Supramolecular polycationic hydrogels with high swelling capacity prepared by partial methacrylation of polyethyleneimine

Abstract

Methacrylation of branched polyethyleneimine (PEI) was performed to generate modified polycationic macromolecules that self-assemble in water into highly swollen supramolecular hydrogels with tunable hydrophilicity and microstructure. The properties of the supramolecular hydrogels, in terms of swelling and porosity, were controlled during synthesis by the extent of methacrylation of the starting PEI macromolecules. The methacrylation reaction was conducted under several conditions, and the methacrylated PEI (PEI-MA) molecules were investigated by FTIR and NMR to relate the reaction parameters to the amount of methacrylate moieties on the modified polymer. The hydrogel morphology and its dynamic nature arising from the non-covalent interactions among the PEI-MA macromolecules were characterized by small-angle X-ray scattering (SAXS) and fluorescence microscopy at different temperatures, which enabled visualization of the large interconnected microcavities of the supramolecular gel network.

---

1 Introduction

Supramolecular polymer hydrogels constitute a fascinating class of soft materials deriving from the three-dimensional assembly of macromolecules via dynamic and specific non-covalent interactions, including those mediated by a guest molecule, in an aqueous environment.^1,2 These materials are the subject of extensive research due to the versatile and often unexpected physical properties that they display under the effect of external stimuli, most notably changes of pH, temperature, ionic strength and electro-magnetic fields, which can drive structural modifications of the supramolecular assembly at several length and time scales.^3–9 Indeed, the stimuli-responsive and self-healing properties¹⁰ of supramolecular hydrogels make them ideal systems for a wide range of relevant applications in many fields, for example in sensing and actuating,^11,12 or in energy transducing systems.^13,14 In the biomedical sector, supramolecular hydrogels of naturally occurring or biomimetic polymers are routinely used as media for controlled drug and gene delivery,^15–18 including injectable drug delivery systems,¹⁹ and for encapsulating cells in tissue engineering applications.^20–22

More recently, the development of hydrogels confined to smaller dimensions, such as microgels and nanogels, has been of increasing interest for injectable polymer-based delivery systems, due to their rheological properties and large surface area available for multivalent bioconjugation. Such multifunctional delivery systems have been designed on the microscale for cell-based therapies with enhanced therapeutic effects,²³ or on the nanoscale for glucose-regulated insulin delivery.²⁴

The chemical design of amphiphilic polymers that self-assemble to form supramolecular networks with multi-responsive behavior is one of the most investigated fields in polymer science. So far, supramolecular chemistry has created many functional architectures through the elegant control of molecular interactions. Suitable modifications of macromolecules can stabilize reversible and directional non-covalent interactions to give supramolecular hydrogel structures with well-defined properties.^25–27 One of the major advantages of supramolecular materials is that appropriate design allows efficient control over the assembled structure and its function.^28,29 In contrast with covalently cross-linked hydrogels, supramolecular polymers generate smart and adaptive materials with the possibility of multiple responses to external stimuli. In some cases, hydrogel responsiveness to four distinct external stimuli has been implemented for a fine tuning of its properties through a combination of the four parameters.³⁰

Cationic polymers such as polyethylenimine (PEI) are widely used as non-viral vectors for gene delivery because they form stable complexes with nucleic acids in aqueous solutions due to electrostatic interactions.³¹ In this context, most of the reported applications of PEI describe its use as a complexing agent for DNA or small-interfering RNA (siRNA) in solution,^32,33 or on the surface of sensor chips for substrate–DNA interaction analysis.³⁴ Indeed, one of the most relevant feature of PEI macromolecules is the significantly high concentration of positively charged nitrogen atoms that can effectively bind with negatively charged groups of linear and plasmid DNA.³⁵ In addition, the high concentration of amine functional groups in PEI makes it an ideal polymeric ligand for complexing heavy metal ions for purification purposes,³⁶ as well as anionic ligands, most notably polysaccharides that also contribute to decrease PEI in vivo toxicity in clinical applications.³⁷

Chemical modifications of PEI to obtain materials with three-dimensional physical structure or specific functionality have been investigated by several groups. Examples include PEI cross-linking by disulfide bonds, through Michael addition between amine groups and N,N′-bis(acryloyl)cystamine, to generate degradable hydrogels for controlled drug delivery.³⁸ Other PEI modifications reported in the literature are based on partial acetylation of its amino groups for modulation of the polymer cationic charge,^39,40 or PEGylation reactions to form biodegradable copolymers with reduced toxicity and decreased non-specific interactions with serum proteins resulting in a prolonged blood circulation time.^41–43 Recently, we have discovered a methacrylated version of branched PEI that self-assembles in water forming a hydrogel that is responsive to two-photon light excitation in the near infrared wavelength range.⁴⁴ In particular, using a two-photon confocal laser scanning microscope with NIR light, we demonstrated that bioactive molecules containing hydroxyl or carboxyl moieties, such as the RGD peptide, can be immobilized at the PEI-MA hydrogel interface, thus generating materials with microscale biofunctional patterns without the need of additional photoinitiators. In addition to their photoactive properties, the PEI-MA hydrogels can be particularly useful for biomolecule immobilization in a three-dimensional structure, and represent a versatile platform for the binding and release of plasmid DNA in a controlled manner.⁴⁵

In this work, supramolecular hydrogels with different microstructure have been synthesized starting from partial methacrylation of branched PEI. A full investigation of the chemical and physical properties of the PEI-MA hydrogels and how these properties are related to the methacrylation conditions are reported. The covalent modification that we describe confers additional hydrogen bonding ability to the PEI macromolecules via the methacrylate moieties, as compared to the unmodified polymer, which drives the formation of a hydrogel supramolecular network in aqueous solutions. To our knowledge, the PEI-MA hydrogels are the first report of supramolecular polycationic hydrogels derived from PEI. This work addresses one of the crucial needs in hydrogel fields, that is to rationally design a hydrogel with tailored architecture and properties starting from its synthesis conditions.

