Pyridine dicarboxylic acid derived polyesters: prospects for developing safe, circular and sustainable materials

Over the last decades, academic and industrial researchers have made several efforts to investigate the efficient synthesis of aromatic base chemicals using lignocellulosic and other biomass as feedstocks.[1] Despite significant progress, none of the strategies explored have reached commercial-scale production for various reasons. The key issues include a lack of technological developments in obtaining target chemicals with a high yield and selectivity and higher associated production costs compared to benchmarked standards due to complex and multi-step downstream processing. Nevertheless, many bio-based drop-in chemicals and renewable alternatives have been explored for various applications ranging from home & personal care products to automotive & aerospace industries.[2], [∗3], [4].

2,5-Furandicarboxylic acid (2,5-FDCA) is an excellent example of a renewable alternative to terephthalic acid (PTA) in polyester synthesis with superior gas barrier properties and mechanical performance.[5] These characteristics enhance the potential of poly(ethylene furanoate) (PEF) polymers for daily life applications such as bottles, packaging materials, textiles, and coatings.[6].

Pyridine dicarboxylic acid (PDC) is a renewable building block currently under investigation, especially in producing polyester polymers and associated applications. Six isomers are known to exist in the PDC family. Among them, quinolinic acid (2,3-PDC), cinchomeronic acid (3,4-PDC), and dinicotinic acid (3,5-PDC) are rarely investigated as no (recent) literature was found describing their polymer properties. Quinolinic acid is known to be highly toxic (excitotoxin in the central nervous system) explaining its negligence in any applications.[7] However, no clear arguments were found for the other two isomers (3,4 and 3,5-PDC). The remaining three isomers in the PDC, namely, lutidinic acid (2,4-PDC), isocinchomeronic acid (2,5-PDC), and dipicolinic acid (2,6-PDC) have gradually been receiving attention lately in polyester synthesis and exploring their potential applications in packaging and other areas.

Among the three isomers, only 2,6-PDC has been investigated thus far in synthesising polyester using EG. The preliminary data reveals that the polyester has a T_g of 75 ^oC which is 2 ^oC higher than PET and 4 ^oC lower than PEF. The authors did not observe the melting transition in differential scanning calorimetry (DSC) measured up to 300 °C.[8] Unfortunately, no information was available whether the crystallinity can be induced by annealing procedure to determine the T_m. Though the polymer properties of the other two isomers (2,5- and 2,4-PDC) have not yet been investigated, it is too early to establish the structure and property correlations between the isomers.

Next to the biotechnological strategies, the electrochemical process to produce 2,4-PDC and 2,5-PDC via the direct carboxylation of methyl pyridine-2-carboxylate in the presence of CO₂ has been described recently.[14] The selectivity of the carboxylation reaction was tuned by an undivided electrochemical cell, favoring the formation of 2,4-PDC and 2,5-PDC in a 6:1 ratio.[14] The catechol is a key intermediate in this strategy, which can be directly derived from lignin or obtained via the decarboxylation reaction of protocatechuic acid.[15], [16] It was shown that the enzyme pseudomonas catechol 2,3-dioxygenase produces an unstable intermediate, 2-hydroxymuconic semialdehyde, from catechol, which is then cyclized in the presence of ammonia, yielding pyridine 2-carboxylic acid.[17] Besides lignin as a feedstock, an improved method to produce both protocatechuic acid and catechol from glucose via microbial synthesis was also reported by Wensheng and coworkers.[15].

Polycondensation reactions are typically performed at higher temperatures in the presence of a catalyst to achieve high molecular weights and the required properties suitable for application demands. PET is one such polyester that is industrially produced via a melt condensation procedure in the presence of antimony or titanium catalyst at temperatures between 275 – 285 °C under vacuum conditions. The high molecular weight PET (bottle-grade) is achieved further via a solid-state post condensation (SSPC) process. The new, renewable PEF is also synthesized under similar conditions, however, at a slightly lower temperature due to the thermal stability limitations of the furan moiety, resulting nonetheless in a polyester with a superior gas barrier and mechanical performance compared to PET.[5].

In 1992, Briehl and Butenuth used thermogravimetric analysis (TGA) and DSC to examine the thermo-analytical behaviour of all six isomeric PDC thoroughly. The authors found that the isomers 2,4-PDC, 2,5-PDC, and 2,6-PDC with the carboxylic substituent in the 2 position, are thermally unstable at high temperatures (235-255 ^oC) and show low enthalpies of decomposition (melting followed by decomposition.[29] The nitrogen atom in the heterocyclic ring is hypothesized to be involved in the cleavage of CO₂ under elevated temperatures. Among the six isomers, it is found that the least stable isomer is 2,3-PDC (165 ^oC) and the most stable ones are 3,4-PDC and 3,5-PDC having a decomposition temperature above 311 ^oC. The higher thermal stability of the latter isomers are due to the unlikely intermolecular hydrogen bond formation between the nitrogen atom and carboxylic groups in the meta and para positions.[29].

