CHAPTER-5 BISMALEIMIDE-ALLYL VLAC LIGMERS: SYTHESIS AD CURE KIETICS This chapter deals with Bismaleimide resin systems, with a novolac back bone, containing both maleimide and allyl functionalities incorporated in the same oligomer. The synthesis, characterization and curing kinetics of the single component resin system have been detailed in the following section. Furthermore, the mechanical, thermomechanical and thermal properties of the single component oligomer/carbon fibre reinforced composites are detailed in the present chapter. 5.1 Introduction The extent of cure and efficiency of the reactive blending depend, among many factors, such as, the accessibility of the reactive sites in the polymer chain and miscibility of the components of the blends. Blends of polymers with dissimilar backbone structure could face problems of immiscibility and consequent phase separation, leading to difficulty in the accessibility of reactive functional groups located on them. This can be obviated if a single component thermoset is employed. Thus, in the present chapter, bismaleimide resin systems, with novolac as back bone, containing both maleimide and allyl functionalities in the same macromolecule is developed and characterized. 5.2 Experimental 5.2.1 Materials Material and characterization methods are detailed in chapter 3, section 3.2 5.2.2 Synthesis 5.2.2.1 Synthesis of single component BMI resin system The synthesis of single component BMI resin system involved synthesis and characterization of allyl novolac, maleimido benzoyl chloride and then the co- reaction between maleimidobenzoyl chloride and allyl novolac to form the final BMI resin system. University of Mysore 99
5.2.2.2 Synthesis of Maleimido Benzoyl chloride (MBC) and Allyl novolac Maleimido Benzoyl chloride (MBC) from maleimidobenzoic acid and allyl novolac was synthesized by a reported procedure and characterized. Synthesis of MBC from Maleimidobenzoic acid and Allyl novolac are represented in Figure5.1 and Figure5.2 respectively. H 2 CH + Ac2 Sodium acetate Chemical imidization HC SCl 2 TEA ClC Fig. 5.1: Synthesis of Maleimido Benzoyl Chloride H H H + CH 2 H + CH 2 [ H H ] n Fig. 5.2: Typical synthesis scheme of allyl novolac University of Mysore 100
5.2.2.3 Synthesis of Maleimide incorporated Allyl ovolac oligomer (ne component BMI resin system) Three compositions of single component BMI resin systems (BMI-1, BMI-2 and BMI-3) were synthesized incorporating varying extents of maleimide in the novolac back bone. The general method for the synthesis of maleimide incorporated single component BMI system is as follows: Calculated quantity of Maleimido benzoyl chloride was reacted with required quantity of allyl novolac (50 % allylated, synthesized as described in the chapter 4) in THF at room temperature. Calculated quantity of TEA was added in to the reaction mixture to remove the HCl formed during the course of the reaction. Reaction was continued at room temperature for 8 hrs in 2 atmosphere. The reaction mixture was filtered, and the filtrate was precipitated in methanol and then dried under vacuum at 50 C. The copolymer was characterized by elemental analysis, FTIR, 1 H MR and DSC. The synthesis of single component alder ene polymer system is depicted in Figure 5. 3. H H C-Cl TEA, 2 + RT H Fig. 5.3: Synthesis maleimide incorporated allyl novolac one component BMI oligomers University of Mysore 101
5.3 Results and Discussion 5.3.1 Characterization of Maleimido benzoic acid Maleimido benzoic acid was synthesised and characterized by FTIR, Acid value and itrogen content. The acid value was found to be 260.5 mg KH/g (theoretical 258 mg KH/g) and the nitrogen content was estimated to be 6.4 % (Purity 97%). From FTIR, the characteristics peaks due to -H stretching appears at 3200 cm -1, asymmetric C= of CH and =C- stretch appears together at 1706 cm -1. Further, the peaks at 1774 cm-1, 1383 cm -1 and 700 cm -1 appear due to the C= stretch,c--c and C=C of maleimides respectively. 5.3.2 Characterization of Allyl novolac The hydroxyl value of the synthesized allyl novolac was around 400-500 mg KH/ g with a number average molecular weight of 500-600. In the FTIR spectrum of allyl novolac, strong absorption at 3419 cm -1 confirms the presence of hydroxyl groups. The presence of allyl functionality is confirmed by absorption peak at 1633 cm -1 due to (- C=C-) stretch. C-H stretching is observed at 2977 cm -1, 2843 cm -1, 3013 cm -1 and absorption due to aromatic ring is observed at 1599 cm -1. From the MR spectra, chemical shift due to allyl protons occur at 5 ppm whereas that due to aromatic protons is observed at 6.5 7.5 ppm and the ratio of allyl phenol to aromatic phenol was found to be 1:1 [The Characterization of A is described in detail in chapter 4] 5.3.3 Characterization of Single component alder ene system Spectroscopic characterization of the single component system was done by FTIR and 13 C MR. Figure 5.4 shows the FTIR spectrum of the single component system synthesized. University of Mysore 102
