CATALYTIC HYDROGENATION OF CARDANOL DERIVED BIFUNCTIONAL MONOMERS FOR POLYMERIZATION WITH DIESTERS

Cardanol extracted from cashew nut shell liquid (agro-waste) was used to synthesize a bifunctional monomer; methyl 16-(3-hydroxyphenyl)hexadecanoate using DTBPMB modified palladium catalyst. Hydrogenation of this monomer using a Ru(acac)3/triphos/MSA catalytic system in aqueous dioxane and ammonia solution gave 3-(16-hydroxyhexadecyl)phenol (75% yield) and 3-(16-aminohexadecyl)phenol (37% yield) respectively. Polymerization of 3-(16-aminohexadecyl)phenol and 3-(16-hydroxyhexadecyl)phenol with dimethyl nonadecanedioate (a di-ester obtained from methoxycarbonylation of methyl oleate) resulted into oligomers consisting of up to nine condensed monomer units (in the volatile fraction) as was revealed by MALDI-TOF MS.


INTRODUCTION
Cashew nut shell liquid; a by-product of cashew nut processing factories is a promising source of environmentally friendly starting materials for production of various valuable materials. The liquid is rich in phenolic compounds having functional groups such as double bonds, carboxylic group and alcohol group which make them suitable starting materials for various reactions [1][2][3][4][5]. The phenolic components of cashew nut shell liquid are also potential compounds for making fire retardant polymers due to the presence of aromatic ring in their structures [6][7][8]. Natural cashew nut shell liquid (which is obtained by cold extraction methods) contains anacardic acid as a major component (60-65%). On the other hand, technical cashew nut shell liquid (extracted from cashew nut shells at higher temperatures) contains cardanol up to 85%. Other components of cashew nut shell liquid are cardol (15-20%) and traces of 2-methylcardol ( Figure 1) [9]. In recent years, various research groups have shown great interest in investigating the possibility of replacing non-renewable by renewable materials in synthesizing materials of interest [10][11][12][13][14][15][16]. The major reason is that non-renewable materials are subject to depletion, which means that overdependence on such materials is a risk. Cashew nut shell liquid is one of the bio-based resources that have been extensively investigated for its use in preparation of useful fine chemicals and materials [17][18][19][20][21][22]. One of the areas that cashew nut shell liquid has been investigated is in the synthesis of bifunctional monomers for making fire resistant polymers. Successful synthesis of bifunctional monomers by the palladium catalyzed methoxycarbonylation of cardanol component of cashew nut shell liquid with good yield have been reported [23]. Combination of metathesis and carbonylation techniques were used to make bifunctional monomers of various chain lengths from cardanol. The fact that these monomers derived from cardanol have both alcohol and ester groups ( Figure 2) means that they could be polymerized by simply heating in the presence of appropriate catalyst without the need of a second monomer. However, the polymerization reaction of these monomers using a standard Ti(OBu)4 condensation catalyst produced either brittle polymers, or lactones contaminated with bifunctional monomers [23]. Brittle polymers are difficulty to process. Purification of lactones (formed by cyclization reaction of bifunctional monomers) was unsuccessful. This paper therefore focusses on improving the method of polymerization of cardanol derived bifunctional monomers. A discussion on the catalytic hydrogenation of methyl 16-(3-hydroxyphenyl)hexadecanoate to produce 3-(16-aminohexadecyl)phenol and 3-(16-hydroxyhexadecyl)phenol is presented. The results obtained from polymerization of 3-(16-aminohexadecyl)phenol and 3-(16-hydroxyhexadecyl)phenol with dimethyl 1,19-nonadecanedioate (a di-ester obtained from a methoxycarbonylation reaction of methyl oleate) are also presented.
Both 1 H NMR and 13 C NMR data (shown in the experimental section) agree well with the structure of compound 6. In the 1 H NMR spectral data, the disappearance of peaks at δ 3.7 (-OCH3) and δ 2.3 (-CH2-COOMe) as well as the appearance of a peak at δ 2.7 (-CH2-NH2) confirm that compound 5 has been hydrogenated to form compound 6. In 13 C NMR spectral data, the disappearance of the methoxy (OCH3) peak at δ 60 and carbonyl (C=O) peak at δ176 with the appearance of a peak at δ 47 (-CH2-NH2) further confirm the formation of compound 6.
The fact that the desired compound 7 was the major product of the reaction explains the big difference in the yield of 3-(16-hydroxyhexadecyl)phenol (75%) when compared to the yield of 3-(16-aminohexadecyl)phenol (37%).
Both 1 H NMR and 13 C NMR spectral data agree well with the structure of compound 7. In the 1 H NMR spectral data, the disappearance of peaks at δ 3.7 (-OCH3) and δ 2.3 (-CH2-COOMe) as well as the appearance of a peak at δ 3.6 (-CH2-OH) confirm that compound 5 has been hydrogenated to form compound 7. In the 13 C NMR spectral data, the disappearance of the methoxy (OCH3) peak at δ 60 and carbonyl (C=O) peak at δ 176 together with the appearance of a peak at δ 65 (-CH2-OH) further confirm the formation of compound 7.