For many decades, there has been an interest in investigating the polycondensation reaction under various conditions, such as low-to-high reaction temperatures, different feed ratios of the monomers employed, types of catalysts and their loading, and use of coupling reagents and solvents. In 1982, Tanaka et al., investigated the use of picryl chloride as a condensation agent for the reaction of 2,5-PDC with different aliphatic diols at room temperature in the presence of pyridine as a solvent (Table 2). A polyester yield of up to 89% and a solution viscosity ca.1.0 was achieved within a reaction time of 5.0 h performed at 30 ^oC.[19] It is believed that the trinitrophenyl ester is an active intermediate that is formed when a carboxylic acid reacts with the salt that is produced by the reaction of pyridine and picryl chloride. The ester intermediate then rapidly reacts with an alcohol to form the new ester linkages building in the polyester chains.[19] Two decades later, Doan et al., followed the same procedure to produce polyesters using 2,6-PDC and various poly(ethylene glycol)s. The authors found that yields were satisfactory when 1.5 equivalents of picryl chloride-to-carboxylic acid groups were used.[21] Isfahani and Faghihi used a different approach to perform the polycondensation reaction. Starting from diacid chlorides of 2,5-PDC, various aromatic diols were reacted in the presence of pyridine as base and N,N’-dimethyl acetamide (DMAc) as solvent. The polyesters had yields above 80% and molecular weights ranged from M_n 14,000 and 22,000 g mol^-1.[22].

The first known melt polycondensation reaction of pyridine dicarboxylic acid was reported 62 years ago. In 1962, Broadhead and coworkers of the Standard Oil company (currently known as ExxonMobil), developed water-soluble resins for the coating industry using 2,5-PDC and diethylene glycol. Without using a catalyst, the two monomers were mixed at a 1:1 ratio and heated to 200 ^oC for 30 min and at 175 ^oC for 8 h resulting in a resin that was used as a protective coating material and eventually dissolved in an aqueous solution at temperatures ranging between 15 °C to 80° C.[18] Using a Titanium(IV) isobutoxide (Ti(O-Bu)₄) or Titanium(IV) isopropoxide (Ti(O-iPr)₄) catalyst in the melt polycondensation reaction between 2,6-PDC and with various aliphatic diols resulted in the formation of the oligomers. The polymerization reaction performed at 260 ^oC using 1,3-propane diol (1,3-PDO) resulted in a material with a glass transition (T_g) 30 ^oC and a melting transition (T_m) of 162 ^oC.[20] In another study, the polycondensation reaction was performed at a slightly lower temperature (180 ^oC). Inherent viscosities in the range 0.29 – 0.30 g dL^-1 were reported when using diols such as EG, 1,4-BDO, and 1,6-HDO.[8].

A patent application from Biome Biopastics, describes the synthesis of a copolyester polybutyrate adipate pyridine dicarboxylate (PBAP) that was prepared using the diethyl ester of 2,4-PDC and 2,5-PDC in combination with diethyl adipate respectively and with 1,4-BDO.[30] In line with previous reports, Ti(O-Bu)₄ was used as a catalyst and the polymerization reaction was performed at 180 ^oC for 17 h. The molecular weights of the resulting polymers were determined by ¹H NMR (using repeating units) and estimated to be 4,000 g mol^-1 for 2,4-PBAP and 4,200 g mol^-1 for 2,5-PBAP. Thermal analyses of 2,5-PBAP showed two distinct T_gs (-35.3 ^oC and 41.9 ^oC) and one T_m at 105.9 ^oC. In the case of 2,4-PBAP, only one T_g at -31.7 ^oC was recorded. The mechanical properties of 2,5-PBAP exhibit a similar Young’s Modulus (90.6 vs. 100.8 MPa) when the polyesters were compared to commercial polybutylene adipate terephthalate (PBAT). However, the tensile strength (expressed in MPa) and elongation at break (%) values of 2,5-PBAP are both far below the benchmark values.[30] Another study describing the use of 2,6-PDC dimethyl ester as a comonomer with feed ratios varying from 20 – 60 mol% to the bis-furoate (main monomer) was reported in 2016 by Bougarech and co-workers.[23] The authors used a second catalyst (Ti(O-Bu)₄) during the polycondensation reaction to address the likelihood of the pyridine moiety deactivating the first catalyst (Zn(OAc)₂ by chelation_. As evidenced by previous reports, increasing pyridinic moieties leads to a decrease in the molecular weight of the resulting copolyesters (Table 2). On the other hand, the amorphous characteristics and high T_g values were attributed to the rigid backbone from the pyridine structure that hampers the crystallization phenomenon.[23].

