Printer Friendly

Polymerization compounding of polyurethane-fumed silica composites*.


The problems encountered during the development of new composite materials are often attributed to the lack of interactions between the substrate and the polymer matrix, an insufficient dispersion of the substrate, or a poor wetting of the particle by the polymer phase. Any of the aforementioned deficiencies will have a negative impact on the mechanical properties of the end-product. For thermoset materials, undesirable components adsorbed onto the solid phase, such as water, may also hinder the satisfactorily completion of the polymerization reaction. While there exists no single solution to all these problems, the polymerization compounding (PC) technique has been found to bring a significant improvement of the overall chemical and mechanical properties of selected composite materials. For example, the effective interfacial shear stress, which represents the ability of the interface to transfer a load from the matrix to the solid particles, can be increased by 50% using the PC approach in nylon-6,6/glass fibers materials [1, 2]. The PC approach can be applied to either thermoset or thermoplastic polymer matrix. With thermoplastic-based materials, improved composites of polyethylene/asbestos fibers, polyethylene/UHDPE fibers, polyethylene/Kevlar fibers, and nylon-6,6/glass fibers [3-7] have been obtained. When applied to thermoset polymers, this technique can also be referred to as an interfacial polycondensation reaction. The basic principle behind the process is to modify the substrate that is later incorporated in the composite materials by successively contacting it with the oligomers and the curing agents used in the formulation of the thermoset matrix. The existing hydroxyl sites on the particle can anchor the oligomer chains, which results in a particle coated with a thin film of thermoset polymer prior to its incorporation in the main polymer matrix. In the case of polyurethane, the surface of the solid substrate is first reacted with a diisocyanate and then with a polyol. The terminal hydroxyl groups on such modified solid substrates show a similar reactivity towards diisocyanate groups than the one observed in bulk between the polyol oligomer and curing agents. Under carefully controlled conditions, the polymer coating on the particle retains its reactive groups and thus participates in the bulk polymerization reaction. Such a process has been successfully used by Wu et al., [8] and Pittman et al [9]. In this work, we apply the polymerization compounding process to improve the properties of ultrafine silica-based polyurethane composites. Fumed or ultrafine silica is a widely used inorganic filler in the composite industries and accordingly, the improvement of the mechanical behavior of the composite has been for years the object of experimental investigations, many of them involving the chemical modification of the particle surface and/or polymer grafting [10-13]. For example, Li et al. [10] studied the chemical modification of silica silanol groups by an organo-functional silane coupling agent to obtain the amine and epoxide sites used for in situ ring-opening polymerization of e-caprolactam towards nylon 6 composites. The grafting of a more topologically complex macromolecule, polyamidoamine dendrimers, to silica particles was also conducted [11]. Other modes of polymerization have also been studied, such as free radical polymerization of 1-vinyl-2-pyrrolidone [12] and vinyl acetate on silica [13]. While the wide range of experimental treatments reported makes it difficult to compare the efficiency of these different treatments, it can be said as a general trend that mechanical strength and toughness were often significantly improved. The thermoset polymers of interest for the present study are made from a polycondensation reaction between a diisocyanate and a low molecular weight polymer. Only polyurethane binders were investigated. The polyurethane formation reaction is of the type:


[A.sub.x] + [B.sub.2] [??] [A.sub.(x-1)][B.sub.1] (1)