Methoxycarbonylation of olive oil
Natural oils differ in the amounts of methyl oleate they contain. For instance, olive, rapeseed, and sunflower oils from supermarket contain 73%, 64%, and 38% of methyl oleate, respectively.
Furst et al. [11] investigated the methoxycabonylation reaction of these oils to synthesize dimethyl 1,19-nonadecanedioate (8). The authors found that the amount of dimethyl 1,19nonadecanedioate (8) produced from methoxycarbonylation reaction of natural oils differed depending on the percentage of methyl oleate they contain [11]. Olive oil which contains 73% methyl oleate gave highest yield compared to rapeseed and sunflower oils. Methoxycarbonylation reaction of 10 mL of each of the oils yielded 6.93 g (74.7%), 6.44 g (69.3%), and 3.36 g (36.8%) for olive, rapeseed and sunflower oils respectively. Thus in this paper, olive oil was used as a starting material for the synthesis of dimethyl 1,19-nonadecanedioate (8) to be used in the polymerization reaction with cardanol derived monomers. The methoxycarbonylation reaction of olive oil was achieved using the Pd/DTBPMB/H + catalytic system to give dimethyl 1,19nonadecanedioate (8) in 74.3% yield. The spectral data (shown in the experimental section) agree well with literature data [11,24], as well as the chemdraw prediction for the structure of the compound. (7) with dimethyl nonadecanedioate (8) Schemes 3 and 4 show the polymerization reactions of monomers (6) and (7) with dimethyl 1,19nonadecanedioate (8). It was previously reported [23], that the polymerization reaction of methoxylated cardanol and its derivatives had two main challenges. One was; the formation of a mixture of oligomers and a lactone in a single polymerization reaction. Oligomers were easily isolated from the product mixture, but for the lactones, it was impossible to isolate them from the starting materials. The second challenge; the formed polymers were very brittle. Interestingly, polymerization reaction of monomers 6 and 7 with dimethyl 1,19-nonadecanedioate reported in this paper (Schemes 3 and 4) produced oligomers that are not brittle. This could be due to the increased number of methylene groups between aromatic groups as a result of including dimethyl nonadecanedioate in the polymerization reaction.   (Figures 3-6).

Polymerization of 3-(16-aminohexadecyl)phenol (6) and 3-(16-hydroxyhexadecyl)phenol
Zooming Comparing the 1 H NMR spectral data of 3-(16-hydroxyhexadecyl)phenol monomer (7) with that of an oligomer (shown in an experimental section), it can be observed that the peak for CH2OH appears at δ 3.5 (for compound 7) while in an oligomer there is no any resonance at δ 3.5 but appears at δ 4.1 which corresponds to -CH2O(CO)R. This agrees well with the chemdraw predictions for the structure of an oligomer.
Zooming the aromatic region in the 1 H NMR of an oligomer (Figure 8) in the experimental section; two different kinds of aromatic rings are observed: terminal ones, which should appear very similar to those in the monomer (7), and the one on the middle of AB2, where the aromatic ring is expected to be significantly different from the terminal ones, but similar as those in any oligomers where the monomer is incorporated through the phenolic OH group. The aromatic ring of the cardanol derived monomer is expected to have 4 signals. If the OH in the aromatic ring of the monomer is defined as being in position 1 and the alkyl chain in position 3, then the H atoms should resonate: in positions 2 (s), 4 (d), 5 (t) and 6 (d). Comparing with the spectral data of the cardanol derived monomer (7) it is evident that t, x, y and z arise from terminal aromatic ring of cardanol derived units whilst s, u, v and w arise from the in chain cardanol units. Thus, the spectrum in figure 8 indicates that there are two kinds of aromatc rings, as it should be expected for any linear chain cardanol derived oligomers.
On the other hand, polymerization reaction of 3-(16-aminohexadecyl)phenol with dimethyl 1,19-nonadecanedioate resulted into formation of trimers as revealed by MALDI-TOF MS spectrum (Figure 7). In this case too, both forms of trimers such as AB2 and A2B are observed (where A stands for 3-(16-aminohexadecyl)phenol and B stands for dimethyl 1,19nonadecanedioate). In Figure 7, [A2B+Na] + has an Mr = 980.8 whereas [AB2+Na] has an Mr =1002.9. The other ion peaks in the spectrum do not relate with any oligomer fragments. They may possibly come from matrix clusters.

Materials and instruments
All reagents and solvents were purchased from Sigma-Aldrich and used as received. Cashew nut shell liquid was obtained from the shells by the solvent extraction method [25]. Cardanol and methoxylated cardanol; methyl 16-(3-hydroxyphenyl) hexadecanoate were synthesized using methods reported previously [3,23]. Standard Schlenk line and glove box techniques were employed to carry out all hydrogenation reactions reported in this paper. 1 H and 13 C nuclear magnetic resonance (NMR) spectra were recorded at 298 K by using Bruker 300 or 400 MHz and 75 or 100 MHz spectrometers for 1 H and 13 C NMR, respectively. Samples for NMR analysis were dissolved in deuterated chloroform as solvent. The chemical shifts of both 1 H NMR and 13 C NMR spectra were reported relative to tetramethylsilane. (6) Dioxane (5 mL) and ammonia solution (5 mL) were placed in a Schlenk tube and bubbled with argon (in the Schlenk line) for 10 min. In the glove box, methyl 16-(3hydroxyphenyl)hexadecanoate (600 mg), Ru(acac)3 (8 mg) and 1,1,1-tris(diphenylphosphinomethyl)ethane, Triphos (28 mg) were measured and added into an autoclave previously charged with magnetic stirrer. The autoclave was degassed three times using the Schlenk line. Dioxane (5 mL) and ammonia solution (5 mL) in the Schlenk tube were transferred into the autoclave using a 20 mL syringe. The solution in the autoclave was flushed with hydrogen three times, and pressurized to 40 bar. The reaction was left to proceed at 220 o C for 24 h. After venting hydrogen, and concentrating the solution (under vacuum), the yellowish solid product (565 mg) was obtained. The product was dissolved in chloroform (10 mL). Some insoluble solids (95 mg) were observed and removed from the solution by filtration. The yellowish solid (obtained after concentrating the solution containing soluble products) was dissolved in dichloromethane (5 mL) and recrystallized by addition of methanol to give white solids; 3-(16-aminohexadecyl)phenol (208 mg, 37%). 1 (7) Dioxane (10 mL) and water (1 mL) were placed in a Schlenk tube and bubbled with argon (in the Schlenk line) for 10 min. In the glove box, methyl 16-(3-hydroxyphenyl) hexadecanoate (400 mg), Ru(acac)3 (8 mg) and 1,1,1-tris(diphenylphosphinomethyl)ethane, Triphos (28 mg) were measured and added into an autoclave previously charged with magnetic stirrer. The autoclave was degassed three times using the Schlenk line. Dioxane (10 mL) and water (1 mL) in the Schlenk tube were transferred into the autoclave using a 20 ml syringe. The solution in the autoclave was flushed with hydrogen three times, and pressurized to 40 bar. The reaction was left to proceed at 220 o C for 24 h. After venting hydrogen, and concentrating the solution (under vacuum), the yellowish solid product obtained was dissolved in dichloromethane and recrystallized by addition of hexane to give a white solid; 3-(16-hydroxyhexadecyl)phenol (275 mg; 75%). 1  Olive oil (10 mL) was dissolved in methanol (10 mL) in a Schlenk tube and bubbled with argon for 10 min. In the glovebox, [Pd2(dba)3] (dba = 1,3-dibenzylideneacetone) (0.549 g) and 1,2bis(ditertbutylphosphinomethyl)benzene (DTBPMB) (2.367 g) were placed into the steel autoclave along with a magnetic stirrer. The autoclave was connected to the Schlenk line and it was degassed three times. The olive oil solution was transferred into the autoclave by means of a 20 mL syringe. Methane sulfonic acid (0.78 mL) was added to the solution in the autoclave. The autoclave was purged with CO before it was pressurized to 30 bar. The reaction progressed at 80 °C for 24 h. After cooling, the CO was vented, and the obtained yellow powder was dissolved by adding dichloromethane (20 mL). The resulting yellow solution was filtered through a Celite bed, and concentrated under vacuum until a white precipitate appeared. Cold methanol (0 °C) was added and the mixture was stirred in an ice bath for 20 min before filtration. The desired product was obtained as a white powder (6.89 g, 74.3%). 1 13 C NMR spectrum of 3-(16-hydroxyhexadecyl)phenol (7) with peak assignments.