Light Processable Starch Hydrogels

Light processable hydrogels were successfully fabricated by utilizing maize starch as raw material. To render light processability, starch was gelatinized and methacrylated by simple reaction with methacrylic anhydride. The methacrylated starch was then evaluated for its photocuring reactivity and 3D printability by digital light processing (DLP). Hydrogels with good mechanical properties and biocompatibility were obtained by direct curing from aqueous solution containing lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) as photo-initiator. The properties of the hydrogels were tunable by simply changing the concentration of starch in water. Photo-rheology showed that the formulations with 10 or 15 wt% starch started curing immediately and reached G’ plateau after only 60 s, while it took 90 s for the 5 wt% formulation. The properties of the photocured hydrogels were further characterized by rheology, compressive tests, and swelling experiments. Increasing the starch content from 10 to 15 wt% increased the compressive stiffness from 13 to 20 kPa. This covers the stiffness of different body tissues giving promise for the use of the hydrogels in tissue engineering applications. Good cell viability with human fibroblast cells was confirmed for all three starch hydrogel formulations indicating no negative effects from the methacrylation or photo-crosslinking reaction. Finally, the light processability of methacrylated starch by digital light processing (DLP) 3D printing directly from aqueous solution was successfully demonstrated. Altogether the results are promising for future application of the hydrogels in tissue engineering and as cell carriers


Introduction
Hydrogels are three-dimensional polymeric networks with the ability of retaining large amount of water within their structures without dissolving. Their crosslinked networks can be formed by either chemical or physical gelation process. The physical gel process involves weak Van der Waals forces, electrostatic interaction, and hydrogen bonding while the chemical gel formation leads to strong chemically bonded networks. Thanks to their hydrophilicity, hydrogels can be utilized by a broad range of sectors including agro-industry, pharmaceutical, medical, and food applications [1][2][3][4][5]. Biopolymer (e.g., polysaccharide and protein) based hydrogels are gaining increasing attention, due to many favorable properties, such as being renewable and bacteriologically degradable [6][7][8]. Starch is an abundant polysaccharide composed of a mixture of two different D-glucose polymers: amylose and amylopectin. Amylopectin is a highly branched polymer, while amylose is linear. In nature, the extracellular matrix (ECM) of mammals is composed of a mixture of polysaccharides, water, soluble factors for signaling and cell-adhesive proteins. Starch based-hydrogels are therefore, partially able parameters using the storage modulus in the frequency independent plateau region (G' P ). In this plateau, the energy applied is stored elastically by the polymer segments between two crosslinks that behave as entropic springs. In fact, the polymer chains are not capable of sliding by one another since the stress velocity is higher than the segment relaxation ability. The network parameters crosslink density (ν e ) and distance between two entanglement points (ξ) were calculated by Equations (1) and (2), respectively.
R is the universal gas constant, T is the temperature, and N A is the Avogadro's number [38].

Compression Test
The compression tests were performed with MTS QTest TM /10 Elite controller using TestWorks ® 4 software (MTS Systems Corporation, Edan Prairie, Minnesota, USA) The unconfined uniaxial compression was performed with cell loaded of 10 N and a test speed of 0.5 mm/min. The samples tested had cylindrical geometry: ϕ = 10 mm X h = 11 mm. The data acquisition rate was set as 20 Hz. The compressive modulus was calculated by TestWorks 4 software. The modulus was estimated as the slope of the linear region of the stress-strain curves.

Swelling Behavior
The swelling behavior was studied by means of conventional gravimetric procedure. Air-dried hydrogels were placed in deionized water at 25 • C. The weight increase was monitored during different time-intervals by taking the samples from the aqueous media and weighting them after the removal of free water on the surface with filter paper. The swelling ratio percentage (S w% ), the equilibrium swelling ratio (S eq ), and the equilibrium water content (EWC) were calculated with Equations (3)-(5), respectively [39,40].
where W s , W d , and W eq are the weight of the sample in the swollen, dry, and swollen at equilibrium states, respectively.

Gel Content
The gel content percentage (G%) was calculated according to Equation (6): where W 1 is the weight of the dry gel after the treatment with deionized water and W 0 is the weight of dry sample before the treatment.

