composites. The exfoliated state is the most interesting for the improvement of properties (4), (9). In thermoset matrix composites, to enhance the intercalation/exfoliation of clays, polymer clay compatibility, shear stress exerted by resin polymerization and molecular diffusion of polymer chains into the silicate interlayers are considered as the key factors (4), (9). The current use of nanoclays has been basically dedicated to improve the fire retardant properties of thermoset resins in general (12-15). In this article, it is also probed that MMT can also be used to improve the mechanical properties of such resins.
Phenolic resins are irreplaceable materials for a wide range of industrial applications such as adhesives, coatings, laminates, and composites (16-22). Phenolic resins are synthesized by the reaction of phenol with aldehydes, especially formaldehyde, and are classified as resols and novolacs depending on phenol/aldehyde ratio. Only a few studies have been performed on clay-based nanocompo-sites based on phenolic resins due to their three-dimensional molecular structure even before cure, which may avoid the exfoliation of the clay (23-33). Moreover, the formation of water as a byproduct of crosslinking is also another problem of this type of resins.
In a previous study (34), resol type phenolic resin/layered silicate nanocomposites were synthesized by the intercalation of monomer between silicate layers to overcome the structural problem of MMT dispersion and exfoliation into phenolic resin matrix. MMT was modified by using an aminoacid, L-phenyl alanine, to induce condensation reactions between its carboxyl end group and the hydroxyl groups of formaldehyde and so, compatibility with the phenolic resin matrix could be increased. In this work, the type of catalyst for curing, as well as the MMT modifier, have been analyzed with the aim of achieving an optimum degree of exfoliation of the layered silicate in the phenolic matrix. Moreover, other parameters as reactivity ratio and condensation temperature during prepolymer synthesis have also been investigated to achieve clay exfoliation. Morphology, thermal behavior, and stability have been studied by means of transmission optical microscopy (TOM), atomic force microscopy (AFM), dynamic mechanical analysis (DMA), wide angle X-ray scattering (WAXS), and thermogravimetric analysis (TGA). Moreover, mechanical properties have been evaluated and correlated to the morphology of the obtained nanocomposites.
Phenol (P), formaldehyde (F) (35-40% aqueous solution), triethylamine (TEA), and 50% aqueous solution of NaOH were purchased from Panreac (Barcelona, Spain) and used without further purification. Untreated [Na.sup.+] MMT and Cloisite 30B, a MMT organically treated with methyl tallow (~ 65% C18, ~ 430% C16, and ~ 5% C14) bis-2-hydroxyethyl quaternary ammonium chloride, were obtained from Southern Clay Products (Texas, EEUU). L-phenyl alanine and 6-aminocaproic acid were purchased from Aldrich (Madrid, Spain) and used for modifying [Na.sup.+] MMT.
Not very bulky aminoacids were chosen for decreasing the effect of steric hindrance during the formation of the prepolymer between MMT layers. L-phenyl alanine montmorillonite (PheMMT) and 6-aminocaproic acid modified montmorillonite (6aaMMT) were prepared through the ion exchange of [Na.sup.+] MMT with the corresponding amino-acids in acidic environment according to the protocol reported in previous work (34). Different amounts of PheMMT and 6aaMMT were sonicated in formaldehyde solution and treated in presence of concentrated sulfuric acid with the aim of promoting the condensation reaction between the carboxyl end group of both aminoacids and the--OH groups of the formaldehyde in aqueous solution. Condensation reaction between resol chains and aminoacid was demonstrated by FTIR technique in a previous work (34). Cloisite 30B and [Na.sup.+] MMT clays were subjected to the same treatment in order to compare all the composites at the same cure conditions (34). In the case of Cloisite 30B with--OH end groups, as seen by FTIR, condensation reactions did not occur with the--OH groups of formaldehyde solution.
To study the influence of cure conditions on the final dispersion of the clay in the resol matrix, several poly-merizations were carried out. In a first stage, prepolymers were synthesized by mixing the previously modified clays-formaldehyde solutions with formaldehyde and phenol in order to work with a formaldehyde/phenol molar ratio 1.4. Then, the pH of formaldehyde/phenol mixture was adjusted to 8 using different catalysts as TEA or 50% aqueous solution of NaOH. Condensation was carried out at 80 [degrees] C under reflux until prepolymers showed around 1/1 g/g solubility in water. Same treatment was used for formaldehyde/phenol molar ratios 2.0 and 1.0 and when using different condensation temperatures (55, 80, and 95 [degrees] C). Water extraction was performed under vacuum at 45-48 [degrees] C to a solid content of 75-85 wt%. Samples were stored at -20 [degrees] C until they were analyzed. Table 1 resumes the starting conditions and the designations for each synthesized prepolymer.