where [A.sub.X] is a polyol and [B.sub.2] the diisocyanate compound. In the case where one wants the filler to participate in a bulk polymerization and formulation of polyurethane composites, it is necessary to have some of the A's or some of the B's present at the outer surface of the particle. Fortunately enough, many substrates such as carbon powder or aluminum powder possess a significant amount of hydroxyl and carboxylic groups. These groups can react with isocyanate compounds to build chemical bonds between small oligomers and the solid substrate. However, under normal conditions of polyurethane composites preparation, this reaction at the silica/polymer interface is in competition with the more easily accomplished bulk polymerization process. As a consequence, only a small fraction of the silanol sites on the particles effectively react with the free isocyanate. To achieve the attachment of polymer chains on silica, the grafting process must be initiated prior to the bulk polymerization reaction. This is accomplished in solvent for a reaction time of a few hours. A typical reaction pathway for a fumed silica particle is depicted in Figure 1. A number of variables will affect the polymerization reaction. The primary one is obviously the density of active sites on the particle surface. These sites, however, can be increased in number by chemical treatments. For example, carbon particles can be soaked by an acid solution and then thoroughly washed with water and dried [1]. One of the possible outcomes of grafting a difunctional isocyanate to the particle is an intramolecular reaction with the free end of the molecule. This can generally be avoided by using smaller isocyanates and large excess of reactants in solution with respect to the amount of potentially reactive sites on the silica particles. If desired, the molecular weight of the grafted oligomers can be increased by the successive repetitions of isocyanate and polyol reaction cycles, and thus step-by-step building the desired polymer chains. The process has to be carried out in solvent in order to facilitate the separation of the particles from the unreacted reagents in excess at each step of the process. At the end of the grafting process, the dandling oligomer/polymer chains will have at least one free end carrying a reactive group (A or B) that will be able to participate in the bulk polymerization of the polyurethane binder during the composites preparation. In this work, investigations were conducted on fumed silica particles in grafting oligomers of toluene diisocyanate (TDI) and three different polyols: hydroxy-terminated polybutadiene (HTPB), bisphenol A (BA) and glycerol. The aim of this study was to compare the effects on the resulting composites of using these three structurally different polyols: a long flexible diol, a low molecular weight rigid diol, and a trifunctional small molecule.


Fumed Silica Grafting

The most critical aspect of this experimental work was the modification of the fumed silica by the interfacial polycondensation process. The fumed silica was obtained from Sigma Chemicals (98.8%) and dried at 60[degrees]C under vacuum for one full day before use. The specific area of the silica was reported as 390 [m.sup.2]/g. A characterization of the active sites on the particles showed that three hydroxyl sites per [nm.sup.2] were available to react with an isocyanate group. This is slightly less than the silanol density of four sites per [nm.sup.2] reported by Gallas et al [14]. In all experiments, the toluene diisocyanate (TDI) (2,4 80% Aldrich) was used as the difunctional isocyanate-terminated compound. The reaction of fumed silica with TDI was conducted in a glass-coated reactor with acetonitrile as a solvent and using dibutylthindilaurate (DBTDL) as a catalyst. The typical reaction time was 6 h at 70[degrees]C. In all syntheses, the concentration of TDI in solvent was calculated to maintain a NCO/OH ratio of 4 based on the amount of fumed silica to be treated. This large excess of isocyanate was required to reduce the probability of having both ends of the diisocyanate molecule reacting with the silica hydroxyl sites. Because the para-located NCO group is at least 2-3 times more reactive than its ortho counterpart, one may assume that most of the primary bonds between the TDI and the silica are formed at this position [15]. After the reaction, the excess of isocyanate was removed by liquid-liquid extraction and particle filtration. The second step of the reaction was also conducted in solvent under a nitrogen blanket to prevent side reaction with moisture. For the TDI-HTPB grafting reaction, one single process cycle produced 32% w/w coated fumed silica while the same process yielded a 23% w/w coating when carried out with BA-TDI. The fumed silica modified by a TDI-glycerol grafting process showed a 40% weight gain after soxhlet extraction. A comparison with theoretical grafting values is reported in Table 1. In each case, the reaction was conducted with an excess of polymer corresponding to a NCO/OH ratio of 0.25. Again, this excess was necessary to favor the reaction of all available free NCO-sites and to minimize the amount of backbiting.

