5.4:. ".215... .5 {$1. . «Ga I 1-. V I}. . 39:. 4x. 2 ‘3. "5:351 I -s 2 . mu .2 s x ‘ 9.4.3. ‘ , , ‘ . 3%.» 2.2.3.2. . , 4.3.x n ‘ , es 1.: .. 3-32.: :51...” . a. .. 2,. .‘LW,....~\. 3. . 21*-monn (so meq)* l Claytone-AF cr-AF [(CH3)2N(C13H37)2]+-mont. (100 meq)* Claytone-PS CF-PS [(CH3)(C6H5CI-12)N(C13H37)2]+-m0nt. ,, Montmorillonite 1-2 MMT 1—2 Freeze-dried Na+-montmorillonite (1-2 pm) Montmorillonite 0.5-1 MMTO.5-1 Freeze—dried Na+-montmorillonitc (05-1 pm) Montmorillonite 0.1-0.5 MMT 0.1-0.5 Freeze-dried Na+-montmorillonite (0.1-0.5 um) Alisibronz Mica-4 ABM-4 Water-ground muscovite *The miliequivalent of the alkylammonium ion used for 100 g of montmorillonite. Specimen Preparation The smectite clays and muscovite mica were preheated at 150 °C overnight to remove most surface and interlayer water. Depending on the weight percent, the appropriate amount of the preheated mineral was portion by portion added to 15.0 g of epoxy resin EPON-828 in a 250 mL plastic beaker on a hot plate at 75 °C. The mineral-epoxy mixture was magnetically stirred at 75 °C for 1 h. The weight of MPDA was included in the total weight used to calculate mineral weight percents. Meanwhile, 2.175 g (or 14.5 parts per 100 g of the resin (phr)) of MPDA (mp, 62 °C) was melted in a 50 mL plastic beaker at 75 °C. The melted MPDA was then poured into the mineral-epoxy mixture and it was again stirred at 75 °C for 2 min. The whole mixture was then degassed in a VWR Scientific 1410 vacuum oven (~ 25 torr) at 75 °C for 25 min. As for HPM-20 and MPCC at 25 and 35 wt%, the processing temperature was gradually decreasing to ~ 60 °C to prevent premature cures during degassing. The degassed mixture was poured into a silicon rubber mold and cured at 75 °C for 2 h and 125 °C for another 2 h, a ramp rate of 3 °C/min being used. After the oven was gradually cooled down to room temperature (~ 1.5 h), the composite specimens (1.50 x 4.00 x 25.40 mm3) were taken from the oven and polished until the deviation of thickness was :l: 0.01 mm. Characterization Tensile testing was performed at ambient temperature by using a United Testing System (UTS). Neither strain gage nor extensometer was used and the strain rate was 0.51 mm/min (or 0.02 in/min). The relative ultimate tensile strengths (RTS) and Young's moduli (RTM) of clay- and 8 mica-epoxy composites were based on those of the cured epoxy resin tested in the same experimental conditions. SEM micrographs of fractured surfaces of the composite specimens were obtained by a JEOL SEM T-330 scanning electron microscope, operating at an accelerating voltage of 15 kV and in the backseattered electron imaging (BEI) mode. The fractured composite specimens after tensile testing were used without any further treatment. The nonconductive composite specimens were coated with ~ 20 nm gold to prevent charging (or imaging artifacts). Gold-coating was performed by a Polaron sputter coater, using argon as the medium gas. The coater voltage and current were set at 2.5 kV and 25 mA. The target-to-specimen distance and exposure time were 50 mm and l min, respectively. Diffuse-reflectance FTIR studies were carried out by using a Perkin- Elmer FTIR 1850 Fourier transform infrared spectrometer, using KBr powder as the reference (or background). Powdered composites were obtained by filing the tested, fractured specimens. To minimize the deviation of the linear intensity-concentration relationship, the powdered composites were ground down to 10-20 um in particle sizes and then diluted to 5 wt% with KBr powder. Spectra subtraction was performed over the frequency range 450 cm‘1 to 4000 cm'l. X-ray powder diffraction (XRD) patterns were obtained by a Philips XRG-3000 X-ray diffractometer equipped with a Ni-filtered Cu-Ka radiation source (1: 1.5406 and 1.5444 A). The tube voltage and current were set at 30 kV and 20 mA, respectively. The preheated clays and mica were evenly dispersed on vacuum grease-coated glass slides, which were then tapped to obtain thin but uniform sample thickness. However, the tested, fractured specimens of the cured epoxy resin and composites were 9 placed on the sample holder (or rotating analysis disk) directly. The samples were then recorded by monitoring the diffraction angle 20 from 2° to 45°. The scanning speed and the step size were 2°lmin and 0.025°, respectively. XPS studies were carried out by using a PHI 5400 X-ray photoelectron spectrometer equipped with a Mg Ka standard source (PHI 04-548) and an Al Kc toroidal monochromatic source (PHI 10-410). The monochromatic Al source operating at 600 W (15 kV, 40 mA) was used for all the sample analyses. The preheated clays and mica were dispersed on analysis stubs by using double-sided tapes. The XPS spectra were obtained at a system base pressure of ~ 10'9 mbar. The lens was set in the large area, small solid angle mode, and the size of the analysis area was set for a 3.3- mm-diameter circle. Data points were collected in the fixed analyzer transmission mode by using a position sensitive detector and a 1800 hemispherical analyzer. Pass energies were set at 178.95 eV for the elemental survey scans (0-1200 eV) and at 35 eV for the higher resolution narrow scans of the individual elemental regions. The C1s binding energy of the carbon peak was set to 285.0 eV for calibration purposes. Surface areas of the clays and mica were measured by a Quantachrome Quantasorb Jr. surface area analyzer, using N2 and He as adsorbate and carrier gas, respectively. The samples were degassed at 125 °C for 12 h, and the surface areas of the samples were determined by using three-point BET plots. Cation-exchange capacities of the clays and mica were determined by a Fisher Scientific Accumnet 750 selective ion analyzer and the method proposed by Busenberg and Clemency (17). 10 The pH values of 7 wt% clay- and mica-aqueous suspensions were measured by the same selective ion analyzer at room temperature, a pH electrode being used instead. The pH value of the slightly acidic deionized water used was adjusted to 7.0 with 0.1 N NaOH. C. RESULTS AND DISCUSSION Survey A survey (Table 1.2) indicates that the tensile strengths of 2 wt% clay-epoxy composites depend on the pre-treatment or more precisely, surface treatment of the smectite clays. Composite strengths increase in the order of surface treatment, freeze-drying < cationic surfactant alkylammonium chloride < anionic dispersant Na-polyacrylate. Obviously, interlayer cations Na+ and Ca2+ in Na-polyacrylate treated montmorillonites do not play an important role in determining composite strengths. This also holds true for various interlayer quaternary alkylammonium ions in alkylammonium-mantmorillonites. Besides, the differences in particle sizes among freeze-dried Na+-montmorillonites (0.1-2 um) are too small to have any significant effects on the composite strengths. Therefore, Volclay HPM-ZO, Claytone-PS and freeze-dried Na+-montmorillonite 1-2 pm have been chosen as representatives of the three types of smectite clays for further investigation. The reference is a water-ground muscovite mica, Alsibronz Mica-4. Tensile Moduli Recently, Ahmed and Jones (18) published a comprehensive review of reinforcement theories for particulate-polymer composites and concluded 11 Table 1.2. Tensile Strengths and Young's Moduli of 2 wt% Clay- and Mica-Epoxy Composites Tensile Strength Rm Mineral (ksi) (96) (ksi) (%) MPCC 12.36 :1: 0.09 96 211 i 4 103 HPM-20 12.18 3:020 95 212:1:4 104 CD40 11.08 i 0.26 86 206 i l 101 CT-AF 10.82 10.18 84 213i7 104 CT-PS 10.67 :1: 0.29 83 212 i 5 104 WT 1-2 9.95 i 0.17 77 208 :l: 6 102 WT 05-1 9.94 :l: 0.35 77 212 :l: 4 104 MNIT 0.1-0.5 9.58 :l: 0.37 74 203 :l: 3 100 ABM-4 10.81 :1: 0.36 84 235 :1: 8 115 *Relative ultimate tensile strengths and Young's moduli are based on those (12.87 :1: 0.01 and 204 :t 6 ksi) of the cured epoxy resin EPON-828 tested at 20 :l: 2 °C. 12 that a fresh theoretical approach together with systematic experimental studies was needed. The previously proposed models with assumptions and oversimplifications often require empirical parameters and/or constants for curve fitting and thus have limited applications. In general, mechanical properties of particulate-polymer composites depend on the filler, polymer matrix, and filler-matrix interface or interphase (19). As for the filler, the size, shape, size distribution, surface area, aspect ratio, orientation (or anisotropy). and aggregation are factors affecting composite mechanical properties. Moreover, the filler-matrix adhesion and particle-particle interactiOns of aggregates in the matrix cannot be experimentally quantified (18). This makes it more difficult to develop a theoretical model that can satisfactorily predict mechanical properties, especially mechanical strengths of particulate-polymer composites. Summarized in Tables 1.2 and 1.3, the relative tensile moduli of the clay- and mica-epoxy composites show that the smectite clays and muscovite mica investigated enhance the tensile modulus of cured epoxy resin EPON-828. Moreover, the composite modulus increases with increasing the mineral concentration. These are because the epoxy matrix is replaced by higher modulus clay (111 GPa or 1.61 x 104 ksi) and mica (172 GPa or 2.49 x 104 ksi) (20). Clay particles and mica flakes restrict molecular motions and reduce the glassy-state free volume of the epoxy matrix. This makes the stress relaxation of the epoxy matrix more difficult and results in. an increase in the modulus. The clay concentration being equal, the composite moduli remain constant regardless of the smectite clays filled. Mica is the better modulus reinforcement than clay because of the greater Young's modulus and aspect ratio of mica, according to Nielsen equations (20) for particulate-filled composites: 13 Table 1.3. Relative Tensile Strengths and Young's Moduli of 7 wt% Clay- and Mica- Epoxy Composites ReTative TensileStrengfh‘ Relative finsileMtfitlus‘ Mineral (%) (%) HPM-20 921:2 108:1:4 CT-PS 80:1:4 lO7:l:5 MMT 1-2 60:1:4 106:1:5 ABM-4 821:3 134i8 *Based on the ultimate tensile strength and Young's modulus of the cured epoxy resin EPON-828 tested in the same experimental conditions. §L=1+AB¢ E, l-BCO E, l E ‘P(1 (P) A=k—l, B=Tu—; C 1+ n _L+A ¢i E where EC, Em, and Ef are the moduli of the composite, matrix, and filler, respectively; k is the Einstein coefficient; (p is the volume fraction; and (pm is the maximum packing volume fraction. The Einstein coefficient describes the shape (or aspect ratio) effect of the filler on viscosity and it can be determined by the rheological method. Other things being equal, the fiber and the sphere have the highest and the lowest k values, respectively. As shown in Figure 1.3, the SEM micrographs of fractured surfaces of the composite specimens show that aspect ratios of embedded mica flakes are in the range of 20 to 50 and those of embedded clay particles are less than 5. Figure 1.4 shows a comparison of HPM-20 and ABM-4 with respect to the relative composite modulus as a function of the mineral concentration up to 25 wt%. The linear relationship of the relative composite tensile modulus (E) and. the mineral weight percent (0) for HPM- 20 and ABM-4 can be expressed by the following empirical equations: %E,,,.,,,o =100+1.33¢ 9615”,, = 101+ 4.540 15 Figure 1.3. Scanning electron microscopy (SEM) micrographs of fractured surfaces of epoxy composites containing 7 wt% (a) Volclay HPM-2, (b) Claytone-PS, (c) freeze- dried montmorillonite 1-2 um, and (d) Alisibronz Mica-4. ODWX 6...: . .. (a _ . OOWX Dxnn a .. 