A quest for polycarbonates provided via sustainable epoxide/CO2 copolymerization processes

Stephanie J. Poland *a and Donald J. Darensbourg b
aDepartment of Chemistry and Biochemistry, Rose-Hulman Institute of Technology, Terre Haute, Indiana 47803, USA. E-mail: poland@rose-hulman.edu
bDepartment of Chemistry, Texas A&M University, College Station, Texas 77843, USA

Received 21st August 2017 , Accepted 2nd October 2017

First published on 2nd October 2017

The present review highlights advances in the copolymerization of carbon dioxide (CO2) and epoxides (oxiranes) to produce polycarbonates. Specifically, focus has been given to epoxide starting materials that have been generated from renewable feedstocks including straight-chain alkylene oxides, cyclohexadiene oxide, and limonene oxide, among several others. These renewable feedstocks are attractive green alternatives to traditional petroleum-based materials, however, some limitations do exist regarding the availability of monomer feedstocks as well as economic factors. Recent advances for each of these polymeric products will be discussed in detail. The use of industrial post-combustion CO2 waste streams will also be commented upon.

image file: c7gc02560b-p1.tif

Stephanie J. Poland

Stephanie J. Poland was born in Terre Haute, IN in 1986. She received her B.S. from the University of Southern Indiana in 2009 and her Ph.D. from Texas A&M University in 2013. Since 2013, she has been employed as an Assistant Professor in the Chemistry & Biochemistry Department at Rose-Hulman Institute of Technology. Her research interests lie in the development of green polymerization techniques and the expansion of contemporary research practices into undergraduate teaching laboratories.

image file: c7gc02560b-p2.tif

Donald J. Darensbourg

Donald J. Darensbourg was born in Baton Rouge, LA in 1941 and received his B.S. and Ph.D. degrees from California State University at Los Angeles and the University of Illinois/Urbana, respectively. Following a nine month period at the Texaco Research Center in Beacon, NY, he was on the faculties of State University of New York at Buffalo from 1969 to 1972 and Tulane University from 1973 to 1982. He has been at Texas A&M University since 1982 where he currently is a Distinguished Professor. Among his current interests are the utilization of CO2 as both a monomer and solvent in copolymerization reactions with oxiranes and oxetanes, and the ring-opening polymerization of renewable monomers such as lactides.


The search for green catalytic processes for the synthesis of useful organic products such as polymers and fine chemicals can only be expected to increase as fossil-based feedstocks decline. Chemical transformations which incorporate the waste material and greenhouse gas carbon dioxide in non-reductive, highly-selective pathways are poised to become extremely important in the coming years. Among the most prominent transformations meeting these green chemistry principles are the coupling reactions of CO2 with epoxides (oxiranes). Numerous epoxides have been shown to effectively copolymerize with the renewable C1 building block CO2 to provide polycarbonates (Scheme 1).1–13 Nevertheless, to improve the sustainability of this process, it will be necessary to incorporate epoxide monomers derived from renewable resources. This report will focus on the synthesis and properties of polycarbonates prepared from the catalytic coupling of CO2 and epoxides derived from renewable feedstocks. In addition, there are a greater number of bio-based cyclic carbonates produced from the cycloaddition of CO2 and epoxides (Scheme 1).9,14–18 Finally, while pure carbon dioxide is typically used for these coupling reactions, several recent reports detail the use of impure and even post-combustion sources of CO2.
image file: c7gc02560b-s1.tif
Scheme 1 Coupling of epoxides and CO2 to yield polycarbonates and cyclic carbonates.

Catalyst development for the copolymerization of CO2 and epoxides

Initially, it is useful to discuss the advances made in uncovering effective metal catalysts for these copolymerization processes. This process was first discovered utilizing a poorly defined heterogeneous catalyst derived from diethyl zinc and water by Inoue and coworkers in 1969.19 The first homogeneous catalyst for the alternating copolymerization of propylene oxide and CO2 involved tetraphenylporphyrin aluminum derivatives (tpp)AlX (X = OMe and Cl).20 Subsequently, the chromium(III) porphyrin complex and its fluorinated analogue have been shown to be effective at copolymerizing cyclohexene oxide and CO2.21,22 Several zinc derivatives have served as excellent catalysts for the coupling of numerous epoxides with CO2 to provide polycarbonates. These include mono- and di-nuclear zinc phenoxides and β-diiminates1,23,24 as well as bimetallic macrocyclic derivatives.25

The most widely studied and effective catalysts for the completely alternating copolymerization of a wide variety of epoxides and carbon dioxide are metal salen complexes [H2salen = bis(3,5-di-tert-butylsaliclidene)-1,2-diamine] of Cr(III), Co(III), and Al(III).26–28 These metal derivatives are most active in the presence of an added nucleophile such as an onium salt or heterocyclic amine (aka, binary catalysts). There are numerous instances where ionic liquids (ILs) are employed as cocatalysts, providing anions as the added nucleophiles. However, the use of ILs is generally confined to the production of cyclic carbonates, with few exceptions.29–31 Saturated versions of the salen ligand in (salan)MX derivatives of chromium and cobalt have also been utilized as catalysts for the copolymerization process.32 An important modification of these metal complexes involves the incorporation of the added nucleophile into the salen ligand (aka, bifunctional catalysts).33,34 These latter catalysts have greatly enhanced the selectivity for copolymer production over cyclic carbonate formation, and they offer some advantages for catalyst recovery as well. Fig. 1 summarizes representative homogeneous metal catalysts for the copolymerization of epoxides and CO2.

image file: c7gc02560b-f1.tif
Fig. 1 Representative homogeneous catalyst systems for the copolymerization of epoxides and CO2. For the bimetallic complexes, M = Zn, Co, Fe, Mg. For salen binary catalysts, M = Cr, Co, Al. For salen bifunctional catalysts, M = Co, Cr.

The development of these well-defined metal catalysts has made it possible to assess the mechanistic aspects of this copolymerization process in great detail. In general for these catalyst systems, the initial ring-opening of a metal bound (activated) epoxide by a nucleophile is fast, and the formed alkoxide undergoes CO2 insertion rapidly at modest CO2 pressures. Commonly, the rate-determining step in this copolymerization process is the propagation step, i.e., the ring-opening of a metal bound epoxide by the propagating anionic organic carbonate nucleophile. For example, in the case involving mononuclear metal catalysts, the growing polymer chain must first dissociate from the metal center to accommodate epoxide binding and subsequent ring-opening by the anionic growing polymer chain. Indeed, this process can lead to a facile backbiting process of the “free” growing polymer chain to produce the corresponding cyclic carbonate. This latter process is inhibited in the presence of bifunctional catalysts, where the cation attached to the salen ligand can form an ion-pair with the dissociated anionic growing polymer chain. On the other hand, dimetallic catalysts provide adjacent binding sites for both the growing polymer chain and the epoxide monomer. These processes are illustrated in Fig. 2.

image file: c7gc02560b-f2.tif
Fig. 2 Epoxide ring-opening transition states for (a) dizinc and (b) dimagnesium catalysts, where P represents the growing polymer chain.

Renewable epoxides

The copolymerization of CO2 has been studied for a great number of epoxides, many of which are shown in Fig. 3. Specifically relevant to this review are epoxides that can be obtained from renewable feedstocks. The epoxides encased in green in Fig. 3 represent instances where this has been achieved. We wish to distinguish two terms, metabolite and precursor (Scheme 2). In this instance, a metabolite is a direct product of metabolism, e.g., ethanol. Thus, while ethanol can be directly obtained from renewable resources, it must go through additional reaction steps to yield an olefin that can be oxidized to the needed epoxide monomer. Each of these extra steps effectively decrease the greenness of the renewable monomer. Alternatively, there exist some precursor olefins such as limonene that can be extracted from renewable resources and converted to the epoxide in a direct, one-step process. The necessary epoxidation reactions can be carried out by any number of methods, including stoichiometric reactions involving organic peracids such as mCPBA (m-chloroperbenzoic acid) or more atom economical catalytic reactions such as those that utilize H2O2. In Fig. 3, the epoxides highlighted in light green are derived from metabolites whereas those highlighted in dark green have been produced directly from their precursor olefins.
image file: c7gc02560b-f3.tif
Fig. 3 Some of the epoxides that have coupled with CO2 to yield polycarbonates to date. Epoxides highlighted in light green are derived from metabolites whereas those highlighted in dark green are directly synthesized from precursors.

image file: c7gc02560b-s2.tif
Scheme 2 Route of biopolymer production from metabolites and precursors.

Copolymerization of alkylene oxides and carbon dioxide

Ethylene is utilized in the production of almost half of all widely-used polymers today, and most of this chemical feedstock is derived from petroleum sources. Nevertheless, biomass can provide an alternative source of this important chemical.35,36 Currently, the largest bio-ethylene plants, located in Brazil and India, account for less than 0.5% of the global ethylene capacity. Bio-ethanol, which is used as a biofuel in transportation, is produced at about 100 billion liters annually. It can undergo catalytic dehydration to yield bio-ethylene, which can later be oxidized to form ethylene oxide (Scheme 3). Bio-ethanol is presently produced in Brazil from sugarcane and in the United States from corn.36 Unfortunately, using these feedstocks for the ultimate production of bio-ethylene directly competes with food production. An alternative source of bio-ethanol is lignocellulosic biomass (e.g., wood, straw, bagasse), and several companies have just recently started to commercialize bio-ethanol production from these waste materials.37 Notably, lignocellulosic bio-ethanol does not compete with food production and requires less arable land and water.
image file: c7gc02560b-s3.tif
Scheme 3 Synthesis of ethylene oxide from bioethanol.

Ethylene oxide, the simplest three-membered cyclic ether, is a colorless, extremely flammable and highly reactive gas which possesses a faintly sweet odor. It is a very toxic substance which is both carcinogenic and mutagenic. The production of ethylene oxide involves the catalytic reaction of ethylene directly with oxygen or air in the presence of a silver catalyst.38 It is widely used as a chemical feedstock for manufacturing numerous products, for example, ethylene glycol and polyethylene glycol (PEG).

Because of ethylene oxide's hazardous nature, the copolymerization of ethylene oxide and CO2 is much less studied academically than its propylene analogue for safety reasons (Scheme 4). However, Yamada and coworkers have reported the use of di(ketoiminato) cobalt complexes (Fig. 4) for the effective copolymerization of ethylene oxide (EO) and CO2 to provide completely alternative copolymers of poly(ethylene carbonate), PEC.39 Utilizing catalyst 1 in the presence of PPNCl, a sizable quantity of the cyclic carbonate (EC) was produced in addition to PEC. By way of contrast, upon using catalyst 2, only trace quantities of EC were provided. An additional study for the production of EC with 100% selectivity utilizing phosphine-bound zinc halide complexes has been reported.40

image file: c7gc02560b-s4.tif
Scheme 4 Copolymerization of ethylene oxide and CO2 to yield poly(ethylene carbonate).

image file: c7gc02560b-f4.tif
Fig. 4 Di(ketoiminato) cobalt complexes for the copolymerization of ethylene oxide and CO2. For 1, X = I. For 2, X = OBzF5.

Interestingly, the completely alternating PEC copolymer synthesized by Yamada and coworkers exhibited different thermal degradation properties than that of a commercial sample.39,41 That is, it started rapid decomposition at 220 °C, with complete decomposition below 260 °C leaving no residual material. On the contrary, the commercial sample of PEC began to decompose at 190 °C leaving 5% by weight residue at 220 °C which remained even up to 350 °C. This observation was attributed to the structural regularity of the completely alternating copolymer, a property of use in numerous applications including evaporative pattern casting.

For many years, Empower Materials in the United States has commercially produced poly(ethylene carbonate) from ethylene oxide and CO2.41 Importantly, polyols produced from ethylene or propylene oxides are receiving much current commercial attention for their use as replacements for polyether polyols in the synthesis of polyurethane. This process is depicted in Scheme 5, where the chain-transfer agent (CTA) is typically water or a diol. Perfectly alternating polycarbonate polyols can contain 50 wt% CO2 for PEC and 43 wt% CO2 for PPC.

image file: c7gc02560b-s5.tif
Scheme 5 Synthesis of polycarbonate polyols.

Some of the industrial companies operating in this area include Aramco Performance Materials (formerly Novomer, Inc.),42 Econic Technologies,43 and Covestro (formerly Bayer MaterialScience).44 This method of polyol synthesis represents a more sustainable, cost-effective option with performance properties comparable to or exceeding those of conventional polyols.45 Indeed, since the market volume of polyurethane is so large, this technology is destined to become the largest utilization of CO2 as a chemical feedstock in the polymer industry.

Other epoxides derived from ethylene which have undergone copolymerization reactions with carbon dioxide to provide polycarbonates include 1-butene, 1-hexene, and 1-octene. Selective routes to 1-butene, 1-hexene, and 1-octene oxides involve dimerization,46,47 trimerization,48 and tetramerization49 of ethylene, where all processes are catalyzed by homogeneous catalysts. The corresponding epoxide can be readily synthesized by the various established pathways for epoxidation.

Copolymerization of these epoxides with pendant linear alkyl chains with carbon dioxide has been achieved by both homogeneous and heterogeneous catalysts. For example, these three epoxides/carbon dioxide copolymerization reactions catalyzed by a nanosized zinc–cobalt(III) double metal cyanide (DMC) complex have provided the regioirregular copolymers as depicted in Scheme 6.50 The glass transition temperatures (Tgs) of these polycarbonates showed the expected steady decrease with alkyl chain lengths, with observed values of 6 °C for poly(1-butene carbonate), −18 °C for poly(1-hexene carbonate), and −27 °C for poly(1-octene carbonate).

image file: c7gc02560b-s6.tif
Scheme 6 Synthesis of poly(alkylene carbonates) using Zn–Co DMC catalysts.

These three copolymers were also successfully synthesized via homogeneous (salen)Co(III) catalyst systems. That is, Nozaki and coworkers prepared the 1,2-butene and 1,2-hexene oxides copolymers with CO2 utilizing their Co(III) salen complex with a piperidinium end-capping arm 3 (Fig. 5).51 In these instances, moderately high molecular weight copolymers were obtained at 31[thin space (1/6-em)]000 and 34[thin space (1/6-em)]300 g mol−1, respectively.

image file: c7gc02560b-f5.tif
Fig. 5 Nozaki's bifunctional (salen)Co(III) catalyst 3.

Similarly, Lee and coworkers have prepared these copolymers using a salen Co(III) catalyst, where the salen ligand is tethered by four quaternary ammonium salts.52 Because of the very high activity of this catalytic process, high molecular weight polycarbonates were obtained, i.e., molecular weights greater than 200[thin space (1/6-em)]000 g mol−1 were attainable. Furthermore, this catalyst system likely leads to regioregular copolymers, hence, the Tgs for poly(1-butene carbonate) and poly(1-hexene carbonate) were reported to be slightly different from those of Zhang et al.50 at 9 °C and −15 °C, respectively.52

While propylene oxide itself is not viewed as a renewable epoxide, it has recently been used in a terpolymerization reaction with a novel epoxide monomer. Propylene oxide, CO2, and epoxidized soybean oil (ESO) were coupled utilizing a cobalt–zinc DMC catalyst (Fig. 6).53 The ESO was produced from soybean oil triglycerides and contains approximately 4.5 epoxides per triglyceride. The structure of the resultant polymers is believed to be poly(propylene carbonate) chains bonded to the ESO. Up to 7.8% ESO incorporation was achieved, although polyether linkages and byproduct cyclic carbonate were also observed. Notably, the ESO's triglyceride core remained intact during the coupling process.

image file: c7gc02560b-f6.tif
Fig. 6 Epoxidized soybean oil (ESO) contains approximately 4.5 epoxides per triglyceride.

Copolymerization of tert-butyl-3,4-epoxybutanoate and carbon dioxide

(S)-3,4-Dihydroxybutyric acid ((S)-3,4-DHBA) is a straight chain fatty acid and human urinary metabolite that results from the degradation of carbohydrates or the metabolism of γ-hydroxybutyrate, a naturally occurring compound found in the human central nervous system.54–57 Recently, (S)-3,4-DHBA has been obtained as a chiral biomass from the microbial (E. coli) glucose metabolism process.58,59 3,4-DHBA can be protected and subsequently converted to the corresponding epoxide for copolymerization with CO2 (Scheme 7).60,61 In a recent study, the synthesis of tert-butyl-3,4-epoxybutanoate (tBu-3,4-EB) was achieved by a two-step synthesis starting from 3-butenoic acid which was coupled with tert-butylacetate to afford tert-butyl-3-butenoate.60 Epoxidation of tert-butyl-3-butenoate with mCBPA provided the protected epoxide in good yield.
image file: c7gc02560b-s7.tif
Scheme 7 Synthesis of tert-butyl-3,4-epoxybutanoate and its copolymerization with CO2.

The copolymerization of tert-butyl 3,4-epoxybutanoate with carbon dioxide was achieved with high selectivity for copolymer employing a bifunctional cobalt(III) catalyst (Scheme 7).60 The copolymerization reaction was shown to proceed in a regioselective manner, with ring-opening of the sterically congested epoxide taking place at the methylene site. The tert-butyl groups protecting the pendent carboxylate substituents can be removed using trifluoroacetic acid. The resulting polymer can be coupled with aspartates, and the subsequent addition of [(DACH)Pt(H2O)2]2+ (DACH = (1R,2R)-diaminocyclohexane) provides platinum–polymer conjugates for use as possible anticancer pharmaceuticals. Of importance, the copolymer was shown to undergo degradation to human-friendly and environmentally benign biomasses including β-hydroxy-γ-butyrolactone and 3,4-dihydroxybutyrate.

Copolymerization of epoxy methyl 10-undecenoate and carbon dioxide

Over one million tonnes of castor oil are produced annually from the castor oil plant (Ricinus communis) for various uses in the chemical industry (2007).62 The major component of castor oil (>87%) is ricinoleic acid which can be converted to epoxy methyl 10-undecenoate (EMU) via a multistep process (Scheme 8).62–64 The coupling of EMU and CO2 has been accomplished with a Zn–Co(III) DMC heterogeneous catalyst.64 Due to the long chain of the EMU, the resulting copolymers had very low Tgs in the range of −38 to −44 °C, but decomposition temperatures remained typical of other polycarbonates with Td,5% = 240 °C. Notably, the copolymer contained >99% carbonate linkages when reaction temperatures were held at or below 80 °C. When temperatures were increased above 80 °C, polyether content increased, as is expected for DMC catalysts. The EMU/CO2 copolymers displayed two hydroxyl chain ends, and it was able to be further used as a macroinitiator to form a lactide-co-EMU/CO2-co-lactide triblock copolymer.
image file: c7gc02560b-s8.tif
Scheme 8 Synthesis of EMU and its copolymerization with CO2.

Copolymerization of epichlorohydrin and carbon dioxide

Epichlorohydrin, ECH, is a commodity chemical utilized in epoxy resins, paints, coatings, and adhesives.65,66 Annual production of ECH is approximately 2[thin space (1/6-em)]000[thin space (1/6-em)]000 tons per year.66 The traditional industrial synthesis of epichlorohydrin utilizes petrochemical propene as its starting material (Scheme 9).66 Dow's Glycerol-to-Epichlorohydrin (GTE)65 and Solvay S.A.'s Epicerol67 processes instead produce ECH from glycerol, the biorenewable byproduct of biodiesel synthesis. Combined annual production of ECH from glycerol at Solvay S.A.'s two plants is 100 tons per year.66
image file: c7gc02560b-s9.tif
Scheme 9 Traditional industrial production of ECH compared with the GTE and Epicerol® processes.

Inoue first coupled ECH and CO2 in the original 1969 paper using diethyl zinc and water, although less than 1% insoluble polymer product was produced after a 48 hours reaction time (Scheme 10).19 In 1994, Shen et al. utilized a set of rare earth phosphonates/triisobutylaluminum catalysts to produce poly(epichlorohydrin carbonate), PECHC, with high (70%) ether content.68 Zinc–cobalt(II) DMC heterogeneous catalysts will couple ECH and CO2 with up to 71% carbonate linkages and only 5 wt% cyclic epichlorohydrin carbonate byproduct being produced.69 Homogeneous cobalt(III) salen complexes have also been employed for the copolymerization of ECH and CO2.70 Perfectly alternating PECHC was produced with >99% selectivity at 0 °C using (salen)Co(2,4-dinitrophenoxide) with PPN(2,4-dinitrophenoxide) cocatalyst, although the molecular weight was low at Mn = 4700 g mol−1. Higher PECHC molecular weights were achieved by switching to a bifunctional cobalt (salen) catalyst featuring an appended quaternary ammonium salt, reaching Mn's of up to 25[thin space (1/6-em)]900 g mol−1. Relatively low temperatures of ≤25 °C are required in order to maintain the high selectivity, as the electron-withdrawing chlorine atom aids the production of cyclic carbonate byproduct.

image file: c7gc02560b-s10.tif
Scheme 10 Coupling of epichlorohydrin and CO2 to yield poly(epichlorohydrin carbonate).

Perfectly alternating, regioirregular poly(epichlorohydrin carbonate) is amorphous with a glass transition temperature of around 30 °C.69 In 2013, Lu, Darensbourg, and coworkers produced highly regio- and stereoregular PECHC utilizing bifunctional cobalt (salen) catalyst 4 (Fig. 7).71 The bulky adamantly group on 4 allows for the preferential attack of the methylene carbon of (R)-epichlorohydrin, retaining the (R)-stereochemistry in the polymer. Up to 97% retention was achieved, and the resulting regio- and stereoregular polymer had an increased Tg of 42 °C as well as the formation of crystalline domains with Tm = 108 °C.

image file: c7gc02560b-f7.tif
Fig. 7 Bulky bifunctional (S,S)-(salen)Co(III)X complex 4 utilized in the stereospecific copolymerization of (R)-epichlorohydrin and CO2. Here, X = 2,4-dinitrophenoxide (DNP). The bulky group helped to promote nucleophilic attack at the methylene carbon of (R)-epichlorohydrin, retaining the stereochemistry in the resulting polymer.

Copolymerization of furfuryl glycidyl ether and carbon dioxide

Furfural is a bioplatform molecule synthesized from feedstocks rich in hemicellulose including oats, sugarcane, and bagasse.72 It is estimated that over 400[thin space (1/6-em)]000 tonnes of furfural are produced every year, and it is currently employed as a biofuel, industrially to make furan resins, and as a starting material in the synthesis of other chemicals including furfuryl alcohol.72,73 Deprotonation of furfuryl alcohol with subsequent reaction with epichlorohydrin yields furfuryl glycidyl ether (FGE), an epoxide capable of reacting with CO2 (Scheme 11).74 The epichlorohydrin used in this process can also be produced from renewable resources (vide supra).
image file: c7gc02560b-s11.tif
Scheme 11 Synthesis of furfuryl glycidyl ether and tetrahydrofurfuryl glycidyl ether.

The copolymerization of FGE and CO2 has been accomplished with both a rare earth Y(ClCl3OO)3-ZnEt2-glycerine catalyst74 as well as (salen)CoCl catalyst with added PPNCl cocatalyst (Scheme 12).75 The pendant furan ring is susceptible to discoloring crosslinking reactions, but this process can be slowed through the addition of antioxidants to the copolymer.74 Further stability can be imparted by instead substituting the fully hydrogenated tetrahydrofurfuryl glycidyl ether (THFGE) in the copolymerization with CO2. Reported Tgs for the copolymer of CO2 with FGE and THFGE were −10 °C75 and −6 °C,74 respectively.

image file: c7gc02560b-s12.tif
Scheme 12 Copolymerization of FGE and CO2.

The diene of the FGE/CO2 copolymer can undergo Diels–Alder reactions with different dienophiles to again impart added stability to the copolymer.74,75 Frey and coworkers produced terpolymers of FGE, methyl glycidyl ether, and CO2 and later used these terpolymers and the FGE/CO2 copolymer to produce insoluble crosslinked networks with improved thermal properties (Tg = 93–96 °C).75 The crosslinks were fully reversible via a retro-Diels–Alder reaction, reproducing the soluble, non-crosslinked polymer.

Copolymerization of 1,2-epoxy-4-cyclohexene and carbon dioxide

The precursor compound 1,4-cyclohexadiene can be obtained as a waste product in the self-metathesis of some polyunsaturated fatty acids derived from renewable plant oils (e.g., linoleic acid).76–78 Epoxidation of 1,4-cyclohexadiene can be readily achieved using either mCPBA or oxone as oxidants to yield 1,2-epoxy-4-cyclohexene (CHDO) (Scheme 13).79 Alternatively, this epoxide can be hydrogenated to cyclohexene oxide (CHO), a well-studied epoxide monomer for coupling with CO2 to selectively provide poly(cyclohexene carbonate).
image file: c7gc02560b-s13.tif
Scheme 13 Synthesis of CHDC and CHO and their subsequent copolymerizations with CO2.

Recently, there have been several reports of the copolymerization of CHDO with CO2 to provide poly(cyclohexadiene carbonate) (PCHDC).79–81 A beneficial quality of using this epoxide, in addition to it being derived from a renewable resource, is that it has an olefinic group for easy post-polymerization functionalization via thiol–ene click chemistry (Scheme 14).82 This feature allows for the modification of the main-chain rigidity of poly(cyclohexadiene carbonate), thereby affecting its physical properties such as toughness and glass transition temperature.

image file: c7gc02560b-s14.tif
Scheme 14 Copolymerization of CHDO and CO2 with subsequent derivatization using thiol–ene click chemistry.

Several catalyst systems have been employed for this copolymerization process. These include chromium(III) and cobalt(III) binary (salen)MX/onium salt catalysts,79,80,83 a bifunctional (salen)CrX catalyst,80 dinuclear zinc and magnesium catalysts,79 and a (porphyrin) CoCl/DMAP catalyst system.81 The latter catalyst was shown to be considerably less effective than the other transition metal systems, where in this instance the reaction time at 40 °C was over 500 hours. Similarly, the dinuclear zinc and magnesium complexes exhibited poor activity for the copolymerization of 1,2-epoxy-4-cyclohexene and carbon dioxide.79

Specifically, the (salen)CoX complexes along with the corresponding PPNX cocatalysts, where X = dinitrophenoxide or chloride, were found to show comparable catalytic effectiveness at 28–40 °C and 2.0 MPa CO2, providing high conversions over a few hours.80 In all instances, polymer selectivity was greater than 99%, affording moderately high molecular weight copolymers with low polydispersities (PDI = 1.10–1.30). The analogous binary (salen)CrN3/PPNN3 system operating at 80–90 °C exhibited much lower copolymer selectivity, typically 37–89% depending on reaction temperature and time. Both cis- and trans-cyclic carbonate isomers were formed in the process, where polymer degradation afforded exclusively the trans-cyclic carbonate. On the other hand, the analagous bifunctional (salen)CrN3 system was 100% selective for copolymer but was less active that its binary counterpart at 90 °C.

Terpolymerization of 1,2-epoxy-4-cyclohexene/cyclohexene oxide/CO2 employing the binary (salen)CoCl/PPNCl catalyst system has also been investigated.79 That is, a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 mixture of the two epoxide monomers in the presence of 2.0 MPa of CO2 at 28 °C provided a terpolymer where the percent conversions of the two monomers was 22/72%, indicative of faster incorporation of the saturated epoxide monomer. A terpolymer with Mn of 11[thin space (1/6-em)]500 g mol−1 and PDI = 1.12 was isolated. In the same work, 1,2-epoxy-4-cyclohexene oxide has also been shown to undergo copolymerization with phthalic anhydride (PA) in the presence of a dinuclear magnesium catalyst as well as (salen)MCl (M = Cr(III) and Co(III)) to provide a perfectly alternating polyester (Scheme 15).

image file: c7gc02560b-s15.tif
Scheme 15 Alternating copolymerization of CHDO and phthalic anhydride.

As mentioned earlier, the presence of the double bond in poly(cyclohexadiene carbonate) offers the possibility for post polymerization modification of the copolymer's backbone via thiol–ene click chemistry. For example, the functionalization of poly(cyclohexadiene carbonate) has been achieved using AIBN (azobisisobutyronitrile) radical-initiated addition of thioglycolic acid (Scheme 16).80 It should be noted that these thiol–ene click reactions occur with anti-Markovnikov regioselectivity.82 Upon deprotonation of the amphiphilic copolymer with an aqueous solution of NH4OH, a water-soluble polymer is produced as demonstrated by dynamic light scattering (DLS) analysis.80

image file: c7gc02560b-s16.tif
Scheme 16 Derivatization of PCHDC.

Modification of poly(cyclohedadiene carbonate) by postpolymerization reactions or by incorporation of an alternate epoxide can lead to small changes in the polymer's thermal properties. The Tg values in several of these modified polymers are summarized in Table 1. For example, the Tgs of the unsaturated poly(cyclohexadiene carbonate) and its saturated analogue poly(cyclohexene carbonate) are essentially the same. On the other hand, the addition of thioglycolic acid to the unsaturated copolymer's backbone led to a drop in Tg by 26 °C.83 Further, upon deprotonation of the acid function with NH4OH the Tg rebounds to 120 °C. A dibrominated PCHDC polymer has also been synthesized with an unexpected low Tg value of 36 °C. Additionally, a terpolymer of 1,4-cyclohexadiene oxide/cyclohexene oxide/CO2 exhibited a Tg of 112 °C, whereas a copolymer of 1,4-cyclohexadiene and phthalic anhydride displayed a Tg value of 128 °C. Notably, the addition of the syn- and/or anti-forms of cyclohexadiene diepoxide in these epoxide/CO2 copolymerizations led to insoluble, crosslinked materials.79

Table 1 Glass transition temperatures of PCHDC-derived polycarbonates
Polymer T g (°C)
a This value is reported for a Mn value of 8700 g mol−1, whereas, the higher values are for Mns of 11[thin space (1/6-em)]500 g mol−1 and 12[thin space (1/6-em)]900 g mol−1.
PCHDC 116 (ref. 80), 115 (ref. 79), 112a (ref. 81)
PCHDC_COOH 90 (ref. 80)
PCHDC_COONH4 120 (ref. 80)
CHDO/CHO/CO2 112 (ref. 79)
CHDO/PA 128 (ref. 79)
PCHDCBr2 36 (ref. 81)

As reported in the literature, it is possible to isomerize 1,4-cyclohexadiene to its 1,3-cyclohexadiene isomer.76 Furthermore, the transformation of this alternative diene to its epoxide has been effectively carried out (Scheme 17). Subsequent copolymerization of 1,3-cyclohexadiene oxide with carbon dioxide has provided a facile route to the copolymer, poly(1,3-cyclohexadiene carbonate).83 Noteworthy, under identical catalytic conditions, the 1,3-cyclohexadiene oxide/CO2 copolymerization process occurs at a significantly faster rate than its 1,4-analogue. This enhancement in reactivity was also seen in the terpolymerization reaction of 1,3-cyclohexadiene oxide/propylene oxide/CO2. The Tg of poly(1,3-cyclohexadiene carbonate) was found to be 104–108 °C, 10 °C lower than that in its 1,4-copolymer counterpart. A terpolymer composed of a 1[thin space (1/6-em)]:[thin space (1/6-em)]1.2 (PO/1,3-CHDO) monomer ratio displayed a Tg value of 69 °C, approximately midway between Tgs of the respective copolymers.

image file: c7gc02560b-s17.tif
Scheme 17 Synthesis of 1,3-cyclohexadiene and its copolymerization with CO2.

Copolymerization of α-pinene oxide and carbon dioxide

For many years, the monoterpenes α-pinene and limonene have both greatly interested CO2/epoxides copolymerization researchers as promising precursors to yield polycarbonates produced from fully renewable sources. As the name suggests, α-pinene has a pine-like odor and is one of the major components of pine tree oil.84 To the best of our knowledge, there are no academic articles that have reported the successful coupling of α-pinene oxide with CO2. In fact, we were only able to find one Chinese patent that briefly details the production of poly(α-pinene carbonate) from α-pinene oxide/CO2 copolymerization utilizing (salen)CrCl/PPNCl (Scheme 18).85
image file: c7gc02560b-s18.tif
Scheme 18 Copolymerization of α-pinene and CO2.

Copolymerization of limonene oxide and carbon dioxide

On the other hand, much more work has been accomplished with limonene. (+)-Limonene is a major component of citrus oils and has a distinct orange odor whereas (−)-limonene is produced by oak and pine trees and smells like turpentine.84,86 In 2013, approximately 70[thin space (1/6-em)]000 tons of (+)-limonene was produced globally.87 As it contains a chiral center, two functionalizable double bonds, and is recognized as a safe substance by the US Food and Drug Association, limonene is quickly being recognized as an important biochemical.88,89 The excellent review by Lopez-Sanchez and Pagliaro provides a detailed analysis of limonene's emerging importance to the bioeconomy.87

Relevant to this study is limonene's 1,2-epoxidation product, limonene oxide (LO), which has a strong resemblance to the well-studied cyclohexene oxide monomer.90 Following epoxidation, there are four possible isomers of limonene oxide, and, for simplicity, we will label and discuss them as shown in Fig. 8. Alternate names for each isomer are also provided. Commercially-available limonene oxide is sold in either its (R)- or (S)-forms which are each mixtures of cis- and trans-isomers. (R)-LO is significantly cheaper due to the higher commercial availability of (+)-limonene. GC-MS analyses of (R)-LO91 and (S)-LO92 purchased from Sigma-Aldrich have revealed that both consist of approximately 54% trans-isomer. Other unidentified commercial sources of (R)-LO have been reported as containing 60% trans-isomer.93,94 Often however, studies desire a single isomer to yield a polymer with high regio- or stereoregularity. Kinetic resolution utilizing either pyrazole or pyrrolidine affords trans- or cis-LO, respectively, with >98% purity.95 Separately, Rieger and Greiner produced an enriched sample of trans-(R)-LO via a two-step process wherein (R)-(+)-limonene is first reacted with N-bromosuccinimide to produce the trans-bromohydrin (Scheme 19).96 The trans-bromohydrin is then ring-closed using aqueous NaOH. This method was scaled up in a 10 L reactor, yielding 1.3 kg of 85% trans-(R)-LO per batch (5% cis-(R)-LO, 10% unreactive side products).

image file: c7gc02560b-f8.tif
Fig. 8 Isomers of limonene and limonene oxide.

image file: c7gc02560b-s19.tif
Scheme 19 Synthesis of limonene oxide via bromohydrin intermediate.

The first report of the successful copolymerization of limonene oxide and CO2 came in 2004 from Coates et al. (Scheme 20).91 By employing β-diiminate (BDI) zinc catalyst [(BDI)ZnOAc], 5, and relatively mild conditions (25 °C, 0.69 MPa CO2), poly(limonene carbonate), PLC, was afforded with Mn = 9300 g mol−1 and PDI = 1.13 after 2 hours (Fig. 9). While the catalytic activity was moderate (TOF = 32 h−1), no production of cyclic limonene carbonate byproduct or polyether linkages were reported. The catalyst's productivity was increased by incorporating an electron-withdrawing cyano group in the catalyst's imine backbone (6), reaching TOF = 37 h−1, Mn = 10[thin space (1/6-em)]800 g mol−1, and PDI = 1.12. Both catalysts preferentially reacted with trans-(R)-limonene oxide over the cis-(R)-isomer (% trans- in copolymer >98%).

image file: c7gc02560b-s20.tif
Scheme 20 Copolymerization of limonene oxide and CO2.

image file: c7gc02560b-f9.tif
Fig. 9 Various [(BDI)ZnX] catalysts utilized in the copolymerization of limonene oxide and CO2.

Following this early triumph, further published research in PLC production stalled for over a decade. Starting in 2015, a torrent of new works has been published on the synthesis and uses of PLC from research groups across the globe. PLC can be produced with catalyst 5 in batches on the kilogram-scale, yielding polymers with Mn's in excess of 100[thin space (1/6-em)]000 g mol−1.96 Despite the successes with 5, many groups switched to [(BDI)ZnN(SiMe3)2] 7 where the acetate group was replaced by the hydrolysable N(SiMe3)2, allowing all PLC chain ends to be hydroxyl-terminated.92,97–101 Related complex 8 displaying two electron-withdrawing CF3 groups in the imine backbone has shown the highest activity for PLC production to date, reaching a maximum TOF of 310 h−1 at 60 °C.102 Preferential reactivity toward trans-(R)-LO isomer is observed for all Zn-based systems.

In 2015, Kleij and coworkers developed amino-trisphenolate aluminum complex 9 that is capable of the copolymerization of LO and CO2 in the presence of PPNX (X = Cl, Br) cocatalysts (Fig. 10).93 Catalyst 9 with added PPNCl cocatalyst produces PLC under relatively mild conditions (42 °C, 1.0 MPa CO2). The polymerization is much slower than with [(BDI)ZnX] catalysts however, reaching 60% conversion after 48 hours. Molecular weights of up to 10[thin space (1/6-em)]600 g mol−1 were achieved (PDI = 1.43). Notably, whereas [(BDI)ZnX] catalysts preferentially react with trans-(R)-LO,919/PPNCl can react with either isomer but reacts faster and with greater selectivity toward cis-(R)-LO.93 Like [(BDI)ZnX] catalysts,96 the 9/PPNX system is believed to operate via a bimetallic mechanism.93,103 It should also be noted that, to the best of our knowledge, there have not been any published studies of LO/CO2 copolymerization utilizing the well-studied binary and bifunctional cobalt(III) and chromium(III) salen catalysts.

image file: c7gc02560b-f10.tif
Fig. 10 Kleij's amino-trisphenolate catalysts. For 9, M = Al. For 10, M = Fe.

Chain-transfer has long been an issue of CO2/epoxides copolymerization reactions, as it can decrease both polymer molecular weight and polydispersity. Even with stringent drying techniques to reduce water present in the limonene oxide and carbon dioxide monomers, other chain transfer agents (CTAs) may still exist. GC-MS analysis of commercially-available LO reveals a small yet significant amount of hydroxyl-containing impurities with high (>170 °C) boiling points (Fig. 11).96 Neither distillation nor column chromatography are effective at removing these OH-impurities. Instead, O-methylation can be employed to render them inactive (Scheme 21). After masking the hydroxyl groups, Zn catalyst 5 coupled (R)-LO and CO2 to achieve PLC with molecular weights of up to 108[thin space (1/6-em)]600 g mol−1 with PDI = 1.13 (Mn, theoretical = 78[thin space (1/6-em)]400 g mol−1). For comparison, a similar reaction without the same pretreatment resulted in PLC with Mn = 17[thin space (1/6-em)]100 g mol−1 (Mn, theoretical = 86[thin space (1/6-em)]500 g mol−1). CTAs can also be purposefully added to the reaction to generate low molecular weight polyols in a controlled manner for use in polyurethane synthesis (vide supra). However, CTAs appear to hinder PLC polyol production. When water and 1,3-propanediol were added to catalyst 7 (Zn[thin space (1/6-em)]:[thin space (1/6-em)]CTA ratios ranged from 1[thin space (1/6-em)]:[thin space (1/6-em)]10 to 2[thin space (1/6-em)]:[thin space (1/6-em)]1), the polymerization reaction was completely inhibited, likely due to catalyst decomposition.92

image file: c7gc02560b-f11.tif
Fig. 11 OH-containing impurities in commercial limonene oxide as identified by GC-MS.96

image file: c7gc02560b-s21.tif
Scheme 21 Removal of limonene oxide CTAs via O-methylation.

Notably, the anticipated byproduct of cyclic limonene carbonate (cLC) is rarely observed in the coupling reaction of LO and CO2. To the best of our knowledge, only two reports detail the synthesis of cyclic limonene carbonate from LO and CO2. Kleij and coworkers reported the selective formation of cLC when cocatalyst PPNCl was used alone in the absence of catalyst at 90 °C.93 In the presence of tert-butyl-substituted-9/PPNCl, cLC can be produced from LO and CO2 at 85 °C with >99% selectivity (1.0 MPa CO2, 66 h).94 This tBu9/PPNCl system shows greater activity toward the trans-epoxide over the corresponding cis-monomer when producing cLC (73% conversion for trans-LO versus 4% conversion for cis-LO), in direct contrast of 7/PPNCl's preference for cis-LO when selectively producing PLC.93trans-(R)-LO reacts with CO2 to form the cis-cyclic carbonate isomer by undergoing a double inversion pathway (Fig. 12).104 The higher temperatures required for the formation of cLC are consistent with previous mechanistic studies that show its formation goes through a higher energy pathway than for that of PLC production.93

image file: c7gc02560b-f12.tif
Fig. 12 Thermal ellipsoid plot of cyclic limonene carbonate.94

Thermal decomposition of PLC yields a variety of products including limonene oxide, limonene, and various alcohols and diols derived from limonene oxide.102 Cyclic limonene carbonate is only observed as a product when maleic anhydride end-groups are employed to cap the polymer chains. Chemical decomposition of PLC was first studied by Koning in 2015.92 When the strong organic base 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD) and a polyol are added to a sample of PLC dissolved in toluene at 100 °C (Mn = 13[thin space (1/6-em)]800 g mol−1, PDI = 1.2), the major products of the reaction included oligomeric PLC (Mn < 4000 g mol−1, PDI 1.5–2.0), 1-methyl-4-(prop-1-en-2-yl)cyclohexane-1,2-diol, and trans-(R)-LO. This base-catalyzed depolymerization of PLC was further investigated in 2017 utilizing 1H NMR.105 High temperatures (≥100 °C) are required to initiate the depolymerization reaction. TBD extracts the hydrogen off the hydroxyl-terminated chain end, and a backbiting reaction commences, however no cyclic limonene carbonate is observed. Notably, cis-LO is a major product early in the depolymerization despite its limited incorporation in PLC catalyzed by Zn complexes (vide supra). As the reaction continues, cis-LO's peaks stabilize, and trans-LO's continue to grow. This indicates that the polymer displays high cis-LO incorporation at the start of the OH-terminated end while high trans-LO incorporation follows through the rest of the PLC polymer. This idea differs slightly from observations in limonene dioxide/CO2 copolymerization, where cis-isomer incorporation is believed to happen later in the polymerization when much of the trans-epoxide has already been consumed.97 Other CO2-based polycarbonates including poly(cyclopentene carbonate)106–108 and poly(indene carbonate)107 also generate their corresponding epoxides when undergoing base-initiated depolymerization, but their cyclic carbonate byproducts are still formed in varying amounts. Whereas the production of both cyclopentene oxide and indene oxide can be limited by increasing the pressure of CO2 in the depolymerization reaction, an increase in PCO2 to 4.0 MPa did not affect the depolymerization products of PLC.105

Koning suggests that the unique depolymerization characteristics of PLC are due to PLC's tetrasubstituted cyclohexane rings.105 That is, on a single polymeric repeat unit, the 4-isopropenyl is the most sterically encumbering group, and it thus preferentially occupies an equatorial position (Fig. 13). Separate energy calculations by Auriemma and Coates affirm this equatorial designation if a chair conformation of the cyclohexane ring is assumed.99 Once deprotonated by TBD, the axial alkoxide chain end is conformationally blocked from attacking the vicinal carbonate carbon to produce cyclic limonene carbonate.105 Instead, the alkoxide attacks the adjacent carbon, eliminating limonene oxide and producing a carbonate-terminated polymer chain. Carbon dioxide is then released, and the backbiting reaction propagates to produce more LO and CO2. In further support of the crucial role of the 4-isopropenyl group in directing LO's depolymerization mechanism, the depolymerization of poly(1-methyl-cyclohexene carbonate) under the same conditions yields 93% cyclic 1-methyl-cyclohexene carbonate and only a small amount of 1-methyl-cyclohexene oxide.105

image file: c7gc02560b-f13.tif
Fig. 13 Depolymerization of PLC to generate (a) trans-(R)-LO and (b) cis-(R)-LO.105 Traditional backbiting to yield cyclic limonene carbonate is conformationally blocked in for both isomers.

Similarly, Rieger recently identified that an equilibrium exists between PLC and the limonene oxide/CO2 monomers (Scheme 22).102 By changing the temperature of the polymerization reaction between 25–60 °C, the ratio of monomers to polymer in solution could be tuned. Again, no cLC was observed via either NMR or IR, consistent with Koning's proposed mechanism.105 At an LO concentration of approximately 3.5 M, the polymerization's ceiling temperature109 is believed to be around 60 °C.102 Poly(limonene carbonate) thus joins the copolymer of CO2 and 1-benzyloxycarbonyl-3,4-epoxy pyrrolidine110 as the two known examples of fully recyclable polycarbonates produced from CO2 and epoxides coupling.

image file: c7gc02560b-s22.tif
Scheme 22 Dynamic equilibrium between PLC and its monomers.

Enantiopure poly(limonene carbonate) is an amorphous polymer.100 The thermal properties reported for PLC have varied widely from paper to paper. Glass transition temperatures have been reported from as low as 71 °C (Mn = 3700 g mol−1, PDI = 2.0)92 to up to 130 °C (Mn = 53[thin space (1/6-em)]600 g mol−1, PDI = 1.10).96 Thermal decomposition has been reported to onset at temperatures as low as 180 °C,93 but onset temperatures of up to 249 °C have been reached in other studies.91 End-capping PLC with agents including tetraethyl orthosilicate or various anhydrides can increase the decomposition temperatures by ∼15 °C.96 A table of the available data reported for PLC thermal properties can be found in the ESI.

A full analysis of PLC's mechanical performance is included in Rieger and Greiner's excellent 2016 Green Chemistry paper.96 Poly(limonene carbonate) has a density of 1.09 g cm−3, similar to polyethylene and polypropylene. Mechanical properties of PLC are believed to be between that of polystyrene and BPA-polycarbonate, displaying E-modulus of 0.95 GPa, 15% elongation at break, and a pencil hardness of B.

Networked polyurethane coatings generated with PLC polyols of MW ∼2000–4000 g mol−1 displayed intermediate solvent resistance (visible damage after 15–35 acetone rubs) and poor mechanical properties (e.g., cracking during reverse impact testing; 100 cm, 1 kg).92 These inferior properties are likely due to incomplete network formation by sterically-hindered secondary and tertiary alcohol chain ends of the PLC polyols. If primary alcohol linkers are instead utilized, PLC-based PU coatings with improved properties can be formed (vide infra).98

Amorphous poly(limonene carbonate) has extremely good optical properties, with light transmission of ∼95% from 400–1000 nm (240 μm thick film).96 A terpolymer of CO2, LO, and PO containing approximately 50/50 incorporation of PLC and PPC again displays good light transmission of 95–99% in the visible range (400–800 nm, 0.70 mm thick film).102 In addition to this high transparency to light, PLC also shows good permeability to gases such as CO2 and O2.111 It is approximately ten times more permeable to gases than BPA-polycarbonate and one thousand times more permeable to gases than poly(methyl methacrylate), both materials used as glass substitutes for window panes. PLC “breathing glass”111 may have uses in homes or greenhouses where this type of gas permeability is desired.

Poly(limonene carbonate) shows some resistance to biodegradation. No changes were observed in PLC quality after 60 days in active compost at 60 °C, whereas poly(L-lactic acid) degrades within 1 week under the same conditions.112 A main contributing factor to the slow degradation of PLC could be its hydrophobicity (contact angle to water = 94°).112 Several post-polymerization additions to the 8,9-olefin have yielded PLC derivatives with increased hydrophilicity (vide infra, contact angle to water 60–80°), but again no changes in molecular weight, mass loss, or visible characteristics in the polymer samples were observed after four weeks in degradation studies at 37 °C under acidic, basic, or enzymatic (esterase) conditions.112 Despite this, researchers still believe that the available carbonate functionality in PLC would lead to it having faster degradation times than commonly employed polymers including PE and PP.

Even when care has been taken to produce regio- and stereoregular poly(R-limonene carbonate) (PRLC) or poly(S-limonene carbonate) (PSLC), PLC remains amorphous. All known attempts to date to crystallize the pure enantiomers PRLC and PSRC have failed.99 However, when the equal amounts of PRLC and PSLC are co-precipitated from hexane solution, a racemic crystalline stereocomplex (SCPLC) is formed (Fig. 14).99,100 SCPLC has the same Tg as both PRLC and PSLC (∼120 °C), but its decomposition temperature is 265 °C, approximately 15 degrees higher than that of either pure polymeric enantiomer.100 The melting temperature of SCPLC is unknown, as the polymer degrades before the endotherm is observed. A detailed structural analysis of SCPLC has recently been published.99 In forming the stereocomplex, the independent PRLC and PSLC enantiomers come together to form a “steric zipper,” owing largely to the alignment of the carbonyl carbons between alternating PRLC and PSRC layers.99,100 Calculations indicate that the barriers for crystallization of pure PRLC and PSLC are too high, and racemic crystals instead form because of a greater kinetic ease.99

image file: c7gc02560b-f14.tif
Fig. 14 Copolymerization of (R)-LO and (S)-LO to yield enantiomerically pure, regio- and stereoregular PLC copolymers. These are then blended in a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 ratio to form the semicrystalline SCLC stereocomplex.

The remaining pendant olefin on limonene oxide can be modified either pre- or post-polymerization to yield PLC derivatives with a wide range of properties (Fig. 15). The outstanding 2016 Nature Communications article from Greiner et al. details many of these post-polymerization modifications to PLC.112 For example, the hydrophilicity of hydrophobic PLC can be increased by adding polar side chains across the double bond.112 PEG-3-OH has been successfully added with 18% conversion to produce PLC-PEG.112 Higher conversions are limited by the concomitant degradation of the polymer backbone that occurs in the acid-catalyzed reaction.

image file: c7gc02560b-f15.tif
Fig. 15 Modifications of limonene, limonene oxide, and poly(limonene carbonate).97,98,112–114 For simplicity, no stereochemistry has been included.

Radical thiol–ene click chemistry can afford many different poly(limonene carbonate) derivatives with thioether linkages (PLC-TEs) without degradation of the polymer backbone. The addition of mercaptoacetic acid across the double bond (PLC-TE_MAc) produces a hydrophilic polymer that readily dissolves in basic (pH > 8) solution.112 Addition of butyl-3-mercaptopropionate (B3MP) with subsequent polymer curing instead produces PLC-TE_B3MP, a rubbery material with Tg = 5 °C, Young's modulus = 1.0 MPa, tensile strength = 6.8 MPa, and percent elongation at break = 228%.112 Each of these properties are tunable by controlling the percent conversion of the B3MP addition reaction. A polymer with pendant tertiary amines can be produced by adding 2-(diethylamino)ethanthiol hydrochloride across the olefin, and further reaction with benzyl bromide affords PLC-NQ with a pendant quarternary amine. PLC-NQ (3% amination, 20% quaternization) displays antibacterial activity versus E. coli after 12 hours of contact.112 Separately, Koning and coworkers used thiol–ene click chemistry to add both 2-mercaptoethanol and 6-mercaptohexanol across the pendant olefin bond of PLC.98 In direct contrast with their earlier study,92 these PLC derivatives displaying available primary alcohols could be successfully incorporated into curable polyurethane coatings.98 Several hard (pencil hardness 2H) yet brittle PLC-PU coatings were produced. Very recently, Koning reported the synthesis of other novel PLC-based thiol–ene networks with favorable properties for coating applications.115

Chemical modifications to PLC can also occur at earlier stages of the production process. For example, hydrogenation of the limonene precursor utilizing Pt/C selectively reduces the 8,9-alkene to yield menth-1-ene. Following epoxidation, [(BDI)ZnOAc] catalyst 5 has been used to couple menth-1-ene oxide with CO2 to yield poly(menth-1-ene carbonate) (PMC), an amorphous thermoplastic with Tg = 130 °C, Td,5% = 240 °C.112 Both menth-1-ene oxide and PMC should have increased long-term stability over LO and PLC, respectively, as neither can crosslink or undergo deleterious oxidation reactions.

Full epoxidation of limonene yields limonene dioxide (LDO),114 a diepoxide additive used in UV-curing applications both industrially and academically.116–120 LDO can be fully coupled with CO2 at high temperatures (120–140 °C) to yield limonene dicarbonate which can be used in polyurethane production.121,122 Alternatively, catalyst 7 preferentially reacts with 1,2-epoxide of the trans-LDO monomer to form poly(limonene-8,9-oxide carbonate) (PLOC, % trans- in copolymer >90%). This activity is consistent with 7's performance in LO/CO2 coupling91,92,98–100 and its known inactivity in propylene oxide/CO2 coupling reactions.123 No branching, cross-linking, or polyether linkages were observed, though there was limited (<5%) formation of 5-membered cyclic carbonate at the pendant 8,9-epoxides.979/PPNCl is also active for LDO/CO2 coupling, though it produced mainly low molecular weight oligomers and some cyclic carbonate species.97 Separately, Kleij et al. produced PLOC via a two-step process, wherein previously synthesized PLC was epoxidized to PLOC with mCPBA.113 Compared to PLC, PLOC displays a slightly increased Tg of 135 °C.97

Like PLC, PLOC can be recognized as a platform for various post-polymerization modifications (Fig. 15). Poly(limonene dicarbonate) (PLDC) is formed by coupling PLOC's remaining epoxide with CO2. PLDCs prepared with different molecular weights, PDIs, and percent conversion of the pendant cyclic carbonates have displayed Tgs of 145–180 °C and Td,5%s of 200–238 °C.97,113 PLOC's pendant epoxide can be ring-opened by various nucleophiles including thiols (PLC-TA) and carboxylic acids (PLC-CA) to yield PLOC derivatives with varying physical and mechanical properties.97 A summary of the thermal properties of selected PLC-derived polymers can be seen in Table 2.

Table 2 Glass transition temperatures of selected PLC-derived polycarbonates
Polymer T g (°C)
a See ESI for more detailed PLC Tg data. b Observed Tg ranges due to differing conversions of post-polymerization reactions. c Observed Tg ranges due to differing polymer molecular weights as well as conversions of post-polymerization reactions. d Degree of functionalization of amine = 23%. e Degree of functionalization of amine = 3%, degree of amine quaternization = 20%.
PLC Ranges from 71 (ref. 92)–130 (ref. 96), 120a (ref. 91)
PMC 130 (ref. 112)
PLOC 135 (ref. 97)
PLDC 145–180b (ref. 97 and 113)
PLC-PEG 91–129b (ref. 112)
PLC-TA_ME 118 (ref. 97)
PLC-TA_DDT 13 (ref. 97)
PLC-CA_AA 39 (ref. 97)
PLC-CA_HA 46 (ref. 97)
PLC-TE_ME 91–117 (ref. 97 and 112)
PLC-TE_MH 56–94c (ref. 112)
PLC-TE_MAc 82 (ref. 112)
PLC-TE_B3MP 5 (ref. 112)
PLC-TE_N ca. 95d (ref. 112)
PLC-NQ ca. 105e (ref. 112)

Several groups have catalytically coupled limonene oxide with different cyclic anhydrides to produce polyesters (Scheme 23). Coates et al. used an asymmetric [(BDI)ZnOAc] catalyst to couple LO with both diglycolic anhydride and maleic anhydride to produce polyesters with Tg = 51 °C and 62 °C, respectively.124 Poly(limonene oxide-alt-phthalic anhydride) has been produced in many different studies with glass transition temperatures ranging from 12.3125 to 135 °C.126 Duchateau et al. successfully added CTAs to their LO/PA copolymerization to produce various shorter-chain poly(LO-alt-PA) polyols.127

image file: c7gc02560b-s23.tif
Scheme 23 Copolymerization of cyclic anhydrides and limonene oxide.

Kleij et al. determined that substituting an iron(III) metal center into the amino-trisphenolate ligand provided 10 which displays higher activity and selectivity for polyester formation.126 Catalyst 10 with added PPNCl cocatalyst successfully coupled LO with PA and also 1,8-napthalic anhydride, producing a polyester with Tgs up to 243 °C. As was observed in their LO/CO2 coupling studies, 10 showed preferential reactivity toward cis-LO monomer, producing polyesters with higher molecular weights and higher glass transition temperatures. For example, 10 coupled LDO with PA to produce a polyester with Tg = 53–59 °C. Again, the pendant olefin represents a rich site for any desired post-polymerization modifications to the preformed polyesters.

Copolymerization of 1-benzyloxycarbonyl-3,4-epoxy pyrrolidine and carbon dioxide

In 2017, Liu, Lu, and coworkers reported the novel copolymerization of epoxide monomer 1-benzyloxycarbonyl-3,4-epoxy pyrrolidine (BEP) with CO2 utilizing several dinuclear (salen)CrX catalysts 11 with added PPNX cocatalysts (Fig. 16 and Scheme 24).110 BEP is a renewable epoxide, as it can be produced from the biorenewable platform chemical furfural72,73via a six-step process (31% overall yield, unoptimized).110 At temperatures of 25–80 °C, the 11/PPNX system produced PBEP with >99% selectivity for polymer regardless of which anion X was employed (X = F, Cl, NO3). The dinuclear framework is important for the high polymer selectivity, as cyclic BEP carbonate production jumped to 34% when the utilizing the mononuclear (salen)CrCl/PPNF system. None of the analogous cobalt(III) catalysts were able to couple BEP and CO2.110
image file: c7gc02560b-f16.tif
Fig. 16 Tethered dinuclear (salen)CrX catalyst 11, where anion X = F, Cl, or NO3.

image file: c7gc02560b-s24.tif
Scheme 24 Synthesis of BEP and its coupling reaction with CO2.

Notably, while the selectivity for PBEP remained both high and consistent, the activity of the catalyst system decreased as temperature was increased. That is, TOF decreased from 42 h−1 at 60 °C to 13 h−1 at 80 °C for Cl11/PPNF.110 At 100 °C, no coupling or degradation (e.g., diol) products were observed, and only BEP epoxide was found in the reaction flask. The BEP/CO2 coupling reaction was followed using in situ ATR-FTIR along with temperature modulation. Subsequent growth and decay of the polymer peak at 1755 cm−1 could be observed as the temperature of the reaction was cycled between 60 °C and 100 °C, respectively. Both 1H NMR and FTIR confirmed that only BEP monomer and PBEP were present in the reaction flask; no cyclic BEP carbonate or diol byproduct were produced. As such, the coupling of BEP with CO2 is completely reversible based only upon temperature control (Scheme 25), similar to separate results reported for poly(limonene carbonate) (vide supra).102 Complete thermal (150 °C) depolymerization of isolated PBEP back to CO2 and BEP can also be accomplished under N2-atmosphere without any added catalyst, so long as no end-capping agents are employed.110 Backbiting is believed to take place from the chain end, and DFT calculations indicate that the backbiting reaction to epoxide (ΔG = 12.3 kcal mol−1) is energetically preferential to backbiting to either cis- (ΔG = 14.5 kcal mol−1) or trans-cyclic carbonate (ΔG = 25.6 kcal mol−1). This is likely due to increased ring strain from the two fused five-membered rings. Notably, however, when the PBEP chains are end-capped using acetic anhydride, a wide range of degradation products are produced.

image file: c7gc02560b-s25.tif
Scheme 25 Temperature-dependent copolymerization and depolymerization of poly(BEP carbonate).

The use of impure sources of CO2

For nearly all the studies involving epoxides and CO2 coupling, high purity carbon dioxide is employed directly from a gas cylinder. For example, Research Grade CO2 supplied from AirGas is 99.999% and is quoted as containing only 3 ppm H2O.128 However, as water is a known chain-transfer agent, several research groups take further measures to dry their CO2 by passing it through molecular sieves prior to adding it to the reaction flask. This highly purified carbon dioxide does not necessarily fit the mold of “waste CO2” that is so often cited, as several steps were necessary to reach this level of purification.

The excellent review by Aresta and coworkers highlights advances made utilizing exhaust CO2 in a wide variety of applications.129 Of note, a growing number of studies have considered the use of impure CO2 as a starting material to be coupled with epoxides. North and coworkers have reported using bimetallic aluminum salen complex 12 with nBu4NBr cocatalyst to successfully couple ethylene oxide, propylene oxide, and styrene oxide with CO2 produced from the sublimation of dry ice pellets (Fig. 17).130 Pressures of 0.1–0.8 MPa CO2 were achieved, but the use of the solid CO2 pellets also concomitantly introduced water into the reaction flask. Fortunately, the 12/nBu4NBr system was tolerant of this small amount of water and achieved good yields at 0–25 °C. However, the production of cyclic carbonates was completely hindered in the presence of excess added water (33 and 100 mol%). Complex 12 and other silica- and polymer-supported aluminum salen catalysts including bimetallic and bifunctional catalyst 13 have produced cyclic carbonates by coupling epoxides and CO2 doped with NOx and SOx contaminants131 as well as methane from incomplete combustion reactions.132 Catalyst 13 has also been utilized in the synthesis of ethylene carbonate, propylene carbonate, and styrene carbonate in a flow reactor process with coal- and gas-generated industrial flue gas as CO2 sources.133 As expected, some catalyst deactivation was observed as the reaction progressed, with coal's flue gas being more detrimental to the catalyst than natural gas's.

image file: c7gc02560b-f17.tif
Fig. 17 North's bimetallic aluminum salen catalysts 12 and silica-supported bifunctional 13.

D'Elia, Basset, and coworkers have reported the use of Lewis acidic early transition metal halides and nBu4NBr for the formation of cyclic carbonates from epoxides and CO2-laden flue gas.134 Specifically, using industrially-collected flue gas containing 10.2% CO2, YCl3/nBu4NBr produced cyclic epichlorohydrin carbonate at 22 °C with nearly the same effectiveness as a stream containing a similar concentration of pure CO2. This low temperature was critical to minimize the impact of known flue gas catalyst poisons including NOx, SOx, and H2O. Only at higher temperatures (>100 °C) were polyepichlorohydrin and other unidentified polymeric impurities observable, likely because of the increased water vapor content at temperatures above 100 °C. Though it was found to have slightly lower activity than YCl3, a cheaper and more widely available zirconium catalyst was chosen for a follow-up study by the same groups.135 ZrCl4·(OEt2)2 mounted on a silica support with nucleophilic nBu4NBr cocatalyst quantitatively produced cyclic propylene carbonate from propylene oxide and flue gas containing approximately 15% CO2. This heterogeneous complex could also be recovered and reused for at least three cycles.

Less work has been reported for selective polycarbonate production using impure CO2 sources. In 2015, Williams and coworkers discussed the successful coupling of cyclohexene oxide and industrially-captured post-combustion CO2 utilizing homogeneous bimetallic dizinc and dimagnesium catalysts.136 Gratifyingly, the selectivity for polymer and carbonate linkages both remain high, at or above 98%, consistent with their previous reports utilizing CO2 of 99.9999% purity.25,137,138 These polymerizations were carried out with only a modest overpressure of carbon dioxide at 1.06 atm.136 The dimagnesium catalyst also displays very good resistance to various solution- and gas-phase contaminants including H2S, SO2, and monoethanolamine, among others. Industrially, Econic Technologies reports using this technology with captured waste CO2 in their commercial polycarbonate and polycarbonate-polyol production processes.139 Covestro, formerly Bayer MaterialScience, and partner Recticel have also produced polycarbonate-polyols since 2016 utilizing waste CO2 from a nearby chemical plant.44

Of note, much of the CO2 that will be captured from waste streams will be at atmospheric pressure, and, though several catalysts are productive at atmospheric pressure,25,104,138,140–144 many other active catalytic systems for CO2/epoxides copolymerization operate more efficiently at higher pressures. It would be useful to have an effective method for capturing CO2 at atmospheric pressure for utilization at higher pressures without the requirement of mechanical compression. This is particularly important for small-scale research studies and where various sources of CO2 are utilized. We have reported the use of the commercially-available metal organic framework (MOF) Cu3(btc)2(H2O)3 (btc = benzene-1,3,5-tricarboxylate), also known as HKUST-1, to adsorb CO2 under aerobic conditions and atmospheric pressure.145 Following maximum absorption of CO2 by the MOF material, the CO2 was subsequently released into a second copolymerization reactor by increasing the temperature of the system to 120 °C. In this manner, the copolymerization of propylene oxide and CO2 was carried out at room temperature and 0.9–1.5 MPa pressure, utilizing the binary catalyst system (salen)CoDNP/PPNDNP (DNP = dinitrophenoxide). Poly(propylene carbonate) was produced with Mn = 8870 g mol−1 and PDI = 1.06. The MOF/copolymerization system was cycled 10 times with an average 49.9% conversion without significant changes in polymer MW or PDI. This same type of system could easily be employed with other catalyst and monomer choices as well.

Outlook and conclusions

For a chemical process to be truly sustainable, it must meet the three pillars of sustainability: economic, societal, and ecological (Fig. 18).146 These are also known as the three P′s: profit, people, and planet. Many excellent reviews and perspectives have recently been published regarding the development of greener, more sustainable polymers.147–149 With regard to the coupling reactions of CO2 and epoxides to produce polycarbonates, these copolymerizations have been accomplished with increasingly high catalytic activity, high regio- and stereoselectivity, and lower pressures of CO2. The advent of several companies industrially producing polycarbonates from CO2 and epoxides speaks to the rich economic landscape available today for this technology.
image file: c7gc02560b-f18.tif
Fig. 18 The three pillars of sustainability.146

Over the past decade there has been a large swell in the development of polycarbonate polyols for incorporation into polyurethanes.42–45 While polycarbonate polyols have long been said to be greener than their polyether counterparts, a 2014 life cycle analysis performed a deep dive cradle-to-gate assessment of poly(propylene carbonate)-based polyols versus traditional polyols for eventual polyurethane production.150 With 20 wt% CO2 incorporation into poly(propylene carbonate)-polyol, greenhouse gas emissions are decreased by 11–19%, and fossil fuel consumption can be reduced by 13–16%. Perfectly alternating polycarbonate polyols can contain a maximum of 50 wt% CO2 for PEC and 43 wt% CO2 for PPC, and so higher incorporation of CO2 into these polyols would likely further increase these benefits over polyether polyols. However, the report also notes that 2.65–2.86 kg CO2 is still produced per 1.0 kg of polycarbonate polyol formed, indicating that this production does not represent a CO2 sink. Regardless, polycarbonate polyols represent an excellent and exciting economic opportunity for CO2/epoxides coupling.

There is currently much excitement due to the excellent material properties of poly(limonene carbonate) (vide supra), and a 2012 life cycle assessment found that (+)-limonene has a lower contribution to global warming, acidification, smog formation, and natural resource depletion than its petroleum-based counterparts.151 However, the ultimate utility of PLC lies heavily on the availability of raw limonene starting material. The price of limonene is notoriously volatile as it relies almost solely on global citrus production.152,153 Brazil, China, and the US are the leading producers of oranges, but all have been impacted significantly by citrus greening in recent years.87,154,155 Despite these issues, global production of limonene has averaged around 70[thin space (1/6-em)]000 tons per year (2013).87 Limonene makes up 4% of citrus peel waste,156 and with greater recovery efforts the production is expected to grow to a maximum of around 100[thin space (1/6-em)]000 tons per year.87 A 2017 study screened the production parameters of several common industrial polymers including polyethylene, polypropylene, and poly(ethylene terephthalate) alongside polymers produced from renewable resources including PCHDC and PLC with the goal of ranking the plastics with regard to environmental and economic impacts.157 The authors state that, when compared to PLC, PCHDC is identified as a more promising candidate for further research because of its higher efficiency in carbon utilization. Of note, though propylene oxide does not come from renewable resources, the authors determined that poly(propylene carbonate) is one of only two polymers from this study that actually captures additional CO2 during its production rather than producing excess.

One looming concern over polycarbonates produced by the coupling of CO2 and epoxides is their biodegradability. Several studies have successfully degraded representative samples of poly(propylene carbonate) in buffered solutions or compost.158–164 However, the conditions are often specific, and it is unclear how quickly poly(propylene carbonate) would degrade in a typical landfill environment. The recent study on poly(limonene carbonate)'s biodegradation112 leads to similar questions. The carbonate unit should be susceptible to hydrolysis, but pendant hydrophobic groups may prevent timely biodegradation. Focused research efforts are still needed to understand the full scope of the biodegradability of these materials.

Finally, some features relative to the use of renewable epoxides for polycarbonate synthesis, especially when employing nonrenewable energy and/or other chemical reagents, are worthy of consideration. Whether utilizing metabolites or precursor sources of epoxides, these materials must not be competitive with food production, as the world's human population is expected to increase by about two billion by 2050. Furthermore, metabolites may be significantly removed chemically from the corresponding precursors, requiring several additional reaction steps to synthesize useable epoxide monomers. These multistep processes can greatly enhance other negative effects on the environment, e.g., use of nonrenewable resources and the production of CO2. As a result, the number of reaction steps should be minimized wherever possible, and further steps should be taken to enhance the recyclability of employed catalysts. Continued exploration of using waste, captured CO2 at ambient pressure is also imperative to decrease energy costs and improve the overall sustainability of the process.

Conflicts of interest

There are no conflicts to declare.


Donald J. Darensbourg gratefully acknowledges the financial support of the National Science Foundation (CHE 1610311) and the Robert A. Welch Foundation (A-0923).

Notes and references

  1. D. J. Darensbourg and M. W. Holtcamp, Macromolecules, 1995, 28, 7577–7579 CrossRef CAS .
  2. G. W. Coates and D. R. Moore, Angew. Chem., Int. Ed., 2004, 43, 6618–6639 CrossRef CAS PubMed .
  3. H. Sugimoto and S. Inoue, J. Polym. Sci., Part A: Polym. Chem., 2004, 42, 5561–5573 CrossRef CAS .
  4. D. J. Darensbourg, R. M. Mackiewicz, A. L. Phelps and D. R. Billodeaux, Acc. Chem. Res., 2004, 37, 836–844 CrossRef CAS PubMed .
  5. M. H. Chisholm and Z. Zhou, J. Mater. Chem., 2004, 14, 3081–3092 RSC .
  6. D. J. Darensbourg, Chem. Rev., 2007, 107, 2388–2410 CrossRef CAS PubMed .
  7. S. Klaus, M. W. Lehenmeier, C. E. Anderson and B. Rieger, Coord. Chem. Rev., 2011, 255, 1460–1479 CrossRef CAS .
  8. M. R. Kember, A. Buchard and C. K. Williams, Chem. Commun., 2011, 47, 141–163 RSC .
  9. X.-B. Lu and D. J. Darensbourg, Chem. Soc. Rev., 2012, 41, 1462–1484 RSC .
  10. X.-B. Lu, W.-M. Ren and G.-P. Wu, Acc. Chem. Res., 2012, 45, 1721–1735 CrossRef CAS PubMed .
  11. D. J. Darensbourg and S. J. Wilson, Green Chem., 2012, 14, 2665–2671 RSC .
  12. S. Paul, Y. Zhu, C. Romain, R. Brooks, P. K. Saini and C. K. Williams, Chem. Commun., 2015, 51, 6459–6479 RSC .
  13. G. Trott, P. K. Saini and C. K. Williams, Philos. Trans. R. Soc., A, 2016, 374 CrossRef CAS PubMed  , 20150085.
  14. A. Decortes, A. M. Castilla and A. W. Kleij, Angew. Chem., Int. Ed., 2010, 49, 9822–9837 CrossRef CAS PubMed .
  15. R. Martín and A. W. Kleij, ChemSusChem, 2011, 4, 1259–1263 CrossRef PubMed .
  16. P. P. Pescarmona and M. Taherimehr, Catal. Sci. Technol., 2012, 2, 2169–2187 CAS .
  17. J. W. Comerford, I. D. V. Ingram, M. North and X. Wu, Green Chem., 2015, 17, 1966–1987 RSC .
  18. M. North, R. Pasquale and C. Young, Green Chem., 2010, 12, 1514–1539 RSC .
  19. S. Inoue, H. Koinuma and T. Tsuruta, J. Polym. Sci., Part B: Polym. Lett., 1969, 7, 287–292 CrossRef CAS .
  20. T. Aida and S. Inoue, Macromolecules, 1981, 14, 1166–1169 CrossRef CAS .
  21. W. J. Kruper and D. D. Dellar, J. Org. Chem., 1995, 60, 725–727 CrossRef CAS .
  22. S. Mang, A. I. Cooper, M. E. Colclough, N. Chauhan and A. B. Holmes, Macromolecules, 2000, 33, 303–308 CrossRef CAS .
  23. D. J. Darensbourg, M. W. Holtcamp, G. E. Struck, M. S. Zimmer, S. A. Niezgoda, P. Rainey, J. B. Robertson, J. D. Draper and J. H. Reibenspies, J. Am. Chem. Soc., 1999, 121, 107–116 CrossRef CAS .
  24. M. Cheng, E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 1998, 120, 11018–11019 CrossRef CAS .
  25. M. R. Kember, P. D. Knight, P. T. R. Reung and C. K. Williams, Angew. Chem., Int. Ed., 2009, 48, 931–933 CrossRef CAS PubMed .
  26. D. J. Darensbourg and J. C. Yarbrough, J. Am. Chem. Soc., 2002, 124, 6335–6342 CrossRef CAS PubMed .
  27. Z. Qin, C. M. Thomas, S. Lee and G. W. Coates, Angew. Chem., Int. Ed., 2003, 42, 5484–5487 CrossRef CAS PubMed .
  28. P. Chen, M. H. Chisholm, J. C. Gallucci, X. Zhang and Z. Zhou, Inorg. Chem., 2005, 44, 2588–2595 CrossRef CAS PubMed .
  29. E. Comin, R. F. de Souza and K. Bernardo-Gusmão, Catal. Today, 2017, 289, 115–120 CrossRef CAS .
  30. N.-Y. Mun, K.-H. Kim, D.-W. Park, Y. Choe and I. Kim, Korean J. Chem. Eng., 2005, 22, 556–559 CrossRef CAS .
  31. J.-Y. Ahn, H.-L. Shim, K.-H. Kim, I. Kim, S.-W. Park and D.-W. Park, Korean J. Chem. Eng., 2008, 25, 693–696 CrossRef CAS .
  32. D. J. Darensbourg, M. Ulusoy, O. Karroonnirum, R. R. Poland, J. H. Reibenspies and B. Çetinkaya, Macromolecules, 2009, 42, 6992–6998 CrossRef CAS .
  33. W.-M. Ren, Z.-W. Liu, Y.-Q. Wen, R. Zhang and X.-B. Lu, J. Am. Chem. Soc., 2009, 131, 11509–11518 CrossRef CAS PubMed .
  34. D. J. Darensbourg and F.-T. Tsai, Macromolecules, 2014, 47, 3806–3813 CrossRef CAS .
  35. M. Zhang and Y. Yu, Ind. Eng. Chem. Res., 2013, 52, 9505–9514 CrossRef CAS .
  36. A. Mohsenzadeh, A. Zamani and M. J. Taherzadeh, ChemBioEng Rev., 2017, 4, 75–91 CrossRef CAS .
  37. A. D. Moreno, P. Alvira, D. Ibarra and E. Tomás-Pejó, in Production of Platform Chemicals from Sustainable Resources, ed. Z. Fang, J. R. L. Smith and X. Qi, Springer Singapore, Singapore, 2017, pp. 375–410 Search PubMed .
  38. M. DiRienzo and A. Laugharne, Silver Institute Press Release, http://www.silverinstitute.org/site/2016/12/15/six-hundred-million-ounces-of-silver-to-be-consumed-in-photovoltaics-and-ethylene-oxide-production-through-2020/, (accessed August, 2017).
  39. A. Okada, S. Kikuchi, K. Nakano, K. Nishioka, K. Nozaki and T. Yamada, Chem. Lett., 2010, 39, 1066–1068 CrossRef CAS .
  40. H. S. Kim, J. Y. Bae, J. S. Lee, O. S. Kwon, P. Jelliarko, S. D. Lee and S.-H. Lee, J. Catal., 2005, 232, 80–84 CrossRef CAS .
  41. QPAC®25, commercially available from Empower Materials, Inc., http://empowermaterials.com/products/qpac-25, (accessed August, 2017).
  42. Converge Polyols, http://www.saudiaramco.com/en/home/our-business/downstream/converge.html, (accessed August, 2017).
  43. Econic Technologies, http://www.econic-technologies.com/, (accessed August, 2017).
  44. Covestro: Beyond Carbon, https://www.co2-dreams.covestro.com/, (accessed September, 2017).
  45. J. Langanke, A. Wolf, J. Hofmann, K. Bohm, M. A. Subhani, T. E. Muller, W. Leitner and C. Gurtler, Green Chem., 2014, 16, 1865–1870 RSC .
  46. D. S. McGuinness, Chem. Rev., 2011, 111, 2321–2341 CrossRef CAS PubMed .
  47. E. D. Metzger, C. K. Brozek, R. J. Comito and M. Dincă, ACS Cent. Sci., 2016, 2, 148–153 CrossRef CAS PubMed .
  48. J. R. Briggs, J. Chem. Soc., Chem. Commun., 1989, 674–675 RSC .
  49. A. Bollmann, K. Blann, J. T. Dixon, F. M. Hess, E. Killian, H. Maumela, D. S. McGuinness, D. H. Morgan, A. Neveling, S. Otto, M. Overett, A. M. Z. Slawin, P. Wasserscheid and S. Kuhlmann, J. Am. Chem. Soc., 2004, 126, 14712–14713 CrossRef CAS PubMed .
  50. X.-H. Zhang, R.-J. Wei, Y. Y. Zhang, B.-Y. Du and Z.-Q. Fan, Macromolecules, 2015, 48, 536–544 CrossRef CAS .
  51. K. Nakano, T. Kamada and K. Nozaki, Angew. Chem., Int. Ed., 2006, 45, 7274–7277 CrossRef CAS PubMed .
  52. J. E. Seong, S. J. Na, A. Cyriac, B.-W. Kim and B. Y. Lee, Macromolecules, 2010, 43, 903–908 CrossRef CAS .
  53. F. W. Shaarani and J. J. Bou, Sci. Total Environ., 2017, 598, 931–936 CrossRef CAS PubMed .
  54. V. Fell, C. R. Lee and R. J. Pollitt, Biochem. Med., 1975, 13, 40–45 CrossRef CAS PubMed .
  55. G. F. Hoffmann, C. K. Seppel, B. Holmes, L. Mitchell, H. J. Christen, F. Hanefeld, D. Rating and W. L. Nyhan, J. Chromatogr., B: Biomed. Appl., 1993, 617, 1–10 CrossRef CAS .
  56. S. Bouatra, F. Aziat, R. Mandal, A. C. Guo, M. R. Wilson, C. Knox, T. C. Bjorndahl, R. Krishnamurthy, F. Saleem, P. Liu, Z. T. Dame, J. Poelzer, J. Huynh, F. S. Yallou, N. Psychogios, E. Dong, R. Bogumil, C. Roehring and D. S. Wishart, PLoS One, 2013, 8, e73076 CAS .
  57. The Human Metabolome Database (HMDB): (S)-3,4-Dihydroxybutyric acid (HMDB00337), http://www.hmdb.ca/metabolites/HMDB00337, (accessed August, 2017) Search PubMed .
  58. C. H. Martin, H. Dhamankar, H.-C. Tseng, M. J. Sheppard, C. R. Reisch and K. L. J. Prather, Nat. Commun., 2013, 4, 1933 Search PubMed .
  59. H. H. Dhamankar, C. H. Martin and K. L. J. Prather, United States Pat, US20140364629A1, 2014 Search PubMed .
  60. F.-T. Tsai, Y. Wang and D. J. Darensbourg, J. Am. Chem. Soc., 2016, 138, 4626–4633 CrossRef CAS PubMed .
  61. L. Faissat, K. Martin, C. Chavis, J.-L. Montéro and M. Lucas, Bioorg. Med. Chem., 2003, 11, 325–334 CrossRef CAS PubMed .
  62. H. Mutlu and M. A. R. Meier, Eur. J. Lipid Sci. Technol., 2010, 112, 10–30 CrossRef CAS .
  63. M. Van der Steen and C. V. Stevens, ChemSusChem, 2009, 2, 692–713 CrossRef CAS PubMed .
  64. Y.-Y. Zhang, X.-H. Zhang, R.-J. Wei, B.-Y. Du, Z.-Q. Fan and G.-R. Qi, RSC Adv., 2014, 4, 36183–36188 RSC .
  65. B. M. Bell, J. R. Briggs, R. M. Campbell, S. M. Chambers, P. D. Gaarenstroom, J. G. Hippler, B. D. Hook, K. Kearns, J. M. Kenney, W. J. Kruper, D. J. Schreck, C. N. Theriault and C. P. Wolfe, Clean: Soil, Air, Water, 2008, 36, 657–661 CrossRef CAS .
  66. C. J. A. Mota, B. Peres Pinto and A. L. de Lima, in Glycerol: A Versatile Renewable Feedstock for the Chemical Industry, ed. C. J. A. Mota, B. Peres Pinto and A. L. de Lima, Springer International Publishing, Cham, 2017, pp. 39–57 Search PubMed .
  67. Epicerol(R), http://www.solvay.com/en/markets-and-products/featured-products/epicerol.html, (accessed September, 2017).
  68. Z. Shen, X. Chen and Y. Zhang, Macromol. Chem. Phys., 1994, 195, 2003–2011 CrossRef CAS .
  69. R.-J. Wei, X.-H. Zhang, B.-Y. Du, Z.-Q. Fan and G.-R. Qi, Polymer, 2013, 54, 6357–6362 CrossRef CAS .
  70. G.-P. Wu, S.-H. Wei, W.-M. Ren, X.-B. Lu, T.-Q. Xu and D. J. Darensbourg, J. Am. Chem. Soc., 2011, 133, 15191–15199 CrossRef CAS PubMed .
  71. G.-P. Wu, P.-X. Xu, X.-B. Lu, Y.-P. Zu, S.-H. Wei, W.-M. Ren and D. J. Darensbourg, Macromolecules, 2013, 46, 2128–2133 CrossRef CAS .
  72. J.-P. Lange, E. van der Heide, J. van Buijtenen and R. Price, ChemSusChem, 2012, 5, 150–166 CrossRef CAS PubMed .
  73. A. S. Mamman, J.-M. Lee, Y.-C. Kim, I. T. Hwang, N.-J. Park, Y. K. Hwang, J.-S. Chang and J.-S. Hwang, Biofuels, Bioprod. Biorefin., 2008, 2, 438–454 CrossRef CAS .
  74. Y. Hu, L. Qiao, Y. Qin, X. Zhao, X. Chen, X. Wang and F. Wang, Macromolecules, 2009, 42, 9251–9254 CrossRef CAS .
  75. J. Hilf, M. Scharfenberg, J. Poon, C. Moers and H. Frey, Macromol. Rapid Commun., 2015, 36, 174–179 CrossRef CAS PubMed .
  76. R. T. Mathers, M. J. Shreve, E. Meyler, K. Damodaran, D. F. Iwig and D. J. Kelley, Macromol. Rapid Commun., 2011, 32, 1338–1342 CrossRef CAS PubMed .
  77. H. Mutlu, R. Hofsa, R. E. Montenegro and M. A. R. Meier, RSC Adv., 2013, 3, 4927–4934 RSC .
  78. J. A. Mmongoyo, Q. A. Mgani, S. J. M. Mdachi, P. J. Pogorzelec and D. J. Cole-Hamilton, Eur. J. Lipid Sci. Technol., 2012, 114, 1183–1192 CrossRef CAS .
  79. M. Winkler, C. Romain, M. A. R. Meier and C. K. Williams, Green Chem., 2015, 17, 300–306 RSC .
  80. D. J. Darensbourg, W.-C. Chung, C. J. Arp, F.-T. Tsai and S. J. Kyran, Macromolecules, 2014, 47, 7347–7353 CrossRef CAS .
  81. S. Honda, T. Mori, H. Goto and H. Sugimoto, Polymer, 2014, 55, 4832–4836 CrossRef CAS .
  82. A. Dondoni, Angew. Chem., Int. Ed., 2008, 47, 8995–8997 CrossRef CAS PubMed .
  83. D. J. Darensbourg, W.-C. Chung, A. D. Yeung and M. Luna, Macromolecules, 2015, 48, 1679–1687 CrossRef CAS .
  84. M. Laska and P. Teubner, Chem. Senses, 1999, 24, 161–170 CrossRef CAS PubMed .
  85. Q. Jia, J. Mu, W. Fan and S. Shan (Kunming University of Science and Technology), Peoples Republic of China Pat, CN103333329A, 2013 Search PubMed .
  86. P. E. Shaw, J. Agric. Food Chem., 1979, 27, 246–257 CrossRef CAS .
  87. R. Ciriminna, M. Lomeli-Rodriguez, P. Demma Cara, J. A. Lopez-Sanchez and M. Pagliaro, Chem. Commun., 2014, 50, 15288–15296 RSC .
  88. A. F. Thomas and Y. Bessiere, Nat. Prod. Rep., 1989, 6, 291–309 RSC .
  89. Exposure and Risk Assessment on Lower Risk Pesticide Chemicals, D-Limonene, US Environmental Protection Agency, Arlington, VA, 2005 Search PubMed .
  90. R. Carlson, N. Behn and C. Cowles, J. Org. Chem., 1971, 36, 3832–3833 CrossRef CAS .
  91. C. M. Byrne, S. D. Allen, E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2004, 126, 11404–11405 CrossRef CAS PubMed .
  92. C. Li, R. J. Sablong and C. E. Koning, Eur. Polym. J., 2015, 67, 449–458 CrossRef CAS .
  93. L. Peña Carrodeguas, J. González-Fabra, F. Castro-Gómez, C. Bo and A. W. Kleij, Chem. – Eur. J., 2015, 21, 6115–6122 CrossRef PubMed .
  94. G. Fiorani, M. Stuck, C. Martín, M. M. Belmonte, E. Martin, E. C. Escudero-Adán and A. W. Kleij, ChemSusChem, 2016, 9, 1304–1311 CrossRef CAS PubMed .
  95. D. Steiner, L. Ivison, C. T. Goralski, R. B. Appell, J. R. Gojkovic and B. Singaram, Tetrahedron: Asymmetry, 2002, 13, 2359–2363 CrossRef CAS .
  96. O. Hauenstein, M. Reiter, S. Agarwal, B. Rieger and A. Greiner, Green Chem., 2016, 18, 760–770 RSC .
  97. C. Li, R. J. Sablong and C. E. Koning, Angew. Chem., Int. Ed., 2016, 55, 11572–11576 CrossRef CAS PubMed .
  98. C. Li, S. van Berkel, R. J. Sablong and C. E. Koning, Eur. Polym. J., 2016, 85, 466–477 CrossRef CAS .
  99. F. Auriemma, C. De Rosa, M. R. Di Caprio, R. Di Girolamo and G. W. Coates, Macromolecules, 2015, 48, 2534–2550 CrossRef CAS .
  100. F. Auriemma, C. De Rosa, M. R. Di Caprio, R. Di Girolamo, W. C. Ellis and G. W. Coates, Angew. Chem., Int. Ed., 2015, 54, 1215–1218 CrossRef CAS PubMed .
  101. M. Cheng, D. R. Moore, J. J. Reczek, B. M. Chamberlain, E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2001, 123, 8738–8749 CrossRef CAS PubMed .
  102. M. Reiter, S. Vagin, A. Kronast, C. Jandl and B. Rieger, Chem. Sci., 2017, 8, 1876–1882 RSC .
  103. J. González-Fabra, F. Castro-Gómez, A. W. Kleij and C. Bo, ChemSusChem, 2017, 10, 1233–1240 CrossRef PubMed .
  104. A. Buchard, M. R. Kember, K. G. Sandeman and C. K. Williams, Chem. Commun., 2011, 47, 212–214 RSC .
  105. C. Li, R. J. Sablong, R. A. T. M. van Benthem and C. E. Koning, ACS Macro Lett., 2017, 6, 684–688 CrossRef CAS .
  106. D. J. Darensbourg, W.-C. Chung and S. J. Wilson, ACS Catal., 2013, 3, 3050–3057 CrossRef CAS .
  107. D. J. Darensbourg, S.-H. Wei, A. D. Yeung and W. C. Ellis, Macromolecules, 2013, 46, 5850–5855 CrossRef CAS .
  108. D. J. Darensbourg, A. D. Yeung and S.-H. Wei, Green Chem., 2013, 15, 1578–1583 RSC .
  109. G. Odian, Principles of Polymerization, John Wiley & Sons, Hoboken, NJ, 4th edn, 2004 Search PubMed .
  110. Y. Liu, H. Zhou, J.-Z. Guo, W.-M. Ren and X.-B. Lu, Angew. Chem., Int. Ed., 2017, 56, 4862–4866 CrossRef CAS PubMed .
  111. O. Hauenstein, M. M. Rahman, M. Elsayed, R. Krause-Rehberg, S. Agarwal, V. Abetz and A. Greiner, Adv. Mater. Technol., 2017, 2, 1700026 CrossRef .
  112. O. Hauenstein, S. Agarwal and A. Greiner, Nat. Commun., 2016, 7, 11862 CrossRef CAS PubMed .
  113. N. Kindermann, À. Cristòfol and A. W. Kleij, ACS Catal., 2017, 7, 3860–3863 CrossRef CAS .
  114. S. Yamazaki, Org. Biomol. Chem., 2010, 8, 2377–2385 CAS .
  115. C. Li, M. Johansson, R. J. Sablong and C. E. Koning, Eur. Polym. J., 2017, 96, 337–349 CrossRef CAS .
  116. N. Gould and J. Ward (Sericol Ltd), Great Britain Pat., GB2461624-A, 2010 Search PubMed .
  117. R. P. Eckberg (General Electric Company), United States Pat, US4547431-A, 1985 Search PubMed .
  118. M.-A. Tehfe, J. Lalevée, D. Gigmes and J. P. Fouassier, Macromolecules, 2010, 43, 1364–1370 CrossRef CAS .
  119. J. V. Crivello and B. Yang, J. Polym. Sci., Part A: Polym. Chem., 1995, 33, 1881–1890 CrossRef CAS .
  120. (a) R. Akiyama, T. Ushirogochi, Y. Kawakami, H. Kiyomoto, K. Otsu and M. Ishibashi (Tokyo Electric Co Ltd), Japan Pat, JP2007137923-A, 2007 Search PubMed ; (b) R. Akiyama, T. Ushirogochi, Y. Kawakami, H. Kiyomoto, K. Otsu and M. Ishibashi (Tokyo Electric Co Ltd), Japan Pat, JP4210284-B2, 2007 Search PubMed .
  121. V. Schimpf, B. S. Ritter, P. Weis, K. Parison and R. Mülhaupt, Macromolecules, 2017, 50, 944–955 CrossRef CAS .
  122. M. Bahr, A. Bitto and R. Mulhaupt, Green Chem., 2012, 14, 1447–1454 RSC .
  123. S. D. Allen, D. R. Moore, E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2002, 124, 14284–14285 CrossRef CAS PubMed .
  124. R. C. Jeske, A. M. DiCiccio and G. W. Coates, J. Am. Chem. Soc., 2007, 129, 11330–11331 CrossRef CAS PubMed .
  125. F. Isnard, M. Lamberti, C. Pellecchia and M. Mazzeo, ChemCatChem, 2017, 9, 2972–2979 CrossRef CAS .
  126. L. Peña Carrodeguas, C. Martín and A. W. Kleij, Macromolecules, 2017 Search PubMed .
  127. E. H. Nejad, A. Paoniasari, C. G. W. van Melis, C. E. Koning and R. Duchateau, Macromolecules, 2013, 46, 631–637 CrossRef CAS .
  128. Carbon dioxide (CO2). A colorless, odorless nonflammable, liquefied gas., http://airgassgcatalog.com/catalog/P13_Carbon_Dioxide_Pure_Gases.pdf, (accessed August, 2017).
  129. M. Aresta, A. Dibenedetto and A. Angelini, Chem. Rev., 2014, 114, 1709–1742 CrossRef CAS PubMed .
  130. W. Clegg, R. W. Harrington, M. North and R. Pasquale, Chem. – Eur. J., 2010, 16, 6828–6843 CrossRef CAS PubMed .
  131. J. Melendez, M. North, P. Villuendas and C. Young, Dalton Trans., 2011, 40, 3885–3902 RSC .
  132. I. S. Metcalfe, M. North, R. Pasquale and A. Thursfield, Energy Environ. Sci., 2010, 3, 212–215 CAS .
  133. M. North, B. Wang and C. Young, Energy Environ. Sci., 2011, 4, 4163–4170 CAS .
  134. A. Barthel, Y. Saih, M. Gimenez, J. D. A. Pelletier, F. E. Kuhn, V. D'Elia and J.-M. Basset, Green Chem., 2016, 18, 3116–3123 RSC .
  135. M. J. Kelly, A. Barthel, C. Maheu, O. Sodpiban, F.-B. Dega, S. V. C. Vummaleti, E. Abou-Hamad, J. D. A. Pelletier, L. Cavallo, V. D'Elia and J.-M. Basset, J. CO2 Util., 2017, 20, 243–252 CrossRef CAS .
  136. A. M. Chapman, C. Keyworth, M. R. Kember, A. J. J. Lennox and C. K. Williams, ACS Catal., 2015, 5, 1581–1588 CrossRef CAS .
  137. M. R. Kember, J. Copley, A. Buchard and C. K. Williams, Polym. Chem., 2012, 3, 1196–1201 RSC .
  138. M. R. Kember and C. K. Williams, J. Am. Chem. Soc., 2012, 134, 15676–15679 CrossRef CAS PubMed .
  139. Carbon dioxide as a feedstock – Econic Technologies, http://www.econic-technologies.com/catalyst-technology/carbon-dioxide-as-a-feedstock/, (accessed August, 2017).
  140. H. Sugimoto, H. Ohshima and S. Inoue, J. Polym. Sci., Part A: Polym. Chem., 2003, 41, 3549–3555 CrossRef CAS .
  141. Y. Xiao, Z. Wang and K. Ding, Chem. – Eur. J., 2005, 11, 3668–3678 CrossRef CAS PubMed .
  142. Y. Xiao, Z. Wang and K. Ding, Macromolecules, 2006, 39, 128–137 CrossRef CAS .
  143. H. Sugimoto and K. Kuroda, Macromolecules, 2008, 41, 312–317 CrossRef CAS .
  144. W.-M. Ren, X. Zhang, Y. Liu, J.-F. Li, H. Wang and X.-B. Lu, Macromolecules, 2010, 43, 1396–1402 CrossRef CAS .
  145. D. J. Darensbourg, W.-C. Chung, K. Wang and H.-C. Zhou, ACS Catal., 2014, 4, 1511–1515 CrossRef CAS .
  146. R. A. Sheldon, Green Chem., 2017, 19, 18–43 RSC .
  147. A. Llevot, P.-K. Dannecker, M. von Czapiewski, L. C. Over, Z. Söyler and M. A. R. Meier, Chem. – Eur. J., 2016, 22, 11510–11521 CrossRef CAS PubMed .
  148. Y. Zhu, C. Romain and C. K. Williams, Nature, 2016, 540, 354–362 CrossRef CAS PubMed .
  149. D. K. Schneiderman and M. A. Hillmyer, Macromolecules, 2017, 50, 3733–3749 CrossRef CAS .
  150. N. von der Assen and A. Bardow, Green Chem., 2014, 16, 3272–3280 RSC .
  151. Life Cycle Assessment - Renewable and Sustainable Citrus Oils Final Report, Chicago, IL, 2012.
  152. Citrus: World Markets and Trade (January 2017), United States Department of Agriculture, http://www.fas.usda.gov/data/citrus-world-markets-and-trade, 2017.
  153. M. Bomgardner, Chem. Eng. News, 2011, 89, 21–21 Search PubMed .
  154. L. Alvarez, N. Y. Times (N. Y. Ed.), 2013, 162, A1–A14 Search PubMed .
  155. A. Harmon, N. Y. Times (N. Y. Ed.), 2013, 162, 1–19 Search PubMed .
  156. M. Pourbafrani, G. Forgács, I. S. Horváth, C. Niklasson and M. J. Taherzadeh, Bioresour. Technol., 2010, 101, 4246–4250 CrossRef CAS PubMed .
  157. D. Zhang, E. A. del Rio-Chanona and N. Shah, ACS Sustainable Chem. Eng., 2017, 5, 4388–4398 CrossRef CAS .
  158. G. A. Luinstra, Polym. Rev., 2008, 48, 192–219 CrossRef CAS .
  159. G. A. Luinstra and E. Borchardt, in Synthetic Biodegradable Polymers, ed. B. Rieger, A. Künkel, G. W. Coates, R. Reichardt, E. Dinjus and T. A. Zevaco, Springer Berlin Heidelberg, Berlin, Heidelberg, 2012, pp. 29–48 Search PubMed .
  160. X. Fang, S. Yang and L. Chen, J. Funct. Polym., 1994, 7, 143–147 CAS .
  161. L. C. Du, Y. Z. Meng, S. J. Wang and S. C. Tjong, J. Appl. Polym. Sci., 2004, 92, 1840–1846 CrossRef CAS .
  162. S. Inoue, T. Tsuruta, T. Takada, N. Miyazaki, M. Kambe and T. Takaoka, Appl. Polym. Symp., 1975, 26, 257–267 CAS .
  163. M. Zhou, M. Takayanagi, Y. Yoshida, S. Ishii and H. Noguchi, Polym. Bull., 1999, 42, 419–424 CrossRef CAS .
  164. M. Ree, Y. Hwang, J.-S. Kim, H. Kim, G. Kim and H. Kim, Catal. Today, 2006, 115, 134–145 CrossRef CAS .


Electronic supplementary information (ESI) available. See DOI: 10.1039/C7GC02560B

This journal is © The Royal Society of Chemistry 2017