Abstract
Biomass is a class of abundant renewable resource. Its efficient use in the field of biobased materials is one of the important ways for implementation of sustainable development strategies. 2,5-Furandicarboxylic acid (FDCA) as a potential alternative of terephthalic acid (PTA) to make alipharomatic polyesters, can be obtained in mass amount from cellulose via bio- or chemical process. For this reason, FDCA-based polyesters have gained high interest recently. This review systematically summarizes recent progress in the making of FDCA-based polyesters (including poly(ethylene 2,5-furandicarboxylate) (PEF), poly(propylene 2,5-furandicarboxylate) (PPF), poly(butylene 2,5-furandicarboxylate) (PBF), poly(hexylene 2,5-furandicarboxylate) (PHF), and their copolyesters), especially highlighting the progress and fundamental aspects for their synthesis and properties. Significant advantages (and also disadvantages) of the FDCA-based polyesters are clearly indicated relative to price, performance, and sustainable development, in reference to traditional petroleum-based polyesters in industrial application. The goal of this review is to provide useful information regarding the synthesis and properties of FDCA-based polyesters.
Download PDF
Full Article
Progress in the Synthesis and Properties of 2,5-Furan Dicarboxylate based Polyesters
Junhua Zhang,* Yao Liu, Zhaohua Qi, Liang He, and Lincai Peng
Biomass is a class of abundant renewable resource. Its efficient use in the field of biobased materials is one of the important ways for implementation of sustainable development strategies. 2,5-Furandicarboxylic acid (FDCA) as a potential alternative of terephthalic acid (PTA) to make alipharomatic polyesters, can be obtained in mass amount from cellulose via bio- or chemical process. For this reason, FDCA-based polyesters have gained high interest recently. This review systematically summarizes recent progress in the making of FDCA-based polyesters (including poly(ethylene 2,5-furandicarboxylate) (PEF), poly(propylene 2,5-furandicarboxylate) (PPF), poly(butylene 2,5-furandicarboxylate) (PBF), poly(hexylene 2,5-furandicarboxylate) (PHF), and their copolyesters), especially highlighting the progress and fundamental aspects for their synthesis and properties. Significant advantages (and also disadvantages) of the FDCA-based polyesters are clearly indicated relative to price, performance, and sustainable development, in reference to traditional petroleum-based polyesters in industrial application. The goal of this review is to provide useful information regarding the synthesis and properties of FDCA-based polyesters.
Keywords: FDCA; FDCA-based polyesters; Copolyesters; Synthesis; Polycondensation
Contact information: BiomassChem Group, Faculty of Chemical Engineering, Kunming University of Science and Technology, Kunming 650500, China; *Corresponding author: zhangjh0026@163.com
INTRODUCTION
Plastics are closely related to the development of human society. Terephthalic acid (PTA) based polyester, as one of the most important plastics, has been widely used in food packaging, plastic toys, plastic furniture, and engineering plastics. In recent years, due to the excessive consumption of fossil resources and environmental problems, the search for new materials with environment friendly and unique properties to synthesize biobased polymers has attracted extensive attention. Simultaneously, European countries and the United States have begun to support the development and use of biomass-based products to reduce the excessive reliance on oil-based products. Against this background, a large number of monomers derived from renewable resources, such as dodecanedioic acid (Green et al. 2000; Zou et al. 2018; Firoozi and Kang 2019; Zhang et al. 2019a;), lactic acid (Duo et al. 2016; Herran et al. 2019; Montava-Jorda et al. 2019; Wu et al. 2019), levulinic acid (Hartweg and Becer 2016; Pileidis and Titirici 2016; Amarasekara et al. 2018; Sinisi et al. 2019), isosorbide (Sousa et al. 2016a; Chebbi et al. 2019; Hu et al. 2019a; Kieber et al. 2019), ethylene diol (Haernvall et al. 2017; Li et al. 2017; Lu et al. 2017; Sukhawipat et al. 2018), and succinic acid (Rizescu et al. 2017; Landim et al. 2018; Garcia-Campo et al. 2019; Hu et al. 2019b; Pasma et al. 2019) have been continuously developed and promoted the synthesis and widespread use of a large number of new biobased polymers products. Among them, 2,5-furandicarboxylic acid (FDCA) is deemed as a promising platform chemical to prepare biobased polyesters.