TABLE 1 Characteristics of neat resols and resol-clay composites. Sample Designation Modifier Catalyst D.W. S.C. (a) (b) (1 (%) g/g) Neal Resol 80 Res -- TEA 1.20 76 [degrees] C (1.4) 2 wt% PheMMT Phe L-phenyl alanine TEA 0.92 81 80 [degrees] C (1.4) 2 wt% C [MATHEMATICAL TEA 0.95 80 Cloisite 30B EXPRESSION NOT 80 [degrees] REPRODUCIBLE IN C (1.4) ASCII] 2 wt% NaMMT Res-Na -- TEA 1.12 81 80 [degrees] C (1.4) 2 wt% 6aaMMT Res-oaa 6-aminocaproic TEA 1.00 84 80 [degrees] acid C (1.4) Neal Resol 80 ResNaOH -- NaOH 1.05 83 [degrees] C (1.4) 2 wt% PheMMT Phe-NaOH L-phenyl alanine NaOH 1.01 84 80 [degrees] C (1.4) Neal Resol T Res-I -- TEA 1.05 68 80 C(1.0) 2 wl% PheMMT Phe-I L-phenyl alanine TEA 1.16 71 T80C (1.0) Neal Resol T Res-2 -- TEA 1.07 82 80 [degrees] C (2.0) 2 wt% PheMMT Phe-2 L-phenyl alanine TEA 1.04 80 T 80 [degrees] C (2.0) Neal Resol T Res-T95 -- TEA 0.66 82 95 C (1.4) 2 wt% PheMMT Phe-T95 L-phenyl alanine TEA 0.41 82 T 95 (c) [degrees] C (1.4) Neal Resol T Res-T55 -- TEA 1.15 80 55 [degrees] C (1.4) 2 wt% PheMMT Phe-T55 L-phenyl alanine TEA 1.66 84 T 55 (c) [degrees] C (1.4) (a.) D.W., dil inability in water (1 g/1 g). (b.) S.C., solid content (%). (c.) not clear measurement. * Methyl lallow (~65% CI8, ~30% CI6, and ~5% C14).
WAXS measurements were carried out with a powder diffractometer Philips, equipped with a graphite monochro-mator and an automatic divergence slit, using an incident X-ray of Cu Ka radiation with wavelength of 1.54 A.
Morphologies of the nanocomposites were investigated by TOM using an Olympus BH-2 optical microscope and by AFM using a Nanoscope Ills, Multimode [TM] from Digital Instruments operating in tapping mode. An integrated silicon tip/cantilever, from the same manufacturer, having a resonance frequency over 300 kHz, was used. The specimens were prepared by ultramicrotoming at room temperature.
Dynamic-mechanical analysis (DMA) was carried out in a Perkin-Elmer DMA-7 analyzer using a three-point bending device. DMA measurements were carried out with 24 x 5 x 1 m [m.sup.3] specimens maintaining a span of 15 mm and using 110 and 100 mN as static and dynamic forces, respectively. All measurements were carried out at a constant frequency of 1 Hz with a heating rate of 5 [degrees] C/ min using helium atmosphere.
Static flexural properties were determined in a three-point bending device using an Instron universal testing machine, model 4206, equipped with a load cell of 1 kN. Tests were carried out at room temperature with a relative humidity of 50 [+ or -] 5% using a crosshead displacement rate of 0.43 mm/min. Measurements were carried out with 25 x 10 x 1 mm3 specimens and at least five measurements were performed.
Thermogravimetric analysis was carried out using a Mettler Toledo TGA/SDTA 851. Samples were scanned from 25 to 1000 [degrees] C at a scanning rate of 10 [degrees] C/min under nitrogen atmosphere.
RESULTS AND DISCUSSION
Influence of Clay Modifier
Two aminoacids (L-phenyl alanine and 6-aminocaproic) were used for surface modification of MMT to study the influence of the nature of the surfactant on the degree of exfoliation of the layered silicate in the phenolic matrix. Furthermore, Cloisite 30B, a commercial clay functionalized with methyl tallow bis-2-hydroxyethyl quaternary ammonium, and untreated [Na.sup.+] MMT were also used for the synthesis of new composites. The dool spacings of the natural [Na.sup.+] MMT and modified clays, as analyzed by X-ray diffraction, are shown in Table 2. In the case of Cloisite 30B, the long length of the chains of the surfactant (~ 1.90 nm in length) justifies the size of the [d.sub.001] spacing. On the other hand, the basal spacings for both PheMMT and 6aaMMT were lower than for Cloisite 30B due to the small dimensions of the modifier (0.79 nm  and 1.22 nm (1) in length, respectively). Furthermore, as Usuki et al. (1) suggested, the carboxyl (--COOH) end group of the 2-aminoacid can bond with the oxygen (--O--) group of the silicate surface through hydrogen bonding.