Fig. 5.4: FTIR spectrum of single component alder-ene resin Characteristic absorptions due to the maleimide ring observed at 689 cm -1,760 cm -1, 827cm -1, 1016 cm -1, 1106 cm -1, 1383 cm -1, 1603 cm -1 and that due to -H stretching appears at 3456 cm -1. C=C stretch was observed at 1510 cm -1 and 1603 cm -1. The peak at 1706 cm -1 due to CH group observed in MBA disappeared and a new peak due to formation of ester appeared at 1720 cm -1 confirming the incorporation of both functionalities in the novolac back bone. Chemical shift at 34.7, 116.0, 137.5, 138.9, 150, 165.2, 168.9, 161.5 and 131.0 ppm in the 13 C MR (Figure 5.5) also confirms the incorporation of maleimide group in allyl novolac backbone. g g h h f e d H a b c i University of Mysore 103
c g e h d f i b a Fig. 5.5: 13 C MR spectrum of single component BMI resin system From the elemental analysis, the percentage of nitrogen and thereby the maleimide content in each resin formulation was determined. Table 5.1 gives the compositions of the BMI oligomer systems synthesized. Table 5.1: Details of the BMI oligomer systems SI.o Ref % maleimide in the chain 1 BMI-1 50 2 BMI-2 60 3 BMI-3 70 University of Mysore 104
5.3.3.1 Cure Studies Curing of the single component system was performed by FT-IR, Rheometry, and DSC. From the FT-IR spectra of the cured and uncured BMI resin system, (Figure 5.6) it can be seen that the absorption band at 1143 cm -1, attributable to the C--C stretching vibration of the maleimides group disappeared during curing, indicating the conversion of maleimides functionality. Another evidence of the consumption of the BMI unsaturation is confirmed by the absence of absorbance band at 3104 cm -1 which is originally due to the =C-H stretching. Also, peaks due to allyl groups (vinyl =CH 2 wagging) at 916 cm -1 diminished due to the crosslinking. CURED 1508 1605 1376 1105 1262 1193 %T UCURED 1716 3477 3104 2924 1782 1512 1605 1718 1377 1266 1195 951 916 852 828 1017 762 1142 699 1071 1175 4000.0 3000 2000 1500 1000 500 400.0 cm-1 Fig. 5.6: Typical FTIR spectrum of the cured and uncured BMI resin system University of Mysore 105
The well defined two stage cure pattern of the single component BMI resin system is clear from the DSC thermogram as seen in Figure 5.7. In 80-100 o C range the monomers react via Ene reaction pathway to form Ene adduct followed by Diels Alder reactions at 180-250 o C yielding the cross linked polymer. Figure 5.8 represents the cure reaction sequence of the BMI resin systems. Endo Exo Fig. 5.7: DSC thermograms of three resin formulations having varying maleimides content at 10 C/ min University of Mysore 106
H H H 80-160 C Ene mono-adduct (Diels -Alder) 225-275 C H Heat Further crosslinking Tri -adduct Fig. 5.8: Crosslinking of BMI resin systems by alder ene reaction 5.3.3.2 Kinetics of cure reaction ormally, Cure Kinetics of resin system is studied by using non isothermal and isothermal technique. In non isothermal method it includes single and multi-heating rate and in isothermal method it involves studying the reaction mechanism at a specific temperature as a function of time. The non-isothermal single heating rate method measures the curing process at a constant heating rate, while the multi-heating rate method is an iso-conversional method and is suitable for systems with multiple reactions. For understanding and predicting the curing behavior of resin systems the two kinetic analysis models are extensively used. [175-176]. The basic assumption for the application of DSC technique to the cure of the thermosetting resins is that the rate of kinetic process (dα/dt) is proportional to the measured heat flow ( ) as given in Equation (5.1). University of Mysore 107