FDCA has two carboxylic acid groups, and this implies that there is good potential as a replacement for PTA due to similarities between PTA and FDCA (Moreau et al. 2004). This aromatic diacid has been considered as a potential replacement and also as a renewable resource-based equivalent of PTA, which is a monomer in poly(ethylene terephthalate) (PET) plastics. The market for PTA exceeds 50,000 kt per year, demonstrating the large potential of FDCA for the preparation of polyethylene furanoate (PEF) as a replacement for PET (Eerhart et al. 2012; Amarasekara et al. 2017). For this reason, the biobased FDCA has attracted extensive attention from academy and industry, including the preparation of FDCA and its precursor, 5-hydroxymethylfurfural (HMF), together with varieties of FDCA-based materials that can be used in the domain of packaging, textile and coating. There have some related publications reviewing the progress and critical aspects in the synthesis and properties of FDCA-based polyesters a few years ago (van Es 2013; Sousa et al. 2015; Papageorgiou et al. 2016). Now, ongoing research and industrial production of FDCA-based polyesters, especially for poly(ethylene 2,5-furandicarboxylate) (PEF), poly(propylene 2,5-furandicarboxylate) (PPF), poly(butylene 2,5-furandicarboxylate) (PBF), poly(hexylene 2,5-furandicarboxylate) (PHF) and their copolyesters, has progressed in a new stage due to the increasing attention paid by researchers in recent years. In order to gain a deeper understanding of the current status of the synthesis and properties of FDCA-based polyesters, this mini review paper summarizes and discusses the recent progress, hoping to help readers to have a deeper understanding and cognition in this field, and to make a greater breakthroughs in this exciting research area.
SYNTHESIS OF POLY(ETHYLENE 2,5-FURANDICARBOXYLATE) (PEF)
Poly(ethylene 2,5-furandicarboxylate) (PEF) is the most eye-catching member of the FDCA-based polyesters family, which is considered a furan counterpart of the engineering polyester, PET (Gandini et al. 2009b; Sousa et al. 2015; Banella et al. 2019). Although fully renewable-based PET has been foreseen, and several routes are under development today, attempts to amplify the process have been limited due to the low yields (Tavener et al. 2003; Collias et al. 2014; Jerpdal et al. 2019). By contrast, PEF is totally based on renewable resources, as its two monomers, ethylene glycol (EG) and FDCA can be obtained from cellulose (Zhang et al. 2015; Zhou et al. 2019). It is expected that PEF will create its own market niche toward its full penetration in existing polyester market and be widely used in all works of life due to its high performance in high barrier, thermal, and mechanical properties (Sousa et al. 2015).
The synthesis of PEF dates back to the middle of the previous century; however, the literature on PEF is scanty after the research about the X-ray study of the FDCA-based polyester’s structure in 1968 (Sousa et al. 2015). Until the last decade, as people have been paying more and more attention to building a sustainable society, a revival of interest in the synthesis and characterization of PEF and related polyesters can be observed (Ji et al. 2008; Gandini et al. 2009a,b; Gomes et al. 2011). For example, Ji et al. (2008) developed a two-stage method for the preparation of PEF from cellulose (Fig. 1). In the first stage, dihydroxyethyl furanate monomers or oligomers can be obtained through the esterification or reverse transesterification between FDCA and EG. Next, the polycondensation reaction allows the polymerization product to reach the corresponding molecular weight. In the process of polycondensation, the excessive and generated EG is removed by raising the temperature and vacuum degree to promote the polycondensation reaction in the direction of PEF formation.
Fig. 1. Typical route for the synthesis of EG and FDCA from cellulose and the two-stage approach for the synthesis of PEF
Gandini et al. (2009a,b) developed two methods for the synthesis of PEF from FDCA and EG. They used a small amount of hydrochloric acid to catalyze the esterification of FDCA with EG, then antimony trioxide (Sb2O3) was used as the catalyst for condensation polymerization to obtain some white PEFs with DPn values of 250 to 300. It was determined that the PEFs exhibited good chemical stability, which was only soluble in strong acid (trifluoroacetic acid, TFA) and strong polar organic solvents (hot tetrachloroethane, TCE), and was thermally stable (around 300°C) by TGA analysis. Further analysis of the differential scanning calorimetry (DSC) and X-ray diffraction (XRD) showed that the PEFs exhibited a high degree of crystallinity, and the melting temperature (Tm) was 210 to 215 °C. Furthermore, the glass transition temperature (Tg) was 75 to 80 °C (Gandini et al. 2009b). In the same year, they developed another method to obtain PEF, which had similar properties to their prepared PEFs. The product was soluble in TFA and 1,1,1,3,3,3-hexafluoro-2-propanol (HFP) but insoluble in TCE; however, the difference was that the DPn was very low (about 70) (Gandini et al. 2009a).
Gomes et al. (2011) used a similar method to get PEF with a number-average molecular weight (Mn) of 22400, weight-average molecular weight (Mw) of 44500 and polydispersity index (PDI) of 1.99. The peak value of X-ray diffraction was obtained at 2θ=16.0, 20.1, and 27.8°. In addition, the thermogravimetric analysis (TGA) and DSC measurements revealed that the thermal decomposition (Tdi), maximum decomposition temperature (Td), Tg, crystallization temperature (Tc), and Tm was 300, 398, 80, 165, and 215 °C. The Tg as well as the Tc only appeared on the second DSC scan trace after quenching the melted polymer, indicating that the obtained PEF had a high degree of crystallinity.