2 s r . v e......._.“._u.\... m... a. «Q ~ a ... s. .N . rt .. 16 240 A 220‘ 9 AlsibIszica-4 5!; o VolclayHPM—20 3 200- 3 “g 'F E 180‘ .2 '2 [2 160i 0 .2 ‘3 140- a P m 120‘ 1w ' l ' l v I v 0 10 20 30 40 Mineral Concentration (wt%) Figure 1.4. Relative tensile modulus as a function of the mineral concentration for epoxy composites containing (.) Volclay HPM-20 and (o) Alisibronz Mica-4. 17 Tensile Strengths The relative tensile strengths (Tables 1.2 and 1.3) of the clay- and mica-epoxy composites show that the smectite clays and muscovite mica investigated are not reinforcements but fillers for cured epoxy resin EPON- 828. Furthermore, the composite strength decreases with increasing the mineral concentration. As shown in the SEM micrographs of fractured surfaces of the composite specimens in Figure 1.3, clean surfaces of clay particles and mica flakes suggest that there is no strong adhesion between the mineral and the matrix. Consequently, the fracture occurs at the mineral-matrix interface and the load can not effectively pass along to the mineral. The clay concentration being equal, the composite strength varies whereas the composite modulus remains constant within experimental errors. This is because the tensile strength is determined by the weakest point of the specimen whereas the tensile modulus is a bulk property. Table 1.3 and Figure 1.3 also show that the clay composite strength is well correlated with the clay dispersion in the epoxy matrix. Arising from poor dispersion, mineral aggregates would act as weak points for mechanical failure and reduce composite strengths. The degree of aggregation should increase with increasing the mineral concentration. Despite their small sizes (1-2 um), freeze-dried Na+-montmorillonite particles flocculate and form aggregates up to 50 um. As a result, MMT 1-2 exhibits the lowest composite strength among the representative smectite clays and muscovite mica. On the contrary, HPM-20 exhibits the best dispersion and the highest composite strength. In Figure 1.5 the SEM micrographs show that even at 25 and 35 wt% most of RPM-20 particles and aggregates are less than 10 um and still relatively well dispersed in the epoxy matrix. Figures 1.6 shows a comparison of HPM-20 and ABM-4 18 Figure 1.5. SEM micrographs of fractured surfaces of epoxy composites containing (a) 2, (b) 7, (c) 25, and (d) 35 wt% Volclay I-IPM-20. ”AOOOO EiOu Daoeax Dxnu ..._... s... egfifié ($3325.94. fl! r? ., , a... . ... .$. .~ » .. . I Q . s .. .- 4 . r. .p..\. .b 0' 19 100 . 0 Volclay RPM-20 A 9 Alsibronz Mica-4 e ”1 A: l H on t: a so- In .2 ‘ U) E o I" 70 - o .2 , 3 E a) q 4 50 ' I ' I V I ' 0 10 20 30 40 Mineral Concentration (wt%) Figure 1.6. Relative tensile strength as a function of the mineral concentration for epoxy composites containing (.) Volclay HPM-ZO and (o) Alisibronz Mica-4. 20 with respect to the relative composite strength as a function of the mineral concentration up to 25 wt%. The relative composite strength of ABM-4 appears to level off at high mica concentrations presumably because mica has relatively high aspect ratios. It should be noted that ABM-4 and MMT 1-2 are not surface-treated with any adhesion-enhancing agents such as surfactants, dispersing agents, and coupling agents. Processing Problems During specimen preparation it was easier to process the muscovite mica than any of the smectite clays. Na-polyacrylate treated montmorillonites (MPCC and HPM-20) accelerated the epoxy-amine cure and reduced the gelation time (or working life). Long-chain alkylammonium-montmorillonites (CTs) produced tenaciously entrapped air in clay-epoxy mixtures. These processing problems are due to the anionic dispersant Na-polyacrylate and cationic surfactant alkylammonium chloride on clay surfaces, respectively. The degassing problem of Claytones might be somewhat alleviated if they were washed with deionized water or alcohol after cation-exchange reactions or cationic surfactant treatment. Freeze- dried Na+-montmorillonites (MMTs) caused demixing problems and low , filling limits. This, once again, reflects effects of clay pre-treatment or' surface treatment. Freeze-drying changes clay morphologies and results in fluffy clays. However, an increase in the clay surface area enhances the adhesion between hydrophilic clay particles instead of the clay and the epoxy matrix. The approximate filling limits of the representative smectite clays and muscovite mica in our experimental conditions are listed in Table 1.4. The data shows that pre-treatment affects not only processability but also the filling limit. Owing to the degassing problem of CT-PS and the 21 Table 1.4. Approximate Filling Limits of the Smectite Clays and Muscovite Mica for Epoxy Resin EPON-828 Filling Limit Mineral (wt%) HPM-20 40 CT-PS 30 MMT 1-2 10 ABM-4 30 Note: The weight of MPDA is included in the total weight. 22 low filling limit (~ 10 wt%) of MMT 1-2, only HPM-20 and ABM-4 at higher concentrations (25 and 35 wt%) have been investigated. Curing Degrees The mechanical properties of polymers and composites depend strongly on their curing degrees (or extent of polymerization). Diffuse- Reflectance Fourier transform infrared spectroscopy (DRFTIR or DRIFTS) provides a fast way to study curing degrees of the cured epoxy resin and mineral—epoxy composites by using tested, fractured specimens in this study. However, transmission FTIR spectroscopy is the only way to obtain the FTIR spectrum of the pure liquid epoxy resin, as shown in Figure 1.7. The infrared characteristic bands of epoxy groups absorbing at 916 and 863 cm'1 diminish gradually as the amine-epoxy reaction proceeds at elevated temperatures. The curing process can be expressed by the following simplified equations: H It‘NH2 + (Sign —> R‘m-mz-gm (1.1) H H R‘I‘IN-CI‘TTg‘lR + 612- —’ R1N(CI'12'8TR)2 (1.2) Byrne et al. (21) and Chalesworth (22, 23) independently concluded that the reaction rates (Equations 1.1 and 1.2) were indistinguishable. Swarin and Wims (24) used differential scanning calorimetry (DSC) to study the amine-epoxy reaction and obtained a single exotherm peak. That is, the two reactions take place virtually at the same time and only the overall reaction is detected. The 916 cm'1 band is the better analytical hand because it is 23 a . Epoy Resin EPON-828 F N Transmission (96) s V 7 T T jifir 7 T T T T j 7 I T T , 7 n I." 16” 12“ am Wave Number (cm!) Figure 1.7 . Transmission Fourier transform infrared (FTIR) spectrum of epoxy resin EPON-828; the analytical and reference bands are 916 cm'1 (epoxy ring, stretching) and 1184 cm‘1 (Pb-H, in-plane deformation) bands, respectively. 24 much less overlapped with another band. The band intensity ratio is defined as the integrated intensity ratio of the 916 cm'1 band (epoxy ring, stretching) to the 1184, cm'1 band (aromatic hydrogen, in-plane deformation). Since no aromatic rings participate in the reaction, the intensity of the 1184 cm'1 reference band should remain constant as the polymerization proceeds. Listed in Table 1.5, the band intensity ratios show that the smectite clays and muscovite mica have little effects on the cure extent of the epoxy resin after a long cure cycle at elevated temperatures (2 h at 75 °C and another 2 h at 125 °C). The curing degree of EPON-828 cured by 14.5 wt% MPDA under the same cure cycle was determined as 93% by Schiering et al. (25), using the cured epoxy thin-film and the 1034 cm'1 reference band (aromatic hydrogen, out-of-plane deformation). Based on that value, the curing degrees of 7 wt% clay- and mica-epoxy composites are ~ 94% regardless of the smectite clays and muscovite mica. It was noted that HPM-20 (e.g., at 7 and 35 wt%) alone could not cure the epoxy resin even at 125 °C for several days, although it accelerated the epoxy-amine reaction. It is possible that the reaction temperature is not high enough for the polyacrylate-epoxy reaction. However, it is more likely that the concentration of 0.1 wt% Na- polyacrylate on the clay surface is too small to cure the epoxy resin completely. Intercalation X-ray powder diffraction (XRD) patterns serve as excellent fingerprints for crystalline samples. More importantly, XRD patterns can often be recognized even when they are superimposed upon one another. 25 Table 1.5. Intensity Ratios of FTIR Bands and Curing Degrees of 7 wt% Clay- and Mica-Epoxy Composites Intensity Ratio“ Curing Degree” Mineral (x 10-2) (%) NONE 7.58 93.0 HPM-20 6.65 93.9 CT-PS 5.89 94.7 MMT 1-2 6.57 94.0 ABM-4 6.86 93.7 “916 cm'1 (epoxy ring, stretching)/1 184 ch (Ph-H, in-plane deformation). 1’2 h cure at 75 °C and another 2 h cure at 125 °C. 26 XRD has proved useful and convenient by using the tested, fractured specimens directly in this study. Namely, basal spacings of pristine clays and micas embedded in the epoxy matrix can be measured directly and no powdered composites are required. As shown in Figure 1.8, the XRD patterns of ABM-4, EPON-828, and HPM-20 indicate that the muscovite mica is crystalline whereas the cured epoxy resin is amorphous and that the smectite clay is semicrystalline or polycrystalline. The relatively low crystallinity of smectite clays arises from their structural defects and small crystal sizes (14). As shown in Figure 1.9 (bottom patterns), the reduced basal spacings (9.8-14.1 A) of dehydrated HPM-ZO, MMT 1-2, and CT-PS are due to the loss of most interlayer water after the smectite clays were preheated at 150 °C overnight. On the contrary, preheat treatment does not reduce the basal spacing of ABM-4 simply because no interlayer water exists in the pristine muscovite mica in the first place. Figure 1.9 (top patterns) also shows that there are small increases in basal spacings (0.2-1.4 A) for the smectite clays and no changes for the muscovite mica when they are embedded in the epoxy matrix. The result suggests that some epoxy monomers may be intercalated and to a lesser extent, cured by MPDA on the clay interlayer surfaces, although it clearly indicates that nothing happens in the mica galleries. According to Nakamae and co-workers' work (26-29) on stress transmission in particulate-epoxy composites, an increase in basal spacing or crystal strain may be attributed to stress transferred from the epoxy matrix to clay particles. The crystal strain of mica flakes may be too small to be detected by our X-ray diffractometer. However, the concept of crystal strain has yet to be proved valid for the composite tensile strength, which 27 a .wuinA—m 53. .989 .83 32358.. v5 628:2 neck—ma 3:38.00 .52»:— .o~-2m= >223, can—EmbEEom .8 «Susan Any—Xv 5:25.26 3e39— bcén dd 953..— 8 a 292 5:22.: 0. O? 0.00 0.0m «-29.. .. . 5...: .3“ I. . 4.... ON-“ .. u a. .um M...”:::::f .. I no. 3:3:::::::. o.a Airman] ”papa 28 Figure 1.9. Increases in basal (dool) spacings for the dehydrated smectite clays and muscovite mica (bottom patterns) embedded in cured epoxy resin EPON-828 (top patterns): Claytone-PS, freeze-dried Montmorillonite 1-2 um, Volclay 1-1PM-2, and Alisibronz Mica-4. 0.00 0.: 0.00 0.0. 0.0 0.00 0.8 0.00 0.0. 0.0 is! 352.1 m" 86 :66 .86 .86 .26 . . _ _ _ 1 a... ..... 86 .86 . .. . .26 86 . .86 a... if}. . . «N .c 1.. 00.0 .. .030 .00.. .. .30 0_ t 0. 0 fl 0H0. » 0W8 » 0H00 p 0H0. » 0. 0 20.0 ..0.0 v0.0 .v0.0 .86 5..., .86 :6 _ . 