Composite Materials

The composite materials were all prepared in batches of 150 g using a 2CV helicone mixer (Design Integrated Technology, Inc, VA, USA). The mixing cycle was conducted under vacuum to ensure a bubble free end-product and thus more reliable measurements of rheological and mechanical properties. Several different composites materials were prepared in a plasticized TDI-HTPB matrix. They were made using a set of four fillers: (a) unmodified fumed silica, (b) 24% TDI-BA-grafted fumed silica, (c) 32% grafted TDI-HTPB fumed silica, and (d) 22% grafted TDI glycerol fumed silica. Three variables were investigated; the effect of solid loading, the NCO/OH ratio, and the silica treatment. A detailed report of the composites prepared is given in Table 2. In each case, 24% of dioctyladipate (DOA) was used as a plasticizer. Once cast, the materials were cured for five days at 60[degrees]C to achieve the completeness of the polycondensation reaction.


Rheology and Mechanical Properties

The rheology of the cured composites was investigated via dynamic mechanical analysis. The measurements were carried out in two different instruments, a RSA dynamic spectrometer (Rheometrics) and a TA Instruments dynamic mechanical analyzer model 2980. In both cases, a dual cantilever configuration was used within the linear viscoelastic region of the materials. A few samples were analyzed on both machines and proved to be well reproducible. The tensile mechanical properties were evaluated according to the ASTM D638-M procedure on an Instron 4301 instrument.


Grafting, Kinetics, and Chemical Characterization

The grafting procedure when carried out as described in the experimental section of this paper lead to the preparation of three different batches of fumed silica, one for each polymerization treatment (i.e. TDI-BA, TDI-HTPB, and TDI-Glycerol). The corresponding weight percentage in polymer grafted on the raw fumed silica was evaluated by a thermogravimetry analysis (TGA) after conducting a Soxhlet extraction on each material. The extraction ensures that any organic material detected by TGA is effectively attached to solid particles. A typical TGA curve from the analysis of a TDI-BA-coated fumed silica is shown in Figure 2. It is seen that the Soxhlet extraction is important to remove non-grafted materials from the particles since the weight loss observed for the particles only washed with solvent after the reaction is much more pronounced. The TGA curve also confirms that in the grafted oligomers have a decomposition process that occurs two times, with on-set temperatures of 250[degrees]C and 325[degrees]C respectively. It is not possible, however, to identify from these results if each step of the decomposition corresponds specifically to either the TDI or BA segment of the grafted molecule. The complete results of the grafting experiments are reported in Table 1. The table also presents the theoretical polymer content for each treatment if one admits a complete conversion of the OH sites found on the silica, based on the accepted value of 4 sites/[nm.sup.2]. It is obvious from these results that only a fraction of the hydroxyl sites effectively participate in the grafting procedure. However, the amount of oligomers grafted depends on the type of polyols used. In the case of HTPB that is a long flexible chain, each of its ends can easily attach on the TDI molecules grafted on the particles at the first step of the reaction. In addition, because of its size, an already grafted HTPB can prevent other HTPB chains to react with free sites around it. On the other hand, the glycerol molecule does not carry this type of hindrance effect and as a result, a complete grafting of the available sites on the silica is observed. In order to confirm this hypothesis, the free hydroxyl content of the fumed silica grafted with the thermoset polymers was evaluated to determine if chemical bonds could really be established during the compounding in the polymer matrix. By virtue of the successive urethane bonding sequence, one should expect to find, after each normal cycle (one cycle is contacting the powder in a diisocyanate solution followed by a treatment in the diol solution), the same density of hydroxyl sites on the grafted powder as conceptually shown in Figure 1. The number of free hydroxyl sites on the silica was evaluated by a chemical procedure based on a reaction with an alkyl aluminum compound [16]. In the case of TDI-BA-coated silica, the hydroxyl functionality at the surface remained the same for the raw silica while for the TDI-HTPB-coated silica, no free hydroxyl group was detected. This difference is probably due to the greater flexibility of the long HTPB chain compared with the shorter BA molecule. Backbiting is thus more likely to occur with the flexible HTPB rather than with the more rigid BA molecules. The reaction rate of the grafting process in acetonitrile was studied using FTIR spectroscopy. The results for the TDI-glycerol reaction are shown in Figure 3. It is found that the primary reaction between the silanol groups on the silica particles and the isocyanate proceeds at a slower pace than the second urethane bonds formation during the reaction between the TDI and the glycerol. For the first reaction, it took more than one day to reach a point where the isocyanate concentration is almost stable, while for the second one the reaction occurred over a few hours only. In both cases, the kinetics was found to be of first order in isocyanate concentration from the trend-lines obtained from the conversion against time data. The longer time required to carry out the grafting of the isocyanate on the silica confirms that under normal mechanical compounding conditions, where the fumed silica is dispersed in a reactive isocyanate/polyol blend, the silanol groups do not have enough time to participate in the bulk polymerization process due to the competition with the more reactive liquid polyol. Further evidence of the presence of a coating on the silica particles was obtained by studying them when placed in a highly humid environment. Being a hygroscopic material, the fumed silica has a great tendency to uptake water from air moisture. The effectiveness of a polymer coating to prevent this process was evaluated by placing TDI-HTPB coated particles in a control chamber regulated at 60% in relative humidity. It was found that the non-coated fumed silica particles absorbed four times more water than the coated particles, with an equilibrium value of 8% w/w instead of 2% w/w for the modified silica.