2 Experimental

2.1 Materials

Branched polyethylenimine (average M_w 25 000 g mol⁻¹ by LS, average M_n 10 000 g mol⁻¹ by GPC), methacrylic anhydride (MAA, 94%, density 1.035 g mL⁻¹ at 25 °C), triethylamine (TEA, ≥99%, density 0.726 g mL⁻¹ at 25 °C), and ethanol were purchased from Sigma-Aldrich (Germany). Dichloromethane (DCM, extra dry stabilized with amylene) and ultrapure water (Super Purity Solvent grade) were from Romil Pure Chemistry (UK). For NMR experiments, deuterated water (100 atom% D) was purchased from ARMAR Chemicals (Switzerland). For fluorescence microscopy, Alexa Fluor 350 carboxylic acid succinimidyl ester was from Molecular Probes/Invitrogen (USA). Reagents and solvents were used without further purification unless otherwise specified.

2.2 Synthesis of methacrylated branched PEI

Branched PEI and methacrylic anhydride were reacted at five different molar ratios to synthesize PEI-MA molecules with different extent of methacrylation. Five round-bottom flasks containing branched PEI (1 g, 4.00 × 10⁻⁵ mol) were first dried under vacuum and then 3 mL of dichloromethane was added to each flask under argon flow. After 30 min, triethylamine (10 μL, 7.17 × 10⁻⁵ mol) was added to each of the five samples, and the mixtures were stirred for 5 min in order to activate PEI primary and secondary amine groups in basic conditions and favor the formation of amide groups. In the end, methacrylic anhydride (73 μL, 4.89 × 10⁻⁴ mol for PEI-MA-3; 120 μL, 8.06 × 10⁻⁴ mol for PEI-MA-4.5; 158 μL, 10.61 × 10⁻⁴ mol for PEI-MA-6; 192 μL, 12.89 × 10⁻⁴ mol for PEI-MA-7.5; 218 μL, 14.64 × 10⁻⁴ mol for PEI-MA-9) was added to each of the five flasks under argon flow and the reaction was allowed to proceed under stirring for 18 h at room temperature. After synthesis, the methacrylated polymer samples were dried under vacuum for 12 h to remove excess dichloromethane. Next, they were purified by dialysis against ultrapure water for six days to remove unreacted methacrylic anhydride and methacrylic acid by-product. Excess triethylamine was removed by dialysis against ethanol. Spectra/Por 2 membranes (MWCO 12 000–14 000 Da, Spectrum Laboratories, USA) were used for the dialysis. After purification, the PEI-MA hydrogels were dried under vacuum at room temperature for 48 h in a glass desiccator connected to a high vacuum pump (RC 6 chemistry-HYBRID pump, Vacuubrand, Germany), and stored at −20 °C until further use.

2.3 Fourier transform infrared spectroscopy

FTIR analysis of dried PEI and methacrylated PEI was performed on a Nicolet 6700 spectrometer (Thermo Fisher Scientific, USA) equipped with a single-reflection attenuated total reflectance accessory (Smart iTR) under ambient conditions. ATR correction (germanium crystal) and baseline correction were applied to the data before analysis. Spectra were collected in the range of wavenumber from 4000 to 600 cm⁻¹ with a resolution of 4 cm⁻¹ and averaging over 128 scans.

2.4 Nuclear magnetic resonance spectroscopy

¹H NMR and ¹³C NMR spectra were recorded using a Varian Unity INOVA 700 MHz spectrometer. Experiments were performed at 25 °C using D₂O as a solvent. NMR spectra were processed using MestReNova Software (Mestrelab Research, Spain). PEI and methacrylated PEI samples were first dried in vacuum for 24 h, then dissolved in D₂O and left to equilibrate for three days before spectra acquisition. The extent of methacrylation of branched PEI was determined by peak integration using the following formulas:

where M′ is the integration of the peaks at δ 1.80–1.73 ppm corresponding to methacrylated primary amines [–NH–COC(CH₂)CH₃], M′′ is the integration of the peaks at δ 2.09–2.08 ppm corresponding to methacrylated secondary amines [[double bond splayed left]N–COC(CH₂)CH₃], and E_bb is the integration of the peaks at δ 3.5–2.5 ppm corresponding to the protons of the ethylene backbone. N′ and N′′ are the mole fractions of PEI primary and secondary amines, 0.307 and 0.395 respectively, based on the analysis of the commercial PEI material by ¹³C NMR available in the literature.⁴⁶ The degree of branching (DB) of PEI was calculated according to:
where D, T and L are the fractions of dendritic units (tertiary amine), terminal units (primary amine) and linearly incorporated units (secondary amine) in branched PEI.⁴⁷

2.5 Hydrogel swelling behavior

For analysing the hydrogel swelling in water, methacrylated PEI samples were dried in vacuum and the dry weights (W_dry) were measured in glass vials. The hydrogels samples were prepared in quintuplicate. Next, the samples were incubated in ultrapure water at ambient temperature for five days to reach their equilibrium swelling state. Excess water was removed gently from the vials using a micropipette and the swollen hydrogel sample weights (W_swollen) were measured. The hydrogel water content and the weight swelling ratio were calculated as:

2.6 Small-angle X-ray scattering (SAXS)

SAXS experiments were performed on a NanoSTAR instrument (Bruker AXS, Germany) using Cu Kα radiation at 45 kV and 650 μA from an Incoatec microfocus source, three-pinhole collimation, and equipped with a VANTEC-2000 position sensitive detector (2048 × 2048 pixels, Bruker AXS). Samples were measured in the high resolution configuration (diameter of first/second/third pinhole = 500 μm/150 μm/500 μm) and at a sample-to-detector distance of 106.8 cm, resulting in a q range of 0.005–0.21 Å⁻¹. The scattering vector (q) was defined as q = 4πλ⁻¹ sin(θ/2), where θ is the scattering angle and λ is the wavelength of Cu Kα radiation (1.542 Å). SAXS data were collected with a measuring time of 7200 s at room temperature in Mylar cells. Before measurements, the hydrogel samples were weighted (20 mg) and swollen in 15 mL of ultrapure water for six days at room temperature to reach the final swelling state. Analysis of SAXS data was performed using the Irena tool suite within the commercial IGOR Pro 6 Software (WaveMetrics, USA).⁴⁸ Data were corrected for background and empty cell scattering before analysis. Sample transmission was measured using glassy carbon as a secondary standard.

