Synthesis of cyclic α-pinane carbonate – a potential monomer for bio-based polymers

This work reports the first known synthesis of α-pinane carbonate from an α-pinene derivative. Pinane carbonate is potentially useful as a monomer for poly(pinane carbonate), which would be a sustainable bio-based polymer. α-Pinene is a major waste product from the pulp and paper industries and the most naturally abundant monoterpene in turpentine oil. α-Pinene is routinely converted to pinene oxide and pinanediol, but no study has yet demonstrated the conversion of pinanediol into α-pinane carbonate. Here, α-pinane carbonate was synthesised via carboxylation of α-pinanediol with dimethyl carbonate under base catalysis using triazabicyclodecene guanidine (TBD). 81.1 ± 2.8% α-pinane carbonate yield was achieved at 98.7% purity. The produced α-pinane carbonate was a white crystalline solid with a melting point of 86 °C. It was characterised using FTIR, NMR, GCMS and a quadrupole time-of-flight (QTOF) mass spectrometer. The FTIR exhibited a C 
<svg xmlns="http://www.w3.org/2000/svg" version="1.0" width="13.200000pt" height="16.000000pt" viewBox="0 0 13.200000 16.000000" preserveAspectRatio="xMidYMid meet"><metadata>
Created by potrace 1.16, written by Peter Selinger 2001-2019
</metadata><g transform="translate(1.000000,15.000000) scale(0.017500,-0.017500)" fill="currentColor" stroke="none"><path d="M0 440 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z M0 280 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z"/></g></svg>
 O peak at 1794 cm−1 confirming the presence of a cyclic carbonate. GCMS showed that the α-pinane carbonate fragments with loss of CO2, forming pinene epoxide. Base hydrolysis of the α-pinane carbonate using NaOH/ethanol/water regenerated the pinanediol with formations of Na2CO3.


Introduction
Monoterpenes are naturally occurring unsaturated hydrocarbons obtained as waste products from the pinewood pulping industry. The most commonly available monoterpenes are pinene and limonene, which are cyclic monoterpene isomers comprising two iso-propene units with a molecular formula of C 10 H 16 (Scheme 1). Approximately 3.5 Â 10 5 tons per year of turpentine oil is produced from the pulping industry globally, 1,2 with the global supply of turpentine oil expected to increase as the pulping industry grows. The pulping process produces between 0.3 and 1.0 kg of turpentine oil per tonne of pulp, 3 of which the predominant monoterpene is a-pinene ($70%) when using the current sulphate processes. 3,4 Limonene accounts for $8.6% (0.3 Â 10 5 tons per annum) of the worldwide turpentine oil production. 1 Crude turpentine oil from steam distillation of pine is typically 75-85% a-pinene, and the remainder were 0-3% b-pinene and 5-15% limonene. 5 A major source of limonene is the peel of citrus fruits which contains about 90 wt% limonene. 6 Limonene is mainly obtained from the waste products of orange harvesting and peel from orange juice production. 7 The large global tonnage of a-pinene and limonene could provide substantial amounts of sustainable platform chemicals for bio-based polymer productions. 8 There have been extensive studies of limonene derivatives such as limonene epoxide [9][10][11] and limonene bis-epoxides, 12,13 which can be copolymerised with CO 2 to obtain poly(limonene carbonates), 9,11,[14][15][16] as shown in Scheme 2(a). Limonene oxides and carbonates can be used as highly valuable platform chemicals for biopolymer synthesis. [17][18][19][20] Conversely, despite a-pinene being the most abundant naturally occurring monoterpene, 17 there are limited studies on transformations of a-pinene into essential derivatives, such as apinene epoxide and a-pinane carbonate.
The a-pinane carbonate is potentially a monomer via a ringopening copolymerisation to produce poly(a-pinane carbonate). The a-pinene epoxide is an important intermediate for other chemicals including campholenic aldehyde used in fragrances 8,21,22 and pinanediol. 23 a-Pinene can also be oxidised to produce essential chemicals such as verbenol and verbenone (Scheme 2(b)). 8,24,25 Unlike limonene epoxide, which can be readily converted to limonene carbonate, for potential industrial productions of poly(limonene carbonate), the applications of a-pinene oxide have been mainly limited to its isomerisation products such as campholenic aldehyde, 22 as no route from apinene oxide to the carbonate has yet been established.
Various attempts to synthesise polyether via homopolymerisation of a-pinene oxide or poly(a-pinene carbonate) from a-pinene epoxide and CO 2 have been unsuccessful. No poly(a-pinane carbonate) or a-pinane carbonate was obtained in the reactions of a-pinene epoxide with CO 2 . 26 A recent review on the copolymerisation of sustainable epoxides with CO 2 also reported that there are currently no available publications on the successful copolymerisation of a-pinene oxide with CO 2 . 27 However, an existing patent claims that poly(pinene carbonate) can be produced by the reaction of pinene epoxide and CO 2 in the presence of a (salen)CrCl/PPNCl catalyst. 27 The lack of reactivity of the a-pinene epoxide in ring-opening polymerisation has been attributed to a higher reaction barrier of the apinene epoxide in the ring-opening process due to steric hindrance. 26 The lack of reactivity is unfortunate, as there is potentially a signicant market for pinane carbonate, given that the current production of cyclic carbonates is 4 million tons per year. 28 This could add value to the paper and pulp industries, whilst introducing a new sustainable platform chemical. The cyclic pinane carbonate could be polymerised using established methods for ring-opening polymerisation of cyclic organic carbonates. [29][30][31] This study reports the rst synthesis of a-pinane carbonate from a-pinanediol. The synthesis route reported here involved a novel and eco-friendly process via reactions of a-pinanediol with dimethyl carbonate (DMC). Reactions of diols with DMC to obtain a cyclic carbonate have been reported for some other polyhydric alcohols, 32 but have not previously been applied for a-pinanediol. The only existing report of pinane carbonate synthesis was via the reaction of 2,10-b-pinanediol (Scheme 2(c)) with ethyl chloroformate in pyridine, followed by a molecular rearrangement. 33 Although 2,10-b-pinanediol could be obtained from hydroxylation of b-pinene, the content of b-pinene in crude turpentine oil is very low, typically about 0-3%, as compared to 75-85% for a-pinene. 5 The DMC route was chosen for this synthesis because it is less toxic compared to the phosgene process. 34 It is anticipated that the use of a-pinanediol would reduce the reaction barrier associated with a-pinene epoxide and overcome the steric hindrance, 26 allowing for formations of a-pinane carbonate. a-Pinanediol can be readily obtained via osmium tetraoxide catalysed hydroxylation of naturally available a-pinene, 35 or from hydrolysis of the apinene epoxide. 23 Furthermore, DMC is a green solvent, which has been synthesised by CeO 2catalysed reactions of CO 2 and methanol (MeOH) to achieve >95% methanol conversions and >99% selectivity to DMC. 36 Similarly, some other main methodologies for DMC preparation currently applied in the industry are oxidative carbonylation of MeOH, urea mediated synthesis and trans-esterication of ethylene carbonate and propylene carbonate. 37

Experimental procedures
The pinane carbonate synthesis was performed in a 150 mL 2neck batch reactor equipped with a condenser and a thermocouple/sampling unit, by reactions of pinanediol and DMC, as shown in Scheme 3. Synthesis of a-pinane carbonate was carried out using 0.18 mol of DMC, 0.03 mol of pinanediol, and 2 mmol of TBD guanidine catalyst. The reaction was carried in the round bottom ask. Pinanediol (5 g) and DMC (15.88 g), corresponding to a 6 : 1 DMC to pinanediol molar ratio, were added to the reactor and heated to the reaction temperature of 90 C using a heater-stirrer (IKA® Basic). This was followed by additions of 250 mg of TBD guanidine catalyst (5 wt% based on the pinanediol) and mixing at 600 rpm. The reaction mixture was reuxed at 90 C for 6 h in the reactor. 0.25 mL samples were collected every 1 h and analysed by gas chromatography (GC) to monitor the progress of the reaction until the pinanediol peak disappeared.
The reaction mixture was then neutralized using 110 mL of acetic acid. Excess DMC and residual methanol by-products in the reaction were removed by rotary evaporation at 60 C at 120 mmHg pressure. An amber-coloured solid was obtained, and this was puried by re-crystallization 3 times in methanol (25 mL). The recrystallized product was dissolved in a mixture of 25 mL n-hexane and 10 mL methanol with vigorous shaking. The hexane extract was a colourless viscous solution, from which white crystals of a-pinane carbonate appeared as the excess hexane evaporated at room temperature. The apinane carbonate was ltered out and dried at room temperature. The reaction mixture was recrystallized three times to remove any traces of the catalyst. About 4 g of apinane carbonate was obtained, with >98.7 wt% purity as conrmed by the GC. The product was characterised using NMR ( 1 H and 13 C), FTIR, GCMS, a differential scanning calorimeter (DSC) and a quadrupole time-of-ight (QTOF) mass spectrometer. The reactivity of the synthesised a-pinane carbonate was investigated, especially to nd a route for regenerating the pinanediol. This is important in ensuring that a suitable process exists for recycling pinanediol from the a-pinane carbonate at the product's end of life. About 0.5 g of the dried apinane carbonate was hydrolysed in the batch reactor with 25 mL of 0.5 M NaOH in an ethanol-water solution containing 10 (v/v)% deionised water. The 25 mL 0.5 M NaOH solution was heated in the reactor to 60 C, followed by the transfer of 0.5 g apinane carbonate into the reactor and vigorous mixing at 600 rpm. The ratio of a-pinane carbonate to NaOH solution (w/ v) in this study was adapted from the commonly used method for saponication of fats and oils. [38][39][40] About 0.5 mL of the reaction sample was collected using a micropipette at various time intervals from 0-60 min. The sample was transferred into a pre-weighed 2 mL vial containing 14 mL of acetic acid to quench the reaction immediately.

Sample analysis
FTIR analysis of the samples was performed using a Mettler Toledo ATR-FTIR spectroscopy (React IR 4000) equipped with a DiComp diamond K6 conduit 16 mm probe. The spectra of the product samples were collected over the range 4000-650 cm À1 . The instrument was initialized by collecting 256 scan background spectra for air and water vapours before collecting the spectra of the samples. 1 H-NMR and 13 C-NMR spectroscopy of the samples were recorded on a Bruker Avance III HD spectrometer at 700 MHz using a nitrogen-cooled cryoprobe prodigy™, with methanol-d (CD 3 OD) as a solvent. Melting points of the pinanediol and the produced a-pinane carbonate were determined using a Reichert platform melting point apparatus. DSC analysis was performed using (TA Instruments, DSC Q20 V24.11 Build 124), with a TA Universal analysis soware to do the DSC analysis using 3.1-3.4 mg of samples that were carefully weighed into Tzero aluminium pans and sealed with Tzero aluminium lids. The samples were scanned over the temperature range of 25-400 C at a heating rate of 5 C min À1 and a 50 mL min À1 ow of oxygen-free nitrogen for the sample purge.
A 6890 Hewlett Packard Series GC equipped with a Nukol™ fused silica column of 30 mm length, 0.32 mm ID and 0.25 mm lm thickness was used to quantify the a-pinanediol and apinane carbonate in the samples. The GC oven temperature was programmed from 120 C initial temperature held for 5 min and ramped at 15 C min À1 to a nal temperature of 210 C and held for a further 10 min (total time of 21 min). The ame ionisation detector (FID) and injector temperatures were set at 260 and 250 C, respectively, while a helium carrier gas was used at 10 PSI pressure. Quantications of the a-pinanediol and a-pinane carbonate were based on a calibration curve using standard reagents where correlation data (R 2 > 0.995) was obtained while methyl heptadecanoate was a GC internal standard. The calibration curve is drawn following concentration ranges from 1.0 to 50 mg kg À1 of the reference standard. About 100 mg of each sample was measured into a 2 mL GC vial and mixed with 500 mL of methyl heptadecanoate (10 mg mL À1 of ethanol), and 1 mL of the sample mixture was injected into the GC using a 5 mL GC syringe (SGE). The sample is run against this calibration curve to estimate its concentration. The a-pinanediol and a-pinane carbonate were also analysed using a GCMS: 7890B Agilent GC coupled to a 5977B mass selective detector (MSD) in full scan mode from 50-520 amu and 70 eV ionisation energy. The 7890B Agilent GC program was held at 50 C for 2 min initially and ramped to 310 C at 5 C min À1 (total time of 75 min). Molecular ions of the analytes were determined with a QTOF mass spectrometer. The chromatogram showed a single peak with a major ion on Rt 11.084 with a maximum area covering the purity content. The exact mass measurement is followed by QTOF with the least error and uncertainty incorporation as the ndings correlated with other analysis techniques and no impurity is highlighted.

Productions and characterisation of the a-pinane carbonate yield
An a-pinane carbonate was synthesised from an a-pinene derivative for the rst time, via reactions of a-pinanediol and DMC. The cyclic a-pinane carbonate yield for the process was 81.1 AE 2.8%, and the pinane carbonate was isolated as a white crystalline solid at 98.7% purity as conrmed by the GC. The melting points of the a-pinanediol and the a-pinane carbonate were determined by DSC, using methods reported elsewhere, 41 and these values were 86 C for the a-pinane carbonate, and 55 C for the 2,3-pinanediol (ESI Fig. 1 †). FTIR data for the apinane carbonate and the a-pinanediol are shown in Fig. 1. Fig. 1(a) shows an FTIR peak around 3200-3300 cm À1 , which is assigned to the O-H stretching vibration for the a-pinanediol. This peak was conspicuously absent in the a-pinane carbonate product. The disappearance of the O-H peak resulted in the formation of a new peak at 1795 cm À1 , corresponding to C]O stretching vibration for cyclic carbonates, as shown in Fig. 1(b) and (c), which is signicantly higher than the absorption wavenumber of about 1750 cm À1 for linear organic carbonate (DMC). The C]O stretching vibration for the a-pinane carbonate compares well with 1800 cm À1 , which has been reported for limonene cyclic carbonate. 11 A previous study has also demonstrated that C]O stretching for cyclic propylene carbonate occurs at a higher wavenumber (1799 cm À1 ) than the C]O stretch for linear poly(propylene carbonate) with wavenumber at 1747 cm À1 . 42 Apart from the characteristic FTIR peaks at 3200-3300 cm À1 for the O-H and 1795 cm À1 for the C]O functionalities, other FTIR peaks observed at 2980-2850 cm À1 for C-H stretching vibrations in methyl and methylene groups are similar for both species, as would be expected. The FTIR data clearly shows that the a-pinanediol was converted to a-pinane cyclic carbonate during the reaction.
The 1 H NMR data for the a-pinanediol and the a-pinane carbonate are shown in Fig. 2. The a-pinanediol has 1 H NMR peaks (700 MHz, CD 3 OD), as shown in Fig. 2(a), at the following chemical shis (d/ppm): 1.30 (3H, s), 3.98 (1H, dd), 1.89 (1H,t) for proton that points out of the a-pinane ring (exo-proton) and 1.97 (1H, dtd) for the proton that points inside the a-pinane ring (endoproton) which is more shielded, 2.18 (1H, dtd), 0.98 (3H, s), 1.27 (3H, s), 2.43 (1H, dddd), 1H (1.47, m) for exo-proton and 1H (2.43, dddd) for the more shielded endo-proton, and 4.88 (2H, s) for the -OH groups. The 1 H NMR peak at 4.88 ppm (2H, s) corresponds to the O-H protons of the pinanediol, which indicates an overlap in the 1 H NMR signals for the two hydroxyl groups on the pinanediol. A total of 9 hydrogen environments corresponding to a total of 18 hydrogen atoms were observed for the a-pinanediol (C 10 H 18 O 2 ).
The formation of a-pinane carbonate resulted in the disappearance of the proton peak at a chemical shi of 4.88 ppm, and slight variations in the chemical shi for some of the proton peaks, as shown in Fig. 2(b 1 H NMR peaks were also observed at 3.4 ppm (3H, s) and 4.9 ppm (1H, s), due to remaining methanol impurities. Overall, 8 hydrogen environments were observed, corresponding with a total of 16 hydrogen atoms for the C 11 H 16 O 2 in the apinane carbonate. The enantiotropic effect was observed in some of the 1 H-NMR peaks due to the different chemical environment of protons. For instance, the two protons attached to C4 were in different chemical environments as endo and exo protons, likewise, the two protons attached to C10. Therefore, the enantiotropic effect accounts for the differences in chemical shis for the C4 and the C10 protons shown in Fig. 2(a). Similarly, protons on the two methyl groups (C7 and C8) attached to carbon number 6 experience different chemical environments due to the enantiotropic effect. Protons on C7 are more shielded (d ¼ 0.98 ppm), indicating an endo-position of the C7 methyl group, while the protons on C8 are more deshielded (d ¼ 1.27 ppm) which suggests that the C8 methyl group occupies an exoposition on the a-pinane ring.  -pinanediol and, (b) a-pinane carbonate.
The chemical shis observed in a-pinane carbonate differed from pinanediol. An additional carbon peak was observed at d (ppm) of 155.1, as shown in Fig. 3(b), corresponding to C]O of the a-pinane carbonate. The carbon peak at 155.1 ppm chemical shi is consistent with C]O carbon for organic carbonates. 9,43 Fig. 3(b) shows that the a-pinane carbonate has 13  The major m/z mass spectra peaks for the a-pinane carbonate (ESI Fig. 3 †) were 55.0, 67.0 (100%), 69.0, 83.0, 109.1, and 137.0. A molecular ion peak for pinane carbonate was also not observed at the expected m/z of about 196.2. Mass spectra for the a-pinane carbonate in ESI Fig. 3 † showed that the m/z peaks are consistent with fragmentation patterns for a-pinene epoxide in the National Institute of Standards and Technology (#21620 in NIST98.1 MS database), indicating that loss of CO 2 and molecular rearrangement to form a-pinene epoxide was the plausible fragmentation mechanism of a-pinane carbonate. Molecular ions of the a-pinanediol (152.1 g mol À1 ) and the apinane carbonate (196.1 g mol À1 ) were subsequently determined as shown in ESI Fig. 4 † with the QTOF mass spectrometer.
It is expected that a-pinane carbonate can be polymerised to obtain poly(a-pinane carbonate), which can be used as a biobased thermoplastic, like poly(limonene carbonates). 9,15 poly(limonene carbonate) has a substantially high glass-transition temperature (T g ¼ 130 C), 15 and this is the expected range for a poly(a-pinane carbonate). Therefore, it is envisaged that a poly(a-pinane carbonate) could be a sustainable replacement for petroleum-derived polycarbonates.

Recovery of a-pinanediol via base hydrolysis of the pinane carbonate
Currently, a major challenge for the plastics industry is the development of efficient strategies for the handling of waste plastic at the end of life. 44,45 Proposed routes for waste plastics handling include recycling 44 or composting, in the case of biodegradable materials, such as bioplastics used in packaging. 45 However, a more acceptable route would be through recovery and recycling of the starting materials for further utilisation. To develop a feasible route for possible recovery and recycling of the a-pinanediol starting material, the reactivity of the a-pinane carbonate towards base hydrolysis was investigated. The results in Fig. 4 show that quantitative recovery of apinanediol was achieved from the base hydrolysis of the apinane carbonate. A poly(a-pinane carbonate) is expected to undergo similar reactions.