d dt (5.1) H Δ being the enthalpy of the reaction [177]. The kinetic rate process can be described using Equation (5.2) d cf ( ) (5.2) dt Where α is the chemical conversion or extent of reaction and f ( ) is assumed to be independent of temperature [177]. The rate constant k (T) is dependent on temperature and is assumed to follow Arrhenius equation Ea k( T) Aexp (5.3) RT Where, A is the pre-exponential factor or frequency factor and Ea is the activation energy. R is the gas constant (8.314J/K/mol) and T is the absolute temperature. The kinetic parameters of the curing reaction, with special reference to E a, can be calculated using different computational methods [178-181]. Ea is determined by the isoconversional methods using the logarithmic form of the kinetic equation (5.2). d ln dt Ea RT Af ln (5.4) The slope of the plot of ln [dα/dt] versus 1/T for the same values of α gives the value of activation energy. This assumption is related to the zawa and Kissinger methods which relate activation energy Ea and heating rate and the peak exotherm temperature Tp with the assumption that, the extent of the reaction at the exotherm peak is constant and independent of the heating rate [182]. University of Mysore 108
5.3.3.3 Calculation of Ea by Kissinger and zawa methods In the study, the activation energy ( Ea ) values are computed using the equations by zawa and Kissinger. The activation energy equation according to zawa model is R ln Ea 1.052 1 Tp (5.5) Where is the heating rate, Ea is the activation energy. Activation energy was calculated from the slope of the plot of 1 against ( ) Tp The activation energy, Ea according to Kissinger Model is based on peak temperature (T p ) corresponding to the exothermal peak position of the dynamic scan. ln Tp 2 1 Tp Ea R (5.6) Where is the heating rate, Ea is the activation energy. The activation energy can be obtained from the slope of the plot of Vs. A was found in both cases by the relation E / RTm Ee RTm 2 A (5.7) The peak maxima (Tm) obtained from DSC for the two step reactions are compiled in Table 5.2. The initial cure temperature and the peak temperature T P increased with increasing heating rate. University of Mysore 109
Table 5.2: Peak temperatures at different heating rates in DSC experiment for the different BMI oligomers Heating Rate ( C/ min) 54% maleimide 65 % maleimide 74 % maleimide Tm 1 Tm 2 Tm 1 Tm 2 Tm 1 Tm2 5 148 239 111 229 101 230 7 149 246 116 238 113 241 10 154 254 122 245 124 248 The peak temperatures obtained by nonisothermal measurements for the BMI oligomers at different heating rates were transformed according to the equations of Kissinger and zawa. In Figure 5.9, the reciprocal temperature 1/Tp is plotted against the logarithm of the heating rate, ln, based on the zawa method and, respectively, against following the Kissinger equation. From the fitted slope one obtains the activation energy E a (tabulated in Table 5.3). Both values are very similar, although the E a by the zawa method is slightly higher than the one calculated by the equation of Kissinger. Such difference was also noticed by other researchers [183]. University of Mysore 110
a b Fig.5.9: Kinetic plots by (a)sawa and (b)kissinger methods The computed activation parameters for the BMI resin systems are given in Table 5.3 below. The rate constant of the reaction was calculated at the respective peak maximum temperatures for each reaction step using equation 5.3. Table 5.4 gives the rate constants at each peak maximum temperatures.. University of Mysore 111
Table 5.3: Kinetic parameters of curing for the different compositions zawa Kissinger Reference Maleimide content 1 st Step 2 nd Step 1 st Step 2 nd Step E a (kj) A(s -1 ) E a (kj) A(s -1 ) E a (kj) A(s -1 ) E a (kj) A(s -1 ) BMI-1 50 130 1.2x 10 15 95 2.9x10 7 105 2 x 10 10 84 3.3x 10 6 BMI-2 60 63 1.2x 10 7 83 3.5x10 6 69 7.4x 10 6 81 2x 10 6 BMI-3 70 35 164 71 2.1x10 5 33 43 69 9.6x10 4 University of Mysore 112
Table 5.4: Rate constant of the two step reaction at different temperatures zawa Kissinger Composition 1 st Step 2 nd Step 1 st Step 2 nd Step k at 425K k at 520K k at 425K k at 520K 50 0.0126 0.0052 0.0088 0.0048 k at 385 K k at 510 K k at 385 K k at 510 K 60 0.0052 0.0047 0.0051 0.0045 k at 385 K k at 510 K k at 385 K k at 510 K 70 0.0033 0.0038 0.0029 0.0036 From the DSC thermograms of resin systems, two exotherms were identified corresponding to the Ene and Diels alder reactions and thus, the network formation (in two steps) during curing (Figure5.7). From the evaluation of the kinetic parameters, it was inferred that, the activation energy, rate constant and frequency factor are higher for the nearly equivalent allyl to maleimide formulation. As the maleimide content increases, especially, the first reaction (ene), takes place at a lower rate. The apparently lower E a is compensated by the low frequency factors as well as with a resultant lowering in the rate of the reaction in maleimide dominated polymers. Enhanced activation energy was observed for all formulations, for the second step (Diels Alder reaction) which is attributed to the prevalence of diffusion control in the kinetics of the cross-linking. University of Mysore 113