In addition to the two-step method, direct esterification is also an important method to prepare PEF. For example, Jiang et al. (2012) prepared the PEF via direct esterification method. The obtained PEF had a specific viscosity (ηsp) of 1.20 dL/g, and its Mn, Mw, and PDI values were 105300, 252000, and 2.39. In addition, the tensile test of PEF samples showed that PEF was a thermoplastic polyester material. Its tensile modulus, tensile strength, and elongation at break of PEF were 2070 MPa. 66.7 MPa, and 4.2%, respectively. Codou et al. (2014) also obtained some melt-crystallized and glass-crystallized PEF samples via the direct esterification method. The products exhibited similar crystallinities and unit cell structures with those reported for PEF obtained by transesterification (Gandini et al. 2009b; Papageorgiou et al. 2014). In addition, the non-isothermal data analysis of PEF was carried out by DSC, and a new calculation method was used to evaluate the Hoffman-Lauritzen theory. The evaluation equation of the method showed that the effective activation energy of the growth rate depended on the melting temperature (Tm) and the assumed temperature (T∞), and results showed the growth kinetics of the crystals were different during the two crystallization processes. Nevertheless, PEF crystals from the glass temperature and melting temperature had a similar structure. Furthermore, compared with melt crystallization rate of PEF, the experimental data of glass crystallization was in good agreement with the predicted value of Hoffman-Lauritzen rate equation (Codou et al. 2014). The main reason for the deviation of melt crystallization was that the secondary nucleation limited the transverse growth rate (Codou et al. 2014).
In order to further understand the unique characteristics of the PEF, its crystal structure, dynamics, structure-property correlations, and penetrant transport have attracted much attention (Burgess et al. 2014, 2015; Tsanaktsis et al. 2015a; Mao et al. 2016; Burgess et al. 2016; Araujo et al. 2018; Rosenboom et al. 2018). For example, Mao et al. (2016) determined the crystal structure of PEF with X-ray fiber diffraction (XRFD) and molecular model, which was composed of monoclinic unit cell with a space group of P21, whose cell parameter was a=5.784 Å, b=6.780 Å, c=20.296 Å, and γ=103.3º, and the density of this crystalline structure was 1.562 g/cm3. Araujo et al. (2018) revealed that the PEF presented strained synFDCA-transEG extended zigzag conformation in the crystalline region, which was favorable in the crystal but was not feasible for single chain due to the formation of a large array of C−H···O bonds. In the amorphous regions, PEF chains tended to the antiFDCA-gaucheEG helical conformation. Compared with PET, PEF with a high energy difference among the amorphous and crystalline chains resulted in a higher crystallization temperature, while strain-induced crystallization required a higher tensile ratio. The use of novel inelastic neutron scattering (INS) spectroscopy further indicated that the PEF chains of several low frequency vibration modes was less flexible than that of PET, so that the gas barrier properties were higher.
What’s more, a preliminary experiment and analysis of uniaxial stretching of PEF were carried out by Menager et al. (2018). It turned out that strain-induced crystallization (SIC) can develop under appropriate stretching conditions with PEF and controlled in a manner comparable to that of PET to promote strain hardening. According to the tensile conditions of PEF, the crystals with different morphologies were formed. On the other hand, the three-phase models can be well applied to the stretched PEF. The mobile amorphous content was between 0.50 and 0.65, and the rigid amorphous content was between 0.15 and 0.30. In addition, when the tensile condition was at the most rubbery state of polyester, the mobile amorphous content in the polymer was much higher, reaching 0.70.
In addition, correlations between molecular weight and solution viscosity for PEF in comparison to PET were studied by van Berkel et al. (2018). Their research revealed that the melt viscosity of PEF was a function of molecular weight, and the temperature-dependence of the melt viscosity showed a higher activation energy for PEF, and the shear rate dependence indicated a lower reputation time. They found that the PEF with sufficient molecular weight displayed ductility but overall a higher tendency to brittleness than PET in the glassy state, which resulted in a significantly higher yield stress combined with a higher strain rate dependence.
However, PEF has some inherent shortcomings in thermomechanical properties, such as slow melt crystallization and brittleness, which may limit the application of PEF. In order to strengthen PEF, some PEF blends or copolymers have been developed. (Sousa et al. 2013; Yu et al. 2013; Matos et al. 2014; Sousa et al. 2016a; Wang et al. 2017a; Poulopoulou et al. 2019b; Xie et al. 2019). For example, Yu et al. (2013) prepared some poly(ethylene 2,5-furandicarboxylate-co-ethylene succinate) (PEFS) copolymers with FDCA and succinic acid via melt polycondensation in the presence EG. Random copolymers with number-average molecular weights exceeding 25600 and thermal stability exceeding 300°C were obtained.
Sousa et al. (2013) prepared a straightforward partial substitution of non-renewable PET by renewable homologous PEF through random copolymerisation of bis(2-hydroxyethyl) terephthalate (BHETP) and bis(hydroxyethyl)-2,5-furandicarboxylate (BHEFDC) (Fig. 2). They found that the isolation yields of all copolyesters were around 80%, quite independent of the stoichiometric ratio of BHETP/BHEFDC used. This was an indication of the successful incorporation of both aromatic and furanic units into the copolyesters backbone.