TABLE 2. Modifiers and [d.sub.001] spacing of the modilied clays. Clay [Na.sup.+] PheMMT 6aaMMT Cloisite 30B MMT Modifier -- [MATHEMATICAL HOOC- [MATHEMATICAL EXPRESSION NOT [(C[H.sub.2]) EXPRESSION NOT REPRODUCIBLE .sub.2] - REPRODUCIBLE IN ASCII] N[H.sub.2] IN ASCII] T = L-phenyl 6-animocaproic Tallow (~65% alanine acid CI8, ~30% C16, ~5% CI4) [d.sub.001] 1.10 1.32 1.26 1.60 spacing (nm)
Figure 1 shows X-ray diffraction patterns of composites and neat resol catalyzed with TEA and using sonication. Res-6aa and Phe composites exhibited almost no diffraction peaks when compared with the composites with the unmodified clay (Res-Na). This suggests that silicate layers of 6aaMMT and PheMMT were better dispersed in the phenolic matrix than the other composites. This fact may be attributed to the induced condensation reaction between the carboxyl end group of the aminoacid and the--OH groups of the formaldehyde in aqueous acidic solution, thus acting like an anchoring point between the layers and the resin, as previously reported (34). Further-more, diffraction peak was hardly noticeable for Res-6aa composite when compared with Phe composite. Interactions between resol reactive molecules and 6-aminocaproic modifier could be more easily formed due to the lack of bulkiness and the higher flexibility of the linear alkyl chain of 6-aminocaproic acid when compared with L-phenyl alanine aminoacid. In the case of composite C, diffraction peaks were also hardly observed, that could indicate that a homogeneous clay dispersion was obtained. Taking into account TOM images (Fig. 2d) and by studying in detail the XRD pattern, it was observed that the basal spacing of the composite C appeared around 1.35 nm, whereas the basal spacing of the Cloisite 30B was 1.60 nm (Table 2). This significant contraction of interlayer spacing from 1.60 nm to 1.35 nm could be caused due to the presence of the bulky modifier of the clay whose steric hindrance avoids the polymerization of the resol inside the layered silicates (intragallery). Thereby, the polymerization could be more favorable outside them (extragallery) (34). The alkylammonium chains of the modifier occupied a large space between the layers and therefore, not much space remained accessible for the polymer chains to diffuse between the layers (36). If no polymerization does occur in the intragalleries, the layers cannot be further separated and polymerization takes place in the extragallery region, leading to shrinkage of the interlayer spacing (34). Furthermore, the lack of interactions between the modifier of the Cloisite 30B and the resol reactive molecules (27), (34) at the synthesis temperature could also favor this behavior. On the other hand, for Res-Na composite, a peak appeared at 1.26 nm. As the interlayer spacing of [Na.sup.+] MMT in the composite increased from 1.10 (Table 2) to 1.26 nm, part of the reactive process could occur inside the layers although this increase was not enough to achieve a complete intercalation (34).
The morphology of the composites was also studied by TOM and AFM to better analyze the dispersion of the organoclays in the phenolic matrix. TOM images are shown in Fig. 2a-e. In Fig. 2a, the morphology of the homogeneous surface of the neat resol matrix is seen. For Phe and Res-6aa systems, a uniform dispersion of clay in the matrix of the composite was observed and no significant variations were seen when compared with TOM picture of the neat resol. In contrast, in composites C and Res-Na (Fig. 2d-e), aggregates of a broad range of sizes were observed thus indicating poorer clays dispersions. In the case of Res-Na composite, the lack of modifier in the clay could explain this behavior. For C composite, as above described, the steric hindrance of the bulky surfactant and the lack of reactive groups in the surfactant of the clay could avoid the polymerization of the phenolic resin inside the layered silicate. Consequently, big agglomerates remained in the cured material. As above shown, the faintness of XRD peaks for C composite could indicate intercalated or nearly exfoliated clay structures. This fact was unexpected since Cloisite 30B clay does not react with resol matrix. However, despite XRD results indicating some extent of intercalation, TOM pictures confirmed the existence of layer agglomerates for composites with untreated (with a size around 3-5 [micro] m) or Cloisite 30B (5-15 [micro] m) clays.