One of the motivations in developing a polymerization compounding process was to improve the mechanical properties of the resulting composites. Some of the dynamic mechanical properties (DMA) of the materials obtained in the linear viscoelastic regime are reported at Figures 4 to 12. Several composites containing a higher weight fraction of solids were first characterized. Figures 4 and 5 show the effect of coating on the storage modulus (E') for 10% and 22% silica-loaded composites respectively. Both types of coated particles produce an improvement in the modulus, but while the TDI-HTPB coating offers only a marginal augmentation, the TDI-BA material gives a ten-fold increase. For TDI-BA composites, the same increase is seen in the glassy and viscous state, while the better properties of TDI-HTPB materials are only observed above the glass transition temperature ([T.sub.g]). This difference demonstrates that the polymer-substrate interactions are not the same for both types of modified silica particles. It confirms that the free hydroxyl sites on the TDI-BA-coated silica participate cooperatively with the liquid HTPB to the bulk polymerization process during the compounding of the thermoset materials. On the other hand, the improvement observed on the TDI-HTPB-coated particles is most likely only due to a better wetting of the substrate by the matrix since very few free hydroxyl groups are found on the surface of the silica after the grafting process. In this case, the polymer coating more or less acts as a wetting agent. Moreover, the comparison of the results between the 10% and 22% loaded materials demonstrates that only a slight increase in the modulus is obtained by using higher concentration of solids. This is probably the consequence of a poor dispersion of the solid phase in the more viscous formulation. In fact, a 22% TDI-HTPB composite even shows a decrease in the modulus and it was also particularly difficult to process in a defect-free material. The variations in [T.sub.g] are presented in Figures 6 and 7. Because of the presence of the DOA plasticizer, all formulations have [T.sub.g] values below -70[degrees]C, which is fairly good for polyurethane materials. However, the composites loaded with TDI-HTPB-coated silica have a slightly higher [T.sub.g] as well as a larger peak of the loss tangent, which is a result of improved wettability and thus dispersion of particles in the binder phase.