Fig. 2. Polytransesterification reaction of BHEFC with BHETP
The Mw of these polyesters ranged between 13000 and 47000, and the polydispersity index (Mw/Mn) was close to 2. They found that the most relevant copolyesters corresponded to PET-ran-PEF 4/1, whose chemical structure contained 20% of aromatic and furanic units. Its Tg, Tc and Tm were 62.4, 125.2, and 220.1 °C, respectively, which were very similar to the properties of commercial PET. Furthermore, the thermal stability of the PET-ran-PEF 4/1 was as high as 260°C and the operating temperature determined by dynamic mechanical analyses ranged from -40.9 to 67.5 °C, which were very close to those of PET.
Matos et al. (2014) developed some poly(ethylene 2,5-furandicarboxylate)-co-poly(lactic acid) copolyesters based on both PEF and poly(lactic acid) (PLA) to enhance their degradability. They found that the obtained copolyesters were amorphous polymers with high glass transition temperatures and had improved degradability with compared with PEF homopolyester counterpart.
In addition, Sousa et al. (2016a) reported the synthesis of fully renewable-based poly((ether)ester)s (PEEs) from FDCA, poly(ethylene glycol) (PEG), and isosorbide by polycondensation reactions (Fig. 3). Poly((poly(ethylene glycol)) 2,5-furandicarboxylate) (PEGF) polymers were obtained by polycondensation of FDCA with PEG or a mixture of PEG and isosorbide, and the isolation yield was about 65%. Furthermore, the PEGF polymers had a number-average molecular weight (Mn) in the range of 3300 to 17300, and the polydispersity values were close to 1.4, which was determined by size-exclusion chromatography in CH2Cl2/CHCl3/HFP. The properties of PEGF material could be easily adjusted by changing the chain length of PEG and adding isosorbide or not. The product had comparable or better thermal properties than that of the fossil-PEE, and the diluted PEGISF200 chloroform solution can be cast directly on the surface of a Teflon plate to form a film. PEGF series polyesters was found to be a single Tg and below room temperature by DSC thermograms analysis. The crystallization domains were formed by a sufficiently long chain (soft PEG segments) modification so that the Tm of the PEGF200 was about 49 °C. However, due to the enrichment of the second amorphous phase in the stiff isosorbide-furandicarboxylate segments, the existence of PEG2000 and the amorphous domains of PEG soft segments, poly((poly(ethylene glycol) 2,5-furandicarboxylate)-co-poly(isosorbide 2,5-furandicarboxylate) (PEGISF2000) exhibited different thermal properties, resulting in a Tg value of 168 °C, while Tm and Tg were about 40 °C and -26°C. In addition, the XRD results revealed that the semicrystalline character of PEGF2000 and PEGISF2000 has two sharp signals at 2θ≈19 and 23°, which was similar to that of PEG2000 polyester. Additionally, XRD analysis of PEGF2000 and PEGISF2000 revealed an amorphous center at about 20°, whereas the more crystalline PEG precursor did not appear.
Fig. 3. Synthesis of two series of FDCA-based poly((ether)ester)s (PEGF and PEGISF2000) by polycondensation reaction
Wang et al. (2017a) prepared a series of multiblock copolymers poly(ethylene 2,5-furandicarboxylate)-poly(ethylene glycol) (PEFEGs) from FDCA, EG, and PEG through a two-step melt polycondensation method (Fig. 4). TGA analysis showed that PEFEGs had good thermal stability. Additionally, by changing the chain length and content of the PEG, the properties of PEFEGs, including melt crystallization peak, melt peak, and enthalpy of PEG block, were improved. It was also found that the elongation values at break of PEFEGs were significantly higher than PEF, such that they exhibited excellent shape memory properties.
Fig. 4. Synthetic route to PEFEGs copolyesters
More recently, Xie et al. (2019) modified PEF with biobased 1,5-pentanediol via melt copolycondensation to improve its tensile strength and O2 barrier property. The obtained poly(ethylene-co-1,5-pentylene 2,5-furandicarboxylate)s (PEPeFs) demonstrated that the amorphous copolyesters had excellent thermal stability. They found that the PEF-rich PEPeFs (ϕPeF 9-47%) showed greatly improved elongation at break, enhanced impact strength, improved O2 gas barrier property, high Young’s modulus and yielding strength in comparison with PEF. Poulopoulou et al. (2019b) revealed that the PEF-PPF copolyester with the belending of PEF and poly(propylene 2,5-furandicarboxylate) (PPF) showed a single composition-related Tg.