AFM phase images are shown in Fig. 3a-c. Lines or scratches in surfaces, appearing after ultramicrotomy cutting due to the resol fragility, made difficult the observation of individual layers. In Fig. 3a, the globular structure of neat matrix can be seen (34). For Phe and Res-6aa composites, though homogeneously dispersed individual layers were observed, layers forming intercalated agglomerates with lateral size around 20-100 nm were also seen for Phe composite. Although in overall, the dispersion of the modified MMT in Res-6aa composite was very similar to Phe composite one, individual layers seemed to be more homogeneously distributed in the matrix (Fig. 3b), which is consistent with the results of XRD. Resol reactive molecules can diffuse into the inner clay layers when agglomerates are thin (34), (36). Indeed, XRD and AFM results showed a fairly good dispersion of PheMMT and 6aaMMT in the phenolic matrix. When clay stacks are thicker, reactive molecules could only insert inside the most superficial layers, remaining some agglomerates as in the case of C and Res-Na composites. Thus, the clay dispersion in C composite appeared to be very poor, as also observed by AFM elsewhere (34). As a conclusion, the nature of MMT modifier and its possible interactions and/or reactions with the polymer matrix result in a key factor in order to achieve clay exfoliation in phenolic composites.
Flexural properties of composites and neat resol catalyzed with TEA were analyzed by both dynamic and static mechanical measurements. As shown in Fig. 4, flexural modulus was higher for all the composites than for the neat matrix. For Phe and Res-6aa composites, the modulus achieved higher values than for the other systems. For flexural strength, the highest value was found for Res-6aa composite with an increase of 15% with respect to the neat matrix. The increase of flexural strength for Phe and Res-6aa composites can be attributed to more homogeneous dispersion of clays, as well as to the interactions with the matrix (37). In the case of Res-6aa composite, these interactions could be more easily reached than in the Phe composite due to the flexibility and the lack of bulkiness of the linear alkyl chain of 6-aminocaproic inside the clay. On the other hand, for poorer dispersion of clay (C and Res-Na composites), a slight increase in modulus was observed, which is usual for polymeric composites even without remarkable interfacial interactions between matrix and inorganic fillers (5). A significant decrease of flexural strength was observed for these composites, especially for Res-Na composite. As stated above, poor clay dispersions were obtained for composite C due to the presence of the bulky modifier of the clay whose steric hindrance avoids the polymerization of the resol in the clay. Therefore, if no polymerization does occur in the intragalleries, the layers cannot be further separated and polymerization takes place in the extragallery region. Consequently, larger clay agglomerates remains without being exfoliated. Further-more, the lack of interactions between the modifier of the Cloisite 30B and the resol reactive molecules (27), (34) could also favor this behavior. As stated by other authors (37), poorly dispersed clay layered silicates serve as stress concentration and flaws for crack initiation, which results in premature failure upon mechanical deformation.
On the other hand, Fig. 5a and b shows the thermal decomposition behavior of neat resol and clay-filled composites. Different stages of degradation were observed for the neat resol resin in the TGA thermograms: in the first stage, from 30 to 350 [degrees] C, the release of formaldehyde due to the breakage of ether bridges, and also phenol, water, and the onset of the degradation of organic modifier of the clay, took place simultaneously. In the second stage, in the temperature range of 350-700 [degrees] C, two zones can be distinguished: at 400-550 [degrees] C, the oxidation of the network and at 550-650 [degrees] C, the formation of the char structure (16). Phenolic composites showed a slightly better thermal stability than the neat resol. Silicate layers would act as a heat barrier, which enhances the overall thermal stability of the system (34) especially when the clay layers are homogeneously dispersed throughout the composite (Phe and Res-6aa composites).
Influence of Catalyst on Composite Synthesis
Different catalysts can be used for phenol-formalde-hyde resol synthesis, being the most used NaOH, Ba[(OH).sub.2], or LiOH and rarely, hydroxides of divalent metals (16), (38-43). Carbonates (sodium carbonates) and oxides (calcium or magnesium oxides) are also employed (38), (39). Tertiary amines, in particular triethylamine, are also used (16), (19-22), (41) being it the selected catalyst throughout this work. NaOH was also chosen to compare its influence on the final properties of the composites for being one of the most worldwide used catalysts in phenolic synthesis and as representative of the hydroxides. As it was studied elsewhere (41), resol curing in presence of NaOH is normally faster and can produce a bigger amount of condensed water than for curing of these resins with TEA. Figure 1 shows X-ray diffraction patterns of Phe composite catalyzed by NaOH where the diffraction peak is hardly observed, possibly indicating that a homogenous dispersion was obtained. The possible more uniform clay dispersion in the matrix of this composite could not be verified by TOM or AFM because a suitable surface could not be obtained due to the big amount of water bubbles formed during the resol condensation in the curing stage. Flexural properties of Res-NaOH and Phe-NaOH composites are reported in Fig. 4. The presence of bubbles in specimens catalyzed by NaOH, slightly decreased the flexural modulus and strength compared with TEA catalyzed systems. Furthermore, thermal stability is shown in Fig. 6. During polymerization, prepolymers catalyzed with NaOH mainly give methylene-type bridges (41) while using TEA dimethylene ether brigdes are formed (41). Consequently, composites with less oxygen content such as those synthesized with NaOH, resulted in more thermally stables mixtures.