Because of the difficulty to obtain a perfect dispersion of the fumed silica at high solids content, a series of composites were prepared with only 2.5% and 5% w/w of silica. The processing of these formulations was easy to carry out and the high specific surface of the silica was preserved. Under such conditions, the effects of the particles coating and the interactions of the latter with the binder phase are more easily noticeable. To conduct a comparison of the rheological properties of these thermoset materials, it is important to consider the results from composites obtained at the optimum NCO/OH ratio. In a polyurethane formulation, the best mechanical properties should theoretically be observed at the stochiometric ratio (i.e. NCO/OH value of 1). The NCO/OH ratios for these experiments have been established only on the basis of the hydroxyl groups found in the HTPB liquid prepolymers. However, because of the presence of traces of water in the particles and polymer, and also due to the fact that available OH groups on the coated particles are not considered, the best set of rheological and mechanical properties should normally be observed at NCO/OH values greater than one. Moreover, beyond a certain level of NCO/OH, a decrease in the properties is expected to occur since the unreacted amount of the excess component, in this case the isocyanate, acts as a plasticizer. This typical behavior is well depicted in Figure 8 where the equilibrium storage modulus of 5% w/w unmodified silica composites gradually increases up to a NCO/OH ratio of 1.4, and then starts to decrease for a NCO/OH of 1.5. In Figure 9, we also report G' values but for 5% w/w TDI-BA-grafted composites. In that particular case, the equilibrium storage modulus continues for values of NCO/OH as high as 1.7, which was the highest ratio used for this series of samples. However, since it is also seen that results for NCO/OH of 1.6 are almost identical, it is likely that these correspond to the optimum formulations and that a reduction of E' would be observed at higher NCO/OH ratios. The fact that this optimum is reached at the NCO/OH value of 1.7 rather than 1.4 for the non-grafted fumed silica confirms that an additional amount of OH groups, available from the grafted BA molecules, participated to the bulk polymerization process. A similar behavior is also observed for composites containing only 2.5% w/w of TDI-BA-grafted silica, as shown in Figure 10. The effect of the grafting process and improved particles/binder dispersion is also depicted by the loss tangent results of 2.5% TDI-Gly composites presented in Figure 11. The grafted glycerol should bring two additional OH groups per molecule to the polymerization process and thus higher optimal NCO/OH ratio is to be expected. It is seen that the higher [T.sub.g] and the broader tan delta peak is found at a NCO/OH value of 1.7, but nothing indicates that this corresponds to a maximum because the results at NCO/OH = 1.5 are significantly lower. Finally, we present in Figure 12 the storage modulus of the 2.5% w/w composites based on each type of silica used in this study. Even at low concentrations of solids, the enhancement due to the grafting process is clearly seen as the equilibrium E' for both types of grafted silica is three times higher than the one observed with nonmodified silica composites.





Tensile Mechanical Properties

The mechanical properties of the fumed silica filled composites are presented in Tables 3 and 4, respectively. Two sets of data are reported, one series for low solids content composites at 2.5% w/w, Table 3, and one series used for highly loaded composites corresponding to 10% and 22% w/w presented in Table 4. The general trend of these results is in agreement with the observations gathered from dynamic measurements. For the composites made from 10% and 22% silica, the use of coated fumed silica translates by a higher elastic modulus, but elongation at break decreases slightly. Accordingly, this behavior is explained by the fact that while the greater stiffness of composite comes from the better interaction between the coated substrate and the polymer matrix, the breaking process is still governed by the strength of the continuous polymer phase [16]. Again, at similar silica concentrations, the BA-TDI-coated particle gives the larger increase in the modulus and thus confirms the ability of the free hydroxyl groups of the grafted BA to bond with the HTPB matrix by urethane links. For these results, the Young moduli reported are average results of eight tensile experiments. The overall variance in the data is by ascending order: silica, BA-TDI-coated silica, and TDI-HTPB-coated silica. From our experimental observations, the formulation using TDI-HTPB-coated particles were also the most difficult to process and thus the increased variance confirms the dispersion problem of these materials. This also explains why the improvement obtained with the modified silica remains modest given the amount of solids used. More enhancements in the results of tensile measurements are found for the composites prepared using only 2.5% w/w of silica. The smaller solids content allowed for a better dispersion, as demonstrated by the fact that even for untreated silica, these composites show a larger stiffness than their 10% w/w and 22% w/w counterparts. For modified silica particles, as observed from dynamic measurements, the storage modulus data demonstrated great improvements with increasing NCO/OH ratios. However, for the tensile strength tests, the increase in the elastic modulus is somewhat larger for the TDI-Glycerol-grafted silica composites than for the TDI-BA silica-based materials. For the former case, the optimum Young's modulus is almost twice as much as the one observed with nonmodified silica composites. This improvement is even better than the effect seen at higher concentration of silica and again, it confirms that the low concentrations of solids allow for a better dispersion of the particles.

