Furan-2,5- and Furan-2,3-dicarboxylate Esters Derived from Marine Biomass as Plasticizers for Poly(vinyl chloride)

Abstract

Esters of furan dicarboxylic acids (DAFs) were synthesized by a one-pot reaction between marine biomass-derived galactaric acid and bioalcohol under solvent-free conditions and were fully characterized. The catalyst amount could be reduced without loss of reaction yields using p-xylene as the material separation agent. Also, a possible mechanism was proposed for the first time. Then the properties of four DAFs as plasticizers on the poly(vinyl chloride) (PVC) matrix were investigated. The experimental results showed that DAFs exhibit competitive efficiencies of plasticization when compared to the most commercialized plasticizer, DOP. It was found that the combination of DAFs and PVC produced homogeneous smooth-surface films, indicating miscibility between them. ATR-FTIR depicted the upshift of carbonyl absorption bands after mixing with the PVC matrix, with a magnitude of at most 18–21 cm^–1. TGA, DSC, and UTM data illustrated equivalent plasticization efficiencies. Due to their small molecular weights, the investigated DAFs are more volatile. However, due to bearing an oxygen atom in the aromatic furan ring, the degree of polarization of DAFs was boosted and helped inhibit leaching into the surrounding media. In brief, these synthetic compounds have promising feasibility as biobased plasticizers. Moreover, another interesting point is that the properties of furan-2,3-dicarboxylic acid derivatives were studied for the first time and herein reported.

Introduction

Defined by IUPAC in 1951, plasticizer is “a substance or material incorporated in a material (usually a plastic or elastomer) to increase its flexibility, workability, or dispensability”.¹ Plasticizers are inserted into polymeric structures without a chemical reaction and remain between the chains to help them separate. As a result, an originally rigid, brittle, and hard polymer then becomes soft, flexible, and easy to process.² In the global market, bis(2-ethylhexyl) phthalate (dioctyl phthalate, DOP), belonging to the phthalate family, is the most commercial. Phthalates dominate the market, encompassing approximately 90% of all plasticizers produced each year.^2,3 The production is so massive, owing to low cost, high demand, and indulging properties exhibited in the polymer matrix. On the contrary, DOP is also well-known for its toxicity in living organisms, eco-unfriendliness, and petroleum-origin.¹⁻⁴ These accumulate and in turn become concerns when DOP plasticizes from the third most-used polymer, poly(vinyl chloride) (PVC). While PVC has a wide range of applications, from industry to medical and housing purposes, on grounds of excellent physical properties, and chemical resistance at a cheap price. A substitute for phthalates, DOP in particular, is inevitable to sustain future demand.²

As the world’s limited amount of fossil resources becomes depleted and the increasing emission of carbon dioxide contributes to global warming,⁵ renewable and sustainable biomass has become a hopeful alternative for fossil resources.^6,7 In particular, biomass-derived carbohydrates have great potential because they are the most abundant among all the biomass-containing carbon components. While they account for 75% of all carbon-based biomass, less than 5% are used for food or nonfood purposes.^8,9 Biomass-derived carbohydrates can be selectively dehydrated into furanic compounds via chemical reaction or biological fermentation pathways,^10,11 and they are highlighted as potential biobased candidates for the replacement of petroleum-based aromatics.¹²⁻¹⁵ Therefore, much effort has been made to transform carbohydrates into furanic fuels and chemicals.¹⁶⁻²¹ Sources for carbohydrates generally come from terrestrial crops and wood (lignocellulose), but recently, another kind of biomass, which is seaweed (macroalgae) from the ocean, has attracted attention in recent years.²² There are several advantages in marine-based macroalgae. First, they have a higher growth rate than land-based crops and trees, allowing for multiple harvests per annum, without feeding additional fertilizer or pesticide or even irrigating.²³ Second, they do not require land area for cultivation, which can prevent land competition with crops and the environmental destruction of forests. Third, the effectiveness of CO₂ remediation, which is crucial to reduce greenhouse gases, is outstanding by comparison with land plants.²⁴ Fourth, their carbohydrate content is high with no lignin; thus, the carbohydrates can be readily extracted from the original feedstock.²⁵ Noticeably, marine-based carbohydrates contain galactose-type monosaccharides such as agarose, while land-based carbohydrates consist of glucose-type monosaccharides such as starch and cellulose. Unlike glucose, galactose is oxidized to the optically inactive galactaric acid (Gal-dA, trivially called mucic acid), and galactaric acid precipitates during oxidation with nitric acid due to its low solubility in water;²⁶ therefore, it can be readily isolated and employed as a starting material for various value-added chemicals.