Influence of FormaldekydelPhenol (F/P) Molar Ratio
The initial F/P molar ratio is one of the most important factors on the formation of phenolic resol resins (20), (38), (41). In the past, many authors reported the influence of the initial formaldehyde to phenol molar ratio in the synthesis of resol resins catalyzed with alkaline catalysts, such as sodium hydroxide and barium hydroxide (38), (39), (43). Our group investigated the influence of F/P ratio in resols synthesis catalyzed with triethylamine (20), (41) but no studies do exist on its effect during clay nanocompo-sites synthesis.
In this study, the range of F/P molar ratio for resol fabrication was covered by analyzing three resols synthesized at 80-C, with three initial formaldehyde to phenol molar ratios (F/P= 1.0, 1.4, and 2.0), catalyzed with triethylamine. Every initial formaldehyde/phenol mixture was adjusted to pH = 8 with a different amount of catalyst, depending on the initial pH of the mixture. Figure 7 shows X-ray diffraction of ta-phenylalanine-modified clay composites taking into account the F/P ratio (Phe-1 and Phe-2). Faint diffraction peaks were observed. In the case of Phe-1 composite, the area of the peak was slightly higher than for the other composites and shifted to higher angles. This behavior seems to indicate that at low formaldehyde content, part of the reactive process could occur inside the layers (intragallery) but the content of formaldehyde was not enough to overcome the attraction forces between the clays and fully separate them. This trend has been also verified below by TOM and AFM techniques (Fig. 8a and b). Thereby, intragallery reactions were favored (34), (44) but as the amount of formaldehyde was quickly finished, reactions were earlier stopped. Figure 8a shows the TOM image of Phe-1 composite where some MMT aggregates of around 5-10 [micro] m are observed in the matrix, whereas in Fig. 8b, AFM image, a few individual clay layers are seen, remaining the most of them in groups forming small stacked aggregates. Moreover, individual layers are hardly separated between them. In the case of composite Phe-2, the behavior seems to be slightly different. In this case, during polymerization, extragallery reactions catalyzed by TEA and intragallery reactions catalyzed by--COOH groups of L-phenyl alanine should proceed simultaneously to achieve the exfoliation state. In the presence of a high content of formaldehyde, both reactions initially might be parallel processes, but as the reaction continued, more easily accessible extragallery reactions would be favored (34), (44), thus leading to larger stacked agglomerates (Fig. 8c). As a result, intercalated aggregates and exfoliated sheets are also observed in Fig. 8d. It seems that the small shear forces exerted on PheMMT agglomerates during polymerization are able to overcome the attraction forces between the layers due to the weak forces that stack them together (9), (45), thus exfoliating the smaller stacks. Thereby, Fig. 8d shows a better dispersion of the clay in the matrix.
On the other hand, flexural properties for Phe-1 and Phe-2 composites are shown in Fig. 9. In neat resols, when formaldehyde content was increased (Res-2), the polymerization took place faster . Therefore, the presence of bubbles increased, thus decreasing flexural modulus and strength. In general, when an uniform clay dispersion was achieved, significantly improved both flexural modulus and strength in all the composites. The differences between neat and modified-clay resol composites were more significant for Phe-2 composite where the dispersion of the exfoliated clay was even better than for Phe composite. Anyway, Phe-1 composite showed slightly higher flexural properties due to the higher homogeneity of the resol network.
Similar conclusions can be extracted from thermal behavior shown in Fig. 10. The presence of the thermally stable MMT can act as a barrier to hinder heat diffusion and migration of degraded volatiles, thus delaying the decomposition rate (34). At increasing reactivity ratio (Res-2 and Phe-2 composites), the oxygen content increased (41), (46) thus resulting in less thermally stable mixtures compared with composites and matrices with lower content in formaldehyde. As above shown, this behavior can be mainly observed in the second stage of thermal decomposition when the char structure is formed. Res-1 matrix was the most thermally stable. In the case of Phe-1, the existence of oxygen groups increased when compared with the neat matrix owing to the clay modifier and its interactions with the reactives. This fact seems to be the responsible for the decrease in thermal stability.
Influence of Temperature of Synthesis
There are different studies concerning the influence of temperature in the resol prepolymer formation. Some of them were carried out employing fixed synthesis temperatures [19-21), (38), (40), (41), (43) and others combined steps during synthesis of the resin (47), (48). No works about its influence on clay-based nanocomposites formation do exist. In this study, triethylamine catalyzed resols with F/ P=1.4 synthesized at 55, 80, and 95 [degrees] C under reflux were investigated.