1,4-cyclohexanedimethanol (CHDM) and 1,4-cyclohexanedicarboxylic acid (CHDA) were also selected to be a copolyester monomer for the modification of PEF in recent years (Fig. 5) (Hong et al. 2016; Wang et al. 2016, 2018c). Wang et al. (2016) modified PEFs with CHDM to improve their mechanical and barrier properties. They found that the composition of the resulted poly(ethylene-co-1,4-cyclohexanedimethylene 2,5-furandicarboxylate)s (PECFs) had the characteristics of random polyester, and the composition of PECFs could be tunable by adjusting the molar amount of CHDM and dimethyl furan-2,5-dicarboxylate (DMFD). Also, the crystallinity, tensile modulus, and strength of the PECFs can be gradually reduced or increased by adjusting the amount of CHDM (0-59% or 76-100%), as revealed by DSC and tensile testing results. Furthermore, PECF-32 and PECF-76 exhibited better properties compared to PEN and PEI by barrier performance studies of O2 and CO2 permeability in PEF, which made PECFs more useful as a packaging material. Hong et al. (2016) also modified PEF with CHDM to obtain some copolyesters via esterification and polycondensation processes. They found that the CHDM had a positive effect on the previously seen drawbacks of furan-based polyesters, and the obtained co-polyester showed a dramatic increase in elongation at break (79±18%) and impact strength (4.0±0.1%). They thought that the cyclohexane ring conformational transition in the co-polyester chains containing CHDM improved the polyester’s chain mobility and resulted in the ductility in the co-polyester. In addition, the results also revealed that the O2 permeability and amounts of acetaldehyde (AA) content of the co-polyester-containing bottle was considerably lower than that of the PET-containing bottle. Therefore, they deemed that the obtained furan-based co-polyester was superior in both sustainability and performance and showed no limits in applicability for bottles and packaging. In addition, Wang et al. (2018c) synthesized a series of poly(ethylene 2,5-furandicarboxylate-co-ethylene 1,4-cyclohexanedicarboxylate)s (PEFCs) from FDCA, EG and CHDA through a two-melt polycondensation. Their results showed that all PEFCs displayed the properties of a random copolymer, and the composition of PECFs could be tunable by adjusting the molar amount of FDCA and CHDA. The obtained PEFCs were amorphous, and the Tg decreased with the increase of CHDA content, but its thermal stability did not change. Furthermore, it was found that the obtained PEFC-30 was a toughened plastic with tensile modulus and elongation of 801 MPa and 667%, respectively.
Fig. 5. Step-growth polycondensation of furan-based co-polyesters via melt transesterification and polycondensation processes
SYNTHESIS OF POLY(PROPYLENE 2,5-FURANDICARBOXYLATE) (PPF)
Polyesters derived from 1,3-propanediol (PD) were just reported in recent years due to the difficulties in the preparation of PD in sufficient quantities and purity (Bikiaris et al. 2008). However, the research of PD polymers has attracted wide attention in academic and industry, as the attractive process for producing PD from renewable resources has been developed (Haas et al. 2005; Papageorgiou and Bikiaris 2005; Tsanaktsis et al. 2015b; Poulopoulou et al. 2018). Poly(propylene 2,5-furan dicarboxylate) (PPF) is a thermoplastic polyester, and its synthesis and structural, thermal, and mechanical properties have been reported in recent years (Gandini et al. 2009a; Gomes et al. 2011; Jiang et al. 2012; Papageorgiou et al. 2015a; Vannini et al. 2015; Sousa et al. 2016b; Terzopoulou et al. 2017b; Guidotti et al. 2018; Papadopoulos et al. 2018; Wang et al. 2019).
Gandinii et al. (2009a) first reported the synthesis of PPF in 2009. They found that the prepared PPF had similar properties to PEF, except for the characteristic resonance related to the additional intermediate CH2 groups, which can be obtained by FTIR and NMR spectra analysis of PPF. The Tm and Tg values of PPF were 174°C and 65°C, respectively, which were lower than their PEF homologues. This was attributed to the introduction of methylene groups in each polymer unit, resulting in higher macromolecular flexibility. Gomes et al. (2011) also synthesized some PPF polyesters with Mn and Mw values of 21600 and 27600. Analysis of the thermal properties revealed that PPF exhibited high crystallization ability, and its Tg as well as Tc appeared only on the second DSC scan trajectory after molten PPF quenching. In addition, Vannini et al. (2015) found that the PPF copolyester displayed excellent barrier properties due to their special crystal phase, which can be used as packaging material due to their properties of exceptionally low oxygen transmission rates and good impermeability to water vapors.
Papageorgiou et al. (2015a) reported the synthesis of PPF, and they compared their thermal behavior and solid-state structure with its terephthalate (PPT) and naphthalate (PPN) homologues. Through the structural analysis, they found that the final properties of the prepared samples were affected by different structural rings and ultimate chain stiffness. Compared with PPF and PPT, the PPN with higher rigidity showed higher Tg and thermal stability. The equilibrium melting point of PPF was obtained by the Hoffman-Weeks method at 199 ºC, and the heat of fusion of pure crystalline PPF polyester was calculated to be 141.8 J/g or 27.8 KJ/mol. In addition, the spherulites of PPFs at different temperatures were observed by polarized light microscopy (PLM), which revealed that PPF polyester formed small spherulites and had a slow growth rate. With the increase of temperature, the growth rate of PPN was faster and the spherulite was larger, and the band appeared in the spherulite grown above 160 °C. Then the PPT exhibited faster growth rates and banded spherulites, and as the crystallization temperature increased, the band spacing increased.