226 00.0 . .. #090 .86 .86 ov.0 20.0 <4. 2 23.0 .30 ..0.0 ill :0.0 NJ .5. 3246: .2 2-8 2..:tm.m: ,... n0-I n0... 29 is more sensitive to stress transmission than the composite tensile modulus. Nevertheless, it is certain that most or all epoxy monomers and oligomers are cured by MPDA outside the clay particles. Surface Composition Mineral particles with high surface free energies are hydrophilic and tend to flocculate in a polymer matrix. Moreover, surface free energy is a function of the surface area and surface elemental/chemical composition. Table 1.6 lists physicochemical properties such as surface area, cation- exchange capacity (CBC) and basicity/acidity of the smectite clays and muscovite mica. However, none of the clay properties measured accounts solely for the clay dispersion in the epoxy matrix or the composite tensile strength. X-ray photoelectron spectroscopy (XPS) has been used to determine relative concentrations and chemical states of surface elements of clay minerals (30-32) and zeolites (33-35). XPS sampling depths range from 10 to 30 A for most inorganic samples, (36) and basal spacings are 10 to 14 A for the dehydrated smectite clays and muscovite mica. Thus, the XPS data obtained are based on the two to three outermost clay and mica surface layers. Table 1.7 indicates that MMT 1-2, RPM-20, and ABM-4 contain l7.6-13.2% contaminant carbon. It is reasonable to assume that CT-PS contains the average concentration of contaminant carbon 14.9%. The extremely high concentration of intrinsic carbon (45.7%) and the presence of nitrogen are due to [(CH3)(C5H5CH2)N(C18H37)2]+ ions on the clay surface and in the two outermost clay galleries. The relative concentrations of surface elements excluding carbon are listed in Table 1.8. Named by the major interlayer ion, Na+- Table 1.6. Physicochemical Properties of the Smectite Clays and Muscovite Mica Sufl'ace Area C.E.C.‘ Mineral (mzlg) (meg/103g) pH Value” MMT 1-2 53.92 i 1.16 82.48 :1: 0.50 9.34 :1: 0.01 HPM-20 30.25 :i: 0.89 61.67 :1: 0.50 9.21 :t 0.01 CT~PS 6.40 :l: 0.29 0.51 :l: 0.14 7.78 :l: 0.02 ABM-4 2.91 d: 0.03 0.88 i 0.15 6.32 :t 0.05 “C.E.C.: Cation exchange capacity. [Based on 7 Wt% mineral aqueous suspensions at 25 °C. 31 Table 1.7. Surface Elemental Concentrations of the Smectite Clays and Muscovite Mica Measured by XPS Relative Concentration (96) Element MMT 1-2 RPM-20 CT-PS ABM-4 C (1s) 17.64 :1: 2.36 13.93 :l: 2.41 60.59 :1: 0.22 13.17 :1: 1.44 Mg (ls) 2.15 $0.15 l.64:l:0.15 0.62:1:0.10 0.55 $0.06 Fe (2p) 0.50 :l: 0.21 0.61 :l: 0.30 0.35 :l: 0.18 0.35 :l: 0.16 Na (ls) 1.60 :l: 0.10 1.55 i 0.08 0.26 :l: 0.11 0.66 i 0.16 Ca (2p) 0.66 i 0.07 0.54 :l: 0.07 ND ND N (Is) ND ND 0.83:0.11 ND K (2p) ND ND ND 5.32 :l: 0.28 A1 (2p) 11.90 :1: 0.25 12.16 i 1.52 6.32 :l: 0.58 18.94 :t 0.44 Si (2p) 28.82 i 0.48 29.81 :l: 0.68 14.32 i: 0.26 23.18 :I: 0.56 0 (1s) 36.73 :t 1.68 39.77 :1: 0.93 16.73 :l: 0.95 37.66 :I: 0.86 ND: Non-detectable. 32 Table 1.8. Relative Concentrations of Surface Elements Excluding Carbon of the Smectite Clays and Muscovite Mica Measured by XPS Relative Concentration (%) Element MMT 1.2 HPM-zo CT-PS ABM-4 Mg (1s) 2.58 10.20 1.87 10.11 1.59 10.22 0.641006 Fe (2p) 0.691035 0.711034 0.98 10.46 04010.19 Na (1s) 1.911020 1.811006 0.741013 0.77 10.20 Ca (2p) 07610.01 0.64:l:0.08 ND ND N (Is) ND ND 2.151.24 ND K (2p) ND ND ND 6.121023 A1(2p) 14.44 10.52 13.97 11.26 162411.49 21.811015 Si (2p) 35.16 1 0.28 34.91 1 1.00 36.05 1 0.53 26.69 1 0.55 o (Is) 44.45 1 0.69 46.10 1 0.65 42.26 1 2.08 43.37 1 0.77 ND: Non-detectable. 33 montmorillonites, MMT 1-2 and HPM-20 contain not only Na+ but Ca2+ interlayer ions. No Ca2+ but some Na+ ions remain in CT-PS after the cation-exchange reaction or cationic surfactant treatment. The presence of K" ion and the relatively low Si“”/A13+ ratio (1.22 vs. 2.22-2.50) verify that ABM-4 is a muscovite mica. As mentioned earlier, a negative layer charge of montmorillonite arises primarily from the partial substitution of Al3+ ions for Mg2+ in the octahedral sheet and the negative charge is balanced by various interlayer cations. The relatively high Mg2+ concentration and CEC value of MMT 1-2 might be somewhat related to the aggregation or poor dispersion of the clay particles in the epoxy matrix. However, the differences in Mg2+ concentrations (OJ-1.0%) among the smectite clays are very small, and the CEC value increases with increasing the surface area as well (14). Thus, the poor dispersion of MMT 1-2 in the epoxy matrix is mainly due to the high surface area of MMT 1-2, resulting from freeze-drying. D. CONCLUSIONS All the smectite clays and muscovite mica investigated reduce the tensile strength but enhance the tensile moduli of epoxy resin EPON-828 cured by MPDA. Among them, Volclay HPM-20 exhibited the highest filling limit and composite strength, whereas Alsibronz Mica-4 exhibited the best processability and composite modulus. For the clay-epoxy composites, surface treatment determines processing problems, filling limits, clay dispersions, and tensile strengths. Composite strengths increase in the order of surface treatment, freeze- drying < cationic surfactant alkylammonium chloride < anionic dispersant 34 Na-polyacrylate. The clay concentration being equal, tensile composite strengths vary whereas tensile composite moduli remain constant within experimental errors. 106892411» 11. 12. 13. 14. 15. 35 LIST OF REFERENCES S. N. Maiti and B. H. Lopez, J. Appl. Polym. Sci., 44, 353 (1992). P. Godard, J. L. Wertz, J. J. Biebuyclt, and J. P. Mercier, Polym. Eng. Sci., 29, 127 (1989). C. G. Bk, .Rheol. Acta, 27, 279 (1988). G. M. Newaz, Polym. Compos., 7, 176 (1986). S. N. Maiti and K. K. Sharma, J. Mater. Sci., 27, 313 (1992). M. A. Ramos and J. P. V. Matheu, Polym. Compos., 9, 105 (1988). J. E. Stamhuis, Polym. Compos., 9, 72 (1988). D. M. Bigg, Polym. Compos., 8, 115 (1987). J. P.Trotignon, L. Demdoum, and J. Verdu, Composites, 23, 2563 (1992). S. F. Xavier and J. M. Schultz, J. Mater. Sci., 25, 2428 (1990). S. Inubushi, T. Ikeda, S. Tazuke, S. Satoh, Y. Terada, and Y. Kumagai J. Mater. Sci., 23, 1182 (1988). S. Cacoutis, R. T. Woodhams, and P. G. K. Van de Poll, Polym. Compos., 9, 51 (1988). J. T. Pinnavaia, Science, 220, 365 (1983). R. B. Grim, "Clay Mineralogy," 2nd ed., McGraw-Hill, New York (1968). C. A. May, Ed., "Epoxy Resins," 2nd ed., Marcel Dekker, New York (1988). 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 36 H. Lee and K. Neville, "Handbook of Epoxy Resins," McGraw-Hill, New York (1967). E. Busenberg and C. V. Clemency, Clays Clay Miner., 21, 213 (1973). S. Ahmed and F. R. Jones, J. Mater. Sci., 25, 4933 (1990). X. B. Chen and X. 8. Wu, Polymer, 33, 3639 (1992). L. E. Nielsen, J. Appl. Phys., 4, 4626 (1970). C. A. Byrne, G. L. Hagnauer, N. S. Schneider, and R. W. Lenz, Polym. Compos., l, 71(1980). J. M. Charlesworth, J. Polym. Sci., polym. Phys. Ed., 17, 1557 (1979). J. M. Charlesworth, J. Polym. Sci., polym. Phys. Ed., 17, 1571 (1979). S. J. Swarin and A. M. Wims, Anal. Calorim., 4, 155 (1976) D. W. Schiering, J. E. Katon, L. T. Drzal, and V. B. Gupta, J. Appl. Polym. Sci., 34, 2367 (1987). A.-R. Xu, T. Nishino, and K. Nakamae, Polymer, 33, 5167 (1992). K. Nakamae, T. Nishino, and A.-R. Xu, Polymer, 33, 2720 (1992). T. Nishino, and A.-R. Xu, T. Miyamoto, K. Matsumoto, and K. Nakamae, J. Appl. Polym. Sci., 45, 1239 (1992). K. Nakamae, T. Nishino, and A.-R. Xu, K. Matsumoto, and T. Miyamoto, J. Appl. Polym. Sci., 40, 2231 (1990). N. Davison, W. R. McWhinnie, and A. Hooper, Clays Clay Miner., 39, 22 (1991). M. L. Occelli, J. M. Stencel, and S. L. Suib, J. Mol. Catal., 64, 221 (1991). 32. 33. 34. 35. 36. 37 M. Stoecker, K. J. Jens, T. Riis, and J. K. Grepstad, J. Chem. Soc., Faraday Trans. 1, 84, 1863 (1988). M. J. Remy, M. J. Genet, P. F. Lardinois, and P. P. Notte, J. Phys. Chem., 96, 2614 (1992). B. L. Meyers, T. H. Fleisch, G. J. Ray, J. T. Miller, and J. B. Hall, J. Catal., 110, 82 (1988). T. H. Fleisch, B. L. Meyers, G. J. Ray, J. B. Hall, and C. L. Marshall, J. Catal., 99, 117 (1986). R. S. Swingle, II and R. S. Riggs, Anal. Chem., 5, 267 (1975). CHAPTER II CLAY-POLYMER NANOCOMPOSITES FORMED FROM ACIDIC MONTMORILLONITES AND EPOXY RESIN A INTRODUCTION Delaminated smectite clays (e. g., montmorillonite) have recently been shown to be useful materials for the design of polymer-based composite materials with novel mechanical and physical properties. One especially intriguing system, recently reported by Okada and co-workers (1, 2), is based on the exfoliation of [H3N(CH2)1lCOOHf—montmorillonite in a semicrystalline nylon-6 (or polyamide-6) matrix. The 9.6-A-thin clay layers greatly improved the tensile strength (107 vs. 69 MPa), tensile modulus (2.1 vs. 1.1 Gpa), and heat distortion temperature (152 vs. 65 °C) relative to nylon-6. Kojima et al. (3) found that replacing protonated lZ-aminolauric acid by protonated e-caprolactam as the clay exchange cation imparted comparable composite tensile strength and modulus, but the higher heat distortion temperature (164 vs. 152 °C) to nylon-6. Fujiwara and Sakamota (4) chose a very specific organoclay-polymer system to demonstrate the swelling of the clay layers in a thermoplastic polyamide matrix. [H3N(CH2)5COOH]+-montmorillonite (9 g) was dispersed in e-caprolactam (20 g) and the mixture was heated at 250 °C to form a clay/nylon-6 composite. The basal spacing of the organoclay in the 38 39 nylon matrix was found to be 20 A. Related experiments have been carried out by using protonated 12-aminolauric acid [H3N(CH2)11COOH]+ as the exchange cation (5, 6). The interlayer spacing was in the range of 51-210 A for organic montmorillonite content 5-30 wt%. For protonated e- caprolactam as the exchange cation, the interlayer spacing was found to be ~ 250 A (3). It has also been reported in the patent literature (7) that swelled montmorillonite (9001 = 50 A) improved mechanical properties of an amine- cured epoxy resin. [H3N(CH2)1ICO’OHr-montmorillonite slightly enhanced the impact resistance (1.08 vs. 0.91 J/m) and the heat distortion temperature (207 vs. 193 °C) of epoxy resin Epicoat-828 cured by p, p'- diaminodiphenylsulfone (DDS). However, it was necessary to use N, N- dimethylformamide as a swelling solvent to achieve clay delamination. Since epoxides are reactive toward self-polymerization to polyethers, it was of interest to us to investigate the direct polymerization of an epoxy resin in the galleries of montmorillonite. The epoxy resin selected for the present study was the same epoxy resin EPON-828 as used before. B. EXPERIMENTAL Materials Source clay SWy-l (Wyoming Na+-montmorillonite) was obtained from the Clay Minerals Depository at the University of Missouri (Columbus, MO). Epoxy resin EPON-828 (Shell), aminocarboxylic acids H2N(CH2),,-1COOH, primary diamines H2N(CH2)nNH2, and primary amines CH3(CH2)nNH2 (n = 6, and 12) (Aldrich) were used as received. Sample Preparation H2N(CH2),,.1COOH, H2N(CH2)nNH2, or CH3(CH2)nNH2 (10 mmol) was dissolved in 1 L of a 0.01 N (0.02 N for [H3N(CH2)nNH3]C12) HCl solution at 60 °C. Source clay SWy-l (10 g) was thoroughly dispersed in the appropriate 0.01 N solution of interest at 60 °C for 3 h. The cation- exchanged montmorillonite was separated from the centrifugate and washed with deionized water. Centrifugation and washing were repeated until no white AgCl precipitate was observed when a drop of 0.1 N AgNO3 was added to the centrifugate. The washed