The first observed influence of the condensation temperature on the formation of the composite was related to the time needed to reach the final value of 1/1 g/g dilutability in water. The higher the condensation temperature, the shorter the synthesis time was. Phe-T55 composite showed a very slow evolution and the condensation time was around 5160 min. Phe-T95 reached the prefixed final point much faster (75 min), whereas Phe spent ~ 290 min. Figure 7 shows X-ray diffraction patterns of Phe-modified composites depending on the temperature of synthesis (Phe-T55 and Phe-T95 composites). As can be observed, weak diffraction peaks were observed and the interlayer spacing for all the composites was very similar (around 1.35 nm), although the area of the peak of Phe-T95 composite was slightly higher. This fact indicates that at higher temperature the amount of stacked clays increased (26), (27). This behavior was confirmed by TOM and AFM (Fig. I la and b). While at 55 [degrees] C and 80 [degrees] C, the curing could be controlled because condensation reactions proceeded slowly, at 95 [degrees] C, they occurred very fast and they were difficult to control due to formaldehyde and water evaporation despite using reflux. Thereby, curing took place less homogeneously and an increasing amount of bubbles and thick clay agglomerates were present in Phe-T95 matrix (Fig. 1 la and b). On the other hand, as can be observed in Fig. 11c and d, the dispersion of the clay in the matrix for Phe-T55 composite was different. When the synthesis was carried out at 55 C and in presence of TEA as catalyst, the polymerization seemed to be favored in the extragallery region. As a consequence, only superficial layers could be separated and thus the silicate layers appeared poorly dispersed in the matrix, remaining big agglomerates, as it is observed in Fig. 11c and d. Thereby, both composites showed poorer dispersions than composites synthesized at 80 [degrees] C.
As shown in Fig. 9, flexural properties were also affected by the condensation temperature. For Phe-T95 composite, the high temperature used during resin synthesis led to an increase in the polymerization rate of the network that generated flaws and the formation of nonhomogeneous resol network. Thus, though the modulus value was fairly constant compared with Res-T95 value mainly due to the presence of the clay, flexural strength was significantly decreased compared with this value for Res-T95 matrix and for Phe composite synthesized at 80 [degrees] C. On the other hand, for resol matrix synthesized at 55 [degrees] C, the low temperature used for the synthesis allowed the formation of a homogeneous network that conducted to better mechanical properties respecting to matrices syn-thesized at higher temperatures. This increase was more significant for the flexural strength values. Comparing Res-T55 matrix with the Phe-T55 composite, the existence of big clay aggregates due to the absence of polymerization in the intragallery region led to a slight decrease in flexural strength values.
Furthermore, the thermal stability of nanocomposites was also examined by TGA. Figure 12 indicates that the Res-T55 matrix was more thermally stable compared with neat matrices synthesized at higher temperatures (Res and Res-T95). The presence of less oxygen groups owing to the lower synthesis temperatures (41), (46) and the homogeneity of the resol network could be responsible for this behavior. During polymerization, prepolymers at 55 [degrees] C could form mainly methylene-type bridges. Consequently, resols synthesized at lower temperatures had less oxygen, thus resulting in more thermally stable mixtures. On the other hand, when high synthesis temperatures were used (95 [degrees] C) and taking into account that clays could lead to oxidation of the network (44), (48), thermal stability was decreased.
Resol-layered silicate composites were synthesized by intercalative polymerization of phenol and formaldehyde in the presence of differently modified clays. A few parameters of the synthesis of these resins and casting of composites were studied to find the optimum conditions to improve the intercalation/exfoliation of montmorillonite layers in phenolic resol matrices. On one hand, the nature of the clay modifier was concluded to be one of the key factors to obtain exfoliated nanocomposites. The choice of L-phenyl alanine and 6-aminocaproic acid as clay modifiers and the reactions between modifiers and phenolic resin resulted, at low clay concentration, an adequate method to obtain exfoliated nanostructures, as verified by different techniques. Thus, the significant improvement in the mechanical and thermal properties of these composites can be justified by homogenous clay dispersion.
Moreover, the catalyst used during the prepolymer synthesis was also assessed. Concerning mechanical properties-morphology relationships, NaOH was not a good suitable catalyst for these systems. In contrast, prepolymers catalyzed with NaOH resulted in more thermal stable composites compared with TEA catalyzed ones.
The influence of the reactivity ratio during composite curing was also investigated. For composites with low content in formaldehyde (F/P = 1 and T L 80 [degrees] C), reactive molecules only could be introduced within the most superficial layers, thus remaining unreacted stacked layers. In the case of composites with higher formaldehyde content (F/P = 2 and T -2= 80 [degrees] C), the polymerization of resol occurred faster. Thus, extragallery reactions could be favored leading to some clay agglomerates. Anyway, a homogeneous dispersion of the individual layers for the whole Phe2 composite was observed being more significant the differences in mechanical properties between neat and composite Phe-2. Nevertheless, Phe-1 composite showed the best flexural properties due to the higher homogeneity of the resol network and the presence of the clay. Thermal properties were also affected, being composite synthesized with F/P = 1 the most thermally stable due to its lower oxygen content and the network homogeneity.
Furthermore, the effect of temperature of the synthesis during the polymerization of the composite was also assessed. For the composite synthesized at the lower temperature (F/P = 1.4 and T = 55 [degrees] C), only most superficial layers were separated from the big agglomerates, thus negatively affecting both flexural and thermal properties. For composites synthesized at 95 [degrees] C and F/P = 1.4, the presence of flaws significantly decreased flexural strength, whereas the thermal stability was similar to the matrix synthesized at same conditions.
Thus, different conditions of curing could be chosen depending on the final application of the composite.
(1.) A. Usuki, M. Kawasumi, and Y. Kojima, J Mater Res., 8, 1174 (1993).
(2.) D. Garcia-Lopez, I. Gobernado-Mitre, J.F. Fernandez, and J.M. Pastor, Polymer, 46, 2758 (2005).
(3.) M. Tortora, G. Gorrasi, V. Vittoria, and E. Chiellini, Polymer, 43, 6147 (2002).
(4.) 0. Becker, R. Varley, and G. Simon, Polymer, 43, 4365 (2002).
(5.) C.H. Dan, M.H. Lee, Y.D. Kim, and J.H. Kim, Polymer, 47, 6718 (2006).
(6.) C.I.W. Calcagno, C.M. Mariani, S.R. Teixeira, and R.S. Mauler, Polymer, 48, 966 (2007).
(7.) Y. Rao, Polymer, 48, 5369 (2007).
(8.) X. Meng, Z. Wang, Z. Zhao, X. Du, W. Bi, X. Tang, Polymer, 48, 2508 (2007).
(9.) R. Pucciariello, V. Villani, F. Langerame, G. Gorrasi, and V. Vittoria, J Polym Sci Part B: Polym Phys., 42, 3907 (2004).
(10.) B. Chen and J.R.G. Evans, Polym Int., 54, 807 (2005).
(11.) K. Wang, L. Chen, J. Wu, M.L. Toh, C. He, and A.F. Yee, Macromolecules, 38, 788 (2005).
(12.) Y. Guan, L.X. Zhang, L.Q. Zhang, and Y.L. Lu, Polym Degrad Stab., 96, 808 (2011).
(13.) C. Kaynak, G.I. Nakas, N.A. Isitman, Appl Clay Sci., 46, 319 (2009).
(14.) S.S. Rahatekar, M. Zammarano, S. Matko, K.K. Koziol, A.H. Windle, M. Nyden, T. Kashiwagi, and J.W. Gilman, Polym Degrad Stab., 95, 870 (2010).
(15.) P. Kiliaris and C.D. Papaspyrides, Prog Polym Sci., 35, 902 (2010).
(16.) A. Gardziella, L.A. Pilato, and A. Knop, Phenolic Resins, 2nd ed., Springer-Verlag, Berlin (2000).
(17.) B.D. Park, B. Riedl, Y.S. Kim, and W.T. So, J Appl Polym Sci., 83, 1415 (2002).
(18.) J.E. Shafizadeh, S. Guionnet, M.S. Tillman, and J.C. Seferis, J Appl Polym Sci., 73, 505 (1999).
(19.) G. Astarloa-Aierbe, J.M. Echeverria, M.D. Martin, A.M. Etxeberria, and I. Mondragon, Polymer, 41, 3311 (2000).
(20.) G. Astarloa-Aierbe, J.M. Echeverria, M.D. Martin, A.M. Etxeberria, and I. Mondragon, Polymer, 41, 6797 (2000).
(21.) G. Astarloa-Aierbe, J.M. Echeverria, M.D. Martin, A.M. Etxeberria, and I. Mondragon, Polymer, 43, 2239 (2002).
(22.) C.C. Riccardi, G. Astarloa-Aierbe, J.M. Echeverria, and I. Mondragon, Polymer, 43, 1631 (2002).
(23.) H. Wang, T. Zhao, L. Zhi, and Y. Yan, Macromol Rapid Commun., 23, 44 (2002).
(24.) H. Wang, T. Zhao, and Y. Yan, J Appl Polym Sci., 92, 791 (2004).
(25.) H.Y. Byun, M.H. Choi, and 1.J. Chung, Chen, Mater., 13, 4221 (2001).