Furan-2,5-dicarboxylic acid (2,5-FDCA) is a natural di-acid found inside the human body and was prominently named among the top 12 priority biobased molecules for the future “green” industry by the U.S. Department of Energy. 2,5-FDCA is encouraged and gradually replaces the fossil-originated terephthalic acid.^27,28 Studies have been conducted to convert biomass into 2,5-FDCA since 1876. At the time, Fittig and Heinzelman were pioneers who dehydrated galactaric acid with fuming HBr.²⁹ In 2008, Taguchi et al.³³ reported that furan-2,5- and furan-2,3-dicarboxylic acid esters (DAFs) could be coproduced from Gal-dA. Such a revelation not only paved the way for synthesizing another regioisomer of 2,5-FDCA directly from biomass but also introduced 2,3-FDCA as a new building block. Generally speaking, FDCAs are capable of substituting phthalate substances whose market size is now worth millions of U.S. dollars. From the strategic perspectives of marine biomass and FDCAs, four furanic compounds were successfully synthesized for our research, namely, di-n-butyl furan-2,3-dicarboxylate (2,3-DBF), di-n-butyl furan-2,5-dicarboxylate (2,5-DBF), diisoamyl furan-2,3-dicarboxylate (2,3-DIAF), and diisoamyl furan-2,5-dicarboxylate (2,5-DIAF) (depicted in Figure 1). During the synthesis, a possible reaction mechanism was proposed for the first time due to the discovery of by-products, which are alkyl 2-furoates. The four investigated compounds with the poly(vinyl chloride) powder were then utilized to cast thin plastic films followed by a series of evaluating methods. As a further matter, 2,3-FDCA and its derivatives have not been well studied so far, making this class of compounds popular among the scientific society. This report, which comprises the detailed information, is worthy of attention.

Figure 1.

Chemical structures of dialkyl furan dicarboxlates (DAFs) originating from marine biomass and bis(2-ethylhexyl) phthalate (DOP).

This study has focused on the synthesis of furanics from natural materials and on the investigation of their properties. It is literally a combination of bio-basis in the green chemistry with plasticizers (PLS) in the polymer industry, meeting the current research progress. Oils from castor, soybean, and tung tree or multi-carboxylic acids (like citric acid) are good examples for synthesis of bio-based plasticizers.^1,3,29 Our study aims at (a) producing FDCA derivatives from galactaric acid (marine biomass^10,11) in alcoholic media (bio-butyl/isoamyl alcohols^30,31) and then (b) utilizing them to plasticize the rigid polymers. As a result, these bio-plasticizers can be alternatives for petroleum-based phthalates, especially DOP.

Experimental Section

Materials

Activated charcoal (4–14 mesh), 1-butanol (BuOH, >99.5%), galactaric acid (Gal-dA, 97%, also known as mucic acid), 3-methyl-1-butanol (98%, also known as isoamyl alcohol), molecular sieves type 3 Å (UOP, pellets, 3.2 mm), poly(vinyl chloride) (PVC, powder, M_w ∼80000, M_n ∼47000), sodium carbonate (Na₂CO₃, >99.5%), sodium chloride (NaCl, 99%), sulfuric acid 98% fume (H₂SO₄), tetrahydrofuran (THF, anhydrous, inhibitor-free, >99.9%), and p-xylene (PX, 99%) were all purchased from Sigma-Aldrich (Korea). Analytical-grade solvents (n-hexane, ethyl acetate, methanol, etc.) were obtained from Samchun Chemicals (Korea). Bis(2-ethylhexyl) phthalate (DOP, >98%) was purchased from TCI Chemicals (Japan), which was used as the reference plasticizer chemical.

Synthetic Procedure

In a Dean–Stark apparatus, galactaric acid (Gal-dA, 3.0 g, 14 mmol) was suspended in an alcohol medium (324 mmol). The suspension was heated at the boiling point of 1-butanol before adding concentrated H₂SO₄ (1.2 g, 12 mmol) and p-xylene (PX, 3.5 g, 33 mmol) in a dropwise manner. Then the temperature was raised to and was kept at 160 °C. After 10 h, the reaction mixture was allowed to cool down and was diluted with ethyl acetate followed by washing with saturated solution of NaCl and aqueous solution of Na₂CO₃ (10%, w/w). The organic phase was dried over the molecular sieve and evaporated under reduced pressure. The viscous crude product was purified by flash chromatography over silica gel (60 Å, 230–400 mesh) from Sigma-Aldrich (USA) under a UV lamp at 245 and 265 nm simultaneously. Reactions were monitored by thin layer chromatography (TLC, silica gel Merck 60 F₂₅₄ plates), and then spots on TLC were seen under a UV lamp at 254 nm and/or were stained by a solution of potassium permanganate in ethanol.

From the above reaction, 2,5-DBF and 2,3-DBF were obtained. The 2,5-DIAF and 2,3-DIAF could be produced under identical conditions, with isoamyl alcohol instead of butanol (Scheme 1).

Scheme 1. Dehydration of Galactaric Acid To Give Diakyl Furandicarboxylates (DAFs).

Preparation of Plasticized PVC Films

The PVC powder (2.5 g) was dissolved in ∼55 mL of THF together with a plasticizer and was stirred for at least 3 h. The homogeneous solution was then cast on a clean, dustless Petri dish. The solvent was allowed to freely evaporate overnight to obtain flexible films. Then they were put in a vacuum oven at 35 °C for 24 h. The complete formulations of the film-preparing process and designation of plastic films are listed in Table S8.

It is known that, after drying, there is little amount of residual THF solvent. The last 2–3%, which is hard to remove completely, could have an impact on plasticization effects.

Characterizations

After purification, the four target furanic compounds were structurally characterized. ¹H- and ¹³C-NMR spectra were recorded using a Bruker Spectrospin 300 (Germany) in chloroform-d (CDCl₃) as the solvent and deuterated tetramethylsilane (Me₄Si) as the internal standard. The chemical shifts (δ) and coupling constants (J) are expressed in parts per million (ppm) and hertz (Hz), respectively. Infrared spectra (FTIR) were recorded using a Nicolet 6700 FTIR spectrometer from Thermo Fisher Scientific Inc. (USA) with 32 scans, the resolution being 4 cm^–1. The mass spectroscopy spectra (EI-MS+) were recorded on a Shimadzu GCMS-QP-2010 Ultra integrated with a Shimadzu GC for mass spectrometer GC-2010 Plus (Japan).

Morphology studies of platinum-coated (1–2 nm thick) film samples were conducted on a field-emission scanning electron microscopic (FE-SEM) system JSM-6701F, provided by JEOL (Japan). Infrared spectra (FTIR) were recorded using the same machine aforementioned but equipped with a Smart MIRacle Accessory ZnSe crystal for the attenuated total reflectance (ATR) technique.

A TGA 4000 system from PerkinElmer (USA) was used to perform thermogravimetric analyses (TGA). About 12–15 mg of each film in the absence of a suitable stabilizer was put into a ceramic cup without a lid, from the same instrument provider. Heating the sample from 25 to 700 °C at 10 °C/min under nitrogen gas flow (20 mL/min) gave information about thermal degradation of the PVC-based material.

Differential scanning calorimetry (DSC) was performed on a DSC 8000 instrument from PerkinElmer (USA). Small film pieces with weights of 7–10 mg were thermally scanned from −50 to 200 °C at a 10 °C/min rate under nitrogen gas flow (20 mL/min).

Tensile properties of polymer-PLS materials were tested on a universal testing machine (UTM) from Zwick Roell (Z005 model, Germany) based upon the ASTM D882-12 standard. At least six or more tensile specimens were prepared from each film. Specimens were preconditioned for not less than 48 h. The standard distance between grips was 25 mm; the rate of grip separation was 2.5 mm/min for E-modulus determination.

PVC-PLS 100phr films were cut into 20 × 20 mm² pieces before migration resistance examination. Standards ASTM D1203-10 (method A) and ASTM D5227-13 were adapted for evaluation of plasticizers. The test results are expressed in parts percent, describing how much in quantity the plasticizer is migrated under the given conditions. The volatility test was performed at 70 °C for 24 h, and detailed conditions are described in the Supporting Information (Section S9).

Results and Discussion

Synthesis and Proposed Mechanism

It is known that the dehydration of galactaric acid (Gal-dA) under acidic conditions gives furandicarboxylic acid.^34,33,32 However, due to its unusually dense structure and large lattice energy,^35,36 the reaction required excessive amounts of dehydrating agent up to 200 wt % or more under certain conditions. This, in turn, made the reaction vigorous, not green, and needed neutralization afterward. A modification is required so that the amount of acid catalyst can be considerably reduced without much influencing the reaction yield. In a Dean–Stark apparatus were added Gal-dA, butanol, p-xylene (PX), and sulfuric acid (H₂SO₄). Gal-dA was not initially dissolved in the alcoholic medium, but later, the suspension gradually turned into a homogeneous solution as the acid-catalyzed reaction progressed under reflux conditions. H₂SO₄ was chosen as an acid catalyst, while PX was considered as a material separation agent (MSA), which can change the molecular interactions, remove water exhaustively, and further eliminate the azeotrope if existing. The reaction resulted in the formation of two regioisomeric dibutyl furandicarboxylates, namely, 2,3-DBF and 2,5-DBF. It is obvious to speculate that Gal-dA could experience (a) an intramolecular etherification of two hydroxyl groups (at C2 and C5 positions) to form a five-membered ring followed by (b) dehydration inside the ring and (c) esterification to create 2,5-DBF. However, the existence of the 2,3-regioisomer is a thing that is hard to give an explanation to since the two −COOH groups of the original Gal-dA are terminal. On top of that, monitoring by TLC and via GC–MS during and after the reaction found butyl 2-furoate, which has not been reported before. The production of either dibutyl 2,3-furandicarboxylate (2,3-DBF) or butyl 2-furoate raised complications to the reaction mechanism but hinted at a chance of growing awareness of its nature.

To optimize the reaction conditions, the effect of acid amounts on product yields was investigated in the BuOH medium (Figure 2a). More than 90% total yield (mixture of 2,5- and 2,3-DBF) after 10 h could be achieved in the presence of 50 wt % H₂SO₄ to Gal-dA. A decreased amount of acid, 20 wt % H₂SO₄ to Gal-dA, caused a decrease in the reaction rate, thus resulting in only 38% total yield after 10 h. In fact, as stated previously, at least 100 wt % H₂SO₄ to Gal-dA was required to achieve a total yield (GC yield) over 90% of DBFs after 9 h, unless MSA was used.³³ Nevertheless, 50 wt % acid was still a large amount. In an effort to improve the reaction conditions, further optimization of the MSA was carried out. As shown in Figure 2b, the composition of MSA (mol % PX to BuOH plus PX) was closely related to the reaction rate. The maximum total yield after 10 h was obtained at the specific composition of MSA. The amount of H₂SO₄ for quantitative yields after 10 h could be reduced to 30 wt % at 9.5 mol % MSA composition.

Figure 2.

DBF yields from Gal-dA according to (a) amounts of acid (H₂SO₄) and (b) composition of material separation agent (PX). Reaction conditions: Gal-dA (5.0 g, 24 mmol), BuOH (40.0 g, 540 mmol), H₂SO₄ (30 wt %) for panel (b), PX (5.0 g, 47 mmol) for panel (a), 160 °C (reflux), 10 h.

In addition, synthetic efficiencies of various C4-C8 alkyl alcohols including n-butyl, n-pentyl, n-hexyl, n-heptyl, n-octyl, i-pentyl, and i-octyl (2-ethylhexyl, in particular) alcohols were evaluated (Table 1). As the chain length of the alkyl group was prolonged, total yields of the DAFs were decreased under identical conditions. When n-butyl alcohol and i-pentyl alcohol were employed (entries 1 and 6), almost quantitative amounts of the desired products were afforded on the basis of Gal-dA. On the other hand, approximately 50% was yielded from C8 alkyl alcohols such as n-octyl and i-octyl alcohols (entries 5 and 7). This implies that the long alkyl chains hamper intramolecular cyclization of dialkyl galactarate due to steric hindrance. Noticeably, other by-products were also detected. They were homologous compounds of butyl 2-furoate, which was mentioned above. This suggested that this class of furoate compounds exists during and after the reaction and plays a role in the events of the reaction mechanism, possibly being a key intermediate in between Gal-dA and DAFs. To generalize, the conversion of Gal-dA led to two regioisomeric furandicarboxylates (2,5- and 2,3-DAFs) as well as 2-furoate. Characterizations are all presented in the Supporting Information.

Table 1. DAF Yields from Gal-dA According to Various Alkyl Alcohols^a.

		yield (%)
entry	R	2,5-DAFb	2,3-DAFb	total
1	CH3(CH2)3–	63	30	93
2	CH3(CH2)4–	62	23	85
3	CH3(CH2)5–	54	21	75
4	CH3(CH2)6–	40	21	61
5	CH3(CH2)7–	35	17	52
6	(CH3)2CH(CH2)2–	62	32	94
7	CH3(CH2)3CH(C2H5)CH2–	35	11	46

^aReaction conditions: Gal-dA (3.0 g, 14 mmol), alcohol (324 mmol), H₂SO₄ (40 wt %), PX (3.5 g, 33 mmol), 160 °C (reflux), 10 h.

^bIsolation yield.

Until now, the reaction mechanism for this transformation has not been unveiled, especially in the case of 2,3-DAF. The mechanism is believed to be sophisticated because many events including esterification and dehydration take place simultaneously. Here, the mechanism for the one-pot multistep reaction of Gal-dA to 2,3- and 2,5-DAF is proposed in Scheme 2. Initially, acid-catalyzed esterification of Gal-dA with an alcohol produces dialkyl galactarate (routes A–D). The diester of Gal-dA is dominantly formed and can be spontaneously crystallized out of the reaction mixture in case that the reaction time fails to meet 10 h long. It is then transformed to 1AB by dehydration at the C4 position (routes A and B) followed by tautomerization from the enol to keto form. Afterward, an intramolecular cyclization leads to a five-membered cyclic compound 2AB. In this stage, two molecules of water from 2AB are eliminated to give 2,5-DAF (route B). Meanwhile, the dehydration of only one molecule of water generates compound 3AB. Dimerization of 3AB by transesterification and then decarboxylation of the dimer compound 4A results in compound 5A. Compound 5A, which is alkyl 2-furoate, is detectable during and after the reaction. Carbonylation of 5A with released CO₂ can generate compound 6A, and esterification of 6A with an alcohol affords 2,3-DAF (route A). On the other hand, dialkyl galactarate can undergo another intramolecular cyclization by etherification between the OH groups at the C2 and C5 positions to afford a five-membered cyclic compound 1CD (routes C and D). Dehydration of 1CD at the C3 and C5 positions forms a furan ring to give 2,5-DAF (route C). Additionally, the carbonyl group at the C5 position of 1CD activated under acidic conditions can promote an intramolecular cyclization to afford the bicyclic compound 2D. From 2D, two molecules of water are released to give 2,3-DAF via 3D (route D). From this mechanism, we can speculate that routes B and C for 2,5-DAFs are preferable to routes A and D for 2,3-DAFs because routes B and C require only a dehydration step from intermediates 1CD and 2AB. Consequently, 2,5-DAFs are yielded as major products.

Scheme 2. Possible Reaction Mechanism To Synthesize 2,3- and 2,5-DAF from Gal-dA.

Miscibility and Interaction with Poly(vinyl chloride)

To examine compatibility of four target bio-plasticizers (PLS) in a PVC matrix, the surface morphologies of the PVC-PLS films were analyzed with FE-SEM images at a magnification of 3000×. There was no obvious phase separation between PLS and the matrix in all of the cases, indicating a compatibility between them. Chain length (C4 or C5) and position (para or ortho) of the alkyl ester groups attached to the furan ring had little effect on the surface morphology. Unlike the blends made by mixing at elevated temperatures,^37,38 the surface of PVC-PLS films cast from solution was homogeneous, smooth, and flat even though the amount of PLS increased to 100phr (Figures S69–S73).

The molecular interactions between the polymer matrix and plasticizers were studied, focusing on the major frequency differences between the absorption bands of the PVC-PLS combinations and the pure form of PLS. The interaction between DOP and PVC has been well researched in which the H atom attached to the chlorine-bearing C atom of the PVC acts as the electron acceptor and the O atom of the carbonyl group of DOP is an electron donor.^39,40 The interaction results in a shift of carbonyl-related bands to a lower frequency. The peak of the carbonyl absorbance of DOP at 1724 cm^–1 slightly moved to 1720 cm^–1 in the PVC-DOP films. Considering that there was no change of peaks corresponding to C=C stretching bands at 1600 and 1579 cm^–1 in the PVC-DOP films, the aromatic benzene ring of DOP seemed unaffected by the C–Cl of PVC (Figure 3a and Table S9). Unlike DOP, with only a carbonyl O atom for interacting with PVC, the four synthetic furanic compounds investigated in this study (namely, 2,5-DBF, 2,5-DIAF, 2,3-DBF, and 2,3-DIAF) bear an extra O atom in the aromatic furan ring as well as the carbonyl O atom(s). As a result, the peaks of the C–O stretching bands, ranging 1300–1000 cm^–1, were more complicated, owing to the additional C–O bonds in the furan ring. High peak intensities of the C=C stretching band in the range of 1700–1450 cm^–1 indicate that the furan ring is polarized. When these furanic compounds were added to PVC, the absorption bands were shifted with various absolute magnitudes from at least 0 to 2 cm^–1. Noticeably, C=O bands of para-substituted 2,5-DBF/2,5-DIAF were even moved to higher frequencies by 8–12 cm^–1, and those of ortho-substituted 2,3-DBF/2,3-DIAF were enhanced greatly to 18–21 cm^–1 and became broader. These observations suggest two aspects. First, these interactions of furanic compounds with PVC were intense and comparably multiple times stronger than that of DOP in terms of the absolute interaction magnitude. Second, with either 8–12 or 18–21 cm^–1 enhancement of C=O bands, it was about a disorder that would happen to the π-electron system of furanic compounds after interacting with PVC since electronic factors (inductive, mesomeric effects) are often responsible for such a shift. Further studies are needed to fully understand the main reasons leading to big shifts of absorption bands.

Figure 3.

Comparison of IR absorption bands of pure PLS and PVC-PLS films containing different amounts of (a) DOP, (b) 2,5-DBF, (c) 2,5-DIAF, (d) 2,3-DBF, and (e) 2,3-DIAF.

Plasticizing Efficiencies on Flexibility

DSC experiments were conducted to understand the thermal behavior of the investigated materials. A prescan was performed to collect thermal properties of DAF products as well as neat PVC powder. Within the scanning range from −20 to 150 °C, melting points (T_m) of DAFs and the glass transition temperature (T_g) of neat PVC were determined, but unfound were the T_g values of DAFs or DOP. Consequently, to the mixture of PVC-PLS, only T_g from PVC was observable during the scanning range.

A good miscibility by a primary plasticizer like DOP^41,42 is expressed through a single T_g representing a new entity. On the other hand, in the case of limited or poor miscibility between an additive and a polymer matrix, numerous T_g values would appear.⁴² Because of unequal distribution of plasticizer content, the matrix would consist of several different parts where they are either well-plasticized or possibly unplasticized, and by that, they are each reflected through a unique T_g value. Practical data from DSC experiments informed an only T_g for any PVC-PLS combination (data plotted in Figure 4). As can be seen from the diagram, the T_g value and the plasticizer content appear unproportional. The growth in the amount of plasticizer downgrades the glass-liquid transition of PVC materials. To put it another way, the insertion of the plasticizer into the matrix provides more free volume for motion of polymer chains, letting the mixture to be more flexible.⁴³ Another observation is that DAFs’ tendency of plasiticization seems familiar with DOP’s. To be more specific, the para-substituted DAFs (2,5-DBF and 2,5-DIAF) exhibited a very similar pattern to DOP. Meanwhile, the ortho-substituted ones (2,3-DBF and 2,3-DIAF) followed the trend but with a relatively higher set of T_g values. This result can be explained based upon the chemical structure. Although alkyl chains of 2,5-DBF/2,5-DIAF (C4 or C5) are shorter than DOP (C8), their molecular volume may be equivalent to that of DOP because the two substituents are spread in opposite directions, whereas 2,3-DBF/2,3-DIAF having both substituents on the same side can just provide smaller free volume for the polymer chains.

Figure 4.

Dependence of T_g on the amount of plasticizers in PVC-PLS films.

From Figure 4, the T_g of neat PVC is 80.77 °C, which agrees reasonably well with the value of 83 °C that has previously been reported for PVC.^44,45 However, the T_g of PVC plasticized with DOP obtained in our experiment is less than that expected from the literature. The literature⁴⁵ shows that the T_g of PVC plasticized with 50phr DOP is 24.6 °C, but our value is 16 °C, which was obtained by interpolation of Figure 4 and Table S14. This is likely due to the presence of residual THF. According to the literature,^46,47 it is very difficult to remove THF from PVC after its use as a solvent in making cast films, and the residual THF can have a plasticizing effect on the polymer. As much as 6–8% THF can remain in the PVC after casting and drying, and at least 2–3% is very difficult to remove.

To sum up at this point, the addition of the plasticizer provides free volume for motion of polymer chains. 2,5-DAFs were able to generate the equivalence of free volume as DOP due to the opposite aliphatic moieties, following an order of DOP ≈ 2,5-DAFs > 2,3-DAFs. A single T_g was found for any PVC-PLS combination, indicating the miscibility in accordance with the homogeneity on the material surface (seen from SEM analysis). A decent reduction of T_g, from the original value of the polymer matrix (∼82 °C), confirms a boost in flexibility.

Tensile testing provides information about the flexible nature of the PVC-PLS films. One of the most important functions of the plasticizers is to reduce the intermolecular attraction between chains of polymer, which leads to flexibility in the processing of materials.⁴⁸ Parameters like elastic modulus (E-modulus), stress at break (so-called tensile strength, δ), and elongation at break (ε) were under investigation and compared to PVC-DOP films. Unplasticized PVC (uPVC) reportedly appears very hard and rigid with a δ of 56.6 MPa, ε of 85%, and E-modulus of up to 3000 MPa (at 30 °C).⁴⁹ Experimental data (depicted in Figure 5) illustrate that all investigated plasticizers in the PVC matrix showed desirable behavior. In detail, their tensile strengths and elastic moduli were halved or much less than the ones of uPVC with a δ of 8–20 MPa and E-modulus of 25–330 MPa, respectively, while the strain at break (ε) was prolonged twice, up to 200%. There was no noticeable difference in comparison between DAFs versus DOP. As the amount of PLS in PVC increased (42.9 to 100phr), the elastic modulus of the PVC-PLS films decreased (Figure 5a). A decrease in the stress at break δ was presumable as well (Figure 5b) since, in principle, δ is proportionally correlated with E-modulus. Rather, plasticizing efficiency of biobased PLS was slightly better than that of DOP. Take 42.9phr plasticization as an example. DOP’s E-modulus of 314.2 ± 6.6 MPa and δ of 19.38 ± 0.92 MPa hardly surpassed the four remainders. The ε values, on the contrary, were not dramatically changed according to the amount and the kind of PLS, and they showed approximately 150–200% in the investigated range (Figure 5c). As a result, at a given plasticization extent, the biobased furanic plasticizers were available to substitute DOP in terms of tensile efficiency. The complete data is summarized in Table S15.

Figure 5.

Comparison of (a) elastic modulus, (b) stress, and (c) elongation at break according to the plasticizers contained in PVC-PLS films.

Migration Resistance and Thermostability

Pure polymer itself is rigid, brittle, and hard for processing. When combined with a plasticizer, flexibility, workability, and distensibility of the material can be obtained.¹ However, a significant disadvantage is that the plasticizer and polymer matrix do not usually form permanent linkages.^3,50 This loose interaction can fail to prevent PLS from migrating out of the matrix over time. Therefore, under a suitable condition, the plasticizer can be released from the matrix.⁵⁰ Because of a need for safe use in end products, the migration degree of PLS to the surrounding environment should be determined. Two tests comprising volatility (ASTM D1203-10) and extractability (ASTM D5227-13) were conducted in this study. The former provides details about how much in quantity the PLS could evaporate under an elevated condition without a suitable heat stabilizer, and the latter reflects the leachable nature of PLS into a (hot) surrounding liquid medium. After volatility tests, a physical difference was not observed in PVC films when no heat stabilizer was used.

A basic difference between DOP and DAFs comes from the core in which furanic plasticizers have an oxygen atom capable of boosting polarity, making DAFs more polarized. Figure 6b depicts results from the extractability test in the hexane medium that all DAFs exhibited a better resistance nature than DOP as expected. Also, the ortho-products (2,3-DBF/2,3-DIAF) performed better than the para-substituted ones. With regard to lyophilicity, hexane acts very attractive to the less-polar migrants.^51,52 Furthermore, the boosted dynamicity and polarity led to an improved dipole moment, together with a better attraction strength for the polymer matrix. As observed, there was a correlation between the attraction and the polarity (shown in Table 2).⁴¹ Accordingly, the Δν difference of carbonyl absorption bands representing the attraction strength from investigated compounds toward PVC was proportional to the polarity that is expressed reversely in terms of the extraction resistance. Providing the facts, PVC-DOP interaction is quantitized by a 4 cm^–1 difference via FTIR observation;^39,40 in contrast, the effects caused by 2,5- and 2,3-DAFs generated 8–12 and 18–21 cm^–1 shifts, respectively, to higher frequencies. DOP turned out easier to be leached, while 2,3-DBF was the most resistant among all. This reaffirmed that the greater the polarity of a plasticizer molecule, the better the attraction it has for the PVC chain, the less extractable content is migrated into hot surrounding media.^41,52,53 In brief, the extraction resistance followed the order of 2,3-DAFs > 2,5-DAFs > DOP.

Figure 6.

(a) Volatility of plasticizers surrounded by activated charcoal at elevated temperature and ambient pressure and (b) extractability of plasticizers in hot hexane for 2 h.

Table 2. Correlation between the Absorption Difference Δν and the Extractability.

	C=O absorption banda (cm–1)			extractabilityd
plasticizer	νob	νc	|Δν̅| ↓	↑ (wt %)
DOP	1724	1720	4.0	14.1%
2,5-DBF	1724, 1709	1736, 1717	10.0	13.5%
2,5-DIAF	1724, 1709	1736, 1718	10.5	13.1%
2,3-DIAF	1712	1722	10.0	7.9%
2,3-DBF	1716	1730	14.0	5.4%

^aApparatus’ resolution: 4 cm^–1.

^bFrequency of unbound C=O stretching band (pure form, before mixing).

^cFrequency of interactive C=O stretching band (PVC-PLS 50%, after mixing).

^dExtraction resistance (ASTM D5227-13): the lower the value, the higher the polarity.

Due to their small structures, it is apparent that furanic plasticizers are more volatile than DOP. Their molecular weights are all below 300 g/mol indeed. The volatility of each furanic plasticizer can be understood in the same way. The volatilities of 2,5-DBF/2,3-DBF having butyl moieties were higher than those of 2,5-DIAF/2,3-DIAF having isopentyl moieties (Figure 6a). To understand more about this behavior as well as the thermal stability of the material, we performed thermogravimetry analysis (TGA) in the absence of a suitable heat stabilizer. As can be seen from Figure 7, TGA curves for all the investigated PVC-PLS films have similarities. All combinations came through three stages of thermal decomposition. The first stage, near 205 °C, is mainly due to the release and evaporation of the plasticizer. The next stage, up to 385 °C, is related to the dehydrochlorination of PVC to form macromolecules with conjugated bonds. In the last stage, up to 520 °C, newly formed macromolecules undergo cracking and pyrolysis. Further information is presented in Tables S16–S19 in the Supporting Information.

Figure 7.

Degradation thermograms (TGA curves) of 100phr PVC-PLS films containing DOP, 2,5-DBF, 2,5-DIAF, 2,3-DBF, and 2,3-DIAF.

PVC is known as a nonvolatile material. Inside the TGA chamber, it is thermally degraded into two stages comprising dehydrochlorination and pyrolysis.^54,55 The introduction of the plasticizer, on the one hand, helps the polymer to stay away from its rigid nature, but on the other hand, it enhances the degree of volatility regarding vapor pressure of the additive. This, in turn, is reflected by an additional process prior to the actual two-stage thermal decomposition of the matrix. Furthermore, it should be noted that the higher the vapor pressure of a chemical, the faster the rate at which a material loses weight and vice versa.⁴¹ Take 2,5-DBF as an example. Its volatility, which was measured by about 12% according to Figure 6a, allowed its combination with PVC to lose 1/10 of its weight (at 250 °C) faster than any others (Figure 7). In contrast, it was hard for DOP to do so because its T_90% is only 9° different from the one of pure PVC (being 286 °C). As a result, DOP was deemed more advanced than the four investigated compounds.

Conclusions

Furan dicarboxylic acid derivatives were successfully synthesized by a one-pot multistep reaction between marine-biomass-originated galactaric acid and bioalcohol. The reaction condition was adjusted less vigorous, but milder and greener regarding reduction in the amount of acid catalyst and addition of a material separation agent. The unexpected detection of a 2-furoate product opened a chance of understanding the reaction mechanism. After that, properties of four DAFs (namely, 2,5-DBF, 2,5-DIAF, 2,3-DBF, and 2,3-DIAF) as plasticizers on the PVC matrix were investigated through various analytical techniques. As an ultimate result, data interfered with one another to highlight DAFs’ promising properties over DOP, except the volatility and the low-temperature performance related to small molecular weight. It was found in detail the compatibility through no phase separation, the formation of intense interactive bonds related to carbonyl band movement, the comparable flexibility, the higher polarity, and the better extraction resistance to the surrounding liquid media. Meanwhile, the volatile resistance and the thermal stability of the material can be superior to DOP’s only if DAFs possess colossal molecular weights,^41,52 simply by prolonging the alkyl moieties.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.9b02448.

• Reaction optimization, product characterization (spectroscopic data and thermal properties), preparation of plasticized PVC films for evaluation, and additional data (SEM, UTM, DSC, TGA, and FTIR observation of PVC-PLS interaction) (PDF)

The authors declare no competing financial interest.