Relationship between Composition and Environmental Degradation of Poly(isosorbide-co-diol oxalate) (PISOX) Copolyesters

To reduce the global CO2 footprint of plastics, bio- and CO2-based feedstock are considered the most important design features for plastics. Oxalic acid from CO2 and isosorbide from biomass are interesting rigid building blocks for high Tg polyesters. The biodegradability of a family of novel fully renewable (bio- and CO2-based) poly(isosorbide-co-diol) oxalate (PISOX-diol) copolyesters was studied. We systematically investigated the effects of the composition on biodegradation at ambient temperature in soil for PISOX (co)polyesters. Results show that the lag phase of PISOX (co)polyester biodegradation varies from 0 to 7 weeks. All (co)polyesters undergo over 80% mineralization within 180 days (faster than the cellulose reference) except one composition with the cyclic codiol 1,4-cyclohexanedimethanol (CHDM). Their relatively fast degradability is independent of the type of noncyclic codiol and results from facile nonenzymatic hydrolysis of oxalate ester bonds (especially oxalate isosorbide bonds), which mostly hydrolyzed completely within 180 days. On the other hand, partially replacing oxalate with terephthalate units enhances the polymer’s resistance to hydrolysis and its biodegradability in soil. Our study demonstrates the potential for tuning PISOX copolyester structures to design biodegradable plastics with improved thermal, mechanical, and barrier properties.


Table of Content
Tetrahydrofuran was purchased from VWR. DMSO-d6 was purchased from Eurisotop.
All chemicals were used without prior purification.

Synthesis procedure of O,O'-(cyclohexane-1,4-diylbis(methylene)) dimethyl dioxalate
Methyl oxalyl chloride (3.675 g, 30.0 mmol, 2.0 equiv.) was dissolved in tetrahydrofuran (THF).To this, a solution of 1,4-cyclohexanedimethanol (2.163 g, 15.0 mmol, 1.0 equiv.)and triethylamine (3.066 g, 30.3 mmol, 2.02 equiv.),dissolved in approximately 30 ml THF, was slowly added with a dropping funnel over the course of one hour.During the addition, a white precipitate formed.Additional THF was added if the stirring of the reaction mixture was inhibited by the precipitate.After complete addition, the reaction mixture was stirred for another hour.The white precipitate was filtered off and the THF was evaporated.The residual, viscous yellow liquid was redissolved in dichloromethane (DCM) and extracted with water and brine.The organic layer was dried over Na2SO4 and DCM was evaporated.Due to some remaining impurities detected by 1 H NMR, the white solid was washed with cyclohexane and diethyl ether, which predominately dissolved the cis-stereoisomer of the reaction product.In order to obtain a representative mixture of both cis and trans isomers of the reaction product, the cyclohexane-diethyl ether washing fraction was partially evaporated until approximately 10 ml of solvent remained.Some crystals were precipitating from the solution, and the flask was placed in the refrigerator overnight.
The crystals were then filtered off and combined with the previously obtained crystals of the trans product.To ensure a homogenous reaction product, the crystals were dissolved in a small amount of DCM.After evaporation of the solvent, the white crystals (2.67 g, 56.3%) were dried in vacuo.A cis/trans ratio of 21/79 was determined by 1 H NMR.         peaks is similar to that of TPA (4.00/4=1.00)(ratio,98:100), which suggests that the amount of OX units decreased significantly after biodegradation comparing to their original ratio before biodegradation (54:100).The shortest lag phase of PISOX100 homopolyester and the fact that PISOX-NPG37 has a shorter lag phase than PISOX-NPG50 suggests that a higher isosorbide content versus neopentyl glycol (NPG) favours biodegradation due to the strong hydrophilicity/hygroscopicity of isosorbide.The small variations in lag phase of PISOX-PrDO, PISOX-PDO, PISOX-HDP show that intermediate chain length of the co-diol (i.e.C3 to C6, and the ratio from 25% to 50%) have little effect on the biodegradability of PISOX copolyesters.The relatively short lag phase of PISOX-DEG37.5 suggests that the presence of oxygen in ether diols could affect the interaction between polymer and enzyme to facilitate biodegradation.The relatively long lag phase of PISOX-NPG50 (especially compared to PISOX-PrDO50) and the slow nonenzymatic hydrolysis may be attributed to the higher steric hindrance (relative to the linear structure) as a result of the two methyl branches.
The cyclic building blocks are too apolar (hydrophobic) and may thus hinder the biodegradation/hydrolysis of the copolyesters.As a result, the lag phase of PISOX-CHDM50 and PISOXT54-PrDO49 increased.Compared to PISOX-PrDO49, the lag phase of the biodegradation curve for PISOXT54-PrDO49 was notably prolonged (over three times to 70 days).This resulted from replacing easily hydrolysable ester bonds (i.e.oxalic esters) with aromatic acid esters (i.e.terephthalate esters).This supports the conclusion that oxalate esters can play a crucial role in faster biodegradation.

Figure S. 14 Figure S. 15 Figure S. 16 3 Figure S. 17
Figure S. 14 Individual yield of monomers, isosorbide and co-diol for PISOX-DEG37.5 (a), PISOX-NPG37 (b) and PISOX-PDO36.4(c) during 6-month hydrolysis at 25 °C in D2O as percentage of their theoretical maximum release.Error bars represent standard deviation of triplicate hydrolysis experiments.* Percentages of dissolved NPG and DEG relative to the total amount of hydrolysed diols in time are shown in Figure S. 16.

Table S 1
. Properties of soil.

Table S 2
. Constituents of mineral salts solution used to adjust soil moisture (OECD, 2014).