Recently, Wang et al. (2019) studied the effects of various PD on the properties of PPF. In this research, the FDCA polyester with different thermal properties was prepared by using FDCA and different substituent PD as the starting material, which had a similar skeleton structure to the PPF (Fig. 6). They found that the Tg and crystallinity of poly(neopentyl glycol furandicarboxylate) (PNF) was increased compared to PPF, whereas the Tg and crystallinity of poly(2-ethyl-2-butyl-1,3-propylene furandicarboxylate) (PEBF) with more -CH3 substituted groups decreased compared with PNF. In this study, PNF exhibited high crystallinity with a ΔHm value of 32.1 J/g, which had the highest Tg (70°C) and Tm (201°C). In contrast, PPF had a low ΔHm value of 0.9 J/g and Tm of 173°C, while poly(2-methyl-1,3-propylene furandicarboxylate) (PMF) was a non-crystalline polyester. In addition, the gas barrier results revealed that PPF polyester had the best gas barrier property, followed by that of PMF and PNF, mainly due to the introduction of side chain -CH3 substituents resulting in an increase in β relaxation and fractional free volume.
Fig. 6. Illustration for the synthesis of polyesters from FDCA and different diols (Reprinted with permission from Wang et al. 2019, American Chemical Society)
In addition, copolyesters based on FDCA and PD were also studied in recent years (Jiang et al. 2015; Lomeli-Rodriguez et al. 2016; Wang et al. 2017b; Jia et al. 2019; Righetti et al. 2019). For example, Jiang et al. (2015) studied the synthesis, structure, and properties of novel copolyesters from TPA, FDCA, and PD via direct esterification method by using tetrabutyl titanate (TBT) as catalyst. They found that the obtained copolyester was random copolymers with an average Mw more than 1×104 g/mol, which had similar thermal stabilities and higher elongation at break to corresponding homopolyesters. Wang et al. (2017b) used 2,2,4,4-tetramethyl-1,3-cyclobutanediol (CBDO) to modify PPF to get the copolyester of PPTF. They found that PPTF had the characteristics of random polyester. On the other hand, the semi-crystalline PPF can be made into a fully crystalline polyester by the addition of 10% CBDO, and it was easily to obtain higher transparency, and the Tg increased from 55.5 °C for PPF to 63.5 °C for PPTF. They also found that the O2 and CO2 barrier properties would decrease by the addition of CBDO units, but the modified copolyester still showed good barrier properties.
SYNTHESIS OF POLY(BUTYLENE 2,5-FURANDICARBOXYLATE) (PBF)
In earlier studies, it was found that poly(butylene terephthlate) (PBT) is an important semi-crystalline engineering thermoplastic with the advantages of high rigidity, low humidity, excellent properties, and favorable for processing and production (Tomar and Maiti 2010a; Sharma and Maiti 2013). Moreover, these characteristics were stable in a wide range of temperature and humidity. Compared with PET, PBT had more advantages, mainly depending on its facile crystallinity and easy dyeability, which made it the most commonly used durable product formed by inject molding (Tomar and Maiti 2010b; Sharma and Maiti 2013). However, in the current study, it can be found that the notched Izod impact strength of PBT was very low (Tomar and Maiti 2007, 2008). Therefore, the synthesis of poly(butylene 2,5-furandicarboxylate) (PBF) as PBT counterpart, from FDCA and 1,4-butanediol (BDO) has attracted much attention (Ma et al. 2012; Morales-Huerta et al. 2016a; Zhu et al. 2013).
Ma et al. (2012) successfully synthesized furan derivative polyester, PBF, and its crystallization properties was explored. The results of 1H NMR at 750 MHz showed that chain-end groups were detected in different polymerization stages, which clearly revealed the process of polymerization. By further testing, it is revealed that the polymer had high tendency to form crystalline structured (Tm=130 °C and 169 °C; Tg=32°C), whose butylene segment is in a folded conformation, which was beneficial to the continuous fiber extraction from the melt of PBF.
Zhu et al. (2013) also prepared bio-based polyester PBF with 2,5-furandicarboxylic acid (FDCA) and BDO as raw materials via a two-stage polymerization. The obtained PBF had good thermal stability, strength, and ductility. Its melting point was around 172 °C, and a maximum degradation rate occurred at 428 °C. In addition, it was found that the obtained PBF was triclinic (a= 4.78(3) Å, b= 6.03(5) Å, c= 12.3(1) Å, α=110.1(2)°, β= 121.1(3)°, γ= 100.6(2)°), which was very similar to the α- and β-forms of PBT.
In addition, Morales-Huerta et al. (2016a) prepared cyclic ethylene and butylene 2,5-furandicarboxylate oligoesters and then converted them to furan-based polyesters of PEF and PBF by ring opening polymerization (ROP). High yields of cyclic oligoesters were obtained by both high dilution condensation and thermal cyclodepolymerization methods, and they were found to be crystalline compounds melting within the 140 to 200 °C range. The obtained PEF and PBF by ROP had weight-average molecular weights between 50000 and 60000 g/mol.