2.7 Confocal laser scanning microscopy (CLSM)

Imaging of the hydrogel samples was done using an inverted Leica TCS SP5 confocal and multiphoton microscope (Leica Microsystems, Germany) equipped with a HeNe/argon/MP laser source for fluorescence images and differential interference contrast optics for transmission images. A water immersion objective Leica HCX IRAPO L 25×/0.95 W was used. Dry PEI-MA samples were hydrated in water for five days at room temperature. After the final swelling state of the hydrogels was reached, the water in excess was removed, and the samples were immersed in a 1 mM solution of fluorescent dye, Alexa Fluor 350 carboxylic acid succinimidyl ester (Molecular Probes, λ_ex = 346 nm, λ_em = 442 nm) or ATTO 647N (ATTO-TEC, λ_ex = 644 nm, λ_em = 669 nm), at room temperature overnight. The samples were washed several times with deionized water to remove the fluorescent probe in excess, and then transferred to FluoroDish Petri dishes (World Precision Instrument, USA) for microscopy observation. Confocal microscopy images were acquired as z-stacks with a z-spacing of 3 or 10 μm between consecutive slices.

3 Results and discussion

3.1 Synthesis of methacrylated PEI-MA hydrogels

Polycationic hydrogels with different swelling behavior were prepared through the self-assembly of partially methacrylated polyethyleneimine (PEI-MA) in water.

For the synthesis of the PEI-MA molecules in dichloromethane (DCM), different molar ratios of methacrylic anhydride with respect to the amount of amine groups in the starting PEI were used, and the reaction was allowed to proceed at room temperature for 18 hours (Scheme 1).

Scheme 1 Synthesis of partially methacrylated branched polyethyleneimine (PEI-MA) in dichloromethane at room temperature.

The reaction conditions that we use are summarized in Table 1. At these conditions, a nucleophilic attack of the primary and secondary amine groups of PEI on the carbonyl group of methacrylic anhydride is favourable, and as a consequence amide bond formation occurs. Triethylamine (TEA) was added as a Lewis base catalyst to enhance the nucleophilicity of the amine groups. Methacrylic acid was formed as a by-product and was easily removed from the hydrogels through dialysis against ultrapure water and ethanol in several cycles over one week. After 18 h of methacrylation reaction, we observed that all PEI samples in DCM were spontaneously gelled into an amber-coloured material, with an apparent volume contraction with respect to the starting material. The DCM solvent was then removed from the methacrylated gel under vacuum, and several dialysis steps against ultrapure water and ethanol were applied to purify the samples from reagents and by-products, resulting in a swollen transparent polymer matrix (Fig. 1).

Table 1 Experimental conditions for synthesis of PEI-MA in dichloromethane^a

PEI-MA sample	Branched PEI (mol L^−1)	MAA (μL)	TEA (μL)	MAA/PEI reactive amines^b	MAA/TEA
a Reaction conditions based on 3 mL of anhydrous DCM. Synthesis conducted at 25 °C under argon atmosphere and in the dark overnight (18 h).b Reactive amines (primary and secondary) calculated assuming that PEI has 581 amine functions (based on a mean Mw of 43 Da for the repeat unit), and a distribution of 31% primary, 39% secondary, 30% tertiary amines calculated from ^13C NMR analysis.^46
PEI-MA-3	0.0136	73	10.2	2.95 × 10^−2	6.82
PEI-MA-4.5	0.0145	120	10.9	4.54 × 10^−2	11.23
PEI-MA-6	0.0142	158	10.7	6.11 × 10^−2	14.78
PEI-MA-7.5	0.0139	192	10.5	7.62 × 10^−2	17.97
PEI-MA-9	0.0133	218	10.0	9.00 × 10^−2	20.40

---

Fig. 1 Photographs of PEI-MA hydrogels synthesized with different ratios of MAA and PEI reactive amines at their final swelling state in 60 mm glass Petri dishes.

The introduction of hydrophobic moieties along the PEI polycationic backbone gives an amphiphilic character to the methacrylated macromolecule, which drives a spontaneous self-assembly according to the surrounding solvent environment. When the PEI-MA macromolecules are solved in DCM, the gel network is contracted due to the intramolecular interactions among the polymer hydrophilic domains. In aqueous solutions, partially methacrylated PEI-MA macromolecules self-assemble to generate hydrophilic gels with a unique microstructure held by a fine balance of non-covalent forces, including hydrophobic and electrostatic interactions due to the PEI-MA methacrylate moieties and protonated amine groups, respectively. All five formulations of the PEI-MA hydrogels (Table 1) were found to swell to a large extent embedding more than 90% of water. In addition to conferring an amphiphilic character to the PEI-MA molecules, the modification of amines by methacrylation acts to decrease the high cationic charge density of branched PEI, therefore lowering its cytotoxicity and generating more biocompatible hydrogels. Other amino group modifications, for example by acetylation or succinylation, have been proven to result in derivatives of branched PEI exhibiting strongly reduced citotoxicity.⁴⁹

3.2 FTIR and NMR characterization of PEI-MA macromolecules

The successful incorporation of the methacrylate groups in the branched PEI molecules was demonstrated by ATR-FTIR and ¹H NMR spectroscopy, which were used to investigate and quantify the modified PEI at the level of the amine groups. In ATR-FTIR spectroscopy, the spectrum of unmodified branched PEI in the fingerprint region shows typical absorbance bands at 1123 cm⁻¹ for C–N stretching vibrations of secondary amines, at 1298 cm⁻¹ for C–N stretching vibrations of primary amines and CH bending vibrations, at 1458 cm⁻¹ for N–H bending of secondary amines and CH₂ scissoring vibrations, and at 1585 cm⁻¹ for N–H bending vibrations of primary amines.^44,50 After reaction with methacrylic anhydride, for each of the PEI-MA macromolecules the appearance of strong absorption bands in the region 1600–1700 cm⁻¹ and at 1542 cm⁻¹ indicated the formation of amide groups (region III in Fig. 2A). Fig. 2B shows the individual absorption bands in the fingerprint region: the peak at 1652 cm⁻¹ and 1615 cm⁻¹ can be assigned to the C=O stretching vibration of the amide groups (amide I band), whereas the peak at 1542 cm⁻¹ is related to the C–N stretching vibration (amide II band). FTIR measurements also show a decrease of the absorption bands at both ∼1300 cm⁻¹ and at ∼1100 cm⁻¹ as compared to the starting PEI material, which confirms that methacrylation occurs at both primary and secondary amines. In addition, the large bands observed between 3600 and 3200 cm⁻¹ refer to the O–H stretching from the intramolecular and intermolecular hydrogen bonds (region I in Fig. 2A), whereas the vibrational bands between 3000 and 2750 cm⁻¹ are assigned to the C–H stretching from alkyl groups (region II in Fig. 2A).

Fig. 2 (A) FTIR absorbance spectra in ATR mode of dried PEI-MA macromolecules with different degrees of methacrylation collected under ambient conditions: (a) PEI-MA-3, (b) PEI-MA-4.5, (c) PEI-MA-6, (d) PEI-MA-7.5, (e) PEI-MA-9. (B) Individual absorbance bands in the fingerprint region between 1900 cm⁻¹ and 900 cm⁻¹ for the PEI-MA samples.

The degree of methacrylation of the branched PEI amines was investigated using NMR spectroscopy (Fig. 3). The starting commercial PEI had been previously analyzed by ¹³C NMR to determine the relative ratio of the different types of amines, which was found to be: 31% primary, 39% secondary, and 30% tertiary amines.^40,46 This ratio corresponds to a degree of branching (DB) of 61%, meaning that nearly every second nitrogen forms a branch. After reaction of PEI with methacrylic anhydride, we determined the extent of secondary and tertiary amide formation by ¹H NMR spectroscopy using a procedure adapted from the literature,⁴⁰ in particular by analysis of the peaks at δ 1.80–1.73 ppm corresponding to methacrylated primary amines [–NH–COC(CH₂)CH₃], and of the peaks at δ 2.09–2.08 ppm corresponding to methacrylated secondary amines [[double bond splayed left]N–COC(CH₂)CH₃]. Results from this analysis are summarized in Table 2. We determined that the overall methacrylation extent of the PEI amines in all PEI-MA samples is between 59.0% and 67.0%. The methacrylation reactions were greatly influenced by the large number and the different reactivity of the PEI functional amine groups, as well as by the highly branched structure of PEI. The percentages of methacrylated secondary amines and of methacrylated primary amines were calculated from the areas of the peaks assigned to the methyl signals (peaks (a) and (b) in Fig. 3) using the formulas reported in the Experimental section. Results from these calculations indicate that PEI primary amines are more reactive than secondary ones, as expected for the conditions used in our synthesis protocol, leading to the formation of a higher fraction of methacrylated primary amines than methacrylated secondary amines in each hydrogel formulation (Table 2).

Fig. 3 ¹H NMR spectra (D₂O, 700 MHz) of PEI-MA hydrogels with different degrees of methacrylation at 25 °C. The peaks between δ 3.0 ppm and 2.5 ppm correspond to the protons of the PEI backbone, the characteristic peak at δ 1.80–1.73 ppm to methacrylated primary amines [–NH–MA] (a), and the peak at δ 2.09–2.08 ppm to methacrylated secondary amines [[double bond splayed left]N–MA] (b). The peaks at δ 4.65 ppm and 5.30 ppm are assigned to solvent D₂O and to residual dichloromethane,⁵¹ respectively.

Table 2 Properties of methacrylated branched PEI as determined by NMR analysis^a

PEI-MA sample	% NH–MA^b	% N–MA^b	Amine mole fractions^c			MA yield (%)
			1°	2°	3°
a ^1H NMR spectroscopy (700 MHz) in D2O at 25 °C.b Percentages of methacrylated primary amines [–NH–COC(CH2)CH3] and methacrylated secondary amines [[double bond splayed left]N–COC(CH2)CH3] of branched PEI as determined by ^1H NMR peak integration.c Mole fractions calculated from percentage of methacrylation starting from the amine content of commercial branched PEI previously reported.^40,46
Branched PEI			0.307	0.395	0.298
PEI-MA-3	42.4	16.6	0.177	0.460	0.363	59.0
PEI-MA-4.5	45.6	18.5	0.096	0.455	0.448	64.2
PEI-MA-6	43.6	16.1	0.054	0.424	0.521	59.8
PEI-MA-7.5	51.4	15.7	0.026	0.385	0.588	67.0
PEI-MA-9	49.9	16.1	0.013	0.336	0.650	66.0

---

3.3 Swelling behavior of PEI-MA hydrogels

In water all the synthesized PEI-MA molecules showed large degrees of swelling. The water content of the equilibrated hydrogels was between 93 wt% and 95 wt% (Table 3), whereas the values of the weight swelling ratio were in the range from 15.3 to 20.6, with the highest value observed for the PEI-MA-6 formulation. The high water-sorption capacity of the PEI-MA hydrogels makes them potential candidates for highly biocompatible materials, as one of the prerequisite of such materials is their high hydrophilicity, which results in weak interactions with the extracellular matrix components.^18,52 PEI-MA hydrogels were left to equilibrate in water for five days before determining their swelling ratio. This equilibration time was chosen after conducting kinetics experiments in which the time required by the hydrogels to reach complete swelling was determined. All PEI-MA hydrogels were fully swollen in water after three to four days (Fig. S1†), and therefore an equilibration time of at least five days was chosen for all investigations.

Table 3 Swelling behavior of PEI-MA hydrogels and relative amounts of PEI-MA amines

PEI-MA sample	Water content^a (wt%)	Swelling ratio^a	Amine ratio^b (1° + 2°)/3°	% NH–MA/% N–MA
a Water content and swelling ratio calculated averaging the weights of five different samples for each hydrogel formulation.b Relative amounts of PEI-MA amines, including methacrylated primary and methacrylated secondary amines, calculated from the values in Table 2.
PEI-MA-3	94.2 ± 0.1	17.3 ± 0.3	1.75	2.55
PEI-MA-4.5	93.6 ± 0.2	15.5 ± 0.4	1.23	2.46
PEI-MA-6	95.1 ± 0.1	20.6 ± 0.2	0.92	2.70
PEI-MA-7.5	93.5 ± 0.1	15.3 ± 0.2	0.70	3.28
PEI-MA-9	93.5 ± 0.2	15.5 ± 0.5	0.54	3.09

---

The swelling behavior of the PEI-MA hydrogels was analyzed in relation to the extent of methacrylation of branched PEI, in particular considering the ratio between the methacrylated primary and secondary amines. We observed that when the amount of primary and secondary amines, including those bearing methacrylate groups, is higher or approximately equal to the amount of tertiary amines, which occurs for the first three hydrogel formulation in Table 3, the value of the swelling ratio appears to be correlated with the ratio between the percentages of methacrylated primary and methacrylated secondary amines. In this context, the PEI-MA-6 hydrogel with a ratio of methacrylated primary amines (–NH–MA) to methacrylated secondary amines ([double bond splayed left]N–MA) of 2.70 achieves the highest values of swelling ratio (20.6) and water content (95.1 wt%). On the other hand, at the reaction conditions of PEI-MA-7.5 and PEI-MA-9 hydrogels, for which the amount of tertiary amines is prevalent with respect to that of primary and secondary ones (amine ratio (1° + 2°)/3° of 0.70 and 0.54 respectively), the hydrogel swelling behavior is similar with values of the swelling ratio around 15.3–15.5 wt%. The composition of the modified PEI-MA macromolecules in terms of amine ratio (1° + 2°)/3° plays a crucial role in the hydrogel network formation and equilibrium structure in water. However, to fully understand the self-assembly of the hydrogel network and its swelling behavior other factors need to be considered, for example the spatial distribution of the methacrylated amines also contributes to a well-defined gel microstructure in water. To further investigate this issue the microstructure of the completely swollen hydrogels was investigated in SAXS experiments. Results from this analysis are illustrated in the following section.

3.4 SAXS analysis of PEI-MA hydrogels

PEI-MA hydrogels with various degrees of methacrylation were investigated by small-angle X-ray scattering (SAXS) in order to gain information on the nature of their supramolecular structure and how that relates to the synthesis conditions. SAXS profiles for the swollen PEI-MA samples in water are shown in Fig. 4. The spectra clearly indicate the presence of structures at different length scales in all hydrogels, with a broad shoulder at q values around 0.07 Å⁻¹ and a power law behavior in the low q regime. The presence of a broad shoulder over the nanometer length scale (∼0.01 Å⁻¹ to 1 Å⁻¹) indicates that these hydrogels have a weakly ordered structure characterized by a broad polydispersity in correlation lengths and random orientation.^53,54 This structure is consistent with a loose and highly swollen polymer network, such as the supramolecular PEI-MA hydrogel under investigation here, whose structure is sustained through a delicate balance between the hydrophobic interaction driving aggregation of the methacrylated side chains in water and the charge repulsion from the protonated amine groups. This interaction balance enables the self-assembly of nanoscale domains with a characteristic period of approximately 9 nm, corresponding to the SAXS shoulder position at 0.07 Å⁻¹ via the equation d = 2π/q*.

Fig. 4 Small-angle X-ray scattering spectra for the fully swollen PEI-MA hydrogels with different degrees of methacrylation obtained at room temperature, showing power law behavior with two distinct regimes at low q values and a broad shoulder at 0.07 Å⁻¹. The intensity values are offset for clarity.

In the low q regime the scattering profiles of all PEI-MA hydrogels show a power law behavior with a distinct crossover between two regions of different slopes around the q value of 0.01 Å⁻¹. A power law behavior is indicative of the presence of large-scale structures in the hydrogels and is often associated to scattering from fractal structures.⁵⁵ In the power law regime, the SAXS spectra can be fitted according to the equation I(q) = I₀q^−α, where the value of the exponent α gives a measure of the fractal dimension D and is related to network density of the hydrogel structure.^55,56 In particular, for mass fractals the power law exponent α is equal to D_mass and varies between 1 and 3, whereas for surface fractals α is equal to 6 − D_surf and varies between 3 and 4. The values of D_surf are between 3 for rough surfaces and 2 for smooth surfaces.

The scattering profiles in Fig. 4 suggest that the PEI-MA hydrogels are characterized by a multilevel fractal structure, where clusters of different densities, hence held together by supramolecular interactions of different nature, coexist. This structure is similar to that occurring in self-assembled associative colloidal systems, where aggregates, rigid clusters of primary particles, coexist with agglomerates, larger structures resulting from weak bonding among aggregates.⁵⁷

The values of the power law exponents (α) obtained from the fit of the scattering curves of the PEI-MA hydrogels in both regions I (0.005 < q < 0.01 Å⁻¹) and II (0.01 < q < 0.03 Å⁻¹) are reported in Fig. 5. The horizontal line at α = 3 denotes the crossover from mass fractal to surface fractal structures. The power law exponents in region II are in the range 1.5–1.9 for all PEI-MA hydrogels, which is consistent with mass fractals made of loosely aggregated clusters. None of the hydrogels have compact structures with well-defined interfaces, with the PEI-MA-6 formulation characterized by the smallest value of the exponent α (1.52), and therefore by the more loosely connected network. This result is in agreement with the higher water content and swelling ratio of the PEI-MA-6 hydrogel with respect to the other PEI-MA gels (Table 3). In addition, it is interesting to note that, when related to the reaction conditions of the hydrogel formulations, the trend of the α exponent in region II closely follows the ratio between the percentages of methacrylation of the primary and secondary amines of branched PEI, which were determined by NMR analysis in the section above (Table 3).

Fig. 5 Power law exponents (α) from the fit of the scattering profiles of the PEI-MA hydrogels in region I (0.005 < q < 0.01 Å⁻¹) and region II (0.01 < q < 0.03 Å⁻¹) (left axis), and relation with the ratio of methacrylated primary and secondary amines (right axis) as determined by NMR analysis (Table 2). The correlation coefficients of all power law fits are higher than 0.99.

The scattering profiles of the PEI-MA hydrogels show a crossover between two distinct power law regimes around the q value of 0.01 Å⁻¹. The power law exponents in region I are always higher than those in region II, indicating that larger agglomerate structures coexist with smaller loose aggregates in all PEI-MA hydrogels. For the PEI-MA-4.5 and PEI-MA-6 gels, such larger structures are still mass fractals with dimension of 2.09 and 2.34, respectively. After the PEI-MA-6 formulation, a crossover from mass fractal to surface fractal can be observed in the power law exponents (Fig. 5). From the power law exponents, the values of D_surf for the PEI-MA-7.5 and PEI-MA-9 gels can be calculated as 2.26 and 2.77, respectively. These values indicate that PEI-MA-7.5 hydrogels, which are characterized by a larger ratio of methacrylated primary to secondary amines, have a smoother surface fractal structure than the PEI-MA-9 gels.

The PEI-MA-3 hydrogel follows a different trend than the other gels. The lacking of the shoulder in the scattering profile indicates that the balance of the interactions is not sufficient to drive an ordered nanophase separation with characteristic network spacing, but rather might form randomly orientated domains. Further, such hydrogel is characterized by a steep slope of the power law behavior at very low q values with an exponent larger than 4. Values of the power law exponents larger than 4 occur in systems with a continuous density change at the boundary of the scatterer.⁵⁵

3.5 Morphology of PEI-MA hydrogels

The hydrogel morphology, microscopic structures and pore network dimension were investigated by confocal laser scanning microscopy. From the observation of the fluorescent images it was possible to visualize the microstructures of the hydrogels, including the extent of the pore connectivity in the network. Confocal laser scanning microscopy images of the PEI-MA hydrogels stained with Alexa Fluor 350 carboxylic acid succinimidyl ester are reported in Fig. 6. The hydrogels were first fully hydrated in deionized water containing the dye overnight, and then rinsed several times with pure water before being transferred to the FluoroDish for microscopy imaging. Fig. 6a and d show the overlay of dye-labeled hydrogel sections with thickness of 10 μm for the PEI-MA-6 and PEI-MA-7.5 gel formulations.

Fig. 6 z-stack of confocal fluorescence microscopy images (a and d) and corresponding white-light transmission images (b and e) of 10 μm-thick sections across PEI-MA hydrogels with different degrees of methacrylation: PEI-MA-6 (a–c), PEI-MA-7.5 (d–f). Images on the right side of the panels (c and f) show a superposition of the two microscopy images on the left. Hydrogels stained by Alexa Fluor 350 dye overnight and imaged at room temperature after rinsing.

The network of the PEI-MA-6 hydrogel is characterized by an open structure with many interconnected cavities having irregular interfaces. The maximum size of the interconnected pores is of the order of 100 μm.

On the other hand, the PEI-MA-7.5 hydrogel has a more compact microstructure with smoother boundaries at the material/water interface. These results by fluorescence microscopy are in agreement with the weight analysis of the hydrogel swelling behavior reported in Table 3. Indeed, the more open and less dense structure of the PEI-MA-6 hydrogel is compatible with the larger water content and swelling ratio than the other PEI-MA materials. Further, the SAXS investigation of the hydrogel structure, through the mass and surface fractal dimensions determined by the power law behavior of the scattering profiles (Fig. 5), also finds confirmation in the fluorescence microscopy results. In fact, the PEI-MA-6 formulation is characterized by the smallest value of the exponent α (1.52), and therefore by the more loosely connected network. In addition, by analyzing the lowest q values (region I) of the scattering profiles, two different fractal regimes, mass fractal and surface fractal with smooth surface, were determined for the PEI-MA-6 and PEI-MA-7.5 hydrogels, respectively. These fractal regimes are consistent with the material/water interfaces that can be visualized in fluorescence microscopy (Fig. 6). Overall, all the investigations reported here point to the fact that the chemical composition of the modified PEI macromolecules, that is the extent of the amine methacrylation, can be related to different supramolecular arrangements leading to self-assembled hydrogel networks with defined microstructure and swelling behavior.

Experiments of confocal fluorescence microscopy were also used to highlight the different morphology of the PEI-MA hydrogels at different temperatures. The variation of the hydrogel structure with temperature is a typical feature of supramolecular systems, where the microstructure is determined by non-covalent interactions that are highly dependent on the temperature. Therefore, significant changes in the hydrogel structure as a function of the temperature can be considered as a sign of its supramolecular nature. In confocal experiments, the PEI-MA hydrogel stained by fluorescent dye (ATTO 647N) was left to equilibrate at the selected temperature, and then z-stacks of 3 μm-thick sections were collected. Fig. 7 shows the microstructure of the PEI-MA-6 hydrogels for one of such sections at the temperature of 4 °C (top images) and 20 °C (bottom images). As expected for supramolecular hydrogels, there is a substantial difference in the hydrogel morphology at the two temperatures. The PEI-MA network at 4 °C appears more compact and arranged in close overlapping layers, whereas at 20 °C its structure is more open and organized in a fractal arrangement with interconnected microcavities filled by water. The supramolecular nature of the PEI-MA hydrogels was therefore verified through ¹H-NMR spectroscopy, where the characteristics peaks that can be related to cross-linking reactions were not present, and through the analysis by confocal laser scanning microscopy of how the hydrogel morphology changes with the temperature.

Fig. 7 Confocal fluorescence microscopy images (a and d) and white-light transmission images (b and e) of the PEI-MA-6 hydrogels stained with ATTO 647N, showing the different microstructure of the hydrogel at 4 °C (a–c) and 20 °C (d–f). Images on the right side of the panels (c and f) show a superposition of the two microscopy images on the left.

4 Conclusions

We have investigated the synthesis and properties of novel supramolecular polycationic hydrogels prepared via the self-assembly of partially methacrylated PEI molecules in water. PEI-MA molecules were shown to generate hydrogels characterized by large swelling capacities, and their hydrophilicity and microstructure was optimized by varying the degree of PEI methacrylation. The SAXS analysis revealed that these hydrogels contain multiscale fractal structures with larger agglomerates coexisting with smaller loose aggregates. Hydrogels up to the PEI-MA-6 formulation are characterized by mass fractal structures in both regions I (0.005 < q < 0.01 Å⁻¹) and II (0.01 < q < 0.03 Å⁻¹) of the scattering vector, whereas PEI-MA hydrogels synthesized at larger values of the ratio MAA/PEI reactive amines show a transition to surface fractal structures in the lower q region. In addition, a correlation was found between the fractal dimension of the hydrogel and the ratio between methacrylated primary (–NH–MA) and methacrylated secondary ([double bond splayed left]N–MA) amines. In the hydrogel series, the PEI-MA-6 formulation, which is characterized by the largest swelling capacity, represents the border between the transitions from mass fractal to surface fractal microstructure. The more open fractal structure of the PEI-MA-6 hydrogel was confirmed in fluorescence imaging experiments by confocal microscopy, which enabled to visualize the interconnected micro-cavities of the hydrogel network. The fractal microstructure of the PEI-MA hydrogels, in combination with their highly hydrophilic nature and the reduced polycationic charge after PEI methacrylation, makes them strong candidates as potential biocompatible materials for biological and pharmaceuticals applications.

Acknowledgements

The authors thank Fabio Formiggini (IIT@CRIB) for technical assistance with fluorescence microscopy, and Livia Chitu and Günther Maier (Materials Center Leoben, Austria) for the SAXS measurements. M. G. S. acknowledges financial support from the Italian Ministry of Education, University and Research (MIUR) through the Montalcini Grant.