Fig. 5.10: Conversion profile of the BMI resin systems The zawa and Kissinger equation were used for the determination of the activation energy at different percentages of conversion (from 20 % to 90%), if the Tp (Peak temperature exotherm) can be considered as the temperature at which the system has reached the respective conversion percentages [184]. Hence, the multi-heating rate technique adopted in this part of the study is an iso-conversional method that assumes the activation energy (E a ) changes with conversion. Secondly, from literature and DSC plots, it is evidenced that BMI homo-polymerization, ene reaction, and Diels Alder reaction occur at significant rates between 180 o C to 300 o C [185-186]. Based on this assumption, the change in the activation energy for the system (65 % maleimide content) was analysed with increasing conversion (Figure 5.10) for the principal curing reaction and is illustrated in Figure 5.11. University of Mysore 114
Ea(k J/ mol) Toughening Procedure, Processing and Performance of Bismaleimide Resin Composites 100 90 80 70 60 50 40 30 20 10 0 Kissinger sawa 0 10 20 30 40 50 60 70 80 90 100 Conversion (%) Fig. 5.11: Plot of Ea versus percentage conversion The plot of activation energy as a function of conversion does not reveal the reaction mechanism but will provide significant information about kinetically controlled and diffusion controlled mechanisms. It can be clearly inferred from Figure 5.11 that, as the percentage of conversion increases, the activation energy also increases. The kinetics of each step is decided by its respective activation parameters, reactant concentration and viscosity of the medium. At lower cure temperature, the crosslinking results in the ene adduct, or the Wagner-Jauregg adduct. At higher temperature, all the reactions are possible irrespective of the stoichiometry. The conversion increases rapidly with the curing time but reaches a maximum asymptotically as seen from the conversion profile (Figure 5.10) of the system. This profile reflects the change of the reaction kinetics from a chemically controlled process to a diffusion-controlled regime. nce the maximum conversion is reached, vitrification occurs, practically stopping the reaction completely. It is obvious from Figure 5.11 that, the onset of diffusion controlled reaction is at around 50% conversion as indicated by the sharp change in the slope of Ea Vs conversion plot. Vitrification also University of Mysore 115
slows down the rate of the reaction in the diffusion controlled regime; making the attainment of 100% conversion practically difficult. 5.3.3.4 Mechanical and Dynamic mechanical properties of the composites Carbon fiber reinforced composites were fabricated using the different BMI resin oligomers and characterized for mechanical and thermomechanical properties. The resin content in the composites was maintained at 35-40 % (by weight). The mechanical properties obtained for the composites are tabulated in Table 5.5.The trend in ILSS values indicates that the resin/reinforcement interaction at the interface is greatly affected by the composition of the matrix. The composites with higher maleimide content resulted in poor ILSS values (around 49% reduction with increase of maleimides composition from 0.5 to 2 mol ratio). However, the retention of ILSS at high temperature (150 C) was superior for the composites with higher maleimide groups in the matrix. The flexural strength of the composites showed a decreasing trend with maleimide rich systems implying higher brittleness for these systems. This trend is also reflected in the impact analysis of the composites, which decreased with maleimide rich BMI oligomer. It can be inferred from the mechanical performance of the composites that, allyl groups induces flexibility to the network structure. Table 5.5: Mechanical properties of the carbon fiber reinforced composites Property Resin composition BMI 1 BMI 2 BMI 3 Flexural Strength (MPa) 480 465 430 ILSS @ RT(MPa ) 95 80 55 ILSS @150 C (MPa) 25 28 39 Impact Strength (k J/ m 2 ) 117 103 79 University of Mysore 116