(26.) D.C. Wang, G.W. Chang, and Y. Chen, Polym Degrad Stab., 93, 125 (2008).
(27.) M.H. Choi, LJ. Chung, and J.D. Lee, Chem Muter., 12, 2977 (2002).
(28.) M. Natali, J. Kenny, and L. Torre, Compos Sci Technol., 70, 571 (2010).
(29.) H. Wang, T. Zhao, and Y. Yu, J Appl Polym Sci., 96, 466 (2005).
(30.) J. Pappas, K. Patel, and E.B. Nauman, J App! Polym Sci., 95, 1169 (2005).
(31.) C. Kaynak and C.C. Tasan, Eur Polym J., 42, 1908 (2006).
(32.) W. Jiang, S.H. Chen, and Y. Chen, J App! Polym Sci., 102, 5336 (2006).
(33.) L.B. Manfredi, D. Puglia, J.M. Kenny, and A. Vazquez / App! Polym Sci., 104, 3082 (2007).
(34.) M. Lopez, M. Blanco, J.A. Ramos, A. Vazquez, N. Gabilondo, J.J. del Val, J.M. Echeverria, and I. Mondragon, J Appl Polym Sci., 106, 2800 (2007).
(35.) A. Fudala, I. Palink6, and I. Kiricsi, Inorg Chem., 38, 4653 (1999).
(36.) X. Kornmann, H. Lindberg, and L.A. Berglund, Polymer, 42, 1303 (2001).
(37.) Y.H. Kim and D.S. Kim, Polym Compos., 30, 926 (2008).
(38.) M.F. Grenier-Loustalot, S. Larroque, and P. Grenier, Polymer, 37, 639 (1996).
(39.) M.F. Grenier-Loustalot, S. Larroque, P. Grenier, and D.I. Bedel, Polymer, 37, 939 (1996).
(40.) G. Astarloa-Aierbe, J.M. Echeverria, M.D. Martin, and 1. Mondragon, Polymer, 39, 3467 (1998).
(41.) N. Gabilondo, M. Larraiiaga, C. Petia, M.A. Corcuera, J.M. Echeverria, and 1. Mondragon, J App! Polym Sci., 102, 2623 (2006).
(42.) B. Machin, D. Hanton, J. Le Goff, and J.P. Tanneur, Fur Polym J., 22, 115 (1986).
(43.) M.F. Grenier-Loustalot, S. Larroque, P. Grcnicr, J.P. Leca, and D. Bedel, Polymer, 35, 3046 (1994).
(44.) M. Lopez, M. Blanco, A. Vazquez, A. Arbelaiz, N. Gabilondo, J.M. Echevenria, and I. Mondragon, Thermochim ACW, 467, 73 (2008).
(45.) T. Holopainen, L. Alvila, J. Rainio, and T.T. Pakkanen, J App! Polym Sci., 66, 1183 (1997).
(46.) H.W. Lochte, E.L. Strauss, and R.T. Conley, App! Polym Sci., 9, 2799 (1965).
(47.) A.W. Christiansen and L. Gollob, J App! Polym Sci., 30, 2279 (1985).
(48.) M.G. Kim, Y. Wu, and L.W. Amos, J Polym Sci Part A: Polym Chem., 35, 3275 (1997).
Correspondence to: I. Mondragon; e-mail: email@example.com
Contract grant sponsor: Ministerio de Education y Ciencia; contract grant number: MAT2006-06331; contract grant sponsor: Basque Country Governments (in the frame of Grupos Consolidados); contract grant number: IT-365-07; contract grant sponsor: SAIOTEK; contract grant number: S-PEO7UN39; contract grant sponsor: ETORTEK-inanoGUNE; contract grant sponsor: Eusko Jaurlaritza/Gobierno Vasco (Programa Realization de Tesis Doctorales en Empresas).
Published online in Wiley Online Library (wileyonlinelibrary.com).
[C] 2011 Society of Plastics Engineers
Marta Lopez, Miren Blanco, Maria Martin, Inaki Mondragon
'Materials + Technologies' Group, Escuela Politecnica, Department of Chemical and Environmental Engineering, Universidad Pais Vasco/Euskal Herriko Unibertsitatea. Pza. Europa 1, 20018 Donostia-San Sebastian, Spain
|Printer friendly Cite/link Email Feedback|
|Author:||Lopez, Marta; Blanco, Miren; Martin, Maria; Mondragon, Inaki|
|Publication:||Polymer Engineering and Science|
|Date:||Jun 1, 2012|
|Previous Article:||Influence of induced shear work on the properties of polyolefine nanocomposite pipes.|
|Next Article:||Model for a two-cavity coating die with pressure and temperature deformation.|