onium ion-exchanged montmorillonite was dispersed in 600 mL of deionized water and freeze- dried in a Labconco Freeze Dryer Model 18. The freeze-dried onium ion- montmorillonite was then sieved to < 325-mesh. The onium ion-montmorillonite (0.79 g) was added to 15 g of epoxy resin EPON-828 and the 5/95 (w/w) clay/epoxy mixture was magnetically stirred in a 250-mL beaker at 75 °C for 30 min. The beaker was sealed with aluminum foil and the temperature was then raised at a rate of ~ 20 °C/min to the clay delamination-epoxy polymerization temperature. The liquid-to- powder transformation of the mixture and the concomitant formation of the clay-epoxy nanocomposite took place within a period of one minute. The transformation was accompanied by an increase in the bulk volume (up to sixfold) and a large amount of heat released. Characterization The clay-epoxy nanocomposite sample for powder XRD studies was firmly pressed onto a Rigaku sample holder. The diffraction pattern was then recorded by monitoring the diffraction angle 20 from 2° to 45° on a Rigaku Rotaflex Ru-ZOOBH X-ray diffractometer. The diffractometer was 41 equipped with a Ni-filtered Cu-Ka radiation source operated at 45 kV and 100 mA. The scanning speed and the step size used were 2°lmin and 0.02°, respectively. Quartz was used as a calibration standard. A small amount of clay-epoxy nanocomposite powder for TEM studies was added to a Spurr's mixture in a silicon rubber mold, which was then placed in an oven at 70 °C overnight. The cured epoxy block was trimmed and thin-sectioned with a diamond knife in a microtome. The resulting ultra-thin sections (~ 90 nm in thickness) were mounted on a plastic-film support on a copper grid and examined by a JEOL 100CX transmission electron microscope operated at an accelerating voltage of 100 IN. The organoclay-epoxy resin mixture for DSC analysis (5 mg) was placed in an aluminum sample pan, which was then sealed hermetically. The sample and reference pans were placed in the cell of a Du Pont 910 differential scanning calorimeter. The DSC cell was then heated under nitrogen (50 mL/min) to 350 °C at a rate of 20 °C/min. Indium (mp, 156.6 °C and AHm, 28.42 Hg) was used for DSC calibrations. In the case of the neat epoxy resin, the temperature was raised to 450 °C under the same experimental conditions. However, the sample size was reduced to 2.5 mg to prevent the build-up pressure from breaking loose the sealed sample pan. ' A [H3N(CH2)llCOOHf-montmorillonite (20 mg) for TGA analysis was placed in a platinum sample container, which was then placed in the thermobalance of a Du Pont 990 thermogravimetric analyzer. The sample weight was tared and the sample was then heated under nitrogen (100 mL/min) to 650 °C at a rate of 20 °C/min. 42 An equivalent set of experiments using [H3N(CH2),,.1CH3]+ (n = 6 and 12), NH4+, and H+ as exchange cations was carried out under the same experimental conditions. C. RESULTS AND DISCUSSION Primary amines and acid anhydrides (equivalent to carboxylic acids) are two types of most commonly used curing agents for epoxy resins because primary amino and carboxylic groups are highly reactive toward epoxy rings. Protonated aminocarboxylic acids and primary diamines [H3N(CH2)n.1COOH]+, [H3N(CH2)nNH2]+, and [113i~1(cr12),,1~1113]2+ (n = 6 and 12) have been used as delaminating agents or catalysts rather than curing agents in this work. It is essential to treat these organic acids and diamines with dilute aqueous HCl solution. Acid treatment converts amino- carboxylic acids and primary diamines into the corresponding water-soluble onium ions, which can easily replace interlayer cations such as Na+ and Ca2+ in montmorillonite. 'More importantly, protonation weakens the carbon-oxygen bond and facilitates the epoxy ring opening. It is not the undissociated acid H2N(CH2)n-1COOH but the dipolar ion structure (I) that accounts for the chemical and physical properties of an aminocarboxylic acid. When an aminoacid dissolves in aqueous HCl solution, protonation converts the dipolar ion 1 into the cation 11. + H+ + O H3N(CH,),,,6-0H + o H3N(CI-12)n,IG-O' I II 43 Evidence of Clay Delamination The delamination of the onium ion-montmorillonites in the cured epoxy resin may be confirmed by X-ray powder diffractometry (XRD). For instance, as shown in Figure II.1c, no clay diffraction peaks are observed for a clay-epoxy nanocomposite containing 5 wt% [H3N(CH2)11COOH]+- montmorillonite. Only a very diffuse scattering peak characteristic of the amorphous cured epoxy resin (or polyether) appears in the XRD pattern. The absence of a 17.011 peak for [H3N(CH2)1ICOOH]+-montmorillonite (Figure 11.1a) suggests that the clay tactoids have been exfoliated and the 9.6-A-thin clay layers dispersed at the molecular level. However, it may also mean that the clay basal spacing is greater than 44.2 A (e.g., 20 = 2°) or the clay layered structure collapses. The clay peaks with reduced intensities (Figure II.1b) show the clay layered structure partially collapses after pristine [H3N(CH2)1ICOOH]+-montmorillonite is heated at 229 °C under nitrogen for 1 min, as explained later. Transmission electron microscopy (TEM) provides unambiguous evidence for the delamination of [H3N(CH2)1ICOOH]+-montmorillonite in the cured epoxy resin. As shown in Figure 11.2, the TEM micrographs of a 5 wt% clay-polyether nanocomposite reveal that the micron-sized clay tactoids indeed have been exfoliated by the polymer. The interlayer spacing ranges up to ~ 0.2 um or 2,000 A. Additional complementary results such as the reaction temperature and the heat of reaction are obtained from differential scanning calorimetry (DSC). Figure 11.3 shows the DSC curve of a 5/95 (w/w) mixture of [H3N(CH2)1ICOOH]+-montmorillonite and epoxy resin EPON-828. The thermogram indicates that the spontaneous clay exfoliation-epoxy polymerization takes place at an onset temperature of A L l (a) m3N(CHz)uooom*-Mont (25 °C) (b) AAC12-Mont. (229 'C, 1 min) _ (c) 5% AAC12-MmtJEpoxy Nanocomposite 17.0 A at .5 _ _ .2 ‘5 '5 g .. _ (a) -— (b) (C) o 10 20 30 40 50 Diffraction Angle (2 9) Figure 11.1. X-ray powder diffracnon (XRD) patterns of (a) [113N(CH2)11COOH]+- montmorillonite (AAC12-Mont.) (25 'C), (b) AAC12-Mont. (229 'C, 1 min) and (c) a clay-epoxy nanocomposite containing 5 wt% AAC12-Mont. 45 Figure 11.2. Transmission electron microscopy (TEM) micrographs of a clay-epoxy nanocomposite containing 5 wt% [H3N(CH2)11COOH]+-montmorillonite: (a) x10,000 and (b) x58,000. 5% AACl2-MonL/EPON-828 239 'C Heat Flow (W/g) 8 10- 0 - i 1 i J 229 'C 572 ll: -10 . , . , . , . . . , . , . 160 180 200 220 240 260 280 300 Temperature ('C) Figure 11.3. Differential scanning calorimetry (DSC) thermogram of the epoxy polymerization catalyzed by 5 wt% [113N(CH2)1ICOOH]+-montmorillonite, a heating rate of 20 'C/min being used. 47 229 °C. Based on the integrated peak area and the heating rate, the heat of reaction is 572 J/g. For comparison, the DSC thermograms of the neat epoxy resin and [H3N(CH2)1lCOOH]+-montmorillonite are shown in Figures 11.4 and 11.5, respectively. It should be noted that the scale of the heat flow in Figure 11.5 is much smaller than that in either Figure 11.3 or 11.4. Without an effective catalyst (acid or base), the homo- or self-polymerization of the epoxy resin involves the initiation reaction (Equation 11.1) and etherification (Equation 11.2) to form polyether at elevated temperatures. R'OH is any impurity solvent (e.g., water) or epoxy monomer possessing one or two hydroxyl groups (EPON-828: 10% of n = 1 and 2% of n = 2). R'OH + 63001 —A-> Roan-$1111 (11.1) Room-gain + 11 61211-111 iv R'o(c112-mno)ncnz.glk (11.2) As shown in Figures 11.5 and 11.6, DSC and TGA curves indicate that [H3N(CH2)1ICOOH]+-montmorillonite loses surface and interparticle pore water between 40 and 200 °C. Interlayer water is liberated in the region of 200-300 °C. The loss of hydroxyl groups in the clay framework (or dehydroxylation) occurs between 300 and 650 °C. It is noted that 12- aminolauric acid (mp, 186 °C) loses the crystal water and totally decomposes over temperature ranges 200-400 °C and 400-500 °C. respectively. The heat absorption and the weight loss shown by the DSC and TGA curves, respectively, substantiate the XRD pattern (Figure 11.1b) that the partial loss of interlayer water results in the reduced peak intensity 48 10 403 'C Epoxy Resin EPON-828 Heat Flow (Wig) A 0 i 1 I W— I 384 'C . 611 J/g 250 300 350 400 450 Temperature (’C) Figure 11.4. DSC thermogram of the uncatalyzed self-polymerization of epoxy resin EPON-828, a heating rate of 20 'C/min being used. 49 0.0 1 [113N(CH2)11COOH]+-Mont. Heat Flow (Wig) 0 100 - 200 300 400 50( Temperature ('C) Figure 11.5. DSC thermogram of [113N(CH2)11C00H]+-montmorillonite, a heating rate of 20 ‘C/min used. 100 013N(CH2)uooom*-Mont Weight (96) i? 85.. ‘r f I ' I ' T I ' f 0 100 200 300 400 500 600 70( Temperature (°C) Figure 11.6. Thermogravimetric analysis (TGA) curve of [H3N(CH2)11COOH]+- montmorillonite, a heating rate of 20 'C/min used. 51 for [H3N(CH2)1ICOOH]+-montmorillonite heated at 229 °c for 1 min. More importantly, the heat of reaction for [H3N(CH2)11COOH]+- montmorillonite (especially 5 wt%) is much smaller than that for either the uncatalyzed epoxy polymerization or the epoxy polymerization catalyzed by 5 wt% [H3N(CH2)11COOH]+-montmorillonite Reaction Mechanism According to the evidence presented above, the delamination of the onium ion-montmorillonites in the epoxy resin or epoxy polymerizations catalyzed by the onium ion-montmorillonites at elevated temperatures can be rationalized by the following general equations: 21101211311911: H 613% + H+A (613311]; + A 1...: H OH + O + 0 H3N(CH2)n,1C-O' + 612- R -'——- H3N(CH2),,,6-o-CH2-bHR H OH R‘NH2 + 61(f-CHR —> R‘HN-CHz-CHR + H+ H OH (3 OH R‘HN-CHz-CHR + 612- HR —> R‘N(CH2-CHR)2 + 11* (11.3) (11.4) (11.5) (11.6) 52 W: OH R’OH + n £1}an —H-+-> R’O(CH2-CHRO)n,1CH2-CHR + 11+ (11.7) where H+A is any protonated aminocarboxylic acid or primary diamine; RINHZ is any molecule containing one or two amino groups; and RZOH is any monomer or oligomer possessing one or more hydroxyl groups. At the delamination-polymerization temperature, epoxy monomers first react with protonated aminocarboxylic acids and primary diamines intercalated on the edges of clay particles. The protonation of epoxy rings (Equation 11.3) takes place first, followed by the cleavage of epoxy rings (Equations 11.4- 6). The reactions rapidly propagate toward clay interlayer space and the propagation causes the expansion of the clay gallery region. In the meantime, various acid-catalyzed etherifications (Equation 11.7) occur to form polyether in the expended region and this results in the complete delamination of clay layers. The cured epoxy resin or polyether in the clay galleries becomes phase-segregated from the uncured resin. This gives rise to the liquid-to-powder transformation characteristic and a substantial (ca. 5-6 fold) increase in the bulk volume upon the formation of clay-epoxy nanocomposites. The rapid and complete clay delamination-epoxy polymerization is accompanied by a large heat of reaction. Reaction Temperature Listed in Table 11.1, the onset temperatures of the epoxy polymerization-montmorillonite delamination reactions increase in the order [H3N(CH2)n.1COOH]+< [H3N(CH2)nNH3]2+ < [H3N(CH2)nNH2]+ (n = 6 and 12). That is, the polymerizationadelamination temperature increases 53 Table 11.1. DSC and XRD Data for Epoxy Polymerizations Catalyzed by 5 wt% Protonated Aminoearboxylic acid- and Primary Diamine-Montmorillonites‘ Polym.