Differential Scanning Calorimetry (DSC)
Differential scanning calorimetry was performed with a Mettler-Toledo DSC 820 (Mettler-Toledo, Stockholm, Sweden). Samples having masses of approximately 7 mg were inserted in 100 µL aluminum pans with pierced lids. The applied heating rate was 10 • C min −1 in a nitrogen atmosphere Polymers 2020, 12, 1359 5 of 14 (rate 50 µL/min). Thermal behavior of the samples was investigated using the following heating-cooling cycles: the temperature was raised from room temperature to 100 • C, then cooled down to −30 • C, and raised again to 200 • C. The glass transition temperature (T g ) was determined from the second heating curve.

Cell Viability
Cell viability assay was performed to evaluate the effect of the degradation products of MA-Starch hydrogels on cell viability. BJ human fibroblast cells (ATCC ® CRL-2522™) were cultured on three different 96-well plates at a cell density of 2 × 10 4 cells/well for 24 h to reach confluence using DMEM-Dulbecco's Modified Eagle Medium enriched with 10% fetal bovine serum and 1% penicillin/streptomycin (Carlo Erba, Milan, Italy). Simultaneously, three hydrogels prepared with three different methacrylated starch concentrations (5, 10, and 15 wt%) were soaked for 24 h in 1 mL of DMEM. In each case, 0.1 g of hydrogel was soaked in 1 mL of the medium. After 24 h, the media were collected and filtered through 0.22 µm filters to guarantee sterility. Then, the culture medium was removed from each well of confluent cells and substituted with the supernatant collected from the hydrogels. Controls (CTRL) were obtained using normal medium. After 24 h of incubation, the supernatant was carefully removed and the cell viability assay was performed using non-fluorescent resazurin, which is converted into a highly red fluorescent dye (resorufin) by cell metabolism. Briefly, a volume of 100 µL of 0.1 mg/mL resazurin solution, obtained by diluting a resazurin working solution (1 mg/mL in phosphate buffered saline-PBS, Sigma Aldrich, Milan, Italy) into DMEM, was added in each well and the cultures were incubated for 1 h at 37 • C. Then, the fluorescent signal was monitored at 530 nm excitation wavelength and 590 nm emission wavelength using a plate reader (Victor X3, Perkin Elmer, Milan, Italy). Cell viability was calculated as a percentage value compared to CTRL. Six samples for each condition were used and the experiments were performed three times. GraphPad Prism ® software was used for one or two analysis of variance (ANOVA). Values * p < 0.05, ** p < 0.01, *** p < 0.001 were considered statistically significant.

Results and Discussion
Starch as a raw material for light processable hydrogels was evaluated. To render light processability, starch was first methacrylated (Scheme 1). The methacrylated starch (MA-Starch) was then evaluated for photocuring reactivity, and DLP 3D printability, followed by establishment of the properties of the light processed hydrogels including mechanical and rheological testing and cell viability.

Cell Viability
Cell viability assay was performed to evaluate the effect of the degradation products of MA-Starch hydrogels on cell viability. BJ human fibroblast cells (ATCC ® CRL-2522™) were cultured on three different 96-well plates at a cell density of 2 × 10 4 cells/well for 24 h to reach confluence using DMEM-Dulbecco's Modified Eagle Medium enriched with 10% fetal bovine serum and 1% penicillin/streptomycin (Carlo Erba, Milan, Italy). Simultaneously, three hydrogels prepared with three different methacrylated starch concentrations (5, 10, and 15 wt%) were soaked for 24 h in 1 mL of DMEM. In each case, 0.1 g of hydrogel was soaked in 1 mL of the medium. After 24 h, the media were collected and filtered through 0.22 μm filters to guarantee sterility. Then, the culture medium was removed from each well of confluent cells and substituted with the supernatant collected from the hydrogels. Controls (CTRL) were obtained using normal medium. After 24 h of incubation, the supernatant was carefully removed and the cell viability assay was performed using non-fluorescent resazurin, which is converted into a highly red fluorescent dye (resorufin) by cell metabolism. Briefly, a volume of 100 μL of 0.1 mg/mL resazurin solution, obtained by diluting a resazurin working solution (1 mg/mL in phosphate buffered saline-PBS, Sigma Aldrich, Milan, Italy) into DMEM, was added in each well and the cultures were incubated for 1 h at 37 °C. Then, the fluorescent signal was monitored at 530 nm excitation wavelength and 590 nm emission wavelength using a plate reader (Victor X3, Perkin Elmer, Milan, Italy). Cell viability was calculated as a percentage value compared to CTRL. Six samples for each condition were used and the experiments were performed three times. GraphPad Prism ® software was used for one or two analysis of variance (ANOVA). Values * p < 0.05, ** p < 0.01, *** p < 0.001 were considered statistically significant.