The existing research shows that PBF is a semi-crystalline furanic polyester with Tg near 40 °C, with a high crystallization rate and a similar structure to PBT, which is a good production soft drink bottle or food packaging material. However, one of the main disadvantages of PBF is that the structure of the aliphatic chain of diol leads to its poor thermal performance, which is one of the main disadvantages of the most common aliphatic polyester. In order to extend the application market of polyester materials, the synthesis of a series of copolyesters has attracted the attention of researchers (Zhou et al. 2013; Wu et al. 2014; Diao et al. 2016; Morales-Huerta et al. 2016b; Kwiatkowska et al. 2017; Hu et al. 2018; Matos et al. 2018; Morales-Huerta et al. 2018; Sousa et al. 2018; Hu et al. 2019c; Poulopoulou et al. 2019a).
For example, some poly(butylene adipate-co-butylene furandicarboxylate) (PBAFs) copolymers with high molecular weight were synthesized (Zhou et al. 2013). They found that all PBAFs presented random structures, and their composition could be adjusted according to the feed molar ratio of the diacid monomers. The crystallization ability and melting temperature of PBAFs decreased firstly with the increasing of FDCA content, and the trend was reversed when the content of FDCA exceeded 70%. The semicrystalline copolyester with high FDCA content had higher Tg, good tensile modulus, and strength. It was worth noting that the obtained copolyesters with 0 to 50% of FDCA content were biodegradability.
Wu et al. (2014) obtained some high molecular weight biobased polyesters of PBF and aliphatic-aromatic copolyesters, including poly(butylene adipate-co-butylene 2,5-furandicarboxylate)s (PBAFs) with adipic acid, BDO and highly purified biobased monomer of 2,5-furandicarboxylic acid (FA) via a two-step melt polycondensation reaction. The solubility of PBAF in 5 mg/mL organic solvents was determined, and the results showed that the solubility was related to the composition. In the case where the ϕBF did not exceed 50%, the PBAFs had high elastic deformation and rebound resilience, and the tensile properties (E=18 to 160 MPa, σb=9 to 17 MPa, εb=370 to 910%) were comparable to those of poly(butylene adipate). When ϕBF was higher, stress hardening occurred in the tensile test, and PBAFs exhibited a low tensile modulus (42 to 110 MPa), a medium strength (30 to 42 MPa), and a high elongation at break (256%).
Along the same idea, some new PBF-based copolyesters had been synthesized, such as poly(butylene-co-1,4-cyclohexanedimethylene 2,5-furandicarboxylic acid) (PBCFs) and poly(butylene 2,5-furandicarboxylate-co-terephthalate) P(BF-co-BT) (Diao et al. 2016; Morales-Huerta et al. 2016b). For PECFs, which were synthesized from FDCA, BDO, and CHDM, displayed the characteristics of random polyester with a triad component, and by changing the molar content of BDO and CHDM, the structural properties of PECF will vary, such as the average sequence length and monomer percentage (Fig. 7). It was found that the BDO was easily removed in the process of polycondensation, which was beneficial to the formation of long chain containing CHDM unit. Compared with PBF, the PBCFs modified by CHDM had a significantly change in thermal performance and crystallinity. The DSC results revealed that the Tg, Tm, and decomposition temperature changed from 45.7 to 74.4°C, 140.1 to 251.9 °C, and 380.6 to 388 °C respectively with the molar percent of CHDM unit increased from 20% to 70%. The crystallizability and crystal structure was changed by CHDM incorporation. When the 1,4-cyclohexanedimethylene fragment (CF) unit increased to 31%, the copolyester was non-crystalline due to the difficulty in forming long regular segments, which had no cold crystallization and melting behavior. Nevertheless, the DSC analysis indicated that these copolyesters had multiple melting behavior with increase of CF content, and the shoulder peaks appeared on the main melting point side, mainly due to partial melt recrystallization and final melting (Diao et al. 2016). The copolymer P(BF-co-BT), which was prepared from cyclic oligomers of butylene 2,5-furandicarboxylate c(BF)n and bytylene terephthalate c(BT)n via ring opening polymerization (ROP), had an average molecular weight in the range of 55000 and 80000, displaying a good thermal stability (Fig. 8). Although the homopolymer PBF was highly crystalline, the crystallinity of the copolyesters was largely repressed due to the presence of furanoate units. It was found that both Tm and Tg of the copolyesters varied steadily according to the furanoate/terephthalate ratio and the values were comparable to those reported for similar copolyesters prepared by melt polycondensation by other authors (Morales-Huerta et al. 2016b).
Fig. 7. (a) Polymer synthesis from FDCA, CHDM and BDO; (b) Synthetic route via ROP toward aromatic copolyesters containing butylene 2,5-furandicarboxylate and butylene terephthalate units (Reprinted with permission from Diao et al. 2016; Morales-Huerta et al. 2016b, American Chemical Society)
In a similar vein, Morales-Huerta et al. (2018) prepared the cyclic oligo(butylene 2,5-furandicarboxylate) and ε-caprolactone copolyester (PCL-PBF) via enzymatic ring-opening polymerization without using organic solvents and mental catalysts, making the polymer greener than conventional polycondensation methods. They found that the PCL-PBF copolyesters with wide and high yield molecular weight were obtained. The copolyesters were semi-crystalline, their monomers were distributed in blocks on the polymer chain, and their Tm and Tg values were between the PCL and PBF polyesters. More importantly, when an appropriate amount of furanoate was added to the chain, the Tg of PCL was significantly increased, which was related to the physical behavior of the polyester. They deemed that the obtained copolyester had the potential use in packaging and biomedicine due to their excellent properties.