Concluding Remarks

It has been shown that a reactive coating of fumed silica particles may be achieved with thermoset polymers by using an interfacial polycondensation technique. Some desirable effects of the coating are improved dynamic mechanical properties, improved elastic modulus under tension conditions and reduction of moisture uptake. When the interfacial polycondensation process is conducted in such a way that active sites are left on the coating surface, these sites participate in the bulk polymerization process and chemically bond with the polymer matrix. This is the case when small polyol molecules such as BA and glycerol are grafted. When a long flexible polymer chain such as HTPB is used, a backbiting reaction during the second step of the grafting process leaves only a few hydroxyl groups to participate in the bulk polymerization process: For the BA- and glycerol-grafted silica, the presence of available OH groups for the composites preparation is confirmed by the larger amount of isocyanate required to achieve optimal properties of the thermoset materials. We have found that even if small silica particles can be coated individually, they nevertheless have a tendency to agglomerate during the process. This dispersion problem is likely to be responsible for the marginal improvement in mechanical properties observed under higher silica contents. This problem is alleviated with less concentrated composites. Accordingly, when only 2.5% of grafted silica is used, the polymerization compounding approach demonstrates its full potential to obtain better composites by an enhancement of the particles-binder interactions.


1. H. Salehi-Mobarakeh, A. Ait-Kadi, and J. Brisson, Polym. Eng. Sci., 36, 778 (1996).

2. H. Salehi-Mobarakeh, A. Ait-Kadi, and J. Brisson, J. Mater. Sci., 32, 1297 (1997).

3. Q. Wang, S. Kaliaguine, and A. Ait-Kadi, J. Appl. Polym. Sci., 44, 1107 (1992).

4. Q. Wang, S. Kaliaguine, and A. Ait-Kadi, J. Appl. Polym. Sci., 45, 1023 (1992).

5. Q. Wang, S. Kaliaguine, and A. Ait-Kadi, J. Appl. Polym. Sci., 48, 121 (1993).

6. W. Feng, A. Ait-Kadi, J. Brisson, and B. Riedl, Polym. Comp., 24, 512 (2003).

7. Z. Yu, A. Ait-Kadi, and J. Brisson, Polym. Eng. Sci., 31, 1228 (1991).

8. Z. Wu, C.U. Pittman, and S.D. Gardner, Carbon., 35, 59 (1996).

9. C.U. Pittman, G.R. He, B. Wu, and S.D. Gardner, Carbon., 35, 333 (1997).

10. Y. Li, J. Yu, and Z.X. Guo, J. Appl. Polym. Sci., 84, 827 (2002).

11. A. Kasseh, A. Ait-Kadi, B. Riedl, and J.F. Pierson, Polymer, 44, 1367 (2003).

12. V. Nguyen, W. Yoshida, J.D. Jou, and Y. Cohen, J. Polym. Sci. Part A: Polym. Chem., 40, 26 (2002).

13. V. Nguyen, W. Yoshida, J.D. Jou, and Y. Cohen, 87, 300 (2003).

14. P. Gallas, J.C. Lavalley, A. Burneau, and O. Barres, Langmuir, 7, 1235 (1991)

15. C. Dubois, S. Desilets, A. Ait-Kadi, and P.A. Tanguy, J. Appl. Polym. Sci., 58, 827 (1995).

16. C. Vajrasthira, T. Amornsakchai, and S. Bualek-Limcharoen, J. Appl. Polym. Sci., 87, 1059 (2003).

Charles Dubois, Mahmoud Rajabian

CREPEC, Department of Chemical Engineering, Ecole Polytechnique de Montreal, Montreal H3C 3A7, Quebec, Canada

Denis Rodrigue

Department of Chemical Engineering, Laval University, Quebec G1K 7P4, Canada

*The authors would like to dedicate this work to the memory of late professor A. Ait-Kadi. His passion and talent continue to inspire us.

Correspondence to: Charles Dubois; e-mail:

Contract grant sponsor: Defence RD Canada--Valcartier; contract grant sponsor: FQRNT--"Support to Research Team" program, Ministry of Education, Province of Quebec.
TABLE 1. Chemical analysis of grafted powders.