Over 80% of the pinane carbonate was saponied within 2 min using 0.5 M NaOH prepared in ethanol/water solution containing 90 : 10 (v/v)% and 60 C. At about 30 min reaction time, all the pinane carbonate had been completely (100%) saponied and converted to pinanediol and sodium carbonate (Scheme 4). Base hydrolysis of the a-pinane carbonate via reactions with hydroxide ions to form carbonic acid salt and apinanediol is shown in Scheme 4. This reaction proceeds similarly to typical ester saponication. It should be noted that such reactions can also be performed using different hydroxide The use of ethanol/water systems (i.e., water and soluble organic solvents) allows for a complete dissolution of the NaOH and the a-pinane carbonate, which eliminates the mass transfer limitations that can occur due to poor immiscibility of carbonate esters and water. To ensure complete saponication of the a-pinane carbonate, the NaOH solution was used in a large excess of the stoichiometric amounts. 39 A proposed reaction mechanism for base hydrolysis of the a-pinane carbonate is shown in Scheme 5, with a rate-determining step involving the bimolecular collision of the pinane carbonate and hydroxide ion to form a tetrahedral intermediate, which decomposes to give the products, as previously reported for saponication of alkyl esters. 51 In this reaction mixture, a molecule of water is required to stabilise the intermediates and the same can also be provided by other protic solvents such as simple alcohols, like base hydrolysis of alkyl esters. 50,51 During the reaction as shown in the Scheme 5, the hydroxyl group acted as a nucleophile and reacted with a-pinane carbonate (a) to yield intermediate (b) which decomposes and the carbonyl group reforms to form 2carboxylate-3-pinacol (d) via intermediate (c) in the basic ethanol/water solution that further rearranges to furnish nal product (e). As one side of a-pinane carbonate is more crowded thus it is most likely that the hydroxyl group as a nucleophile attacks on the opposite side on the most electropositive position however work is in progress to determine the exact mechanism of the reaction. Overall, the a-pinane carbonate hydrolysis was rapid, leading to the quantitative recovery of the a-pinanediol within 30 min. Previously, studies have shown that the reactivity of carboxylic acids decreases with the length of alkyl substituent due to a combination of polar and steric inuences of the alpha substituent on the carboxylic group. [52][53][54] The carbonic acid moiety on the a-pinane carbonate has no alkyl substituent, hence, its base hydrolysis is rapid. It is envisaged that poly(a-pinane carbonate) would follow a similar reaction pattern at slower reaction rates due to higher molecular weight.

Conclusions
This study is the rst case of synthesis of a-pinane carbonate from a-pinanediol, a derivative of the most abundant naturally occurring monoterpene, a-pinene. Previous attempts to produce pinane carbonate by reactions of a-pinene epoxide with CO 2 have been unsuccessful due to the lack of reactivity of the epoxide in ring-opening polymerisation because of steric hindrance. Here, a-pinane carbonate was synthesised by the reactions of a-pinanediol with DMC in the presence of a TBD Scheme 5 Proposed reaction mechanisma bimolecular collision of solvated hydroxide ion with the alkyl esters carbonyl carbon. guanidine catalyst. The a-pinane carbonate is a white crystalline solid with 86 C melting point. The maximum a-pinane carbonate yield achieved was 81.1 AE 2.8%, at 98.7% purity, aer recrystallization. The a-pinane carbonate produced was characterised using FTIR, NMR ( 1 H and 13 C), GC and GCMS and QTOF mass spectrometer. Reactivity of the a-pinane carbonate towards base hydrolysis with NaOH in ethanol/water was investigated to develop a feasible route for possible recovery and recycling of the a-pinanediol starting material. It was shown that a-pinane carbonate could be completely saponied within 30 min to the a-pinanediol and sodium carbonate. The apinane carbonate can likely be polymerised to obtain poly(apinane carbonate), which would be a bio-based thermoplastic, like poly(limonene carbonates) in an industrial scale process. The used poly(a-pinane carbonate) could be hydrolysed to recycle the a-pinanediol starting feedstock in the same way as the a-pinane carbonate monomer.

Conflicts of interest
There are no conicts to declare.