Results and Discussion
Starch as a raw material for light processable hydrogels was evaluated. To render light processability, starch was first methacrylated (Scheme 1). The methacrylated starch (MA-Starch) was then evaluated for photocuring reactivity, and DLP 3D printability, followed by establishment of the properties of the light processed hydrogels including mechanical and rheological testing and cell viability.

Confirmation of Starch Methacrylation
The methacrylation reaction was evaluated and confirmed by 1 H-NMR, 13 C-NMR, and FTIR spectroscopy. Firstly, the 1 H-NMR spectrum ( Figure 1) recorded for Starch-reference, treated with the same process but without methacrylation, shows typical starch peaks: δ = 3.59 and δ = 3.66 represent the CH protons at 2, 3, 4, and 5 positions; the peak at δ = 5.11 represents the C1 anhydroglucose carbon; and δ = 4.57, δ = 5.40, and δ = 5.49 represent the OH protons at 2, 3, and 6 positions, respectively. The huge peak at δ = 3.35 is due to the inevitably absorbed water in the starch structure [41]. The 1 H NMR spectrum recorded for MA-Starch shows the presence of characteristic peaks that correspond to the protons of methacrylic double bond = CH2 (δ 5.66−6.07 ppm) and methyl Scheme 1. Schematic illustration of starch methacrylation and photo-crosslinking reaction. The structure presented for MA-Starch presents the most likely structure where some of the primary alcohol groups were methacrylated.
The huge peak at δ = 3.35 is due to the inevitably absorbed water in the starch structure [41]. The 1 H NMR spectrum recorded for MA-Starch shows the presence of characteristic peaks that correspond to the protons of methacrylic double bond = CH 2 (δ 5.66−6.07 ppm) and methyl groups -CH 3 (δ 1.9 ppm). These spectral changes are consistent with the structural modification of starch. However, as shown in Figure 1a, the peaks of the MA-Starch spectrum are broader compared to Starch-reference. This can be explained by the partial methacrylation of the hydroxyl-groups in the starch chains, leading to different structures due to mixture of free and methacrylated alcohol groups in the repeating units.
Polymers 2020, 12, x FOR PEER REVIEW 6 of 14 groups -CH3 (δ 1.9 ppm). These spectral changes are consistent with the structural modification of starch. However, as shown in Figure 1a, the peaks of the MA-Starch spectrum are broader compared to Starch-reference. This can be explained by the partial methacrylation of the hydroxyl-groups in the starch chains, leading to different structures due to mixture of free and methacrylated alcohol groups in the repeating units. The degree of substitution (DOS) for the hydroxyl groups was obtained as the ratio of integrals of the 1 H-NMR peaks using Equation (7): where I1.9 and I5.11 are the integral intensity of CH3 protons (named b in the spectra) and the proton of the anhydrous glucose unit (AGU) unit in α-carbon (1H position). The formula is divided by three since there are three -OH groups in each AGU. The degree of substitution obtained was 0.08 (I1.9 ≈ 0.75, I5.11 ≈ 1), which can be translated into one methacrylated alcohol-group in every fourth glucose ring ( Figure 1). This degree of substitution is very similar to what was reported for starch methacrylated with glycidyl methacrylate [29]. The 13 C-NMR spectra of Starch-reference and MA-Starch are shown in Figure 1b. All the carbon atoms of starch repeating unit can be detected and assigned. The signals at δ = 100.54 ppm, δ = 79.37 ppm, δ = 73.74 ppm, δ = 72.55 ppm, δ = 72.12 ppm, and δ = 61.09 ppm in the Starch-reference spectrum can be attributed to C1, C4, C3, C2, C5, and C6 of the glucose unit, respectively. In the 13 C-NMR spectrum of MA-Starch, there are additional signals from the methacryl-groups. These are seen at δ = 18.45 ppm (methyl carbon), δ = 136.61 ppm and δ = 127.75 ppm (double bond carbons) and δ = 170.38 ppm (C = O carbon).
The methacrylation reaction was further confirmed by FTIR. Figure 2 presents the FTIR spectra of Starch-reference and MA-starch. The FTIR spectrum of Starch-reference displays its characteristic functional groups. The broad peak at 3320 cm −1 is attribute to the -OH vibration. The peak at 2925 cm −1 is the -CH2-stretching vibration. The band at 1638 cm −1 is the deformation vibration of the hydroxyl group of H2O absorbed in the amorphous regions of starch. The peak at 1148 cm −1 is the C-O-C symmetric stretching of cyclic ether group, while the peak at 1076 cm −1 is the C-H bending vibration. The peaks at 994 and 927 cm −1 are assigned to the stretching and skeletal vibration of the C-O-C α-1-4 glycosidic linkage, respectively. Finally, the peaks at 855 and 762 cm −1 can be assigned to the C(1)-H, CH2 deformation and C-C stretching vibration, respectively [42,43]. In the FTIR spectrum of MA-Starch, it is possible to observe four new peaks that strongly support a successful methacrylation of some of the starch hydroxyl groups. The peaks at 1709 and 1300 cm −1 can be The degree of substitution (DOS) for the hydroxyl groups was obtained as the ratio of integrals of the 1 H-NMR peaks using Equation (7): where I 1.9 and I 5.11 are the integral intensity of CH 3 protons (named b in the spectra) and the proton of the anhydrous glucose unit (AGU) unit in α-carbon (1H position). The formula is divided by three since there are three -OH groups in each AGU. The degree of substitution obtained was 0.08 (I 1.9 ≈ 0.75, I 5.11 ≈ 1), which can be translated into one methacrylated alcohol-group in every fourth glucose ring ( Figure 1). This degree of substitution is very similar to what was reported for starch methacrylated with glycidyl methacrylate [29]. The methacrylation reaction was further confirmed by FTIR. Figure 2 presents the FTIR spectra of Starch-reference and MA-starch. The FTIR spectrum of Starch-reference displays its characteristic functional groups. The broad peak at 3320 cm −1 is attribute to the -OH vibration. The peak at 2925 cm −1 is the -CH 2 -stretching vibration. The band at 1638 cm −1 is the deformation vibration of the hydroxyl group of H 2 O absorbed in the amorphous regions of starch. The peak at 1148 cm −1 is the C-O-C symmetric stretching of cyclic ether group, while the peak at 1076 cm −1 is the C-H bending vibration. The peaks at 994 and 927 cm −1 are assigned to the stretching and skeletal vibration of the C-O-C α-1-4 glycosidic linkage, respectively. Finally, the peaks at 855 and 762 cm −1 can be assigned to the C(1)-H, CH 2 deformation and C-C stretching vibration, respectively [42,43]. In the FTIR spectrum of MA-Starch, it is possible to observe four new peaks that strongly support a successful methacrylation of some of the starch hydroxyl groups. The peaks at 1709 and 1300 cm −1 can be attributed to the C=O and C-O stretching vibrations, respectively, while the two peaks at 1640 and 815 cm −1 are assigned for the C=CH 2 out of plane bending vibrations [28,44,45].

Photo-Rheology of the Photo-Crosslinking Reaction for Forming Starch Hydrogels
The methacrylated-starch was dispersed in water with LAP photo-initiator and the rheological properties of the starch/water/LAP formulations were recorded during the photo-crosslinking reaction to evaluate the reaction kinetics. Initially, an amplitude sweep experiment was performed on the liquid formulations to determine the linear viscoelastic region (LVR). The reaction is considered as completed when a G' plateau is reached. Figure 3 shows that the formulations with 10 and 15 wt% starch started reacting immediately and reached a G' plateau after 90 s, while the 5 wt% starch formulation showed a delay in the curing process and reached a G' plateau after 120 s. This is probably explained by the low starch concentration in water, which in combination with the overall low DOS decreased the velocity of the radical reaction.

Characterization of the Photo-Crosslinked Starch Hydrogels
The photocured hydrogels were fully characterized for their mechanical, rheological, and

Photo-Rheology of the Photo-Crosslinking Reaction for Forming Starch Hydrogels
The methacrylated-starch was dispersed in water with LAP photo-initiator and the rheological properties of the starch/water/LAP formulations were recorded during the photo-crosslinking reaction to evaluate the reaction kinetics. Initially, an amplitude sweep experiment was performed on the liquid formulations to determine the linear viscoelastic region (LVR). The reaction is considered as completed when a G' plateau is reached. Figure 3 shows that the formulations with 10 and 15 wt% starch started reacting immediately and reached a G' plateau after 90 s, while the 5 wt% starch formulation showed a delay in the curing process and reached a G' plateau after 120 s. This is probably explained by the low starch concentration in water, which in combination with the overall low DOS decreased the velocity of the radical reaction.

Photo-Rheology of the Photo-Crosslinking Reaction for Forming Starch Hydrogels
The methacrylated-starch was dispersed in water with LAP photo-initiator and the rheological properties of the starch/water/LAP formulations were recorded during the photo-crosslinking reaction to evaluate the reaction kinetics. Initially, an amplitude sweep experiment was performed on the liquid formulations to determine the linear viscoelastic region (LVR). The reaction is considered as completed when a G' plateau is reached. Figure 3 shows that the formulations with 10 and 15 wt% starch started reacting immediately and reached a G' plateau after 90 s, while the 5 wt% starch formulation showed a delay in the curing process and reached a G' plateau after 120 s. This is probably explained by the low starch concentration in water, which in combination with the overall low DOS decreased the velocity of the radical reaction.

Characterization of the Photo-Crosslinked Starch Hydrogels
The photocured hydrogels were fully characterized for their mechanical, rheological, and swelling properties. The compressive stress-strain curves of MA-Starch10 and MA-Starch15

Characterization of the Photo-Crosslinked Starch Hydrogels
The photocured hydrogels were fully characterized for their mechanical, rheological, and swelling properties. The compressive stress-strain curves of MA-Starch10 and MA-Starch15 photocured cast discs are shown in Figure 4a. The compressive Young's modulus values (E c ) are reported in Table 1. The simple increase of the MA-Starch concentration in water from 10 to 15 wt% increased the compressive stiffness of the hydrogels from 13 to 20 kPa. These values are similar to the ones reported for photo-crosslinked hemicellulose and silk fibroin networks applied for tissue engineering [46,47]. This range also covers the stiffness of different native body tissues, which is promising for utilization of the hydrogels in tissue engineering applications, where the cell microenvironment can be mimicked by matching the stiffness of different body tissues [48].
For rheological evaluation, the polymer solutions (300 µL) were crosslinked directly on the parallel plate for instant measurements. To determine the linear viscoelastic region of the hydrogels, an amplitude sweep experiment was initially performed. After that, a frequency sweep experiment was performed on a freshly prepared sample with a strain amplitude of 1%. The results of the frequency sweep experiment, including G', G", and |η*|, are shown in Figure 4b.
The complex viscosities of the three hydrogels have approximately the same slope of −0.9, which means that all the networks exhibit a pseudo-plastic behavior [38]. All the hydrogels also show linear and stable behavior for G' and G" values in the ω range tested. Their progressions are parallel, but the G" values lie in one order of magnitude lower region. This means that the elastic component of the material is dominant over the viscous behavior, which is common for different gel structures [49]. The values of G' and G" are also comparable with the ones obtained in a previous work for carboxymethylated starch hydrogels [50,51]. The network parameters were evaluated on the G' p at ω = 10 rad/s and collected in Table 1. Increasing the MA-Starch concentration decreased the distance between two entanglements points and increased the crosslink density.
reported in Table 1. The simple increase of the MA-Starch concentration in water from 10 to 15 wt% increased the compressive stiffness of the hydrogels from 13 to 20 kPa. These values are similar to the ones reported for photo-crosslinked hemicellulose and silk fibroin networks applied for tissue engineering [46,47]. This range also covers the stiffness of different native body tissues, which is promising for utilization of the hydrogels in tissue engineering applications, where the cell microenvironment can be mimicked by matching the stiffness of different body tissues [48].
For rheological evaluation, the polymer solutions (300 μL) were crosslinked directly on the parallel plate for instant measurements. To determine the linear viscoelastic region of the hydrogels, an amplitude sweep experiment was initially performed. After that, a frequency sweep experiment was performed on a freshly prepared sample with a strain amplitude of 1%. The results of the frequency sweep experiment, including G', G'', and |η*|, are shown in Figure 4b.
The complex viscosities of the three hydrogels have approximately the same slope of -0.9, which means that all the networks exhibit a pseudo-plastic behavior [38]. All the hydrogels also show linear and stable behavior for G' and G'' values in the ω range tested. Their progressions are parallel, but the G" values lie in one order of magnitude lower region. This means that the elastic component of the material is dominant over the viscous behavior, which is common for different gel structures [49]. The values of G' and G'' are also comparable with the ones obtained in a previous work for carboxymethylated starch hydrogels [50,51]. The network parameters were evaluated on the G'p at ω = 10 rad/s and collected in Table 1. Increasing the MA-Starch concentration decreased the distance between two entanglements points and increased the crosslink density. Figure 5 reports the swelling curves for the three hydrogels in water, while Table 1 collects the swelling at equilibrium Seq, equilibrium water content EWC, and gel percentage G% values for the three hydrogel formulations. As seen in the graph, the swelling capability of MA-Starch5 and MA-Starch10 were very similar, while the swelling capability of MA-Starch15 was lower. However, the values obtained from MA-Starch5 does not corresponds to the maximum swelling capability but only to the maximum swelling before breaking. The values of MA-Starch10 and MA-Starch15 are in good agreement with the values obtained from the rheological measurement on the crosslink density, as higher crosslink density leads to lower swelling degree. The gel content of the dried samples was measured as described in the Experimental Section 2.10 and it reached 100% for all three formulations (Table 1). Starch-reference, MA-starch and the three MA-starch hydrogels (5, 10, and 15%wt of MAstarch) were analyzed via differential scanning calorimetry analysis to investigate the glass transition temperatures (Tg) (see Table 1). There was no significant difference between the Tg of the methacrylated starch and the Tg after photo-crosslinking of the materials.   Figure 5 reports the swelling curves for the three hydrogels in water, while Table 1 collects the swelling at equilibrium Seq, equilibrium water content EWC, and gel percentage G% values for the three hydrogel formulations. As seen in the graph, the swelling capability of MA-Starch5 and MA-Starch10 were very similar, while the swelling capability of MA-Starch15 was lower. However, the values obtained from MA-Starch5 does not corresponds to the maximum swelling capability but only to the maximum swelling before breaking. The values of MA-Starch10 and MA-Starch15 are in good agreement with the values obtained from the rheological measurement on the crosslink density, as higher crosslink density leads to lower swelling degree. The gel content of the dried samples was measured as described in the Experimental Section 2.10 and it reached 100% for all three formulations (Table 1). Starch-reference, MA-starch and the three MA-starch hydrogels (5, 10, Polymers 2020, 12, 1359 9 of 14 and 15%wt of MA-starch) were analyzed via differential scanning calorimetry analysis to investigate the glass transition temperatures (T g ) (see Table 1). There was no significant difference between the T g of the methacrylated starch and the T g after photo-crosslinking of the materials.  105 ± 2 115 ± 1 106 ± 2 106 ± 9 101 ± 8 * Maximum value obtained before the sample failure. Ec, compressive Young's module values; e, crosslink density; ξ, distance between two entanglements points; Seq, swelling at equilibrium; EWC, equilibrium water content; G%, gel percentage; Tg, glass transition temperature.

Cell Viability of the Photo-Crosslinked Starch Hydrogels
Human fibroblast cells were cultured on the supernatants from the three hydrogel formulations. Starch has previously been utilized in many biomedical applications for fabrication of scaffolds, hydrogels, and drug delivery systems [52,53] showing a good biocompatibility. The starch content in the present study had no negative effects on the cell viability ( Figure 6). The cell viability of all three formulations was also very similar to the control, indicating no negative effects from the methacrylation, photo-crosslinking reaction and LAP photo-initiator. These first results indicate that the prepared starch-based photo-crosslinked hydrogels could have potential in biomedical applications, such as tissue engineering and also as cell carriers through dispersion of cells into the photocurable formulations.  T g ( • C) 105 ± 2 115 ± 1 106 ± 2 106 ± 9 101 ± 8 * Maximum value obtained before the sample failure. Ec, compressive Young's module values; ν e , crosslink density; ξ, distance between two entanglements points; S eq , swelling at equilibrium; EWC, equilibrium water content; G%, gel percentage; T g , glass transition temperature.

Cell Viability of the Photo-Crosslinked Starch Hydrogels
Human fibroblast cells were cultured on the supernatants from the three hydrogel formulations. Starch has previously been utilized in many biomedical applications for fabrication of scaffolds, hydrogels, and drug delivery systems [52,53] showing a good biocompatibility. The starch content in the present study had no negative effects on the cell viability ( Figure 6). The cell viability of all three formulations was also very similar to the control, indicating no negative effects from the methacrylation, photo-crosslinking reaction and LAP photo-initiator. These first results indicate that the prepared starch-based photo-crosslinked hydrogels could have potential in biomedical applications, such as tissue engineering and also as cell carriers through dispersion of cells into the photocurable formulations. Polymers 2020, 12, x FOR PEER REVIEW 10 of 14 Figure 6. Cell viability of BJ cells culture using the supernatants collected from the starch-based hydrogels compared to the control conditions (black bars).

The 3D Printability of the Methacrylated Starch Hydrogels by DLP
The fast photo-curing reactivity of the formulations and the obtained mechanical properties indicated that the MA-Starch could be suitable photo-curable resin for the vat 3D printing technology [54]. The 3D printability of the methacrylated starch by DLP was therefore evaluated. The formulation with 10 wt% MA-Starch was selected since the hydrogel prepared with this concentration showed a compression Young's modulus (Ec = 13 kPa) close to the elasticity of muscle tissue (12 kPa) [48]. Simple geometries were designed by CAD and printed in order to evaluate the light processability of the bio-based formulation by DLP printing and to investigate the achievable resolution. The common photocurable 3D printable formulations usually exploit a dye or colorant to avoid the light diffusion in the vat to achieve a better resolution [55]. In the present case, different tests were performed. First, neat formulation without the addition of any dye was tested giving promising results with good resolution on the x-y plane. The smallest feature that could be obtained had dimensions of 1.5 mm × 0.5 mm × 2 mm, presenting sharp edges and limited over-polymerization in the vat (Figure 7a). Subsequently, the addition of a methyl red dye (0.2 wt% based on the weight of the dry starch) was evaluated (Figure 7b,c). The addition of the dye implied the need of changing the printing parameters. Several trials were performed to achieve lower over-polymerization along the profile of the printed shapes. However, since the cytotoxicity of methyl red was not assessed, other alternatives should be searched for. The cell viability test showed that the pH of the starch solution in the cell medium was neutral, which led to a solution with pink color. This suggested the possibility to substitute the aqueous dye solution with DMEM solution in the 3D printable MA-starch formulation. A honeycomb structure was then designed and successfully printed from the DMEM containing formulation, showing that millimeter-thick wall structures could be built from this formulation (Figure 7d). The use of such a light-dye led to a certain degree of over polymerization in the x-y plane, demonstrating the need to correctly balance the different ingredients and printing parameters for further printing studies. All the printed objects required a post-curing process, which was regulated with the help of the photo-rheology test to achieve complete crosslinking reaction. All the 3D printed hydrogels were mechanically stable for easy handling.

The 3D Printability of the Methacrylated Starch Hydrogels by DLP
The fast photo-curing reactivity of the formulations and the obtained mechanical properties indicated that the MA-Starch could be suitable photo-curable resin for the vat 3D printing technology [54]. The 3D printability of the methacrylated starch by DLP was therefore evaluated. The formulation with 10 wt% MA-Starch was selected since the hydrogel prepared with this concentration showed a compression Young's modulus (E c = 13 kPa) close to the elasticity of muscle tissue (12 kPa) [48]. Simple geometries were designed by CAD and printed in order to evaluate the light processability of the bio-based formulation by DLP printing and to investigate the achievable resolution. The common photocurable 3D printable formulations usually exploit a dye or colorant to avoid the light diffusion in the vat to achieve a better resolution [55]. In the present case, different tests were performed. First, neat formulation without the addition of any dye was tested giving promising results with good resolution on the x-y plane. The smallest feature that could be obtained had dimensions of 1.5 mm × 0.5 mm × 2 mm, presenting sharp edges and limited over-polymerization in the vat (Figure 7a). Subsequently, the addition of a methyl red dye (0.2 wt% based on the weight of the dry starch) was evaluated (Figure 7b,c). The addition of the dye implied the need of changing the printing parameters. Several trials were performed to achieve lower over-polymerization along the profile of the printed shapes. However, since the cytotoxicity of methyl red was not assessed, other alternatives should be searched for. The cell viability test showed that the pH of the starch solution in the cell medium was neutral, which led to a solution with pink color. This suggested the possibility to substitute the aqueous dye solution with DMEM solution in the 3D printable MA-starch formulation. A honeycomb structure was then designed and successfully printed from the DMEM containing formulation, showing that millimeter-thick wall structures could be built from this formulation (Figure 7d). The use of such a light-dye led to a certain degree of over polymerization in the x-y plane, demonstrating the need to correctly balance the different ingredients and printing parameters for further printing studies. All the printed objects required a post-curing process, which was regulated with the help of the photo-rheology test to achieve complete crosslinking reaction. All the 3D printed hydrogels were mechanically stable for easy handling.

Conclusions
Methacrylated maize starch was successfully light processed to hydrogels either by photocuring with UV-lamp or by DLP 3D-printing directly from aqueous solution. First, starch was methacrylated and the reaction was confirmed by 1 H-NMR, 13 C-NMR, and FTIR spectroscopy. The methacrylated starch was dispersed in water at different concentration in the presence of LAP photoinitiator. Photo-rheology confirmed rapid in-situ gel formation by following the evolution of elastic storage modulus G' with time. The measured G' plateau values were 1.5, 39.9, and 67.6 kPa for the hydrogels obtained in the presence of 5, 10, and 15 wt% of MA-starch. These values are comparable with other methacrylated biobased hydrogel systems reported in the literature. Rheological characterization of the crosslinked materials was performed and the network parameters were evaluated on the G'p at ω = 10 rad/s. Increasing the MA-Starch concentration increased the crosslink density, which led to decreased swelling capability. The supernatants of the methacrylated starchbased photo-crosslinked hydrogels did not exhibit cytotoxicity against human BJ fibroblast cells in any of the three tested hydrogel concentrations. Finally, the 3D printability of 10 wt% MA-Starch was successfully demonstrated. These findings are promising for future applications of light processed starch in tissue engineering and as cell carriers.

Conclusions
Methacrylated maize starch was successfully light processed to hydrogels either by photo-curing with UV-lamp or by DLP 3D-printing directly from aqueous solution. First, starch was methacrylated and the reaction was confirmed by 1 H-NMR, 13 C-NMR, and FTIR spectroscopy. The methacrylated starch was dispersed in water at different concentration in the presence of LAP photo-initiator. Photo-rheology confirmed rapid in-situ gel formation by following the evolution of elastic storage modulus G' with time. The measured G' plateau values were 1.5, 39.9, and 67.6 kPa for the hydrogels obtained in the presence of 5, 10, and 15 wt% of MA-starch. These values are comparable with other methacrylated biobased hydrogel systems reported in the literature. Rheological characterization of the crosslinked materials was performed and the network parameters were evaluated on the G' p at ω = 10 rad/s. Increasing the MA-Starch concentration increased the crosslink density, which led to decreased swelling capability. The supernatants of the methacrylated starch-based photo-crosslinked hydrogels did not exhibit cytotoxicity against human BJ fibroblast cells in any of the three tested hydrogel concentrations. Finally, the 3D printability of 10 wt% MA-Starch was successfully demonstrated. These findings are promising for future applications of light processed starch in tissue engineering and as cell carriers.