Kwiatkowska et al. (2017) prepared poly(butylene 2,5-furanoate)-block-(dimerized fatty acid) (PBF-b-FAD) polymers with different properties by polycondensation of FDCA, 1,4-butanediol (1,4-BD), and dimerized fatty acid diol (FAD) under relatively mild and excessive diol conditions. A multi-block structure with ester groups was formed between the PBF blocks by the introduction of FAD and had a randomly distributed crystallizable rigid and non-crystalline flexible segment. It was found that the thermal performance of the PBF-b-FAD polymers was related not only to the molar amount of introduced FAD, but it also was affected by the thermal treatment of material. In addition, the mechanical properties of the macromolecular structure PBF-b-FAD polyesters varied with the molar ratio of PBF to FAD, which had a medium tensile strength and a relatively high deformation and damping at room temperature.
Matos et al. (2018), based on the synthesis of polyesters from FDCA, prepared a series of novel nanocomposite materials by introducing bacterial cellulose, such as acetylated bacterial cellulose/poly(butylene 2,5-furandicarboxylate) (Ac-BC/PBF) and acetylated bacterial cellulose/poly(butylene 2,5-furandicarboxylate)-co-(butylene diglycolate)s (Ac-BC/PBF-co-PBDG). The obtained copolyesters had high strength due to the adding of bacterial cellulose fibers. The Tg and Tm of the copolyesters were -25 to 46 °C and 61 to 174 °C, respectively. Furthermore, the nanocomposite copolyesters had better thermal stability (239 to 324 °C), and compared with pure copolyesters, the Young’s modulus of nanocomposite copolyesters were as high as 1239 MPa.
In addition, Hu et al. (2018) prepared the poly(butylene furandicarboxylate)-b-poly(ethylene glycol), poly(butylene furandicarboxylate-co-glycolate) (PBFGA), and poly(butylene furandicarboxylate)-b-poly(glycolic acid) (PBF-b-GA) copolyesters (Hu et al. 2019c), which displayed good thermal, mechanical, barrier and degradable properties.
More recently, the PBF blends were also reported by Sousa et al. (2018) and Poulopoulou et al. (2019a), which showed some advantageous processing features as revealed by the lower melting temperature and higher thermal stability as compared with PBF.
SYNTHESIS OF POLY(HEXYLENE 2,5-FURANDICARBOXYLATE) (PHF)
Poly(hexylene 2,5-furandicarboxylate) (PHF) is a furan-type polyester that is synthesized by monomers of renewable resources and has a similar structure to poly(hexamethylene terephthalate) (PHT). PHF was first synthesized as reported by Moore and Kelly (1978), but in recent years, a series of poly(alkylene 2,5-fruandicarboxylate)s including PHF have been synthesized and characterized (Jiang et al. 2012; Papageorgiou et al. 2015b; Terzopoulou et al. 2016b; Haernvall et al. 2017; Zhang et al. 2019b).
For example, Jiang et al. (2012) reported the synthesis and characterization of some poly(alkylene 2,5-fruandicarboxylate)s including PHF. The obtained PHF with a Mw of 66700 had a tensile modulus, tensile strength, and elongation at break of 493 MPa, 35.5 MPa, and 210%, respectively. The DSC results showed that the Tg and Tm were 28.1 and 148.2 °C, respectively. Papageorgiou et al. (2015b) studied the crystallization and thermal degradation behavior of PHF in 2015. They found that the Tm and Tg of the obtained PHF were 145 °C and 7 °C, respectively. Though the DSC analysis, it was revealed that PHF exhibited multiple melting behavior after the crystallization process, which was mainly due to partial melting, recrystallization, and final melting. It was found that the heat of fusion of pure crystalline PHF polyester was 34 kJ/mol, and further calculation gave an equilibrium melting temperature of 157 °C.
More recently, PHF was synthesized from FDCA and 1,6-hexanediol (HDO) via direct esterification (Fig. 8) (Zhang et al. 2019b). The obtained PHF had typical (110) plane, (010) plane, and (111) plane at 2θ = 13.78° (d=6.42 Å), 17.06° (d=5.19 Å), and 24.9° (d=3.58 Å) with intrinsic viscosity of 0.803 dL/g, average Young’s modulus of 479 MPa, and maximum tensile strength pf 36.5 MPa. Its elongation at break was 216%, which was significantly higher than that of PET (90%) and PEF (3%). It was found that the high intrinsic viscosity caused the increase in the enthalpy of fusion with respect to as-synthesized PEF, in this case, the melting peak was particularly sharp, and the Tg and Tm was about 48 and 144 °C, which were relatively lower when compared to those of PEF.