R Yadav, D Srivastava - Studies on cardanol-based epoxidized novolac resin and its blends - страница 1
CHEMISTRY & CHEMICAL TECHNOLOGY
Vol. 2, No. 3, 2008 Chemistry
Ranjana Yadav and Deepak Srivastava
STUDIES ON CARDANOL-BASED EPOXIDIZED NOVOLAC RESIN
AND ITS BLENDS
Department of Plastic Technology, H. B. Technological Institute Kanpur - 208 002 (U.P.), India, firstname.lastname@example.org
Received: May 05, 2008
Abstract. Cardanol-based novolac-type phenolic resin was synthesized with a mole ratio 1.0:0.5 of cardanol-to-formaldehyde using a dicarboxylic acid catalyst such as succinic acid. The cardanol-based novolac-type phenolic resin may further be modified by epoxidation with epichlorohydrin excess at 393 K in a basic medium to duplicate the performance of such phenolic-type novolacs. Carboxyl-terminated butadiene acrylonitrile copolymer (CTBN) has been studied by various researches with diglycidyl ether of bisphenol-A (DEGBA) epoxy resin and epoxidized phenolic novolac resins. The epoxidized novolac resin was blended with different weight ratios of carboxyl-terminated butadiene acrylonitrile copolymer (CTBN) and cured with a stoichiometric amount of polyamine curing agent. The formation of various products during the synthesis of cardanol-based novolac resin, epoxodized novolac resin and blending of epoxidized novolac resin with CTBN has been studied by Fourier-transform infrared (FTIR) spectroscopic analysis. Further, the products were confirmed by a proton nuclear magnetic resonance ('H-NMR) spectroscopic analysis. The number average molecular weight was determined by a gel permeation chromatography (GPC) analysis. The blend sample, having 15 wt % CTBN concentration showed minimum cure time and most thermally stable systems.
Key words: cardanol, formaldehyde, novolac, epichlorohydrin, epoxy resin, butadiene acrylonitrile copolymer , spectral methods.
Cashew Nut Shell Liquid (CNSL), an agricultural renewable resource and the byproduct of the cashew industry, holds a considerable promise in the different directions because it is the source of unsaturated hydrocarbon phenol, an excellent monomer for thermosetting polymers production. Vacuum distillation of CNSL yields a pure cardanol. Cardanol is n-pentadecadienyl phenol, the aliphatic side chain usually consists of mixture of one, two and three double bonds in a linear chain with a saturated side chain. CNSL and cardanol are used in the manufacture of special phenol resins for coatings, lamination and as friction materials. These polymers are synthesized from CNSL or cardanol either by polycondensation with electrophylic compounds such as formaldehyde, furfuraldehyde or by chain polymerization through unsaturation presented in the side chain using an acid and base catalyst. The cardanol based novolac type phenolic resins may further be modified by epoxidation with epichlorohydrin to duplicate the performance of such phenolic type novolacs [1-4]. Having several outstanding characteristics, epoxy resins show low impact resistance in their cured state which limits the applications of epoxy resins. To alleviate this deficiency, epoxy resins are modified by the incorporation of a reactive liquid rubber without a significant loss in other properties [5-9]. In this way, a carboxyl-terminated copolymer of butadiene and acrylonitrile (CTBN) has been used by various researches [10-13] with diglycidyl ether of bisphenol-A (DGEBA) epoxy resin and epoxidized phenolic novolac resins. But, carboxyl-terminated polybutadiene (CTBN) is nowhere used with cardanol based epoxy resins. Therefore, we have tried to produce the modified epoxy matrices, based on cardanol, by physical blending with CTBN and we have studied the effect of CTBN addition on thermal and morphological changes in the blends.
Cardanol (M/s Satya Cashsew Pvt. Ltd., Chennai), formaldehyde (40 % solution), succinic acid, sodium hydroxide, epichlorohydrin (All from M/s Thomas Baker Chemicals Ltd., Mumbai), polyamine (M/s Ciba Specialty Chemicals Ltd., Mumbai) with amine value 12401400 mgKOH/g and carboxyl-terminated butadiene -acrylonitrile copolymer (CTBN) (Hycar 1300x13) were the initial reagents. CTBN was kindly supplied by M/s Emerald Performance Materials, LLC, Hong Kong having molecular weight Mn of 3500 and acrylonitrile and carboxyl contents 27 and 32 percent, respectively.
2.2. Synthesis of cardanol- based novolac-type phenolic resin
Novolac resin with mole ratio 1:0.5 of cardanol (C) to formaldehyde (F) was prepared using bicarboxylic acid, viz., succinic acid, as a catalyst by a method similar to that adopted by Knop and Schieb  for a phenol based novolac. The catalyst (1 % based on cardanol) was, first, dissolved in methanol at 333 K. Half of a catalyst solution was added to cardanol (about 30 g), charged in a three-necked round bottom flask fitted with a Leibig's condenser and mechanical stirrer. The remaining half of the methanolic solution of the catalyst was added to the formaldehyde (40 %) and this was added to the cardanol dropwise within one hour, once the temperature of the reaction kettle was maintained at 393 K. The initial pH of the reaction mixture was 6.0 reduced to a value of 4.8 after 5 h of reaction at 393 K. Free-formaldehyde and free-phenol contents were checked after every 45 min to see the completion of the methylolation reaction . The reaction product was cooled and dried under vacuum at 333 K overnight before purification by column chromatography. A resin solution prepared with и-hexane, charged to the silica gel column chromatographic purification, was adopted mainly to remove unreacted components, impurities etc., from the methylolated cardanol. Purification was effected using the elutent mixture of ethyl acetate - benzene (60:40). The purified resin was analyzed by infra-red (IR) spectroscopic, nuclear magnetic resonance (1H-NMR) spectroscopic, mass spectroscopic and gel permeation chromatographic (GPC) analysis.
2.3. Epoxidation of cardanol-based novolac-type phenolic resin
The cardanol-based novolac-type phenolic resin was epoxidized by a method similar to that given in . Approximately 1.0 mole of a novolac resin was taken in a 500 ml three-necked round-bottomed flask and 10 moles of epichlorohydrin were added to it while stirring. Then 40 % sodium hydroxide solution was added dropwise to the above mixture for a period of 5 h at 393 K. The reaction mixture was then subjected to distillation under vacuum for removal of unreacted epichlorohydrin. The resulting viscous product was stored for a further analysis.
2.4. Preparation of blends of epoxidized novolac resin and carboxyl-terminated butadiene acrylonitrile copolymer (CTBN)
The epoxy resin was mixed physically with varying concentration of CTBN ranging between 0-20 wt % with an interval of 5 wt % (Table 1).
Epoxy, wt %
CTBN, wt % Sample code
Cardanol-based epoxidized novolac resin (prepared from novolac resin CF52)
2.5. Fourier-transform infra-red (FTIR) spectroscopic analysis and Nuclear magnetic resonance (1H-NMR) spectroscopic analysis
Fourier-transform infra-red (FTIR) spectra of the prepared samples were recorded on a Perkin-Elmer (Model 843) infra-red spectrophotometer, using KBr pellet, in the wave length range of 500-4000 cm-1. 1H-NMR spectra of cardanol-based novolac and epoxidized novolac resins were recorded on Bruker DRX - 300 NMR spectrophotometer in the temperature range of 183 to 353 K.
2.6. Gel permeation chromatographic (GPC) analysis and mass spectroscopic analysis
Gel permeation chromatograph (GPC) (Instrument procured from E Merck consisting of a pump, L-7350 column oven and L-7490 R.I. Detector) was used for determination of the average number molecular weight of the synthesized cardanol-based novolac resin. A small quantity of the resin was dissolved in THF; which acted both as a mobile and stationary phase and was injected into the instrument. Mass spectra of the prepared samples were recorded on Micromass TofSpec2e MALDI TOF mass spectrophotometer.
2.7. Differential scanning calorimetric (DSC) analysis
Cure temperature of the prepared samples were measured by taking very little quantity of blend samples into a shallow aluminum pan sealed by an aluminum cover of differential scanning calorimetry (DSC) (TA, Instrument, USA; Model DSC Q20). It was placed in a sample cell of the instrument. The starting temperature,
programmed rate and final temperature were taken at a heating rate of 283 K/min.
2.8. Thermogravimetric analysis
The thermal stability of the blend samples was determined by a comparison of the onset degradation temperature (up to 5 % weight loss) of the cured samples with that of thermogravimetric analyzer (TGA) of TA Instruments (Model Q50 TGA) at a heating rate of 283 K/min in the nitrogen atmosphere from ambient to 923 K.
The morphological changes due to addition of CTBN into the epoxy matrix were studied by JOEL scanning electron microscope (SEM) (Model JSM 5800). The rubber domains distributed in the matrix and interaction of these domains with the epoxy matrix specimen surface could be observed by SEM. For this, the fractured samples were coated with a thin layer of gold-palladium alloy by sputtering to provide conductive surfaces
3. Results and Discussion
3.1. Synthesis of cardanol-formaldehyde novolac type phenolic resin
The methylolation of cardanol was carried out with formaldehyde in the presence of dicarboxylic acid, viz., succinic acid. The initial pH of the reaction mixture found to be 6.0. The formylation reaction was carried out with 0.5 mole ratio of cardanol-to-formaldehyde. Therefore, under these experimental conditions the complete formylation might be expected to yield resin with a high ortho-ortho linkages for the phenolic novolac resin. The completion of the methylolation reaction was checked by
1,000 3500 Э00О 2500 2 000 1500 '000 500 cm
Fig. 1. FTIR spectrum of novolac resin CF'
where, a = 1; b = 5
periodic withdrawal of reaction mixture to analyze formaldehyde using hydroxylamine hydrochloride . The final pH of the reaction mixture was found to be 4.8. The decrease of pH in the methylolated cardanol might be ascribed to the formation of monohydroxyl substituted cardanol .
3.2. FTIR and 1H-NMR spectroscopic analysis of cardanol-based novolac-type phenolic resin
Cardanol is a monoene meta-substituted phenol. On the basis of FTIR, NMR and mass spectroscopic analysis and gel permeation chromatography (GPC) analysis elucidated the structure of cardanol. Therefore, the empirical formula can be taken as C21H34O [19, 20] and the structure of cardanol may be proposed as shown in Scheme 1.
A shift of the peak from 1075 to 1102 and appearance of the peak near 1708 cm-1 (Fig. 1) was observed in a methylolated cardanol due to the C=O stretching from CH2OH. It has also been found that the intensity of peaks at 1594 cm-1 (C=C, str), 3009 cm-1 (C-H str of alkene) and 779 cm-1 (C-H out-of-plane deformation) remained almost unaffected which indicated that the polymerization has taken place through substitution of CH2OH and not through the double bonds in the side
*ti ................;■..................і...................і1 ■-.............і...............'T
ppm 8 6 4 2 0
Fig. 2. 1H-NMR spectrum of novolac resin CF52
This ion acted as a hydroxy alkylating agent by reacting with a cardanol via electrophylic aromatic substitution. A pair of electrons from the benzene ring attacked electrophylic element forming an intermediate followed by deprotonation and regained the aromaticity of the ring. The reaction is shown in Scheme 3.
The methylol group of the hydroxy methylolated cardanol was unstable under acidic condition and might loose water readily to form a benzylic carbonium ion as shown in Scheme 4.
The products formed in Scheme 4 reacted with another cardanol molecule to form a methylene bridge in another electrophylic aromatic substitution. This process repeated until all formaldehyde was exhausted. The related reaction may be shown in Scheme 5.
CH2 = O + H2O , ' HO CH2 OH
CH2 - OH + H2O
3.3. Molecular weight of cardanol-based novolac-type phenolic resin by gel permeation chromatographic (GPC) technique and mass spectroscopic analysis
The average number molecular weight (M) of CF52 calculated from the number of phenolic units (p) in the novolac resins (proposed structure in Scheme 6), which was determined from the ratio of aromatic-methylene
chain. The band at 3395 for the sample CF52, may appear due to the presence of hydroxyl groups in the methylolated cardanol. Small peaks near 946 and 973 cm-1 indicated the substitution in benzene nuclei. The small peaks near 913 and 697 cm-1 might appear due to three adjacent hydrogen atoms in the benzene nuclei. The peak appearance near 722 cm-1 (Fig. 1) indicated the ortho-and para-substitution in the benzene nuclei. The preceding spectral data were found to be identical with those given in literature [21, 22].
In the 1H-NMR spectra CF52 (Fig. 2) novolac resin, the peak near 6.6-6.8 8 might appear due to the aryl protons of the benzene nuclei. The peaks around the region of 7.17.3 8 appeared due to the phenolic hydroxyl in the novolac resin. The peak at 5.3 8 might be attributed to the methylene protons whereas the peaks between 0.8-2.8 8 appeared due to the presence of long alkyl aliphatic side-chain, originally observed in cardanol. The terminal methyl group of the alkyl side chain may also be seen as a small peak at 0.8 8. The peak at 3.7 8 indicated the presence of methylene protons of C6H5-CH2-C6H5 for the bridge between the phenyl rings . All these spectral data indicated that the condensation of methyolated cardanol has been completed under experimental conditions and was fully consisted of the proposed structure (Scheme 6) resulting due to reaction mechanism shown in Schemes 2-5.
On the basis of the preceding discussions, the following reaction mechanism, similar to that proposed by Kuriakose and Manjooran , may be proposed. The mechanism of formation of novolac oligomers, in acidic media, using excess of cardanol over formaldehyde might proceed in four steps. First, a methylene glycol was protonated by an acid from the reaction medium, which then released water to form a hydroxyl methylene carbonium ion as shown in Scheme 2.
OH oh oh
c15h29 C15H29 c15h29 +
H „ -
[CH2] to aromatic protons,[AR], protons of the 1H-NMR spectra as the following equation [25 ]:
[CH2MAR] = (2p - 2)/(3p + 2) (1)
The calculated result was Mn = 670 g/mol and p = 2.17 for phenolic groups per molecule for a novolac resin CF 2. The value of molecular weight was further confirmed by GPC trace and mass spectroscopic analysis (Figs.3 and 4).
Finally, the structure of the cardanol-based novolac resins may be proposed as following:
3.4. Epoxidation of novolac prepolymer
The novolac based epoxy resin was synthesized by the reaction with epichlorohydrin (ECH). The number
of glycidyl groups per molecule in the resin depend upon the number of phenolic hydroxyls in the starting novolac, the extent to which they were reacted and the extent to which the lowest molecular species were polymerized during synthesis. Theoretically, all the phenolic hydroxyls might be reacted, but in practice all of them did not react because of steric hindrance . The reaction between ECH and novolac resin might be thought to proceed in a similar fashion as in the work given by Lee and Neville . The epoxide group of ECH reacted with phenolic hydroxyls under the alkaline medium and formed the chlorohydrin ether which underwent dehydrochlorination reaction and resulted in to glycidyl ether. The structure of the epoxy resin may be proposed as in Scheme 7.
O-CHj-ch — ch2
O-CHj-ch — ch2
O-CHj-ch — сн2
O—СН^сн — сн2
The FTIR spectra of uncured cardanol-based epoxidized novolac resins, i.e. sample ECF52, the characteristic band of the oxirane ring was observed near 911 and 856 cm-1, as shown in Fig. 5. The 1H-NMR spectrum (Fig. 6) further confirmed the formation of cardanol-based epoxidized novolac resins. The epoxide equivalent weight of the prepared cardanol-based epoxidized novolac resins, i.e. sample ECF52, was found to be 305 eq/g, as determined by a titration method, whereas the molecular mass of the sample was 791 g/mol.
3.5 FTIR analysis of uncured and cured blend samples
FTIR spectral analysis of uncured blend sample containing 15 wt % CTBN in epoxy resin (ECF52) is shown in Fig. 7. It is clear from the figure that the addition of CTBN into the neat epoxy resin shifted the peaks related to the oxirane functionality from 911 and 856 cm-1 (refer to Fig. 5) to 913 and 851 cm-1, respectively (Fig. 7). A minor change in the intensity of the peaks was also seen in the spectrum. These observations clearly indicated that there occurred a chemical reaction between a carboxyl group of CTBN and an epoxide group of the epoxy resin resulting in epoxy end-capped which would be capable of reacting with the hardener (refer to Fig. 8 for the FTIR spectrum of cured product) in the same way as diepoxy [27-28]. This was further confirmed by the appearance of a new stretching peak at 1492 cm-1 due to the formation of carboxylate anion and near 1737 cm-1 due to the formation of functional ester group and С-С multiple bond stretching (Fig. 7).
3.6. DSC analysis of blend samples
Figs. 9 and 10 show the dynamic DSC scans of cardanol-based epoxidized novolac resins without and with CTBN (i.e. samples ECF52 and ECF523) at a heating rate of 283 K min-1. The effect of CTBN concentration on cure parameters of different epoxy matrices has also been compared to that of non-modified epoxy matrices in Table 2. It is evidenced from the results that the exotherm peak was shifted to lower temperatures due to enhanced reaction rate which, finally reduced the cure time of the CTBN-modified blend system (refer to Table 2). The initial addition of CTBN in the epoxy resin decreased the cure time sharply and this trend remained up to 15 wt % CTBN addition and increased thereafter. The enhanced rate behavior could be interpreted in terms of intermolecular transition state for the epoxy-amine reaction. According to this mechanism [29, 30] strong hydrogen bonding species, such as acids and alcohols, stabilized the transition state and strongly accelerated the curing reaction. The decrease of cure time could also be explained by the fact that during the reaction of CTBN with epoxy resin, some of the exothermic energy released during epoxy crosslinking it might have been consumed by CTBN resulting in decrease of cure time . The AH values (Table 2) related to the cure process were determined from the area of the exotherm peak obtained from DSC analysis (Figs 9. and 10) taken in a dynamic mode. In contrast, the presence of CTBN did not affect significantly the heat of polymerization (AH), indicating insignificant influence on the crosslinking density of the epoxy matrix [32, 30].
Fig. 5. FT-IR spectrum of epoxidized novolac resin ECF 2
ppm ~S 6 T 2 □
Fig. 6. 1H-NMR spectrum of epoxidized novolac resin ECF52
DSC results of unmodified and CTBN-modified cardanol-based epoxidized novolac resin cured with polyamine
AH, J mol-1
a Cure time obtained by curing the sample in air oven at 393 K.
T - kick-off temperature, where the curing starts; Tonset - temperature where the first detectable heat is released; Tp - temperature of peak position of exotherm; T' - temperature of end of curing exotherm; AH - heat of curing; tmjn - cure time in minutes.
4000 3500 3000 2500 2000 1500 1000 cm 500
Fig. 7. FTIR spectrum of uncured blend sample BECF523
Fig. 8. FTIR of cured blend sample BECF523