Recently, Liu and coworkers examined a different strategy to address the challenges associated with using a metal catalyst and the poor thermal stability of pyridine carboxylic acids, i.e., melting followed by decomposition at temperatures typically used in the polycondensation reactions. Instead of starting from the diacid or diester of PDCs (T_{m diacid} 235 – 250 ^oC, T_mdiester 130 – 170 ^oC), the authors used dichlorides of PDCs as the monomers that have lower melting points (T_{m dichloride} 55 – 65 ^oC). The dichloride derivatives are reactive substrates that can be used without a catalyst in polycondensation reactions at temperatures between 115 – 140 ^oC to produce reasonably higher molecular weight polyesters (M_n 24,500 g mol^-1). Using 2,6-PDC dichloride as a comonomer in the copolyester synthesis reduces the molar weight but positively affects the T_g, consistent with previous reports. The liquid water and dynamic vapor sorption exhibited higher values for the copolyesters prepared using pyridine monomers. This could be due to the hydrophilic nature of the pyridine structure in the polymer backbone. Next to this, the presence of a pyridine moiety showed a positive influence in enhancing the water and gas barrier properties of the copolyester.[27] These values were comparable to PET (semicrystalline) but lower than that of PEF. The homo polyester poly(decylene 2,6-pyridinedicarboxylate) – PDe26PD prepared from 2,6-PDC dichloride with 1,10-decane diol (1,10-DDO) is semicrystalline (T_m 110 ^oC) and displayed a molecular weight M_n 24,500 g mol^-1. These values are slightly higher for the 2,5-isomer. (M_n 25,700 g mol^-1 and T_m 124 ^oC). Despite lower thermal stabilities compared to the aromatic analogues (e.g. polydecylene isophthalate – PDeIP), the mechanical properties of PDe26PD such as Young’s modulus (E) were over five times, and the strength (σ_b) 2-4 folds higher compared to PDeIP.[26] The gas barrier properties for the polyester prepared using 2,6-PDC dichloride are found to be superior compared to the one obtained from the 2,5-isomer. The ring-flipping in the non-symmetrical 2,6-isomer is hindered similar to 2,5-FDCA (PEF). Thus the molecular motions are restricted causing lower diffusion coefficients of gases like O₂ and CO₂.[26] The effect of the incorporation of the rigid diol (isosorbide-Is) was found to positively influence the thermal and mechanical properties of the resulting copolyester poly(decylene 2,6-pyridinedicarboxylate isosorbide) – PDe26PDIs.[28] For example, as low as 8 mol% incorporation of Is could increase the thermal stability by 8 ^oC and T_g by 6 ^oC. The mechanical properties were significantly improved in the copolyester compared to the homopolyester (PDe26PDIs vs. PDe26PD), Young’s modulus (E) 1390±170 vs. 770±25 MPa, and mechanical strength (σ_b) 31.6±2.7 vs. 12.5 ± 4.8 MPa. The increase in these values was attributed to the intrinsic rigidity of the isosorbide structure in the polymer backbone and its stronger intramolecular interactions.[28] These results indicate that the physical and chemical properties of pyridinic polyesters can be tuned, expanding the scope of these materials in a wide range of applications.

Though not a polyester, Lubrizol Corporation patented a trimer comprised of 2,5-PDC and 2,6-PDC with 2-ethylhexanol respectively, as suitable additives in formulating lubrication oils for internal combustion engines. The anti-wear and/or extreme pressure performance of PDC-derived trimers was found to be superior compared to the conventional zinc dialkyldithiophosphate (ZDDP) additive which has a detrimental effect on the fuel economy and efficiency.[31].

The potential of PDC to be used as an effective monomer is hindered by the chelation of metal catalysts typically used in polymer synthesis and the poor thermal stability of carboxylic acid groups especially adjacent to the nitrogen atom. Researchers have developed various polymerization strategies to overcome these challenges and improve their effectiveness as monomers in producing polyesters with interesting properties.

• •

  The enzyme CaLB-mediated polymerization, a promising method that uses a lower temperature and solvent, produces pyridinic polyesters with molecular weights up to M_n 14,300 g mol^-1 without metal catalysts.

• •

  The polycondensation reaction, which can be performed at room temperature using reactive picryl chloride as a condensation reagent, is not attractive due to the challenges of handling the highly explosive nature of picryl chloride in large quantities.

• •

  The diacid or diester of PDCs typically requires an esterification or transesterification catalyst and higher temperatures for oligomerization and subsequent polymerization reaction. However, the innovative strategy of employing dichloride derivatives, reactive monomers, is an interesting approach as they do not require a catalyst and can be performed at moderate temperatures.

The rigid structure of PDC provides several advantages when incorporated into a polymer backbone. These advantages are.

• 1.

  Enhanced biodegradability

• 2.

  Chelating agent for extraction of metal ions from the compatible medium

• 3.

  Enhancement in water and gas barrier properties

• 4.

  Higher T_g values

• 5.

  Influence on the stacking/crystallization behavior

• 6.

  Positive effect on liquid water and dynamic vapor sorption due to hydrophilic nature

• 7.

  Tuneable morphology of the polymer