-Del. Heat of Heat of Basal Interlayer Cation Temp.” Reaction ponmc Spafing (°C) (J/g) (kcal/mol) ( ) [113N(C112)11COOH]+ 22911 572116 54.4115 17.05 10.14 [1131~1((:112)5coom+ 248 11 5651 6 53.7105 13.33 10.01 [H3N(C117)12NHg]2+ 271 1 1 566 1 8 53.8 1 0.8 13.39 1 0.08 [H3N(CH;)6NH3]2+ 273 1 2 568 1 7 54.0 1 0.7 13.12 1 0.05 [113N(C1-12)12NH2]+ 281 1 2 563 1 7 53.5 1 0.7 13.48 1 0.03 01319012519112? 287 1 2 557 1 3 53.0 1 0.3 13.19 1 0.07 “DSC data were obtained by using a heating rate of 20 °Clmin. [The onset epoxy polymerization-clay delamination temperature. ‘The heat of polymerization for the epoxy equivalent is halved. 54 with decreasing acidity of the interlayer cation. Protonation weakens the carbon-oxygen bond and facilitates epoxy ring opening. For a homologous series of the onium ions, the polymerization-delamination temperature is correlated with the basal (9001) spacing of the onium ion-montmorillonite. A larger basal spacing gives the epoxy monomers easier access to the interlayer onium ions and this leads to a lower polymerization-delamination temperature. In general, the basal spacing of an organoclay depends primarily on the alkyl chain length (or carbon number) and the spatial arrangement of the onium ion intercalated on the gallery surfaces (8). It is noted that all the onium ion-montmorillonites used in this study, except [H3N(CH2)12COOH]+-montmorillonite (17.0 A), have similar basal spacings (13.1-13.5 A). The small basal spacings indicate that the short- and long-chain protonated aminocarboxylic acids and primary diamines lie flat and form monolayers in the clay galleries. As shown in Figure 11.7, the Spatial arrangement of divalent [H3N(CH2)12NH3]2+ in montmorillonite is due to strong ionic bonding between ammonium groups and negatively- charged clay layers. The unusually small basal spacing of [H3N(CH2)12NH2]+-montmorillonite might arise from the sharing of a proton between two adjacent amino groups. Heat of Reaction Table 11.1 shows that the heat of the polymerization-delamination reaction (557-572 J/g) remains constant within experimental errors regardless of the exchange cations used. Since the heat of reaction arises primarily from the polymerization of the epoxy resin, it should remain unchanged provided the epoxy concentration (wt%) is constant. The average heat of reaction is 565 J/g for 5/95 (w/w) organoclay/epoxy resin, —x ‘. . 55 .Lamznfimovzma 3 es. rummznfimovzma as (50002 53873.: 3 ”mezzo—Eogeofi 5 2280 his—e25 ..o acoEowSEa 33mm 8.: 8:2..— _ - . _ ADV «=2:ZNI...E...N=Z\(I)\I\I)\Z~=.:E:.~=Z\I\I\I\I\(I\711:."—.NIZEW—n: _ - - _ 2.: 3v ‘ .11 a... .11 56 or 595 Hg for the neat epoxy resin. This value is equivalent to a value of 53.7 kcal/mole or 26.9 kcal/mole for the epoxy equivalent, which is a typical value for the epoxy resin cured by primary amines (9, 10). However, Table 11.2 shows that the amount of the onium ions used (1.80- 3.57 mmol) to form the nanocomposites is sufficient to cure only 1.4-2.9 % of the epoxy resin. Therefore, the major reaction is the acid-catalyzed epoxy etherification. Effects of Clay Concentration [H3N(CH2)11COOH]+- and [H3N(CH2)5COOH]+-montmorillonites have been chosen for further investigation because they have relatively low reaction temperatures. Figure 11.8 shows that the polymerization- delamination temperature rapidly decreases and levels off as the clay concentration reaches 1 wt% for [H3N(CH2)1ICOOH]+-montmorillonite or 5 wt% for [H3N(CH2)5COOH]+-montmorillonite. That is, only a very small organoclay concentration is sufficient to effectively catalyze the epoxy polymerization and dramatically reduces the epoxy polymerization temperature. For instance, the uncatalyzed epoxy self-polymerization temperature is 384 °C. compared to 241 and 264 °C for epoxy, polymerizations catalyzed by 0.2 wt% [H3N(CH2)“COOH]+- and H3N(CH2)5COOH]+-montmorillonites, respectively. However, in Figure 11.9 the broad DSC curves indicate that such epoxy polymerizations are relatively sluggish. Moreover, the relatively small values of the heat of reaction (560 and 545 J/g) indicate that a clay concentration of 0.2 wt% is too small to achieve the complete epoxy polymerization. Based on the extrapolated heat of reaction obtained below, the curing degrees are 93.8 and 91.1% for epoxy polymerizations catalyzed by 0.2 wt% 57 Table 11.2. Estimated Epoxy Cure Percents Due to Protonated Aminocarboxylic Acids and Primary Diamines Formula Equiv. Cure W Interlayer Cation Weight Weight“ # mmolb Equiv. Cureb [113N(CH7)11COOH]+ 216 115 3.32 3 1.98 [113N(CH7)5COOH]+ 132 108 3.52 3 2.10 [113N(C112)12NH3]2+ 202 106 1.80 4 1.43 [113N(CHZ)6NH3]2+- 118 103 1.85 4 1.48 [113N(CHZ)12NH2]+ 201 114 3.35 4 2.67 [113N(C112)§_N§2]+ 117 107 3.57 4 2.83 “The resulting weight of 100 g of montmorillonite SWy-l (cation exchange capacity, 76.4 meq/100g) after the cation-exchange reaction. (Based on 100g of 5/95 (w/w) organoclay/epoxy resin (251 mmol or 503 meq). 58 400 } ' [143N(C32)1 1C00H]+-Mont. A o [113N(CH2)5COOH]+-Mont. P 360 . E a .. E g. 320- 0 [- 3‘ E 280 E ’4: g a = 240 O ’_C O 2“) I T I ' fl ' 1 ' 1 v 5 10 15 20 25 Clay Concentration (wt%) Figure 11.8. Onset temperature of the epoxy polymerization-clay delamination as a function of the clay concentration for (.) [H3N(CH2)11COOH]+- and (o) [113N(C1-Iz)5COOH]+-montmofillonites. 59 6 (a) 0.2% AACl2-Mont. 5 ~ (8) (b) (b) 02% AACS-Mont. 4 d I / 3" , E 3 ~ E 5 2 q I 1 .. 0 _ 1 241°C 264 °c 560 II: 545 1/3 -1 V I v t t ' I 150 200 250 300 350 40( Temperature (°C) Figure 11.9. DSC thermograms of epoxy polymerizations catalyzed by 0.2 wt% (a) [H3N(C1~12)11COOH]+- and (b) [113N(C1-17)5COOH]+-montmofillonites, a heating rate of 20 'C/min being used. 60 [H3N(CH2)11COOH]+- and [H3N(CH2)5COOH]+-montmorillonites, respectively. In Figure 11.10 the heat of reaction is plotted against the clay concentration for [H3N(CH2)11COOH]+- and [H3N(CH2)5COOH]+- montmorillonite. In each case the heat of reaction is proportional to the epoxy concentration. The extrapolated values are 603 and 604 J/g or 54.5 kcal/mol for the neat epoxy resin. The linear relationship again shows that the heat of reaction is primarily due to the polymerization of the epoxy resin. The greater slope for [H3N(CH2)5COOH]+-montmorillonite suggests that the polymerization of the epoxy resin is relatively incomplete, especially at high clay concentrations. The ratio of the heat of reaction for [H3N(CH2)1ICOOH]+-montmorillonite to that for [H3N(CH2)5COOH]+- montmorillonite decreases from 1.00 to 0.94, as the clay concentration increases form 1 to 20 wt%. At higher clay concentrations, the entrapment of the epoxy resin between clay particles may limit the formation of long- chain polyether. Thus, 5 wt% is an optimal clay concentration to obtain both high curing degree and low reaction temperature of the epoxy polymerization catalyzed by the onium ion-montmorillonite. Kinetics Parameters Non-isothermal or dynamic DSC have been used to obtain kinetics parameters such as the activation energy and the pre-exponential factor. Borchardt and Daniels (11) developed one of two basic methods for dynamic DSC, using a single DSC curve generated at a constant heating rate. However, compared with isothermal DSC, this method often overestimated activation. energies and pre-exponential factors for most 61 650 - Hi3N(CH2)11COOPfl+-Mont o W3N(CH2)5COOH]+-Mont. 600 - g. 1 g 550 - '8 8 4 a: ‘5 500 . 8 a 450 - 400 ' I ' T V I ' I v o 5 10 15 20 25 Clay Concentration (wt%) Figure 11.10. Heat of reaction as a function of the clay concentration for (.) [H3N(C112)11COOH]+- and (o) [H3N(CH2)5COOH]+-montmorillonites. 62 reactions (12-14). Kissinger ( 15) derived the following equation to evaluate the activation energy (15,.) regardless of the reaction order (n), which was assumed to remain constant throughout the reaction: ___‘“"‘P =-_I._2T (11.8) where Tp is the peak maximum temperature (K) and (p is the heating rate ('C/min or K/min). In case EalR is much greater than 2T1), the slope of the plot of ln 0) vs. 1/Tp is equal to - EalR, where R is the gas constant (1.987 cal/Kmol). A typical plot of multiple DSC curves generated at various heating rates (e.g., 15-25 °C/min) for the epoxy polymerization catalyzed by the cation-exchanged montmorillonite (e.g., 5 wt% [H3N(CH2)11COOH]+- montmorillonite) is shown in Figure 11.11. The peak maximum temperature increases with increasing the heating rate. Despite the increased peak area (AP) and the heating rate, the heat of reaction (q) remains constant, as predicted by the following equations: 4=IIIW= 208411144584 where W is the heat flow and T is the temperature. Figure 11.12 clearly shows excellent linear relationships between 1n (1) and HT}, for epoxy polymerizations catalyzed by 5 wt% [H3N(CH2)11COOH]+- and [H3N(CH2)5COOH]+-montmorillonites, respectively. However, the ratios 63 45 5% [113N(CH2)11C00H]+-Mont. (d) (a) 10 'C/min (c) (b) 15 'C/min 35 . (b) (c) 20 °C./min (d) 25 °C/Inin * (a) 3 E 25 -1 E 4 E 3' 15 - a 5‘ ,1 150 175 200 225 30 275 300 Temperature (°C) Figure 11.11. DSC thermograms of the epoxy polymerization catalyzed by 5 wt% [113N(c112)ucoom*-montmorinonite at various heating rates: 10, 15, 20, and 25 'C/min. - maN(CHa)ncoom*-Mont o maN(CH2)scoom*-Mmt In 03 (K/min) 2.2 . . . . I . 1.85 1.90 1.95 2.00 2.05 I m, x 1000 (UK) Figure 11.12. Linear relationship between the natural logarithm of the heating rate (0 and the reciprocal of the peak maximum temperature Tp for epoxy polymerizations catalyzed by 5 wt% 6) [H3N(CHz)nCOOH]+- and (o) [H3N(CH2)5COOH]+- montmorillonites. 65 of 2T1, to EJR range 7.7-7.9% and 8.2-8.5%, respectively, so that 2Tp is too large to be omitted. Prime (16) reviewed several papers and concluded that the Ozawa equation (17) together with Doyle's data (18) should be used to obtain more accurate activation energies. dlno ”10525 1 —E—l- (11.9) 411-, The difference in Ea obtained between the Ozawa and Kissinger methods is 5.2%. According to Equation 11.9, the activation energies of epoxy polymerizations catalyzed by 5 wt% [H3N(CH2)11COOH]+- and [H3N(CH2)5COOH]+-montmorillonites are 25.9 :1: 0.1 and 25.2 :1: 0.3 kcal/mol, respectively. Compared to typical activation energies (18-21 kcal/mol) for other epoxy systems reported in the literature (19-21), the higher values might be attributed to steric hindrance caused by the intercalation of the onium ions (confined between clay layers). Kissinger also derived an equation to evaluate the pre-exponential or frequency factor (A) for nth-order reactions: - E E E ‘ E e ‘ ,, ham] 9 . at") A = RT’na-a )H ' RT’ (“'10) P I P where up is the extent of reaction at the peak exotherm and is independent of the heating rate. The value of n(1-tzp)"'1 is equal to unity by definition for first-order reactions. Prime (13) showed that n(1-ap)"'l was only 2-4% 66 greater than unity for nth-order epoxy cure reactions, and Kissinger reported that the value was independent of the heating rate as well. By using Equation 11.10 and the activation energies obtained above, the pre- exponential factors of epoxy polymerizations catalyzed by 5 wt% [H3N(CH2)11COOH]+- and [H3N(CH2)5COOH]+-montmorillonites are calculated as 1.80 i: 0.02 x 109 and 3.06 :l: 0.06 x 108 s'l, respectively (Table 11.3). The relationship between the pre-exponential factor and the activation entropy in the standard state (AS°*) is given by the following equation: Ah(c°)""l as“ = Rln kTe" (11.11) where m is the molecularity of an elementary reaction or the number of molecules that come together to form the activated complex; (c°)m’l is the factor required to keep the equilibrium constant (K*) dimensionless; and h and k are the Plank and Boltzman constants, respectively. The acid- catalyzed epoxy ring opening has been proved to be a substitution nucleophilic bimolecular (8N2) reaction, although it has considerable substitution nucleophilic unimolecular (8N1) character. Based on Equation. 11.11 and m = 2, the activation entropies of epoxy polymerizations catalyzed by 5 wt% [H3N(CH2)11COOH]+- and [H3N(CH2)5COOH]+- montmorillonites are -21.2 i 0.1 cal/Kmol (at 229 °C or 502 K) and -24.8 :t 0.1 cal/Kmol (at 248 °C or 521 K), respectively. In each case, the negative activation entropy implies that the activated complex has a more ordered structure with less rotational and vibrational freedom than the reactants . 67 Table 11.3. Pre-exponential Factors for Epoxy Polymerizations Catalyzed by 5 wt% Protonated Aminocarboxylic Acid-Montmorillonite Peak Max. Pre-exponential Interlayer Cation Heating Rate Temperature Factor (°C/min) (°C) (8'1) 0131001,), 1coon]+ 10 226.8 1.82 x 109 15 234.6 1.78 x 109 20 239.6 1.81 x 109 25 244.3 1.77 x 109 [113N(CH;)5COOH]+ 10 246.2 3.07 x 108 15 253.9 3.13 x 108 20 260.2 3.07 x 108 25 265.4 2.98 x 108 Note: The activation energies of epoxy polymerizations catalyzed by 5 wt% [113N(CH2)11COOH]+- and [H3N(CH2)5COOI-I]+-montmorillonites are 25.9 1 0.1 and 25.2 :I: 0.3 kcal/mol, respectively. 68 Furthermore, by using the kinetics data obtained above and the following equations: AH°*=E.-mRT (11.12) AG°*= AH°*-TAS°* (11.13) the activation enthalpies (AH°*) and free energies (AG°*) of epoxy polymerizations catalyzed by 5 wt% [H3N(CH2)11COOH]+- and [H3N(CH2)5COOH]+-montmorillonites are calculated as 23.9 :I: 0.1 and 34.6 :1: 0.1 kcal/mol (at 502 K) and 23.1 :I: 0.3 and 35.6 :1: 0.4 kcal/Kmol (at 521 K), respectively. Clearly, the activation entropies, as reflected in the pre-exponential factors, account for the difference in the polymerization reactivity between the two organoclay-epoxy resin systems, although the activation enthalpies are the major parts (69.3 and 64.9%) of the activation free energies. Extended Work Since the clay delamination-epoxy polymerization is proton- catalyzed, other cations such as [H3N(CH2),,.1CH3]+ (n = 6 and 12), NH4+, and H+ may also be used as. delaminating agents or catalysts for the formation of clay-polyether nanocomposites. As shown in Table 11.4, these alkylammonium ions and inorganic cations cause clay delamination-epoxy polymerization over a wide range of onset temperatures (198-287 °C). It appears that basal spacing and cation acidity are two of the factors affecting the delamination-polymerization temperature. The lowest and the second lowest temperatures correspond to [H3N(CH2)11CH3]+- 69 Table 11.4. DSC and XRD Data for Epoxy Polymerizations Catalyzed by 5 wt% Proton-, Ammonium ion, and Protonated Primary Amine-Montmorillonites‘ Polymfijel. Heat of Heat of Basal Interlayer Cation Temp! Reaction Polym! Spaiing (°C) (J/ g) (kcaI/mol) ( ) [H3N(CH,)11CH3]* 198 1 l 550 1 3 52.3 1 0.3 15.89 1 0.19 11131001730113? 287 1 1 554 1 6 52.7 1 0.6 14.90 1 0.13 NH.” 247 1 1 554 1 5 52.7 1 0.5 12.52 _1 0.12 11+ 231 1 l 555 1 12 52.8 1 1.1 13.95 1 0.08 “DSC data were obtained by using a heating rate of 20 °C/min. ”The onset epoxy polymerization-clay delamination temperature. c'I'he heat of polymerization for the epoxy equivalent is halved. 70 montmorillonite with the largest basal-spacing and H+-montmorillonite with the most acidic cation. However, in Figure 11.13 the DSC curves show the contrast between [H3N(CH2)11CH3]+-and [H3N(CH2)5CH3]+-montmorillonites in terms of the delamination-polymerization temperature and the peak shape (or reaction rate). The great difference in reactivity makes the alkylammonium- montmorillonites a unique class among all the cation-exchanged montmorillonites studied. The reason is that chemical and physical properties of the primary amines (e.g., basicity), alkylammonium ions (e.g., acidity), and alkylammonlam-montmorillonites (e.g., basal spacing), respectively, are similar to each other. Moreover, among all the cation- exchanged montmorillonites (5 wt%) studied, only [H3N(CH2)5CH3]+- montmorillonites gives a very broad DSC peak (or sluggish reaction). Table 11.4 also shows that the heat of reaction (550-555 J/g) remains constant within experimental errors regardless of the exchange cations used. The average heat of reaction is 553 J/g for 5/95 (w/w) organoclay/epoxy resin or 582 Hg for the neat epoxy resin. This value is equivalent to 52.6 kcal/mole, which is essentially the same as that (53.7 kcal/mole) derived from the values listed in Table 11.1. D. CONCLUSIONS In this work we have achieved the direct nanoscopic delamination of smectite clay layers in a polyether matrix derived from an epoxy resin. The exfoliation of the acidic montmorillonites in the epoxy resin takes place instantaneously at the delamination-polymerization temperature. In addition . . . . + . . . to simple ac1dlc catlons such as H and NH4+, varlous onlum lons 71 25 . (a) . (a) 5% [H3N(CH2)11CH3] Mon!- 20- (b) 5% memcngscnaf-Mont A 154 i” E 10- E (b) 8 = 5 . 1 L 0 .. ' _ __. q 198 °c 287 °c 550 1/3 554 Hz 150 200 250 300 350 400 Temperature (°C) Figure 11.13. DSC thermograms of epoxy polymerizations catalyzed by 5 wt% (a) [113N(CH2)11CH3]+- and (b) [113N(C1-12)5CH3]+-montm01illonites, a heating rate of 20 'C/min being used. 72 (protonated primary amines, diamines, and aminocarboxylic acids) are suitable delaminating agents. The delamination-polymerization temperature depends on the heating rate and cation-exchanged montmorillonite. The heating rate being equal, cation acidity (or proton availability) and interlayer gallery height (or steric hindrance) are two important factors affecting the delamination- polymerization temperature. Within experimental errors, the heat of epoxy polymerization is independent of the cation exchanged form of montmorillonite, suggesting that differences in the clay exfoliation, energy are small relative to the epoxy polymerization energy. The activation entropy, as reflected in the pre-exponential factor, is important in determining the reactivity of the organoclay-epoxy system, although the activation enthalpy is the major component of the activation free energy. Owing to their powdery texture, the clay-polymer nanocomposites of the present work are not conducive to mechanical measurements. Power formation is caused by phase segregation of the clay-bound polyether and the epoxy resin. It should be possible, however, to circumvent this limitation using more compatible cyclic ethers as polymer precursors. Also, powder processing studies currently in progress may offer a convenient route to the formation of monolithic structures. 10. 11. 12. 73 LIST OF REFERENCES A. Okada, M. Kawasumi, A. Usuki, Y. Kojima, T. Kurauchi, and O. Kamigaito, Mater. Res. Soc. Symp. Proc., 171, 45 (1990). A. Okada, M. Kawasumi, T. Kurauchi, and O. Kamigaito, Polym. Prepr. , 28 (2), 447 (1987). Y. Kojima, A. Usuki, M. Kawasumi, A. Okada, T. Kurauchi, and O. Kamigaito, J. Polym..Sci.: Part A: Polym. Chem., 31, 983 (1993). S. Fujiwara and T. Sakamota, Japan Patent 51,109,998 (1976). Y. Fukushima, A. Okada, M. Kawasumi, T. Kurauchi, and O. Kamigaito, Clay Miner., 23, 27 (1988). Y. Fukushima and S. Inagaki, J. Incl. Phenom., 473 (1987). A. Usuki, T. Mizutani, Y. Fukushima, M. Fujimoto, K. Fukumori, Y. Kojima, N. Sato, T. Kurauchi, and O. Kamigaito, U. 8. Patent 4,889,885 (1989) R. E. Grim, "Clay Mineralogy," 2nd ed., McGraw-Hill, New York (1968). C. A. May, Ed., "Epoxy Resins," 2nd ed., Marcel Dekker, New York (1988). H. Lee and K. Neville, "Handbook of Epoxy Resins," McGraw-Hill, New York (1967). H. J. Borchardt and F. Daniels, J. Am .Chem. Soc, 79, 41 (1957). L. J. Taylor and S. W. Watson, Anal. Chem., 42, 297 (1970) 13. 14. 15. 16. 17. 18. 19. 20. 21. 74 R. B. Prime, Polym. Eng. Sci., 13, 365 (1973) A. A. Duswalt, Thermochim. Acta., 8, 57 (1974) H. E. Kissinger, Anal. Chem., 29, 1702 (1957). R. B. Prime, ”Thermal Characterization of Polymeric Materials," E. Turi, Ed., Academic Press, New York (1981). T. Ozawa, J. Therm. Anal., 2, 301 (1970). C. D. Doyle, Anal. Chem., 33, 79 (1951). P. Peyser and W. D. Bascom, J. Polym. Sci., Polym. Phys. Ed., 18, 129 (1975). S. J. Swarim and A. M. Wims, Anal. Calorim., 4, 155 (1976). M. Cizemciogly and A. Gupta, SAMPLE Q., 16 (April, 1982) CHAPTER III CLAY-POLYMER NANOCOMPOSITES FORMED FROM BASIC HYDROTALCITES AND EPOXY RESIN A. INTRODUCTION As mentioned earlier, delaminated montmorillonite has been shown very useful in the design of polymer-based composite materials with improved mechanical and physical properties. Recently, Yano and co- workers (1) reported that delaminated montmorillonite (e.g., 5 wt%) reduced the gas permeability (25 vs. 110 g.mm/m2, 24 h) and thermal expansion (50 vs. 67 x 10'5/°C, 250 °C) of polyimide. In addition, Usuki and co-workers (2) used protonated 12-aminolauric acid as the exchange cation and showed that delaminated montmorillonite was a better reinforcement for nylon-6 than delaminated saponite (another smectite clay). However, no delaminated clay-polymer nanocomposites formed from non-smectite clays have been reported in the literature. In general, clay minerals are divided into cationic and anion clays by their layer charges and interlayer ions. Mica, talc, kaolinite, and montmorillonite are typical cationic clays. Figure 111.1 shows the general structure of anionic clays or layered double hydroxides (LDHs). A LDH results from the partial isomorphous substitution of M2+ ions for M3+ ions in octahedral M(OH)2 layers. The resulting positive layer charge is 75 76 d003 Figure 111.1. General structure of anionic clays or layered double hydroxides (LDHs); e.g., hydrotalcite Mg6A12(OH)16(CO3)t4H20, where 112* = Mg“; M3+ = A1“; and A” = C0321 77 balanced by interlayer anions An" solvated by water molecules. The general formula of LDHs can be written as [M1_,2+M,3*(0H)2]A,,,,“'-yH20 (0.15 < x < 0.44), where M2+ = Mg“, Zn“, Coz+, N12+, Cu2+, etc.; M3+ = A1“, Cr3+, Fe“, etc.; and An" = CO3'2, 804-2, OH', NO3', Cl', etc. (3- 7). For instance, hydrotalcite Mg6A12(OH)15(CO3)-4H20 is the most well- known natural and synthetic LDH. Miyata and Kimura (8) first reported a series of dicarboxylate ions intercalated in LDHs. However, Drezdon (9) used the c0precipitation method but failed to duplicate the sample preparation. Several groups used the anion-exchange method (10) and especially, the calcination method (11- 13) instead and succeeded in preparing the same organic anion-LDHs. Among them, Dimotakis and Pinnavaia (13) reported a new route to well- ordered organic anion-LDHs, although this method required glycerol and meixnerite as a swelling agent and a precursor, respectively. Calcined synthetic hydrotalcite Mg6A1203(OH)2 or mixed metal oxide (MgO)6-A1203 has been used as a catalyst for polymerizations of B- propiolactone (14) and propylene oxide (15-16). The catalytic property of calcined hydrotalcite arises from its extremely high basicity (pKa ~ 35) and surface area (~ 200 m2/g). Tanaka et al. (17) used NO3'-hydrotalcite as a precursor to prepare polyacrylate-hydrotalcite (d003, = 13.4 A). However, the preparation of acrylate-hydrotalcite (9003 = 13.8 A) required very high acrylate concentrations (e.g., 1.69 M for 5 g of NO3'-hydrotalcite) and long reaction time (1 week, 25 °C). Our interest in the design of new clay-polymer nanocomposites led us to examine aminocarboxylate- and dicarboxylate-hydrotalcites as catalysts for the epoxy polymerization. In this study we emphasized the synthesis and delamination of aminocarboxylate-hydrotalcite, neither which 78 has been reported in the literature. The epoxy resin selected for this study was the same epoxy resin EPON-828 as used before. B. EXPERIMENTAL Materials Epoxy resin EPON-828 (Shell), aminocarboxylic acids H2N(CH2)n- ICOOH, dicarboxylic acids HOOC(CH2)n.2COOH (n = 6 and 12) (Aldrich), and hydrated metal nitrates Mg(NO3)2-6H20 (EM Science) and Al(NO3)3-9H20 (J. T. Baker) were used as received. Sample Preparation Hydrotalcite was prepared by using a slightly modified Reichele et at. method (18). Mg(NO3)2-6H20 (0.50 mol) and Al(NO3)3-9H20 (0.25 mol) were dissolved in 350 mL of deionized water. The solution was dropwise added to a solution of NaOH (1.75 mol) and NazCO3 (0.472 mol) in 570 mL of deionized water. With vigorous agitation, the addition was carried out at 35 °C for 4 h and the reaction continued at 65 °C for another 18 h. The heavy white slurry was centrifuged and the hydrotalcite precipitate was washed with deionized water. Washing and centrifugation were repeated until no or little brown A g20 precipitate was observed when a drop of 0.1 N AgNO3 was added to the centrifugate. The washed hydrotalcite was dried in an oven at 125 °C for 18 h. The oven-dried hydrotalcite was ground and then calcined at 500 °C for 3 h. During calcination, hydrotalcite lost CO32' ions as C02 and became the mixed metal oxide. 79 . H2N(CH2)n-1COOH, H2N(CH2)n-1COOH or HOOC(CH2),,-2COOH (30 mmol) was dissolved in 1.5 L of a 0.02 N (0.04 N for Na2[OOC(CH2)n.2COO]) NaOH solution at 80 °C. The deionized water used for sample preparation was boiled for an hour to remove dissolved C02. Fresh calcined hydrotalcite (3.00 g) was thoroughly dispersed in the 0.02 N solution of interest at 80 °C for 12 h. The organic anion-exchanged hydrotalcite was centrifuged and washed with deionized water. Washing and centrifugation were repeated until no or little brown Ag20 precipitate was observed when a drop of 0.1 N AgNO3 was added to the centrifugate. The washed organic anion-hydrotalcite was dispersed in 250 mL of deionized water and freeze-dried in a Labconco Freeze Dryer Model 18. The freeze-dried organic anion-hydrotalcite was then sieved to < 325-mesh. H2N(CH2)1ICOO'-hydrotalcite (0.79 g) was added to 15 g of epoxy resin EPON-828 and the 5/95 (w/w) clay/epoxy mixture was magnetically stirred in a 250-mL beaker at 75 °C for 30 min. The beaker was sealed with aluminum foil and the temperature was then raised at a rate of ~ 20 °C/min to 280 °C. The partial liquid-to-powder transformation of the clay-epoxy mixture took place within 13 min. The clay-epoxy nanocomposite powder was separated from the solidified polyether and sieved to < 325-mesh, with _ grounding if necessary. No other clay-epoxy nanocomposite was prepared from the other anion-exchanged hydrotalcites because their polymerization-delamination temperatures are well above the flash point of the epoxy resin, ~ 325 °C. Characterization The hydrotalcite-epoxy nanocomposite sample for powder XRD studies was firmly pressed onto a Rigaku sample holder. The diffraction 80 pattern was then recorded by monitoring the diffraction angle 20 from 2° to 45° on a Rigaku Rotaflex Ru-200BH X-r'ay diffractometer. The diffractometer was equipped with a Ni-filtered Cu-Ka radiation source operated at 45 kV and 100 mA. The scanning speed and the step size used were 2°lmin and 0.02°, respectively. Quartz was used as a calibration standard. A small amount of hydrotalcite-epoxy nanocomposite powder for TEM studies was added to a Spurr's mixture in a silicon rubber mold, which was then placed in an oven at 70 °C overnight. The cured epoxy block was trimmed and thin-sectioned with a diamond knife in a microtome. The resulting ultra-thin sections (~ 90 nm in thickness) were mounted on a plastic-film support on a copper grid and examined by a JEOL 100CX transmission electron microscope operated at an accelerating voltage of 100 kV. The mixture of organic hydrotalcite-epoxy resin for DSC analysis (4 mg) was placed in an aluminum sample pan, which was then sealed hermetically. The sample and reference pans were placed in the cell of a Du Pont 91'0 differential scanning calorimeter. The DSC cell was then heated under nitrogen (50 mL/min) to 450 °C at a rate of 20 °C/min. Indium (mp, 156.6 °C and AHm, 28.42 J/g) was used for DSC calibrations. H2N(CH2)1lCOO'-hydrotalcite (20 mg) for TGA analysis was placed in a platinum sample container, which was then placed in the thermobalance of a Du Pont 990 thermogravimetric analyzer. The sample weight was tared and the sample was then heated under nitrogen (100 mL/min) to 650 °C at a rate of 20 °C/min. 81 C. RESULTS AND DISCUSSION Calcined hydrotalcite or amorphous mixed metal oxide (MgO)5-A1203. (Figure 111.2c) was prepared just before use to prevent it from absorbing water and C02 in the air. The extremely basic mixed metal oxide easily converts absorbed C02 and water into CO32' ions to form hydrotalcite (Figure 111.2a). For instance, Figure 111.2b shows the XRD pattern of the partially reconstituted hydrotalcite formed from calcined hydrotalcite stored in a plastic bottle for a month. Deprotonated aminocarboxylic acids H2N(CH2)n.1COOH,. and dicarboxylic acids HOOC(CH2)n.2COOH (n i- 6 and 12) were used as delaminating agents'or catalysts rather than curingagents in thisawork.~.1t is essential to‘ treat ‘thesewor’ga’nic- acids with dilute aqueous NaOH (or NH4OH) solution. Alkaline treatment converts aminocarboxylic acids and dicarboxylic acids into the corresponding water-soluble anions, which can be intercalated in reconstituted hydrotalcite. In the case of an amino- carboxylic acid, deprotonation converts the dipolar ion 1 into the anion 11. + OH‘ + o o H,N(c11,),,,d-o- 112N(c11,),.,d-0' +11+ I 11 Evidence of'Clay Delamination As shown in Figure 111.3d, no clay diffraction peaks are observed for a clay-epoxy nanocomposite containing 5 wt% H2N(CH2)“COO°- hydrotalcite. Only a very diffuse scattering peak characteristic of the amorphous cured epoxy resin (or p01yether) appears in the XRD powder 82 7.62 A (a) Hydrotalcite ' .1 (b) Month-01d Calcined HT (c) Fresh Calcined HT J .. b . b '5 '1 - “as 1 3.81 A r i . , _ fl 2.60 ‘ 254 A ' '1 ,. ‘ .4. .m M (a) ‘ A _ “A W (b) q f _" __ .4- _ WW4 - __ 4.3:- 1 ...A (C) .. O 10 20 3O 40 50 Diffraction Angle (2 0) Figure 111.2. XRD patterns of (a) hydrotalcite (HT), (b) month—old calcined HT, and (c) fresh calcined HT (500 'C, 3 h) (i.e., Mg5A1208(OH)2 or mixed metal oxide (Mame/“203)- 83 l a l 1 l 4 l A l a J a 1 a l 1_ l L 22.8 A (a) H2N(CH2)11COO'-HT (25 °C) (b) AAA12-HT (322 'c, 1 min) _ (c) AAA12-HT (280 'c, 13 min) (d) 5% AAA12-HT/Epoxy Nanocomposite Relative Intensity Diffraction Angle (2 0) Figure 111.3. XRD patterns of (a) H2N(C112)11COO'-hydrotalcite (AAA12-HT) (25 'C), (b) AAA12-HT (322 ’C, 1 min), (c) AAA12-HT (280 'C for 13 min), (d) a clay— epoxy nanocomposite containing 5 wt% AAA12-HT. 84 pattern. The absence of a 22.8-A peak for H2N(CH2)1ICOO'-hydrotalcite (Figure 111.3a) suggests that the clay tactoids have been exfoliated and the 4.8-A-thin clay layers dispersed at the molecular level. However, the clay peaks with reduced intensities (Figures 111.3b and 111.3c) show that the partial thermal decomposition of pristine H2N(CH2)1ICOO'-hydrotalcite occurs when it is heated at 322 °C for 1 min and to the greater extent, at 280 °C for 13 min under nitrogen. As shown in Figure 111.4, the TEM micrograph of a clay-epoxy nanocomposite containing 5 wt% H2N(CH2)11C00°~hydrotalcite reveals that the cured epoxy resin infiltrates submicron-sized clay tactoids and expands interlayer spacing up to ~ 200 A. Apparently the epoxy resin is responsible for the retention of the clay layered structure. That is, the epoxy resin reacts with the organoclay and affects its degree of thermal decomposition. Figure 111.5 shows the DSC thermogram for a 5/95 (w/w) mixture of H2N(CH2)11C00'-hydrotalcite and epoxy resin EPON-828. The thermogram indicates that the clay delamination-epoxy polymerization takes place at an onset temperature of 321 °C and the heat of reaction is 572 J/g. For comparison, the DSC thermograms of the neat epoxy resin and H2N(CH2)1ICOO'-hydrotalcite are shown in Figures 11.4 and 111.6, respectively. It should be noted that the scale of the heat flow in Figure 111.6 is much smaller than that in either Figure 11.4 or 111.5. The DSC and TGA (Figure 111.7) curves indicate that H2N(CH2)1ICOO'-hydrotalcite loses surface and interparticle pore water between 40 and 200 °C. Interlayer water is liberated in the region of 200- 300 °C. The losses of hydroxyl groups in the clay framework (or dehydroxylation) and C032- ions as C02 (or decarbonation) occur between 85 Figure 111.4. TEM micrograph of a clay-epoxy nanocomposite containing 5 wt% H2N(CH?)11COO'-hydrotalcite: x72,000. 86 5 4- 5% H2N(CH2)11COO°-H'1‘/EPON-828 357 'c 3 _ 3 . E 2 - é . <0 1 "‘ 8 = -t 0 _ 1 -1 Y A : 1 321 ‘C V 569 J/g -2 . , r , r , . . . . . 150 200 250 300 350 400 450 Temperature ('C) Figure 111.5. DSC thermogram of the epoxy polymerization catalyzed by 5 wt% H2N(CHz)uCOO'-hydrotalcite, a heating rate of 20 'C/min being used. 87 0.0 H2N(CH2)1 yCOOZHT -0.5 u A E’ E -l.o -4 E. {5 -15 - I -2.0 .. ‘15 v 1 W 1 t I v I fl 0 100 200 300 400 500 Temperature (°C) Figure 111.6. DSC thermogram of H2N(CH2)11COO'-hydrotalcite, a heating rate of 20 'C/min being used. 88 100 112N(CHZ)1 10001111 70‘ Weight (%) I ' I ' I o - 100 - 200 - 300 400 500 600 700 Temperature (°C) Figure 111.7. TGA curve of H2N(CH2)1ICOO'-hydrotalcite, a heating rate of 20 'C/min used. 89 300 and 650 °C (18, 19). The heat absorption and the weight loss shown by the DSC and TGA curves, respectively, substantiate the XRD pattern (Figure 111.3b) that the partial loss of interlayer water results in the reduced peak intensity for H2N(CH2)1ICOO'-hydrotalcite heated at 322 °C for 1 min. More importantly, the heat of reaction for H2N(CH2)11COO'- hydrotalcite (especially 5 wt%) is much smaller than that for either the uncatalyzed epoxy polymerization or the epoxy polymerization catalyzed by 5 wt% H2N(CH2)1ICOO'-hydrotalcite. Reaction Mechanism According to the evidence presented above, the delamination of the organic anion-hydrotalcites in the epoxy resin or epoxy polymerizations catalyzed by the organic anion-hydrotalcites at elevated temperatures can be rationalized by the following general equations: I .. . : O O\CH O O’ R‘C-O‘ + 61,- R —-—» 1210-0012-0111 (111.1) 9 0,: 9 0H -O-dxCthlNHz + 612' HR —.' -O'dXCthlflN'CHZ'éHR (1112) O OH O 9 OH 'O-C(CH2)11HN-CH2-CHR + 641th —> 'O-C(CH2)11N(CH2-CHR)2 (111.3) W: 90 - 0H 0H RZOH "l" I) 61% 1"" R’O(CHz-CHRO),,,1CH2-CHR (H104) where RICOO‘ is any molecule containing one or two carboxyl groups and RZOH is monomer or oligomer possessing one or more hydroxyl groups. At the delamination-polymerization temperature, epoxy monomers first react with aminocarboxylate and dicarboxylate ions intercalated on the edges of hydrotalcite particles. Through an 8N2 mechanism, the cleavage of epoxy rings (Equation 111.1) takes place first. In the case of H2N(CH2)”COO' ion, the other two epoxy ring openings or amine-epoxy reactions (Equations 111.2-3) also occur. The reactions propagate toward clay interlayer space and the propagation causes the expansion of the clay gallery region. In the meantime, various base-catalyzed etherifications (Equation 111.4) occur to form polyether in the expended region and this results in the complete delamination of clay layers. The cured epoxy resin or polyether in the clay galleries becomes phase-segregated from the uncured resin. This gives rise to the liquid-to-powder transformation characteristic and an increase in the bulk volume upon the formation of hydrotalcite-epoxy nanocomposites. A large heat of reaction also accompanies the clay delamination-epoxy polymerization. Reaction Temperature Listed in Table 111.1, the onset temperatures of the epoxy polymerization-hydrotalcite delamination reactions increase in the order H2N(CH2)11COO' < nooc '5 .2! 0 a: Diffraction Angle (2 0) Figure 111.8. XRD patterns of (a) "H2N(CH2)5COO"'-hydrotalcite or more accurately, meixnerite Mg6A12(OH)16(OH)2-4H20 and (b) hydrotalcite. 94 Table 111.2. Estimated Epoxy Cure Percents Due to Aminocarboxylate and . Dicarboxylate Anions Formula Equiv. Cure W Interlayer Anion Weight Weight“ 41 mmolb Equiv. Cureb H2N(C1-12)11COO' 214 988 10.12 3 6.04 H2N(C1-17)5COO' 130 820 12.20 3 7.28 HOOC(CH2)10COO' 229 1018 9.28 2 3.91 HOOC(CH2)4COO' 145 850 11.76 2 4.68 [OOC(CH2)10000]2' 228 778 6.34 2 2.52 [OOC(CH2)4COO]2' 144 704 7.10 2 2.83 “The resulting weight of 1 mol of hydrotalcite Mg6A12(OI-1)16(CO3).4H20 (F.W., 604) after the anion-exchange reaction. bBased on 100g of 5/95 (w/w) organoclay/epoxy resin (251 mmol or 503 meq). 95 Effects of Clay Concentration H2N(CH2)1ICOO’-hydrotalcite has been chosen for further investigation because it has a relatively low polymerization-delamination temperature. Figure 111.9 shows a plot of multiple DSC curves for epoxy polymerizations catalyzed by H2N(CH2)1ICOO'-hydrotalcite at various clay concentrations (120 wt%). As the organoclay (or organic anion) concentration increases, the initiation reactions (Equations 111.1-3) become more comparable to the propagation reaction (Equation 111.4) with respect to the amount of heat released. At a high clay concentration (e.g., 20 wt%), high-temperature side reactions complicate the DSC curve, which is composed of several peaks with relatively small heat flows. As shown in Figure 111.10, the polymerization-delamination temperature is inversely proportional to the clay concentration up to 20 wt%. The base-catalyzed epoxy ring opening is an 8N2 reaction of which the reaction rate depends on the concentrations of the reactants. The linear relationship suggests that the pre-exponential factor increases with increasing the organoclay or the organic anion concentration. In Figure 111.11 the heat of reaction is plotted against the clay concentration for H2N(CH2)1ICOO'-hydrotalcite. That is, the heat of reaction is proportional to the epoxy concentration. The extrapolated values are 607 Hg or 54.9 kcal/mol for the neat epoxy resin. The linear relationship again shows that the heat of reaction is mainly due to the polymerization of the epoxy resin. Kinetics Parameters Figure 111.12 shows a plot of multiple DSC curves generated at various heating rates (15-25 °C/min) for the epoxy polymerization 96 14 12 q H2N(CH2)11C(X)--HT l 10-1 3 '1 2 s- g F E .. g 4 4 i1 2‘ .i! o I fi 1 I I I fi 100 150 200 250 300 350 400 450 Temperature (°C) Figure 111.9. DSC thermograms of epoxy polymerizations catalyzed by H2N(CH?)11COO°-hydrotalcite at various clay concentrations: 1, 5, 10, and 20 wt%. 97 400 0 a H2N(CHz)uCOO'-HT :53 350- 8 E E 3. 300‘ E [- 5, 250- 3‘: 8 200- E O 150 V 1 V I V I V 1 V 0 5 10 15 20 25 Clay Concentration (wt%) Figure 111.10. Onset temperature of the epoxy polymerization-clay delamination as a function of the clay concentration for H2N(CH2)11COO'-hydrotalcite. 98 660‘ .lflflNazflihflCCKYJTT o maNlcnarrcoom*-Mont Heat of Reaction (J/g) m V I V T V I V I V 0 5 10 15 20 25 Clay Concentration (wt%) Figure 111.11. Heat of reaction as a function of the clay concentration for (.) H2N(C1-17)11COO'-hydro1alcite (o) [113N(CH2)1ICOOH]+-montmorillonite. 10 - H2N(CH2)11C00--HT Heat Flow (W/g) 4 20 °C/min 15 °C/min 2 _ 10 °C/min 0 . j . , . . . r . r . 150 200 2.50 300 350 400 45( Temperature (°C) Figure 111.12. DSC thermograms of the epoxy polymerization catalyzed by 5wt% H2N(C117)11COO°-hydrotalcite at various heating rates: 10, 15, 20, and 25 'C/min. 100 catalyzed by 5 wt% H2N(CH2)1[COO'-hydrotalcite. The peak maximum temperature increases with increasing the heating rate. Despite the increased peak area and the heating rate, the heat of reaction remains (3011818111. Figure 111.13 shows a linear relationship between 1n (p and l/Tp for the epoxy polymerization catalyzed by 5 wt% H2N(CH2)11COO'- hydrotalcite. However, the ratio of 2Tp to EJR ranges from 9.45 to 9.84%, so that 2T9 is too large to be omitted. According to Equation 11.9, the activation energy of the basic organoclay-catalyzed epoxy polymerization is ‘ 25.7 :1: 0.5 kcal/mol. By using Equation 11.10 and the activation energy obtained above, the pre-exponential factor is calculated as 8.66 :l: 0.38 x 106 s'1 (Table 111.3). Moreover, based on Equations 11.11-13 and m a: 2, the activation entropy, enthalpy, and free energy of the catalyzed epoxy polymerization are -32.1 :1: 0.2 cal/Kmol and 23.3 i 0.5 and 42.5 :I: 0.7 kcal/Kmol (at 595 K or 322 °C), respectively. Comparison of Organoclays Table 111.4 summarizes DSC, XRD, and TEM data for epoxy polymerizations catalyzed by 5 wt% H2N(CH2)1ICOO'-hydrotalcite and [H3N(CH2)1ICOOH]+-montmorillonite, respectively. The polymerization- delamination temperature is 93 °C higher for H2N(CH2)11COO'- hydrotalcite, although its basal spacing is 5.8 A greater than that of [H3N(CH2)1ICOOH]+-montmorillonite. The reason is that the steric effect is much more important in the base-catalyzed epoxy ring opening, considering the fairly bulky epoxy monomer and the organic anion confined between hydrotalcite layers. Both organophilic anionic and cationic clays 101 3.4 32 .1 C H2N(CHz)ucm.-HT o 1113100112), ,coorn*-lviont. 3.0 " 2.8 " In it) (K/min) 2.6 ' 2.4 '1 2.2 ' I ' I ' I ' I ' T ' 1.5 1.6 1.7 1.8 1.9 2.0 2.1 1/1‘, x 1000 (UK) Figure 11L13. Linear relationship between the natural logarithm of the heating rate (p and the reciprocal of the peak maximum temperature Tp for epoxy polymerizations catalyzed by 5 wt% (.) H2N(CH2)11COO'-hydrotalcite (o) [H3N(CH2)11COOH]+— montmorillonite. . " ' if" 0 102 Table 111.3. Pre-exponential Factors for Epoxy Polymerlntions Catalyzed by 5 wt% lZ-AminolaurataHydrotalcite Peak Max. Pre-exponential Interlayer Anion Heating Rate Temperature Factor cumin) (°C) (,1, H2N(c11,)ncoo 10 337.4 9.08 rt 10‘ 15 350.8 8.29x106 20 358.5 8.39x106 25 363.1 8.89x106 Note: The activation energies of the epoxy polymerization catalyzed by 5 wt% H2N(C1-12)11COO'-hydrotalcite is 25.7 i 0.4 WI. 628595 888830: ..: «88% Baa—:38 8:838 2t.c .8806 an .8 88 08.8: a 08m: .3 8:85: 803 88 Dana 103 008 l :6 H 0.2 _.c H 0.3” _.c H QMN fie H N._Nt :0? 8.0 H cw; fic H awn n4 H Yen _ H aNN 4:02-30: com 1. v.0 H ”mm ad H ndv v.0 H QMN Nd H _.~m- 02x and H we.» «.0 H Emu «A H new N H «mm 5.2% A 0 e8 8 e 8 A .60 8 ...:—=8 e eufiwanm gmfinm Afiw—flé $2898”th CEfiW—hmv Lewes..— Agymé A829“: ...:—m“? 83280.5 8.3.88? 88¢ 8...: 8w. :e_.:>=u< 85:380. 85:98:89.5 8:81.84. he 28% ..:—2:“— eo._o_e.e._o:=.o§=e_o=_5<fi 2... 3.882582 .23. 8588...": 8.232: e... m 3 88.38 26:85:56: :8: ..o. an: SE. ...a .88 .8: .6 .E as; 104 give rise to virtually the same heat of reaction, Arrhenius activation energy, and activation. enthalpy. These nearly identical reaction energies strongly suggest that the-heat of epoxy polymerization is:. much larger than» that of ‘ either clay delamination, and that the activation entropy accounts for the difference in the reaction reactivity between the two organoclay-epoxy systems. .Thermore negative activation entropy reflects the smaller pre- exponential factor‘and imparts. the higher activation free energy to the anionic clay-epoxy system. The activation free energies indicate that basic H2N(CH2)1ICOOT-hydrotalcite is a less effective catalyst for the epoxy polymerization than acidic [H3N(CH2)'11COOH]+-montmorillonite. D. CONCLUSIONS In this work we have achieved the direct nanoscopic delamination of hydrotalcite layers in a polyether matrix derived from an epoxy resin, although deprotonated lZ-aminolauric acid is the only delaminating agent or catalyst that substantially reduces the delamination-polymerization temperature. we also first used a relatively simple method to prepare a well- ordered aminocarboxylate-LDH (12-aminolaurate-hydrotalcite) with a super gallery ((1003 = 22.8 A). The delamination-polymerization temperature depends on the heating rate, clay concentration, and anion-exchanged hydrotalcite. The heating rate and the clay concentration being equal, interlayer gallery height or steric hindrance is an important factor affecting the delamination-polymerization temperature. Within experimental errors, the heat of epoxy polymerization is independent of the anion exchanged form of hydrotalcite, suggesting that 105 differences in the clay delamination energy are small relative to the epoxy polymerization energy. Compared with acidic [H3N(CH2)1ICOOH]+-montmorillonite, basic ' H2N(CH2)11C00'-hydrotalcite is a less effective catalyst for the epoxy polymerization. The activation entropy, as reflected in the pre-exponential factor, accounts for the difference in reactivity between the two organoclay- epoxy systems. 12. 13. 14. 15. 106 LIST OF REFERENCES K. Yano, A. Usuki, A. Okada, T. Kurauchi, and O. Kamigaito, Polym. Prepr. , 32, 65 (1991). A. Usuki, Y. Kojima, M. Kawasumi, A. Okada, T. Kurauchi, and O. Kamigaito, Polym. Prepr. , 31, 651 (1990). H. C. B. Hansen and R. M. Taylor, Clay Miner., 26, 311 (1991). K. A. Carrado, A. Kostapapas, and S. L. Suib, Solid State Ionics, 26, ‘77 ( 1988). W. T. Reichel, Solid State Ionics, 22, 135 (1986) W. T. Reichle, CHEMTECH , 16 , 58 (1986). S. Miyata, Clays Clay Miner., 31, 305 (1983). S. Miyata and T. Kumura, Chem. Lett., 843 ( 1973). M. A. Drezdzon, Inorg. Chem. , 27, 4628 (1988). M. Meyn, K. Beneke, and G. Lagaly, Inorg. Chem., 29, 5201 (1990). K. Chibwe and W. Jones, J. Chem. Soc., Chem. Commun., 926 (1989). E. Narita and T. Yamagishi, Proc. 9th Inter. Clay Conf., 145 (1990). E. D. Dimotakis and T. J. Pinnavaia, Inorg. Chem., 29, 2393 (1990). T. Nakatsuka, H. Kawasaki, s. Yamashita, and s. Kohjiya, Bull. Chem. Soc. Jpn., 52, 2449 ( 1979). S. Kohjia, T. Sato, T. Nakayama, and S. Yamashita, Makromol. Chem., Rapid Commun., 2, 231 (1981). 16. 17. 18. 19. 107 . D. E. Laycock, R. J. Collacott, D. A. Skelton, and M. F. Tchir, J. Catal., 130, 354 (1991). M. Tanaka, I. Y. Park, K. Kuroda, and C. Kato, Bull. Chem. Soc. Jpn., 62, 3442 (1989). ’ W. T. Reichle, S. Y. Kang, and D. S. Everhardt, J. Catal., 101, 352 (1986). F. Rey, V. Fornes, and J. M. J. Rojo, Chem. Soc., Faraday Trans., 88, 2233 (1992). GR STRTE NV HIMTIHIIII unifigulgngullfijflifll“ 129301 nrcut 33