RAFT-Mediated 3D Printing of Polylactones/Itaconate Elastomers with Polypeptide Surface Functionalization

Abstract

Reversible addition–fragmentation chain transfer (RAFT) polymerization has gained interest in vat photopolymerization, particularly for enabling postprinting surface functionalization via reactivation of the RAFT agent. In this work, we report the development of RAFT photopolymerizable resins containing up to 50% renewable content using sustainable dimethyl or dibutyl itaconate as primary monomers combined with hydroxyethyl acrylate as a reactive comonomer. A 4-arm polyester cross-linker end-functionalized with itaconic acid (IA), poly­(caprolactone-co-valerolactone)-IA, was synthesized and incorporated into the resin formulation. Photorheology confirmed efficient polymerization, and mechanical characterization revealed elastomeric properties for networks derived from dimethyl itaconate. Digital light processing (DLP) of this formulation enabled the 3D printing of flexible structures, including microneedles. The presence of pendant carboxylic acid groups in the cross-linker imparted pH-responsiveness to the printed objects, allowing for reversible swelling and size changes in response to environmental pH, demonstrating 4D behavior. Leveraging the controlled nature of RAFT polymerization, a two-stage printing approach was employed. After printing with the itaconate-based ink, a switch to a methacrylated polylysine ink enabled surface biofunctionalization. Successful grafting of polylysine was confirmed by atomic force microscopy (AFM) and FTIR spectroscopy. Preliminary results demonstrate antimicrobial activity of the cationic surfaces, as well as the ability to spatially control surface functionalization, exemplified by patterned attachment of fluorescent polylysine.

Introduction

Additive manufacturing (AM) through the polymerization of photoreactive resins upon UV exposure, commonly known as vat 3D printing, has drawn significant attention because of its fast printing speeds and high-resolution capabilities. − Due to its fast reaction kinetics and versatility, free radical photopolymerization (FRP) is widely used in vat 3D printing to fabricate materials with diverse properties. − This technique enables the production of 3D materials with tailored functionalities, including shape memory, self-healing, and conductivity, to meet specific application requirements. − However, FRP-based materials often exhibit inferior mechanical properties due to the inherent limitations of their chain-growth mechanism. To overcome this, photoinduced thiol–ene cross-linking has been explored, as its step-growth mechanism leads to more uniform network formation, improving mechanical performance. − Similarly, reversible addition–fragmentation chain transfer (RAFT) polymerization has been integrated into vat 3D printing to improve network homogeneity. − Unlike FRP, RAFT promotes dynamic exchange between dormant and active species, providing enhanced control over polymerization kinetics and network architecture. This results in more uniform polymer networks with precisely tuned cross-linking densities, ultimately improving the mechanical performance of printed structures. Moreover, in contrast to the permanently terminated chains in FRP, the presence of a chain transfer agent (CTA) in RAFT allows their postprinting activation to enable controlled surface polymerization for the functionalization of 3D-printed materials to, for example, modulate the interaction with biological environments. ,

Irrespective of the technology used, with the projected growth of AM, concerns surrounding sustainability, particularly with regard to raw materials, are becoming increasingly critical. As a result, the use of renewable feedstocks and circular design principles is gaining traction to replace petroleum-based reagents in vat 3D printing. − Despite these efforts, all RAFT 3D printing reports to date utilize (meth)­acrylate-based resins. Itaconic acid (IA), a biomass-derived feedstock, presents a promising alternative. − Produced industrially via the fermentation of carbohydrates by filamentous fungi, IA offers a green and cost-effective solution at approximately $0.5/kg. In the context of vat 3D printing, IA has primarily been explored through the FRP of its unsaturated derivatives. − However, when compared to (meth)­acrylate monomers, IA and its derivatives exhibit significantly slower propagation rates due to steric hindrance. ,

In this work, we present an IA-based resin formulation specifically designed for 3D RAFT 3D printing. We demonstrate that this formulation enables the fabrication of elastic materials while preserving the advantages of RAFT polymerization for surface functionalization. By reinitiating the RAFT process, the ability to pattern the surface and render it antimicrobial is demonstrated, highlighting its potential for biomedical applications. The proposed resin is based on two commercially available IA-derived monomers: dibutyl itaconate (DBI) and dimethyl itaconate (DMI). As a cross-linker, an IA-tetrafunctional copolymer was used, obtained from ε-caprolactone (CL) and δ-valerolactone (VL), poly­(CL-co-VL)-IA. The results highlight that renewability and functionalization can be seamlessly integrated into the 3D printing of advanced materials and structures.

Experimental Section

Materials

All chemicals were purchased from Sigma-Aldrich unless otherwise noted. 4-Cyano-4-[(dodecylsulfanylthiocarbonyl)­sulfanyl]­pentanoic acid (CDTPA), dibutyl itaconate (DBI), diphenyl­(2,4,6-trimethylbenzoyl)­phosphine oxide (TPO), ε-caprolactone, δ-valerolactone, propylene oxide, 4-dimethylaminopyridine (DMAP), 1-pyrenecarboxaldehyde, malonic acid, and hydrogen bromide (30% in acetic acid) were purchased from Tokyo Chemical Industry. Dimethyl itaconate (DMI), N6-benzoyl-l-lysine (N-Cbz-l-Lysine), 1-Ethyl-3-(3-(dimethylamino)­propyl)­carbodiimide hydrochloride (EDC·HCl), Na₂SO₄, pyridine, lithium phenyl-2,4,6-trimethylbenzoylphosphinatem (LAP) were purchased from Fluorochem. N-Cbz-Lysine N-carboxyanhydride (Lys-NCA) was synthesized following a literature procedure.

Methods

Nuclear Magnetic Resonance (NMR)

¹H NMR and Diffusion Ordered Spectroscopy (DOSY) spectra were recorded by using a Bruker Avance 400 MHz spectrometer at room temperature. All chemical shifts were reported in parts per million (ppm) relative to the solvent reference peak δ at 7.26 ppm for chloroform-d, 4.79 ppm for deuterated water, 11.50 ppm for trifluoroacetic acid-d. Diffusion coefficients are reported in units of cm²·s^–1. All the spectra were analyzed using MestReNova by Mestrelab Research (mestrelab.com). DOSY NMR were analyzed using the Bayesian method, resolution factor 0.1–1, and repetitions 1.

Gel Permeation Chromatography (GPC) in CHCl₃

Molecular weight distributions and polydispersity indexes (Đ _M) of the itaconic acid copolymers were determined by a CHCl₃ Agilent Technologies LC 1200 Series equipped with an Agilent 1260 ISO pump, Agilent refractive index detector, SDV 5 μm 8 × 50 mm precolumn, and 2 SDV 5 μm 8 × 300 mm columns in series. Samples were dissolved in CHCl₃, and their chromatograms were recorded with a flow of 1.0 mL·min^–1 at 40 °C. The system was calibrated against PSS Polymer Standards Service GmbH linear poly­(methyl methacrylate). All GPC samples were prepared at a concentration of 4 mg·mL^–1 and were filtered through a 0.2 μm Millipore filter before injection.

Gel Permeation Chromatography (GPC) in Hexafluoro-2-propanol (HFIP)

Molecular weight distributions and polydispersity indexes (Đ _M) of the poly­(lysine) were determined by a hexafluoro-2-propanol (HFIP) Agilent Technologies 1260 Infinity Series equipped with an Agilent 1260 ISO pump, Agilent refractive index detector, PFG 7 μm 8 × 50 mm precolumn, PFG 1000 Å 7 μm 8 × 300 mm, and PFG 100 Å 7 μm 8 × 300 mm columns in series. Samples were dissolved in HFIP, and their chromatograms were recorded with a flow of 1.0 mL·min^–1 at 35 °C. The system was calibrated against PSS Polymer Standards Service GmbH linear poly­(methyl methacrylate). All GPC samples were prepared at a concentration of 4 mg·mL^–1 and were filtered through a 0.2 μm Millipore filter before injection.

Photorheology and Viscosity

An Anton Paar MCR 301 instrument equipped with a parallel plate of 25 mm diameter and a gap length of 0.05 mm was used to conduct photorheology and viscosity experiments at room temperature. Photorheology experiments were carried out by equipping the machine with a Thorlabs UV LED light 405 nm (M405L3-C1) and a sample glass plate, allowing the passage of light. The experiments were conducted by using a Peltier hood to protect the sample from ambient light. The intensity of the light was 9 mW·cm^–2 (measured on the surface of the glass plate). Each time point was taken every 5 s through a time sweep experiment with constant oscillations at a fixed frequency of 10 rad·s^–1 with a strain of 0.1%. Shear rate sweep experiments were carried out at room temperature to measure the viscosity in the shear range from 0.01 to 100 s^–1, using a Peltier hood to protect the sample.

UV–vis

UV–vis absorption spectra were recorded using an Agilent Cary 3500 Cary Multicell Peltier UV–vis spectrophotometer equipped with a quartz cuvette with a path length of 1 cm. Measurements were carried out over a wavelength range of 200–800 nm at room temperature.

Dynamic Mechanical Analysis (DMA)

Experiments were performed in shear mode using shear clamps on a Mattler-Toledo DMA/SDTA 861e machine. The samples were subjected to an oscillatory deformation at a fixed frequency of 1 Hz, while the temperature was ramped from −100 to 120 °C at a heating rate of 3 °C·min^–1. An amplitude force of 5 N with a displacement amplitude of 2 μm was used. Rectangular-shaped samples were prepared by film casting using molds, resulting in films with dimensions of 20 × 10 × 0.5 mm (L × W × H). The resins were cured overnight under 405 nm LED light at an intensity of 2 mW·cm^–2 at room temperature.

Tensile Tests

Tensile testing was carried out using a Testometric M100–1CT machine equipped with a 50 N cell load (LC5). Cross-linked sheets were used to perform the measurement, prepared with rectangular molds 20 × 10 × 0.5 mm (L × W × H) as described above. A gauge length of 8 mm, pretension of 0.1 N, and test speed of 10 mm·min^–1 were set as parameters for the test. Young’s modulus, elongation at break, and ultimate strength were determined as averages of five independent drawing experiments performed at the same conditions at room temperature.

Differential Scanning Calorimetry (DSC)

DSC measurements were performed using a TA Instruments DSC Q200 and TA Instruments RSC FC-100 immersion cooler, with 5–10 mg of the dry cross-linked polymer network as a sample, or dry polymer. A heating and cooling rate of 10 °C/min was used for 2 cycles. The second heating and cooling cycles were used to analyze the thermal properties. Each sample was measured in an aluminum Tzero pan under nitrogen flow using an empty pan as a reference.

Swelling Tests

Swelling tests were performed using square-shaped samples with dimensions 10 × 10 × 0.5 mm (L × W × H), prepared by film casting using molds and curing the resin overnight under 405 nm LED (2 mW·cm^–2) at room temperature. Chloroform was used as a swelling solvent. Gel fraction and swelling ratio were determined using cross-linked samples. The samples were weighed (W _d) before swelling for 48 h at room temperature. Next, the swelled samples were weighed (W _s) and dried in a vacuum oven (40 °C) overnight. The dry samples after swelling were weighed (W _a) and the gel fraction and swelling ratio were calculated by using the equations below. All measurements were conducted in triplicate at room temperature.

$$\mathrm{gel}\mathrm{fraction}(\%)=\frac{W_{\mathrm{a}}}{W_{\mathrm{d}}}\times 100$$

$$\mathrm{swelling}\mathrm{ratio}=\frac{W_{\mathrm{s}}-W_{\mathrm{a}}}{W_{\mathrm{a}}}$$

Atomic Force Microscopy (AFM)

Samples for AFM imaging were prepared by attaching printed films to a clean coverslip. Imaging and analysis were performed using a JPK NanoWizard 4 system in the Quantitative Imaging (QI) mode. AFM tips (PPP-NCHAuD) with a tip radius of 2–12 nm were purchased from NANOSENSORSTM. Scanning was performed with a set point range of 2–30 nN, a Z length of 50–1000 nm, a pixel time of 3–10 ms, and a pixel resolution ranging from 128 × 128 to 512 × 512 within a scanning window of 200 × 200 nm² to 50 × 50 μm². Data acquisition and analysis, including AFM image and Young’s modulus calculations, were conducted using the JPK Data Processing software in QI mode. Samples were prepared by DLP printing of squares with dimensions 5 × 5 × 0.7 mm (L × W × H) and postcuring for 20 min (405 nm, 9 mW·cm^–2).

pH Response Tests

The test was performed on a cross-linked disc of size H 3 × Ø 7 mm. Samples were prepared by pouring the resins into a disc-shaped mold and irradiating with UV light 405 nm (2 mW·cm^–2) for 6 h at room temperature. The pH-responsive behavior was evaluated by sequentially immersing the cross-linked network in buffer solutions of different pH at room temperature for 24 h in triplicate. The sample was first immersed in a pH 10 and allowed to swell. Next, the hydrogel was removed, the excess surface water removed, and weighed (W ₁). The same sample was then transferred to a fresh pH 7 solution before being weighed again (W ₂). The hydrogel was then placed in a pH 4 buffer, followed by weighing (W ₃). Finally, the hydrogel was dried under a vacuum oven (40 °C) and then weighed (W ₀). At each step, water uptake was calculated using the formula (W ₀ is the dry weight, W _n is the swollen weight at each pH):

$$\mathrm{water}\mathrm{uptake}(\%)=\frac{W_n-W_0}{W_0}\times 100$$

Confocal Microscopy

All samples were placed in a 12-well plate and sterilized by ultraviolet irradiation in an ultraclean cabinet. Then 10 μL of bacterial suspension (1 × 10⁶ CFU·mL^–1) was added at the center region of the sample. The samples were covered with the pristine PE membranes. After that, a small amount of PBS was added into the wells around the samples, and the 12-well plate was placed in a 37 °C incubator to ensure the appropriate temperature and humidity required for normal growth of bacteria. After cultivation for 24 h, the samples were stained in SYTO 9 (10 μg·mL^–1) and PI (10 μg·mL^–1) for 20 min before being imaged by confocal microscopy.

Synthesis of 4-arm-poly­(CL-co-VL)-IA

ε-Caprolactone (11.41 g, 100 mmol, 12 equiv), δ-valerolactone (6.66 g, 67 mmol, 8 equiv), pentaerythritol (1.14 g, 8.4 mmol, 1 equiv), and Tin­(II) octoate (170 mg, 0.42 mmol, 0.05 equiv) were added into a Schlenk flask. The reaction mixture was stirred under a nitrogen atmosphere at 110 °C for 24 h. Subsequently, Itaconic acid (4.37 g, 34 mmol, 4 equiv) was added to the reaction mixture, and the temperature was increased to 140 °C for 16 h with water distilled from the reaction. The polymer was used without further purification for the formation of cross-linked networks. P­(CL-co-VL) ¹H NMR (400 MHz, CDCl₃) δ: 4.13–3.99 (m, 39H, −C H ₂ –O–C­(O)−), 3.66–3.47 (m, 8H, −C–C H ₂ –O−), 2.40–2.24 (m, 40H, −C H ₂ –C­(O)−), 1.74–1.31 (m, 104H). P­(CL-co-VL)-IA ¹H NMR (400 MHz, DMSO-d ₆) δ: 12.30 (s, 1H, −COO H ), 6.25–6.08 (m, 3.4H, −CC H ₂ ), 5.83–5.68 (m, 3.4H, −CC H ₂ ), 4.10–3.09 (m, 47H, −C H ₂ –O–C­(O)−), 2.33–2.14 (m, 40H, −C H ₂ –C­(O)−), 1.60–1.20 (m, 105H).

Synthesis of Poly­(dimethyl itaconate-co-2-hydroxyethyl acrylate) and Poly­(dibutyl itaconate-co-2-hydroxyethyl acrylate) via Photo-RAFT (Target DP 25)

HEA (100 mg, 0.86 mmol, 12.5 equiv), dimethyl itaconate (136 mg, 0.86 mmol, 12.5 equiv), CDTPA (28 mg, 0.069 mmol, 1 equiv), and TPO (48 mg, 0.14 mmol, 2 equiv) were added to a vial. The reaction mixture was stirred under UV light 405 nm (9 mW·cm^–2) for 6 h at room temperature. The polymer appeared transparent, yellow, hard, and brittle, and was characterized by ¹H NMR and GPC. For poly­(dibutyl itaconate-co-2-hydroxyethyl acrylate), the same ratios were used. The polymer appeared transparent, yellow, and as a viscous oil and was characterized by ¹H NMR and GPC.

Synthesis of Poly­(dimethyl itaconate-co-2-hydroxyethyl acrylate) and Poly­(dibutyl itaconate-co-2-hydroxyethyl acrylate) via Photo-RAFT (Target DP 50)

HEA (100 mg, 0.86 mmol, 25 equiv), dimethyl itaconate (136 mg, 0.86 mmol, 25 equiv), CDTPA (14 mg, 0.034 mmol, 1 equiv), and TPO (24 mg, 0.069 mmol, 2 equiv) were added to a vial. The reaction mixture was stirred under UV light 405 nm (9 mW·cm^–2) for 6 h at room temperature. The polymer appeared transparent, yellow, hard, and brittle, and was characterized by ¹H NMR and GPC. For poly­(dibutyl itaconate-co-2-hydroxyethyl acrylate), the same ratios were used. The polymer appeared transparent, yellow, and as a viscous oil and was characterized by ¹H NMR and GPC.

Digital Light Processing (DLP) by Photo-RAFT Polymerization

A custom MONO3-2K40 from Monoprinter (U.S.) equipped with a UV LED 405 nm was used for digital light processing 3D printing. The projector resolution was 1902 × 1080 pixels with an in-plane resolution of 15 μm. A layer thickness of 50 μm and light intensity of 22 mW·cm^–2 (measured on the tank surface) were used for all the printing at room temperature. The CAD designs for the structures were generated using Tinkercad (tinkercad.com) by Autodesk. All the structures were postcured for 20 min (405 nm, 9 mW·cm^–2). The following resin formulations were used:

$$\mathrm{D}\mathrm{M}\mathrm{I}‐\mathrm{R}50:\mathrm{DMI}(25\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v}),\mathrm{HEA}(25\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v}),\mathrm{P}(\mathrm{CL}‐\mathit{co}‐\mathrm{VL})‐\mathrm{IA}(1\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v}),\mathrm{CDTPA}(1\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v}),\mathrm{TPO}(2\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v})$$

$$\begin{gathered} \mathrm{D}\mathrm{B}\mathrm{I}‐\mathrm{R}50:\mathrm{DBI}(25\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v}),\mathrm{HEA}(25\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v}),\mathrm{P}(\mathrm{CL}‐\mathit{co}‐\mathrm{VL})‐\mathrm{IA}(1\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v}),\mathrm{CDTPA}(1\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v}),\mathrm{TPO}(2\mathrm{eq}\mathrm{u}\mathrm{i}\mathrm{v}) \\ \mathrm{p}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{l}\mathrm{y}\sin \mathrm{e}‐\mathrm{b}\mathrm{a}\mathrm{s}\mathrm{e}\mathrm{d}\mathrm{r}\mathrm{e}\sin \mathrm{f}\mathrm{o}\mathrm{r}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h}\mathrm{P}\mathrm{L}\mathrm{L}\mathrm{M}\mathrm{A}:\mathrm{PLLMA}15\mathrm{wt}\%,\mathrm{LAP}1\mathrm{wt}\%(\mathrm{p}\mathrm{h}\mathrm{o}\mathrm{t}\mathrm{o}\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{o}\mathrm{r}),{\mathrm{DI}\mathrm{H}}_2\mathrm{O}74\mathrm{wt}\% \end{gathered}$$

$$\mathrm{p}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{l}\mathrm{y}\sin \mathrm{e}‐\mathrm{b}\mathrm{a}\mathrm{s}\mathrm{e}\mathrm{d}\mathrm{r}\mathrm{e}\sin \mathrm{f}\mathrm{o}\mathrm{r}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h}\mathrm{P}\mathrm{L}\mathrm{L}\mathrm{M}\mathrm{A}‐\mathrm{P}\mathrm{y}:\mathrm{PLLMA}‐\mathrm{Py}15\mathrm{wt}\%,\mathrm{LAP}1\mathrm{wt}\%,{\mathrm{deionized}\mathrm{water}\mathrm{H}}_2\mathrm{O}74\mathrm{wt}\%$$

Surface Antimicrobial Activity Assay

The antibacterial activity of the printed film surfaces was evaluated according to the JIS Z2801 standard using an agar plate colony counting method. The bacterial suspension was diluted with LB broth to achieve a working concentration of 1 × 10⁶ CFU·mL^–1. Two sample sets were prepared: The control group comprised 3D-printed square specimens (10 × 10 × 0.5 mm) fabricated from DMI-R50 resin without surface modification. The experimental group was identically prepared but received an additional in situ surface modification through the direct printing of a polylysine layer (20 nm thickness) onto the DMI-R50 substrate. After printing, the samples were washed and postcured under a 405 nm LED light for 20 min (9 mW·cm^–2). Then, the samples were placed in a 12-well plate, and then 10 μL of Staphylococcus aureus suspension (1 × 10⁶ CFU·mL^–1 (Colony-Forming Unit)) was added at the center region of the surfaces. The surfaces were covered with pristine PE membranes. After cultivating for 24 h, 2 mL of PBS buffer was added to each well directly to ensure the samples were entirely immersed. Then the plate was ultrasonicated for 5 min to release the bacteria adherent on the sample and PE membrane. After that, PBS buffer containing bacteria was then subjected to a serial dilution, plated on agar, and incubated at 37 °C for 24 h. Each sample was carried out at least in triplicate.

Results and Discussion

Itaconic acid and its derivatives DMI and DBI are classified as sterically hindered monomers, resulting in reduced reactivity and lower propagation rates in radical polymerization compared to acrylates. , Under oxygen-free conditions, conventional RAFT polymerization has demonstrated high conversion rates and narrow dispersities for these monomers at extended reaction time. However, incorporating DMI and DBI into a photo-RAFT 3D printing process presents several challenges. First, vat 3D printing requires rapid polymerization, which DMI and DBI alone cannot achieve. Beuermann et al. reported enhanced propagation rates of itaconates in FRP by copolymerization with acrylates, a strategy also adopted here by the addition of 2-hydroxyethyl acrylate (HEA) as a comonomer. Second, unlike typical RAFT polymerization, oxygen exclusion is not easily feasible in conventional vat-based printing, potentially compromising the level of polymerization control. Finally, the choice of a chain transfer agent (CTA) is critical. For the homopolymerization of DMI and DBI, 4-cyano-4-(phenylcarbonothioylthio)­pentanoic acid (CPADB) has been previously reported as the CTA to provide satisfactory control over molecular weight and polymer dispersity (Đ). However, the phenol group in CPADB can cause polymerization retardation, particularly at higher concentrations, making it less suitable for highly reactive monomers, such as acrylates and acrylamides. Since HEA is used as a comonomer in this study, 4-cyano-4-[(dodecylsulfanylthiocarbonyl)­sulfanyl]­pentanoic acid (CDTPA) was selected as the CTA. , CDTPA shares the same leaving group as CPADB, ensuring efficient reinitiation, while the dodecylsulfanyl Z-group minimizes retardation by providing less stabilization to the macroradical compared to the phenol group.

To establish the polymerization characteristics and test the resin formulation, photo-RAFT polymerization was trialed prior to vat 3D printing (Figure A). Photo-RAFT copolymerization of DMI and DBI with HEA was conducted in bulk without degassing to mimic vat 3D printing conditions. The total molar ratio of monomer to CTA was set to 25:1 and 50:1, with an itaconate monomer to HEA ratio of 1:1. The photoinitiator diphenyl­(2,4,6-trimethylbenzoyl)­phosphine oxide (TPO) was used at two equivalents to the CTA to accelerate the polymerization of the low-reactivity itaconate monomers. After 6 h >97% conversion of both itaconate monomers and HEA was confirmed by ¹H NMR analysis (Figure B). For poly­(HEA-st-DMI), both monomers displayed comparable conversion rates (Figure S3). In contrast, for poly­(HEA-st-DBI), DBI converted more slowly than HEA (Figure S4). Nevertheless, ¹H NMR analysis confirmed that both monomers reached a nearly quantitative conversion by the end of the reaction. ¹H and DOSY NMR spectra of the purified copolymers further verified the presence of CTA (Figures S1 and S2), although signal integration suggested a partial loss of RAFT end groups, likely due to the high photoinitiator content. These findings were corroborated by UV–vis analysis (Figure S5). Size-exclusion chromatography (SEC) analysis confirmed good control over the polymerization, as evident from the symmetrical traces of the four targeted copolymers and their relative alignment with the targeted DP of 25 and 50 (Figure C). However, relatively high dispersities around Đ = 2.5 were observed (Table S1), likely due to the presence of oxygen and low [CTA]/[initiator] ratio. The latter was not deemed problematic for the 3D printing of cross-linked materials. Overall, these experiments validate the effectiveness of the selected photo-RAFT conditions for the IA monomers.

1.

(A) Reaction scheme of photo RAFT polymerization of itaconate monomers and HEA at a CTA to monomer ratio of 25 and 50 using TPO as a photoinitiator (λ_max= 405 nm, I = 9 mW*cm^–2). (B) ¹H NMR spectra of poly­(HEA-co-DMI) and poly­(HEA-co-DBI) obtained from the reaction without purification. The yellow regions represent residual TPO. (C) Size-exclusion chromatograms of poly­(HEA-co-DMI) and poly­(HEA-co-DBI) from bulk RAFT polymerization at different CTA to monomer ratios.

To formulate a 3D-printable resin, a cross-linker must be incorporated. In our previous work, we explored thiol-ene cross-linked resins using IA difunctional poly­(caprolactone) (PCL). However, the semicrystalline nature of PCL necessitated the use of a solvent in printing formulations. To mitigate this issue, here we synthesized a four-arm copolymer from ε-caprolactone (CL) and δ-valerolactone (VL) at a 3:2 ratio (M _n = 2250 g·mol^–1), subsequently end-functionalized with IA in a one-pot, solvent-free process, yielding poly­(CL-co-VL)-IA (Figure A). ¹H NMR and SEC analysis confirmed the copolymer’s composition and the successful incorporation of IA end groups (Figure S6) as well as a monomodal molecular weight distribution with a dispersity of Đ = 1.8 (Figure S7). Importantly, DSC analysis verified the amorphous morphology of this cross-linker (Figure S8), which remains a viscous, transparent liquid at room temperature, making it well-suited for solvent-free vat 3D printing formulations. For all subsequent experiments, five photo-RAFT resins were formulated containing the monomers DMI or DBI and HEA at a molar ratio of 1:1, and monomer to CTA ratios of 25:1 and 50:1 (denoted as DMI-R25, DBI-R50, etc., Table ). The molar ratio of P­(CL-co-VL)-IA: CTA:TPO ratio was kept constant at 1:1:2. For comparison, a resin containing only HEA was also investigated (HEA-R50).

2.

(A) Synthesis of the cross-linker poly­(CL-co-VL)-IA. (B) DLP process and resin composition.

1. Resin Formulation and Gel Properties of the Networks from RAFT Photopolymerization.

resin	P(CL-co-VL)-IA (wt %)	monomers (wt %)	CTA (wt %)	TPO (wt %)	gel fraction (%)	swelling ratio (%)
HEA-R50	26.7	61.6	4.3	7.4	95.5 ± 1.3	1.7 ± 0.1
DMI-R25	21.8	59.2	6.9	12.1	77.5 ± 0.7	4.4 ± 0.1
DBI-R25	18.4	65.4	5.9	10.3	66.3 ± 1.4	6.5 ± 0.3
DMI-R50	24.0	65.4	3.9	6.7	88.1 ± 1.6	3.5 ± 0.1
DBI-R50	20.0	71.2	3.3	5.5	72.3 ± 3.2	5.2 ± 0.3

^aMixture of DBI or DMI with HEA at a molar ratio of 1:1.

^bMolar ratio PI:CTA at 2:1 for all formulations.

^cStandard deviation of n = 3.

Photorheology experiments under 405 nm (9 mW·cm^–1) irradiation revealed distinct gel points for all resin formulations (Figure A,B). The onset of gelation for resins containing DMI or DBI with HEA occurred within 70–120 s, whereas the resin containing only HEA exhibited a significantly faster gel point of approximately 15–20 s. This difference aligns with the lower reactivity of the itaconate monomers compared to that of HEA. To validate monomer reactivity, double bond conversion was monitored using Fourier transform infrared spectroscopy (FTIR). The disappearance of the characteristic C = C stretching peak at 1637 cm^–1 (corresponding to HEA, DMI, DBI, and the itaconic acid chain ends of the P­(CL-co-VL)-IA cross-linker) confirmed high double bond conversion across all resin formulations (Figure C,D). When comparing the gel fractions of the itaconate-containing resins, DBI-based resins exhibited a ca. 10% lower gel fraction than DMI-based resins, regardless of the monomer-to-CTA ratio (Table ). This suggests a higher degree of double bond conversion in DMI resins, consistent with its greater reactivity, due to its less bulky ester groups with reduced steric hindrance. Swelling ratio measurements further support this trend, with DMI resins showing lower swelling ratios due to a higher cross-link density.

3.

Photorheology graphs of resins listed in Table . UV 405 nm, 9 mW·cm^–1, light turned on at 60 s. (A) Monomers to CTA molar ratio 25:1. (B) Monomers to CTA molar ratio 50:1. FTIR spectra of (C) liquid itaconate resins and (D) cross-linked itaconate networks. Peaks in the region of 1650–1600 cm^–1 represent the double bond groups.

Dynamic mechanical analysis (DMA) was conducted to evaluate the mechanical properties of RAFT-polymerized resins as a function of temperature (Figure A). All networks exhibited a distinct glass transition, with glassy storage moduli ranging from 57 to 130 MPa, influenced by the T _g values of the respective polymers. The HEA-R50 network showed a glass transition temperature (T _g) of 1.6 °C, influenced by the incorporation of the cross-linker, which restricts the polymer chain mobility. The HEA copolymer networks DBI-R50 and DBI-R25 exhibited a T _g in the range −1 to −5 °C, which is lower compared to the HEA-R50 network and poly­(DBI) (T _g = 15 °C). In contrast, DMI-R50 and DMI-R25 networks have a T _g around 30 °C, higher values compared to the DBI networks, derived from the higher T _g of poly­(DMI) (98 °C). The glass transition behavior of the DMI networks falls near room temperature, suggesting that these networks retain elasticity at room temperature, making them promising candidates for flexible material applications. To verify this, tensile tests were carried out to analyze the mechanical properties of RAFT networks at room temperature (Figure B). The lowest Young’s modulus (0.50 ± 0.05 MPa) and strain at break (43.8 ± 5.0%) were observed at 23 °C for DBI networks, a temperature at which it has reached the rubbery state. The HEA-R50 network shows a slightly higher Young’s modulus (1.42 ± 0.15 MPa) and strain at break (48.8 ± 4.9%) at its rubbery state at 23 °C. A significantly higher Young’s modulus (13.48 ± 1.39 MPa) and strain at break (223.5 ± 13.7%) were observed for DMI-R50 and DMI-R25 networks (5.26 ± 0.61 MPa and 351.1 ± 29.0%). Encouraged by this, we set out to explore the elasticity of DMI networks by cyclic stretching. With an increasing strain from 50 to 200%, the elastic hysteresis can be clearly observed with an increased energy dissipation (Figure C). However, after each cycle, the material did not return to 0% strain after unloading. This means that it underwent irreversible plastic deformation due to the applied strain exceeding the elastic region. The progressive increase in strain after unloading across successive cycles indicates mechanical fatigue or creep, suggesting that the material is gradually undergoing permanent deformation with each cycle.

4.

(A) Thermal mechanical test of RAFT networks. (B) Statical tensile test of RAFT networks at room temperature (23 °C). (C) Cyclic stretching with increased strain for the DMI-R50 network.

Next, the printability of the developed RAFT-based itaconate DMI and DBI resins by DLP was assessed using the formulations listed in Table . We first compared the printability of the DMI-based resin with and without the RAFT agent under identical conditions (50 s exposure time and 22 mW cm^–2 light intensity). A simple square geometry was selected to accentuate differences in curing behavior. The resin without RAFT exhibited pronounced overcuring across the entire structure, whereas the RAFT-containing formulation yielded a sharp, well-defined square with no signs of overcuring (Figure S9). This clear visual contrast highlights the beneficial role of the RAFT agent in controlling the curing process. For the actual printing, each layer (50 μm) was irradiated for 70 s at an intensity of 22 mW/cm². A “HEISE GROUP” patterned object with the dimensions of 10 × 10 × 1.5 mm³ (L × W × H) was successfully fabricated from both DMI and DBI resins (Figure A). Encouraged by the toughness and elastic property of the DMI-R50 network, a microneedle model was successfully printed at a speed of 2.5 mm/h (Figure B). The overall structure measures 10 × 10 × 1.4 mm³, while individual microneedles are 0.95 mm high with a base diameter of 0.5 mm and a tip diameter of ca. 50 μm. These dimensions have previously been reported for transdermal delivery. , Moreover, 4D printing via RAFT polymerization was successfully demonstrated with DMI resin. A clover with 3 leaves was printed with original dimensions of 10 × 10 × 1.5 mm³ (L × W × H) at a speed of 2.5 mm/h from the DMI-R50 resin. The printed clover was then swollen in buffer solutions of different pH until reaching equilibrium. The dimension was found to vary to 14 × 14 mm (L × W), 12 × 12 mm, 11 × 11 mm, respectively, at pH 10, 7, and 4 due to the presence of the carboxylic acid groups in the P­(CL-co-VL)-IA cross-linker (Figure C). A repeatable pH-responsive cycle can be observed by reswelling the structure into pH 10 (Figure S10).

5.

(A) CAD file and 3D-printed structure of “HEISE GROUP” using DMI-R50 resin (50 μm layer thickness, 22 mW·cm^–2, 70 s per layer). Scale bar 3 mm. (B) CAD file and 3D-printed structure of microneedles using DMI-R50 resin (50 μm layer thickness, 22 mW·cm^–2, 70 s per layer). Scale bar 3 mm. Zoom ×5 of microneedles. Scale bar 0.5 mm. Both the 3D models were designed using Tinkercad (tinkercad.com) by Autodesk. (C) 3D-printed structure using DMI-R50 resin over graph paper (50 μm layer thickness, 22 mW·cm^–2, 70 s per layer). First pH-responsive cycle. The 3D model was designed using Tinkercad (tinkercad.com) by Autodesk.

We then investigated the surface functionalization of the DMI resin by leveraging the controlled nature of RAFT polymerization with the goal of rendering the surface bioactive. Specifically, we synthesized methacrylate-functionalized polylysine (PLLMA) for the surface modification of printed itaconate-based resins. A square of 5 × 5 × 0.7 mm³ was first printed at a speed of 2.5 mm/h from DMI resin. The resin was then switched to an aqueous solution of methacrylate-functionalized PLLMA (Figure A). One layer of PLLMA (thickness 20 μm) was then printed on the surface of the DMI film with the same dimensions, at a speed of 3.6 mm/h. No photo absorber or inhibitor was added into the PLLMA resin, and the controlled RAFT polymerization was facilitated by the CTA on the surface of the printed DMI film. Upon surface polymerization, characteristic FTIR bands of polylysine at 1651 cm^–1 (CO stretching) and 1539 cm^–1 (C–N bending) were clearly observed on the PLLMA-DMI film (Figure S11), confirming a successful reaction. In addition, atomic force microscopy (AFM) supported the surface grafting by showing a substantial increase in the surface Young’s modulus from 547 MPa to 3.1 GPa after PLLMA was printed onto the DMI network (Figures B and S12). A preliminary qualitative study further revealed antimicrobial activity attributed to the cationic polylysine-modified surface. Both PLLMA-DMI and unmodified DMI films were tested against Gram-positive methicillin-resistant S. aureus (MRSA). The PLLMA-DMI film demonstrated significant bactericidal activity in contrast to the DMI control, where no dead bacteria were observed (Figure C). This is consistent with the reported mechanism of electrostatic membrane disruption by cationic polymers interacting with MRSA.

6.

(A) Surface modification of DLP printed DMI films with PLLMA and PLLMA-Py by resin switch. (B) Surface Young’s modulus of DLP printed DMI films before and after surface modification with PLLMA obtained by AFM. (C) Antibacterial activity assay of an LB agar plate of DMI films before (left) and after surface modification with PLLMA (right). (D) Photograph of the PLLMA-Py surface-modified DMI film under visible light (left) and 365 nm UV light (right).

To visualize the surface modification achieved via RAFT polymerization, PLLMA was further functionalized with 3-(pyren-1-yl)­acrylic acid to impart fluorescence (PLLMA-Py). A rectangular DMI structure (5 × 10 × 0.7 mm³) was first printed, followed by switching to an aqueous solution of pyrene-functionalized PLLMA (PLLMA-Py) for surface patterning. The word “Lys” was printed onto the surface, and a clearly defined fluorescent pattern was observed under 365 nm UV light (Figure D), further validating the successful RAFT-mediated surface functionalization of the DMI network.

Conclusion

In this work, we demonstrate the use of sustainable itaconate-based monomers for RAFT-mediated 3D printing. To address the inherently low reactivity of itaconates in radical polymerization, hydroxyethyl acrylate was employed as a reactive comonomer. By incorporating a branched polyester-itaconic acid functional cross-linker, the resulting resin formulations contained approximately 50% renewable content. Networks derived from dimethyl itaconate exhibited elastomeric properties and proved compatible with Digital Light Processing (DLP), enabling the fabrication of both 3D and pH-responsive 4D structures. Leveraging the controlled nature of RAFT polymerization, a two-stage printing process, switching from the itaconate-based ink to a methacrylated polylysine ink, enabled the surface modification of printed materials to impart a bioactive functionality. This study expands the capabilities of RAFT 3D printing by integrating sustainable monomer systems with surface biofunctionalization strategies. The combination of renewable feedstocks, controlled polymerization, and biofunctional interfaces highlights the potential of this approach for advanced biomaterials and medical device applications.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acspolymersau.5c00117.

• Synthetic procedures; NMR and UV–vis analysis of RAFT copolymers; NMR, DSC, and GPC analysis of cross-linker; images of curing with and without RAFT agent; characterization of polylysine and polylysine grafted surfaces (PDF)

CRediT: Gianluca Bartolini Torres formal analysis, investigation, methodology, writing - original draft; Tianlai Xia formal analysis; Dengwei Yu formal analysis; Quinten Thijssen formal analysis, writing - review & editing; Sandra Van Vlierberghe resources, writing - review & editing; Bo Li conceptualization, investigation, methodology, writing - original draft; Andreas Heise conceptualization, funding acquisition, project administration, resources, supervision, writing - review & editing.

The authors declare no competing financial interest.