5.3.3.5 Dynamic Mechanical analysis of the composites Dependence of storage modulus and glass transition temperature on composition of cured BMI- Carbon fabric composites is shown in Figure 5.12. All the composites showed a drop in modulus around the glass transition temperature, Tg, of the polymer as expected. It can be observed that in the high maleimide-content systems, the drop in storage modulus with increasing temperature is significantly lower when compared to the low maleimide content systems. For the composition BMI- 3, the temperature at which modulus shift occurred is as high as 280 C indicating the increased crosslinking density of the system with high maleimide content. Fig.5.12: DMA for the different compositions of BMI oligomer / Carbon composites University of Mysore 117
5.3.3.6 Glass transition behaviour of the composites During DMA analysis, Tan δ peak of the composites shifted to higher temperature regime with increase in maleimide content in the blends (Figure 5.13). The shift in Tg (from 207 C to 280 C) with increase in maleimide content confirms the formation of good networks with higher crosslink density. The change in Tg and ILSS at 150 C with increasing maleimide content is represented in Figure 5.14. A larger area under the tan δ curve for the matrix systems with higher allyl content implies better damping properties and more energy absorbed and dissipated. It is also noted in Figure 5.13 that the area under the tan δ curve decreased when the maleimide content increased. This corroborates the assumption that, mobility of molecules in the resins got declined and material exhibited embritttlement at higher degrees of maleimide content in the network. Fig. 5.13: Tan δ peaks from DMA for BMI oligomers /Carbon fabric composites University of Mysore 118
Fig. 5.14: Dependency of Tg and ILSS on maleimide content at elevated 5.3.3.7 Thermal characterization by TGA temperature (150 C) Figure 5.15 depicts the thermo gravimetric analysis of the cured one component BMI resin systems. The thermal characterization of the three-resin system reveals that as the percentage of maleimide content increases the initial degradation temperatures as well as the char yield increase systematically as reported previously by other researchers [186-187]. From the thermogram, almost similar decomposition pattern was observed for all the three compositions. The dependence of BMI content in initial decomposition temperature (Ti) and char residue at 800 C is represented in Figure 5.16. University of Mysore 119
Fig. 5.15: TGA thermogram of the cured resin systems University of Mysore 120
Fig. 5.16: The dependence of BMI content in initial decomposition temperature (Ti) and char residue at 800 C 5.3.4 Morphology of the composites The SEM images of the typical fractured surfaces of the composites are shown in Figure 5.17. It can be seen that as the maleimide content increased, the smooth matrix surface in BMI 1 (Fig.5.17 a) gets transformed to a brittle pattern in the case of BMI -3 (Fig.5.17 c). The fiber wetting is better in the case of less maleimide dominated composites where fiber debonding is not detected. This implies a stronger interface bonding between matrix and fiber which is substantiated by higher ILSS value of BMI-1 and BMI-2(Fig.5.17b) composites.(refer Mechanical characterization part) Matrix appears to be of brittle nature University of Mysore 121
in the case of maleimide dominated composites, (Fig.5.17c). The impact properties are also commensurate with the fracture morphology. As a result of weaker fiber matrix interaction, a small amount of fibre pullout is also observed for the maleimide- dominated composites (Fig.5.17 c). a b c Fig. 5.17: SEM images of fracture surface of composites with different maleimide content University of Mysore 122
5.4 Conclusion BMI resin oligomers were synthesized by co-reacting allylated novolac and maleimido benzoyl chloride. The oligomers manifested a two stage curing in DSC. Broad cure exotherms starting at around 60 C and 180 C are assigned to ene and Diels-Alder reaction respectively. The cure kinetics of the system was determined by zawa and Kissinger models and Ea was found to be increasing linearly with conversion. From DMA, the highest Tg (280 C) was observed with highest maleimide content system. Mechanical properties and impact strength of the composites at ambient temperature were found to be high for the less dominated blends, whereas the retention of ILSS at elevated temperature was found to be better for the high maleimide dominated system. As the maleimide content increased the thermal stability and char yield of the blends increased marginally. Less maleimide content resin composites showed better matrix-reinforcement interface whereas blends with higher maleimide content resulted in brittle matrix and poor interface. University of Mysore 123