 Grafted polymer Expected amount
 after Soxhlet of grafted
 extraction polymer
 (% w/w) (% w/w)

TDI-BA-grafted fumed silica 23 51
TDI-HTPB-grafted fumed silica 32 86
TDI-Glycerol-grafted fumed silica 41 40

TABLE 2. Composites prepared to study the effect of the silica

 Ratio [NCO]/[OH]
Formulation 1.1 1.2 1.3 1.4 1.5 1.6 1.7

HTPB X X X X X -- --
HTPB/2.5% Si X X X X X -- --
HTPB/5% Si X X X X X X --
HTPB/2.5% Si-TDI-BA X X X X X X --
HTPB/2.5% Si-TDI-Glycerol -- -- X X X -- X
HTPB/5% Si-TDI-Glycerol X X X X X -- X

TABLE 3. Tensile mechanical properties of low concentration fumed silica

 Elasticity Maximum
 modulus Elongation at stress
 (MPa) break (%) (MPa)

Composite 2.5% Si, [NCO]/[OH] = 1.1 0.548 231.4 0.572
Composite 2.5% Si, [NCO]/[OH] = 1.2 1.53 136.6 0.792
Composite 2.5% Si, [NCO]/[OH] = 1.3 1.37 132.2 0.928
Composite 2.5% Si, [NCO]/[OH] = 1.4 1.72 85.8 0.906
Composite 2.5% Si, [NCO]/[OH] = 1.5 1.68 97.4 0.974
Composite-grafted 2.5% Si/TDI-BA, 0.367 352.3 0.561
 [NCO]/[OH] = 1.1
Composite-grafted 2.5% Si/TDI-BA, 1.35 128.5 0.919
 [NCO]/[OH] = 1.2
Composite-grafted 2.5% Si/TDI-BA, 1.39 113.6 0.931
 [NCO]/[OH] = 1.3
Composite-grafted 2.5% Si/TDI-BA, 1.49 114.9 0.991
 [NCO]/[OH] = 1.4
Composite-grafted 2.5% Si/TDI-BA, 2.06 87.0 1.093
 [NCO]/[OH] = 1.5
Composite-grafted 2.5% Si/TDI-BA, 2.18 87.7 1.246
 [NCO]/[OH] = 1.6
Composite-grafted 2.5% Si/TDI-Gly, 1.66 165.5 2.740
 [NCO]/[OH] = 1.3
Composite-grafted 2.5% Si/TDI-Gly, 2.50 95.47 1.422
 [NCO]/[OH] = 1.5
Composite-grafted 2.5% Si/TDI-Gly, 2.88 92.3 1.581
 [NCO]/[OH] = 1.7

TABLE 4. Tensile mechanical properties of highly loaded fumed silica

 Elasticity Elongation Maximum
 modulus at break stress
 (MPa) (%) (MPa)

Composite 10% Silica, [NCO]/[OH] = 1.2 0.11 465 0.42
Composite 10% HTPB-coated Silica, 0.64 67 0.25
 [NCO]/[OH] = 1.2
Composite 10% BA-coated Silica, [NCO]/ 0.99 68 0.43
 [OH] = 1.2
Composite 22% Silica, [NCO]/[OH] = 1.2 0.9 400 3.27
Composite 22% HTPB-coated Silica, 1.83 123 1.82
 [NCO]/[OH] = 1.2
Composite 22% BA-coated Silica, [NCO]/ 2.28 41 0.71
 [OH] = 1.2
COPYRIGHT 2006 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Dubois, Charles; Rajabian, Mahmoud; Rodrigue, Denis
Publication:Polymer Engineering and Science
Geographic Code:1CQUE
Date:Mar 1, 2006
Previous Article:Homopolymerization of epoxy monomers initiated by 4-(dimethylamino)pyridine.
Next Article:The draw ratio-Deborah number diagram: a useful tool for coating applications.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters