Isocyanate-free urethanediol itaconates as biobased liquid monomers in photopolymerization-based 3D printing

The basic principle of additive manufacturing, also known as 3D printing, is building up an object layer by layer; in a first stage, the design of the desired object is created by computer-aided design (CAD) and the virtual model is then digitally sliced into horizontal layers for a printable command sequence. Via a command file usually named G-code, the virtual coordinates are sent to the 3D printer, which deposits the material in predesigned locations at the x–y plane and the entire process is repeated along the z-axis. Compared to traditional manufacturing techniques, modifications in the shape of the 3D printed manufactures can be easily done by modifying the digital object, without the need for unique tools for a single modification [1]. Furthermore, 3D printing is a sustainable manufacturing process, as additive fabrication results in much less material waste [2]. Among the different technologies, vat photopolymerization (VP) allows to produce three-dimensional objects by selectively photocuring monolayers of a monomeric resin into layered 3D objects [3‐5]. The use of high-resolution UV light sources for triggering the spatially controlled photopolymerization of the resins enabled the manufacturing of objects with small details at resolutions of the order of tens of microns, depending on the available implementation [6]. Moreover, the limited price of the required equipment has made VP the first high-resolution 3D printing technique available for private consumers’ and everyday use [7, 8]. In addition, by selecting the appropriate monomers to be formulated into the photocurable resin, it is possible to obtain materials spanning over a large range of mechanical, thermal, optical and electrical properties, allowing to employ VP in a great variety of high added value applications, such as for biomedical, dental, jewelry and aerospace [9‐11], with potential applications spanning from stimuli-responsive materials to tissue engineering [12, 13]. Thanks to its important contribution to the good mechanical properties of 3D printed materials, one of the most commonly employed monomers for producing VP resins is (7,9,9-trimethyl-4,13-dioxo-3,14-dioxa-5,12-diaza-hexadecane-1,16-diylbis(2-methylacrylate), known simply as urethane dimethacrylate, produced by reacting a difunctional aliphatic isocyanate with 2-hydroxyethyl methacrylate [14‐16]. In the industrial sector, the synthesis of isocyanates poses a significant hazard not only in terms of the isocyanates themselves, but also concerning the substrates used in their production; on an industrial scale, in fact, isocyanates are obtained through the phosgene route [17]. Owing to the increasing concerns related to the environmental impact of isocyanates and their production processes, the scientific community has developed alternative routes for the production of urethanes and polyurethanes by aminolysis of cyclic carbonates, obtained in turn by reacting bioderived epoxides with CO₂ [18, 19]. Considerable research efforts have been devoted to exploring the potential of carbon dioxide as a renewable C1 building block for chemical synthesis. The rationale behind this stems from the fact that CO₂ is abundant, readily available, non-toxic, and non-flammable, making it an attractive and cost-effective source for the production of chemical products [20]. On the other hand, biobased epoxides can be synthesized from waste vegetable oils via epoxidation, which not only provides a sustainable alternative to traditional petroleum-based epoxides, but also utilizes a waste product, helping to reduce environmental impact [21]. By combining bifunctional primary amines with monofunctional cyclic carbonates such as ethylene carbonate (EC) and propylene carbonate (PC), it is possible to produce diurethanediols (UDOs), which have been previously employed for the preparation of poly(ester urethane)s [22, 23]. The diamines can be selected in the pool of biobased diamines which include 1,3-propanediamine, 1,4-butanediamine, and 1,5-pentanediamine, obtained by decarboxylation of lysine, arginine, and ornithine, respectively [24]. Furthermore, UDOs have been recently employed as additives in VP after functionalizing the alcoholic termini with methacrylate residues [25]. However, the efforts towards an increase in the sustainability of VP resins were limited by the use of methacrylic acid as the photocurable moiety, and the incorporation of diamines from non-renewable resources. Disappointingly, the produced methacrylated UDOs were solid at room temperature, thus limiting their applicability for applications in VP. Currently, the most promising candidate that is emerging as a biobased alternative for replacing (meth)acrylic acid functionalities in photocurable mixtures is indeed itaconic acid, or 4-methylenesuccinic acid. Traditionally produced by distillation of citric acid, itaconic acid is nowadays produced by direct fermentation of glucose-containing biomasses by specifically engineered bacterial strains. [26]. Many examples of the incorporation of itaconic acid in resins for VP have been recently reported in the literature, mainly employing its diesters, polyesters, poly(ester-amide)s and poly(ester-thioether)s [27‐30]. In a single example available in the literature, biobased UDOs have been polymerized with itaconic acid to produced fully biobased photocurable poly(ester-urethane)s, but their physical properties did not make them suitable for the formulations of 3D printable resins, and they were explored mostly as coating materials [31]. This paper introduces the synthesis and characterization of liquid biobased diurethanediols (UDOs) that have been functionalized with itaconic acid monomethyl ester for use in high precision additive manufacturing applications. The liquid state of the UDO backbones allows them to be combined with high percentages of (meth)acrylic and itaconic acid-based co-monomers, resulting in the development of highly biobased photocurable resins. The addition of UDOs-I has been thoroughly investigated for its impact on the mechanical properties of these resins, including tensile, flexural, and hardness properties, revealing mechanical properties comparable to those of commercially available non-renewable urethane monomer-based alternatives. The biobased content of the prepared formulation has also been evaluated and discussed in relation to the extracted mechanical properties. This research presents a fully biobased building block that can be used in photocurable mixtures for high precision additive manufacturing applications to increase their sustainability without significantly affecting their mechanical properties.

All chemicals were purchased from Sigma-Aldrich (St. Louis, MO, USA) and used as received. Chloroform was dried by distillation over CaCl₂ and used promptly. Grindsted Soft-N-Safe (9-hydroxystearic acid monoglyceride triacetate, SNS) was purchased from Danisco (Brabrand, Denmark).

Diurethane diols UDO₂ and UDO₃ were prepared by aminolysis of cyclic carbonates (ethylene carbonate, EC, and propylene carbonate, PC, respectively) using the double-functionalized primary amine 1,4-diaminobutane without the requirement for any solvent. This diamine can be in theory considered as biobased, thanks to the widely explored possibility to obtain it by decarboxylation of the amino acid l-arginine [24, 32]. The aminolysis reaction involves the nucleophilic attack of the amine on the carbonate carbon, leading to the opening of the five-membered ring and the formation of a linear carbamate bearing a free alcoholic group (Scheme 1). To minimize the possibility of the formation of monourethane amines, the diamine was added dropwise into the cyclic carbonate, leading to the immediate formation of urethanediols thanks to the large excess of carbonate at every addition of diamine. Due to the non-selective nature of the aminolysis, the opening of the cyclic carbonates can take place on either side of the C=O; while thanks to the symmetric structure of EC the two possible mechanisms lead to the same product, UDO₂, PC aminolysis is expected to lead to the formation of two structural isomers in comparable concentrations [33].

This effect, combined with the bifunctional nature of the selected amine, leads to the formation of UDO₃ as three regioisomers: one bearing two secondary OH moieties (UDO₃-1), one bearing two primary OH groups (UDO₃-2), and one bearing one primary and one secondary OH functionalities (UDO₃-3) as depicted in Scheme S1. Due to their similar physical–chemical properties, the isomers were not separated and underwent successive functionalization with itaconic acid as a mixture. With respect to UDO₂, the product was characterized by ESI–MS and NMR spectroscopy (Figure S1), which confirmed the predicted molecular weight and chemical structure, respectively. Similarly, the outcome of PC aminolysis was confirmed by ESI–MS, which suggested the presence of the diols as the only reaction products, with the expected molecular ion peak common to all regioisomers. ¹H–, ¹³C–, ¹H–¹H–COSY and ¹H–¹³C–HSQC NMR spectroscopies revealed the presence of all isomers in comparable concentrations, allowing also for the unambiguous assignment of NMR peaks (Fig. 2a, Figure S2-S4 and Table S1). Moreover, by comparing the NMR integrals, it was possible to assess that the aminolysis of PC took place with a preference toward the formation of the primary alcohol (63%) compared to the secondary alcohol (37%), in agreement with previously published results [34]. Assuming that the regiochemistry of the aminolysis of PC performed by the intermediate monohydroxyurethanes was not dependent on the type of alcoholic functionality formed at first, the UDO₃ mixture can be expected to be composed of 14% UDO₃-1, 40% UDO₃-2, and 46% UDO₃-3. Furthermore, ATR–FTIR spectroscopy confirmed the presence of main functional groups expected in the product: N–H/O–H stretching at 3314 cm⁻¹; C–H stretching at 2930–2868 cm⁻¹, carbamate C = O stretching at 1677 cm⁻¹, and N–H bending at 1521 cm⁻¹ (Figure S5). Diurethanediol bis(methyl itaconate) UDO₂-I and UDO₃-I were synthesized by acylation of the corresponding UDO using monomethyl itaconoyl chloride, produced by chlorination of itaconic acid monomethyl ester, as recently reported [35]. The chlorination of monomethyl itaconic acid took place in an excess of oxalyl chloride, and the unreacted chlorinating agent was removed by distillation and recovered for further use. Furthermore, the obtained monomethyl itaconoyl chloride was purified by distillation, allowing for its isolation with high purity. Once isolated, monomethyl itaconoyl chloride was added dropwise to a solution containing triethylamine and the selected UDO to be functionalized, and the esterification reaction took place instantaneously between the itaconic acid chloride and the urethanediols. Such esterification required anhydrous conditions to prevent undesired hydrolysis of the acyl chloride into the corresponding carboxylic acid and the presence of triethylamine that could trap the produced HCl. The expected product structures were confirmed by ESI–MS and NMR spectroscopy (Fig. 2b and Figure S6-S7). Since monomethyl itaconoyl chloride prepared with this method is composed of a mixture of roughly 95% α,β-unsaturated acyl chloride and 5% α,β-unsaturated methyl ester, UDO₃-I is expected to be formed as a mixture of nine isomers where propanediol and itaconic acid residues display all possible orientations. This reflected into highly split NMR signals, which still allowed to confirm the presence of the photocurable moiety of itaconic acid residues and the absence of monosubstituted UDOs derivatives. In addition, ATR–FTIR spectroscopy (Fig. 3a) was further employed to characterize the functional groups present in UDO₃-I. The first important difference between the prepared itaconic acid-modified diurethanediols lies in their physical state around room temperature. In fact, while UDO₂-I is a solid with a melting temperature significantly above 25 °C, UDO₃-I is a liquid. This is believed to be related to the presence of nine isomers in the product mixture that hinders its assembly and prevents its solidification. To further characterize UDO₃-I on this aspect, its rheological properties were evaluated by measuring its viscosity both over a range of temperatures (10–40 °C) at constant shear rate (1 Hz) and at constant temperature (27 °C) over a range of shear rates (0.1–100 Hz) (Fig. 3b,c). Rheological analysis confirmed the stability of UDO₃-I as a liquid over all the explored range of temperatures, with viscosities ranging from over 100 Pa s at 10 °C to around 2 Pa s at 40 °C. At room temperature, UDO₃-I displayed a viscosity of 10 Pa s which was observed to be not dependent on the shear rate, suggesting an ideal Newtonian behavior characteristic of small molecular liquids. Unlike previously reported methacrylated diurethanediols and the herein presented UDO₂-I, the liquid form of UDO₃-I makes it possible for its formulation in photocurable resins with virtually no limit related to its solubility with the other monomeric components, and therefore expanding its applicability for such purposes. In fact, the high polarity and strong H-bonding capabilities of the carbamate group strongly limit its solubility in organic mixture such as those composed of photopolymerizable monomers. While this is not the first ever reported example of liquid bifunctional photocurable diurethane, it has to be underlined that such examples are always produced using isocyanates and are characterized by very low biobased content and sustainability. Therefore, the presented UDO₃-I represent the first biobased liquid bifunctional photocurable diurethane ever reported, so far. To support the claimed advantages of liquid diurethanes compared to their solid counterparts, the prepared UDO₂-I and UDO₃-I were then formulated with photocurable monomers for the preparation of resins for VP. However, it was immediately clear that UDO₂-I displayed very little solubility in the selected monomer mixtures, while UDO₃-I could be mixed virtually in all proportions (Table 1). In fact, no more than 5 wt.% of UDO₂-I could be efficiently dissolved into the resins; since the increase in biobased content of photocurable mixtures would not be significantly altered by such a small amount of added UDO₂-I, its applicability as VP monomer was not explored any further.

A first set of resin was prepared by mixing UDO₃-I with a 1:1 mixture of glycerol dimethacrylate (GDMA, mixture of isomers) and glycerol propoxylate triacrylate (GPT, average M_n = 428), to evaluate its effect on the mechanical properties of fossil-based (meth)acrylate formulations while increasing their biobased contents (resins A0–A3). Then, resins with higher biobased contents were prepared by employing the bifunctional itaconic acid-based reactive diluent 1,4-butanediol bis(methyl itaconate) (named I₂B₁) as a partial replacement of the mixture of glycerol-based photocurable monomers (resins B0 and C0). Finally, UDO₃-I was introduced at 50 wt.% on such formulations, to evaluate once again its effect on their mechanical properties (resins B1 and C1). This is the highest attempted loading of non-isocyanate urethane-based photocurable components for VP, and it has been possible thanks to the liquid state of UDO₃-I at room temperature. In addition to the photopolymerizable component, all formulations contained a photoinitiator (ethyl phenyl (2,4,6-trimethylbenzoyl) phosphinate Et-APO, 2 wt.%), a radical inhibitor (2,6-diterbutyl-4-methylphenol, BHT, 0.5 wt.%), and a photosensitizer (2-isopropyl thioxanthone, ITX, 0.3 wt.%). Moreover, a biobased plasticizer (Grinsted Soft-N-Safe, SNS, 7.2 wt.%) was added to each mixture. The biobased content of the prepared formulations was evaluated by calculating for each component of the resin the percentage of their molecular weight that is composed of building block of renewable derivation. In particular, itaconic acid-based monomers and the plasticizer SNS were characterized by 100% biobased derivation, while the photoinitiating system was considered fully non-renewable. For the glycerol-based monomeric components GDMA and GPT, their biobased contents were calculated to be equal to 32.4% and 17.3%, respectively, considering as fully biobased their glycerol core. This approach has allowed to calculate the biobased contents for each formulation (Table 1). All resins were efficiently printed into 3D shapes with high spatial accuracy and resolution, also supporting the addition of small amounts of a natural dye (purpurin, 0.005 wt.%) to print colored 3D objects (Fig. 4). Furthermore, with each formulation, tensile and flexural test specimens were printed and tested according to the specifications for ISO 37 Type 2 and ISO 178, respectively. Due to the well-known lower reactivity of itaconic acid residues when compared to acrylates and methacrylates with respect to radical polymerization, the replacement of commercially available non-biobased urethane methacrylates with UDO3-I, and other itaconic acid-based reactive diluents require a re-optimization of instrumental printing parameters. In our case, the employed LCD 3D printing technology limited the explorable parameters to the exposure time per layer, which was increased to up to 2 min every 0.1 mm of object height (Table 2). However, more advanced implementations allow for the optimization of additional parameters such as light power, with the aim of reducing the lengthening of the 3D printing process caused by the presence of biobased components.

From the tensile and flexural stress–strain curves (Figure S8-S9), the main mechanical properties of the 3D printed materials were extracted, collected in Table 2, and summarized in Fig. 5. By taking into consideration the first group of 3D printed materials (resins A0–A3) composed of increasing amounts of UDO₃-I in a 1:1 mixture of GPT and GDMA, its effect is immediately evident. The gradual replacement of the glycerol (meth)acrylate mixture with UDO₃-I lead to a continuous decrease in the elastic modulus, with subsequent increase of the elongation at break and an overall improvement of the mechanical properties represented by the increasing tensile strength, up to a UDO₃-I loading of 30 wt.% (corresponding to resin A2). However, when the loading was increased to 50 wt.% (resin A3), a drop in elongation at break and tensile strength was recorded, revealing a limit in the performances of UDO₃-I as a monomer, when formulated with GPT and GDMA. To verify this assumption, GPT and GDMA were partially and almost completely replaced with the biobased cross-linker I₂B₁ and printed with and without the presence of 50 wt.% UDO₃-I (resins B0–C1). The tensile testing of such materials revealed that the addition of the butanediol-based reactive diluents was able to amplify the effect of UDO₃-I on the mechanical properties; in fact, the addition of 50 wt.% UDO₃-I to a resin composed of GPT, GDMA, and I₂B₁ in 3:3:2 proportion (resin B1 vs B0) caused a limited decrease in the elastic modulus (− 15%) and a significant increase in elongation at break (+ 226%) and tensile strength (+ 129%). This effect is even more evident when the resin composition is changed to higher I₂B₁ contents (resin C1 vs C0), where the GPT, GDMA, and I₂B₁ are changed to 1:1:6. In this case, the addition of UDO₃-I caused a less notable decrease in the elastic modulus (− 11%) accompanied by a remarkable increase in elongation at break (+ 400%) and tensile strength (+ 264%). It is worth underlining that resin C1 is the first ever reported resin with a biobased content as high as 89.7 wt.% displaying tensile modulus almost reaching 1 GPa and tensile strength above 30 MPa, being therefore able to serve as a promising candidate for the formulation and the production of biobased resins for VP able to display improved properties compared to many fossil-based counterparts. The flexural properties, on the other hand, do not seem to depend strongly on the resin’s composition, except for two isolated cases. Resin A3 displayed a strong reduction in flexural strength related to a significant decrease in its flexural modulus, while resin C1 still showed reduced flexural strength but, this time, related to a reduction in the deformation at break, with a slight observed increment in the flexural modulus. It is crucial, however, to keep in mind the intrinsic anisotropy of 3D printed materials, especially when evaluating their mechanical properties. In fact, during tensile testing the mechanical stress is applied along the photocured layers, with virtually no component laying perpendicular to it, and therefore the interlayer adhesion gives little to no contribution to the tensile properties. On the other hand, in this configuration, three-point flexural testing generates compressive and tensile stresses perpendicular to the layer planes and consequently the flexural properties are more related to interlayer adhesion. Therefore, the comparison between tensile and flexural properties is not trivial, as they describe different aspects of the materials’ mechanical behavior. While UDO₃-I demonstrated to act as a softening agent that reduces rigidity and increases deformability under tensile stress with an overall improvement of the materials’ strength in all tested formulations except A3, its effectiveness on the flexural properties is limited. Furthermore, the formulation in which UDO₃-I showed the most relevant improvement in tensile properties (resin C1) is also the one that suffered most significantly from the flexural point of view. Finally, with respect to hardness, all materials displayed hardness in the 80–90 Shore D range, following the trend observed for the elastic modulus, as can be expected from photocured thermosets. Finally, when compared to commercially available resins based on urethane diacrylates, the proposed biobased resins display comparable elastic modulus, generally lower elongation at break and often higher tensile strength, therefore underlining the potential of UDO₃-I as a viable replacement for isocyanate-based urethane photocurable monomer for high-performance additive manufacturing. Because of their similarity with commercially available alternatives, their typical applications are those associated with this class of materials, ranging from general merchandise to high-value products such as jewelry and textiles, but representing a significant step toward the improvement of sustainability in the 3D printing industry.

In conclusion, the presented approach described the preparation of biobased and photocurable diurethanediols bis(methylitaconate) named UDOs-I by aminolysis of cyclic carbonates and further functionalization with itaconic acid monomethyl ester, to be employed as monomers for the formulation of biobased resins for VP. While the use of symmetric ethylene carbonate led to a solid compound (UDO₂-I) that could not be efficiently formulated into photocurable mixtures, our objective was the introduction of asymmetry using propylene carbonate that hindered the solidification of the mixture, allowing for the obtainment of a liquid monomer (UDO₃-I) that could be easily formulated for the first time with (meth)acrylic and itaconic acid-based co-monomers in all proportions. By analyzing the materials in terms of their tensile and flexural properties, it was possible to conclude that resins formulated with up to 50 wt.% UDO₃-I and I₂B₁ displayed mechanical properties that are better or equal to the ones of the (meth)acrylate-based formulations, this suggests the remarkable potential of this kind of biobased photocurable monomers for increasing the sustainability of the VP manufacturing process. Finally, resins B1 and C1 are the first ever reported example of a resin for VP associated with a biobased content as high as 75 wt.% and 90 wt.% that are able to be photocured into solid materials showing tensile strength around 1 GPa. At present, there are no biobased resins that can entirely supplant acrylate- and methacrylate-based formulations in terms of both printability and mechanical properties. Nonetheless, researchers are making headway in this domain by creating fresh monomers, for instance, UDO₃-I, which can enable the formulation to have a higher biobased content while maintaining its performance. This advancement presents a viable avenue for industrial producers to enhance their sustainability practices.