References

1. E. A. Appel, J. del Barrio, X. J. Loh and O. A. Scherman, Chem. Soc. Rev., 2012, 41, 6195–6214.
2. S. Seiffert and J. Sprakel, Chem. Soc. Rev., 2012, 41, 909–930.
3. J. D. Hartgerink, E. Beniash and S. I. Stupp, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 5133–5138.
4. A. Aggeli, M. Bell, N. Boden, J. N. Keen, P. F. Knowles, T. C. B. McLeish, M. Pitkeathly and S. E. Radford, Nature, 1997, 386, 259–262.
5. M. von Gröning, I. de Feijter, M. C. Stuart, I. K. Voets and P. Besenius, J. Mater. Chem. B, 2013, 1, 2008–2012.
6. F. Peng, G. Li, X. Liu, S. Wu and Z. Tong, J. Am. Chem. Soc., 2008, 130, 16166–16167.
7. N. E. Shi, H. Dong, G. Yin, Z. Xu and S. H. Li, Adv. Funct. Mater., 2007, 17, 1837–1843.
8. A. Wang, J. Huang and Y. Yan, Soft Matter, 2014, 10, 3362–3373.
9. G. Wilhelmina de Groot, S. Demarche, M. G. Santonicola, L. Tiefenauer and G. J. Vancso, Nanoscale, 2014, 6, 2228–2237.
10. G. M. van Gemert, J. W. Peeters, S. H. Söntjens, H. M. Janssen and A. W. Bosman, Macromol. Chem. Phys., 2012, 213, 234–242.
11. L. Ionov, Adv. Funct. Mater., 2013, 23, 4555–4570.
12. S. Yamaguchi, I. Yoshimura, T. Kohira, S.-i. Tamaru and I. Hamachi, J. Am. Chem. Soc., 2005, 127, 11835–11841.
13. J. Mao and M. J. McShane, Adv. Mater., 2006, 18, 2289–2293.
14. J.-H. Lee, J. H. Lee, Y. J. Lee and K. T. Nam, Curr. Opin. Biotechnol., 2013, 24, 599–605.
15. E. J. Howe, B. O. Okesola and D. K. Smith, Chem. Commun., 2015, 51, 7451–7454.
16. K. Y. Lee and D. J. Mooney, Prog. Polym. Sci., 2012, 37, 106–126.
17. J. Li and X. J. Loh, Adv. Drug Delivery Rev., 2008, 60, 1000–1017.
18. T. R. Hoare and D. S. Kohane, Polymer, 2008, 49, 1993–2007.
19. J. Li, X. Ni and K. W. Leong, J. Biomed. Mater. Res., Part A, 2003, 65, 196–202.
20. K. Ren, C. He, C. Xiao, G. Li and X. Chen, Biomaterials, 2015, 51, 238–249.
21. G. A. Silva, C. Czeisler, K. L. Niece, E. Beniash, D. A. Harrington, J. A. Kessler and S. I. Stupp, Science, 2004, 303, 1352–1355.
22. D.-Q. Wu, T. Wang, B. Lu, X.-D. Xu, S.-X. Cheng, X.-J. Jiang, X.-Z. Zhang and R.-X. Zhuo, Langmuir, 2008, 24, 10306–10312.
23. P.-H. Kim, H.-G. Yim, Y.-J. Choi, B.-J. Kang, J. Kim, S.-M. Kwon, B.-S. Kim, N. S. Hwang and J.-Y. Cho, J. Controlled Release, 2014, 187, 1–13.
24. Z. Gu, A. A. Aimetti, Q. Wang, T. T. Dang, Y. Zhang, O. Veiseh, H. Cheng, R. S. Langer and D. G. Anderson, ACS Nano, 2013, 7, 4194–4201.
25. W. P. Appel, M. M. Nieuwenhuizen and E. Meijer, in Supramolecular Polymer Chemistry, ed. A. Harada, Wiley-VCH, Weinheim, 2012, pp. 3–28.
26. F. Li, Y. Zhu, B. You, D. Zhao, Q. Ruan, Y. Zeng and C. Ding, Adv. Funct. Mater., 2010, 20, 669–676.
27. F. Huang and O. A. Scherman, Chem. Soc. Rev., 2012, 41, 5879–5880.
28. L. Brunsveld, B. Folmer, E. Meijer and R. Sijbesma, Chem. Rev., 2001, 101, 4071–4098.
29. Y.-L. Chiu, M.-C. Chen, C.-Y. Chen, P.-W. Lee, F.-L. Mi, U. S. Jeng, H.-L. Chen and H.-W. Sung, Soft Matter, 2009, 5, 962–965.
30. H. Komatsu, S. Matsumoto, S.-i. Tamaru, K. Kaneko, M. Ikeda and I. Hamachi, J. Am. Chem. Soc., 2009, 131, 5580–5585.
31. M. A. Mintzer and E. E. Simanek, Chem. Rev., 2008, 109, 259–302.
32. B. Urban-Klein, S. Werth, S. Abuharbeid, F. Czubayko and A. Aigner, Gene Ther., 2004, 12, 461–466.
33. Z. Dai, T. Gjetting, M. A. Mattebjerg, C. Wu and T. L. Andresen, Biomaterials, 2011, 32, 8626–8634.
34. X. Cui, F. Yang, A. Li and X. Yang, Anal. Biochem., 2005, 342, 173–175.
35. D. D. Dunlap, A. Maggi, M. R. Soria and L. Monaco, Nucleic Acids Res., 1997, 25, 3095–3101.
36. Y. Chen, B. Pan, S. Zhang, H. Li, L. Lv and W. Zhang, J. Hazard. Mater., 2011, 190, 1037–1044.
37. M. Ahmed and R. Narain, Biomaterials, 2013, 34, 4368–4376.
38. S. C. Han, W. D. He, J. Li, L. Y. Li, X. L. Sun, B. Y. Zhang and T. T. Pan, J. Polym. Sci., Part A: Polym. Chem., 2009, 47, 4074–4082.
39. S. Nimesh, A. Aggarwal, P. Kumar, Y. Singh, K. Gupta and R. Chandra, Int. J. Pharm., 2007, 337, 265–274.
40. M. L. Forrest, G. Meister, J. Koerber and D. Pack, Pharm. Res., 2004, 21, 365–371.
41. Y. Akiyama, A. Harada, Y. Nagasaki and K. Kataoka, Macromolecules, 2000, 33, 5841–5845.
42. C.-H. Ahn, S. Y. Chae, Y. H. Bae and S. W. Kim, J. Controlled Release, 2002, 80, 273–282.
43. G. P. Tang, J. M. Zeng, S. J. Gao, Y. X. Ma, L. Shi, Y. Li, H. P. Too and S. Wang, Biomaterials, 2003, 24, 2351–2362.
44. A. Paciello and M. G. Santonicola, J. Mater. Chem. B, 2015, 3, 1313–1320.
45. A. Paciello, A. M. Cusano and M. G. Santonicola, Eur. Cells Mater., 2013, 26(S6), 52.
46. A. von Harpe, H. Petersen, Y. Li and T. Kissel, J. Controlled Release, 2000, 69, 309–322.
47. D. Hölter, A. Burgath and H. Frey, Acta Polym., 1997, 48, 30–35.
48. J. Ilavsky and P. R. Jemian, J. Appl. Crystallogr., 2009, 42, 347–353.
49. A. Zintchenko, A. Philipp, A. Dehshahri and E. Wagner, Bioconjugate Chem., 2008, 19, 1448–1455.
50. S. Tan, M. Erol, A. Attygalle, H. Du and S. Sukhishvili, Langmuir, 2007, 23, 9836–9843.
51. G. R. Fulmer, A. J. M. Miller, N. H. Sherden, H. E. Gottlieb, A. Nudelman, B. M. Stoltz, J. E. Bercaw and K. I. Goldberg, Organometallics, 2010, 29, 2176–2179.
52. N. Angelova and D. Hunkeler, Trends Biotechnol., 1999, 17, 409–421.
53. S. K. Agrawal, N. Sanabria-DeLong, P. R. Jemian, G. N. Tew and S. R. Bhatia, Langmuir, 2007, 23, 5039–5044.
54. D. J. Waters, K. Engberg, R. Parke-Houben, C. N. Ta, A. J. Jackson, M. F. Toney and C. W. Frank, Macromolecules, 2011, 44, 5776–5787.
55. P. W. Schmidt, J. Appl. Crystallogr., 1991, 24, 414–435.
56. G. Beaucage, J. Appl. Crystallogr., 1996, 29, 134–146.
57. D. W. Schaefer, T. Rieker, M. Agamalian, J. S. Lin, D. Fischer, S. Sukumaran, C. Chen, G. Beaucage, C. Herd and J. Ivie, J. Appl. Crystallogr., 2000, 33, 587–591.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra16576h

---