A simple approach to hybrid inorganic–organic step-growth hydrogels with scalable control of physicochemical properties and biodegradability

We prepared new and scalable, hybrid inorganic–organic step-growth hydrogels with polyhedral oligomeric silsesquioxane (POSS) network knot construction elements and hydrolytically degradable poly(ethylene glycol) (PEG) di-ester macromonomers by in situ radical-mediated thiol–ene photopolymerization.

vacuum and quickly transferred into the cryo-preparation chamber at a temperature of T = -90 ºC. Inside this chamber, the frozen samples were fractured to expose a virgin surface and ice was allowed to sublimate for 1 hour in vacuum before sputtering a thin layer of gold and subsequent imaging.
Thermal gravimetric analysis (TGA) of the dried hydrogels were measured in a TA Q5000 instrument with heating ramps of 10 ºC min -1 from T = 50 ºC to 800 ºC under air atmosphere.
Compression tests of the hydrogel discs were performed in a Zwick/Roell Z0.5 mechanical tester using a load cell of 500 N and a compression velocity of 500 µm min -1 . The waterequilibrated hydrogel discs were objected to compression tests in an unconfined setting.
Dependent upon water-uptake of the respective sample, the dimensions of the discs tested varied from approximately 5 to 6 mm in thickness and from 11 to 15 mm in diameter. The compression strain limit was set to 90%. The compressive moduli were determined from the slope of the initial near-linear part of the stress-strain curve until 10% strain.

Preparation of Homotelechelic Thiol PEG Di-Esters (PEG Macromonomers).
Thioglycolic acid, the respective PEG and concentrated sulphuric acid were dissolved in benzene and allowed to stir for 1 hour under nitrogen atmosphere. The solution was then heated to the refluxing temperature and allowed to react overnight. The molar equivalents (equiv.) of thioglycolic acid to that of hydroxyl functionality of PEG varied from 2.5 equiv.
for the PEG with M n = 600 and 1000 g mol -1 to 4 equiv. and 30 equiv. for M n = 2050 and 6000 g mol -1 , respectively. Water from the esterification reaction formed an azeotrope with benzene and in this fashion was removed from the system with a Dean-Stark trap. After reaction, benzene was removed by vacuum distillation and the reaction product was subsequently taken up in chloroform. The PEG M n = 748.24 g mol -1 macromonomer (which appeared as a transparent liquid) was washed twice with water in a separation funnel with the organic phase being concentrated thereafter by rotary evaporation. The other PEG macromonomers (which typically appeared as white solids) were precipitated twice from chloroform by addition of diethylether. All PEG macromonomers were then placed in a desiccator for drying and storage before use or further characterization. The above-mentioned procedure resulted in the following yields and NMR results of PEG macromonomers: PEG M n = 748.24 g mol -1 : 89% -1 H NMR (CDCl 3 , 300 MHz): δ ppm 2.03 (t, 2H, J = 8.4 Similar peak shifts, integration amplitudes as well as coupling constants were also reported by other authors working with related PEG macromonomers. 1 Hydrogel Preparation. For hydrogel preparation, the vinylPOSS cage mixture and radical initiator (DMPA) (1 wt% with respect to PEG macromonomer) was first dissolved in DMF and then the PEG macromonomer was added.
In a first and second series of experiments, this procedure was followed with an equimolar concentration of functional groups while varying the PEG macromonomer chain length (Gels 1-4), and with a disparity of functional group concentration with one example PEG macromonomer chain length (M n = 1148.24 g mol -1 ) (Gels 5-7). Additionally, in a third series of experiments the PEG macromonomer chain length remained constant at equimolar concentration of thiol and vinyl functional groups and we applied the co-solvent dodecanol that was added after the dissolution of the gel precursor monomer mixture in respective amounts of DMF (Gels 8-10). The homogeneous liquid precursor solutions were transferred to 4 mL glass vials, and sealed with the cap. UV-initiated reactions were triggered by placing the vials in a Rayonet Chamber reactor. Following gel formation, the resulting disc-shaped materials were placed in glass vials with a frequent exchange of solvent DMF in order to remove any non-reacted monomers and initiator. Subsequently, DMF was gradually exchanged to deionized water, in which the hydrogels were allowed to equilibrate for one day under frequent water exchange before further characterization.
Hydrogel Functionalization. Hydrogel discs prepared with a stoichiometric disparity of functional groups, resulting in gels containing vinyl-or thiol-pendant moieties, underwent further functionalization. Vinyl-pendant hydrogels (Gel 5) were functionalized in a solution of 80:20 water/DMF (%, v/v) containing 5 % (w/v) of cysteine or thioglycolic acid with 1 wt% DMPA (with respect to the thiol). Reactions were performed in the UV reactor for one hour under stirring. The functionalized gels were purified by frequent exchange of a solution of 80:20 water and DMF (%, v/v) before gradual exchange to deionized water, followed by further characterization. An example gel containing thiol-pendant functionality (Gel 7) was functionalized with fluorescein-5-maleimide. Therefore, the DMF containing hydrogel was placed in 4 mL DMF containing 10 mg of the dye. The sealed reaction vessel was placed in an oil bath overnight at 60 ºC under stirring. Thereafter, the hydrogel was thoroughly washed with DMF to remove non-reacted species. Then, the solvent was exchanged to deionized water and the gel further characterized.

Determination of Gel Fractions and Water Uptake. The gel fraction is a useful estimate
to judge on the efficiency of transformation of monomeric precursors to actually incorporated amount of these precursors in the final gel. Therefore, after formation of gels, washing, and subsequent exchange to water, they were dried, and the gel fraction calculated via the following equation: For characterization of water content under equilibrium conditions, the weights of the water-containing hydrogel discs at room temperature were recorded after quickly wiping off the excess of water on their surfaces with moistened tissue. Thereafter, control samples were allowed to dry at room temperature, while others were placed either in a water bath at T = 37 ºC or in a cold room at T = 5 ºC for one day. Afterwards, the corresponding weights were determined as mentioned before. Finally, also these materials were allowed to dry at room temperature until their weight did not reduce any further. Percent swelling ratios were calculated according to the following equation.
(eq. 2) Additional Tables   Table S1: Comparison of theoretical and experimentally determined ceramic yields (SiO 2 ) from TGA analysis of all hydrogels. b Values calculated based on the assumption that all POSS precursors are incorporated in the polymer in correct stoichiometric amounts and the SiO 3/2 contained in the respective hydrogels is oxidized resulting in amorphous SiO 2 as indicated by FTIR spectroscopy ( Figure   S4).  no rupture no rupture a 1031.6 (± 40.9) a For Gels 9 and 10 no rupture was observed in the stress-strain curves until 90 % strain (maximum deformation applied due to security reasons). Therefore, the compressive strength at 90 % strain is reported. Figure S1. Experimentally found swelling ratios against M n of PEG macromonomer used to prepare Gels 1-4 at three temperatures (T = 5°C, red squares; T = 25 °C, blue squares; T = 37 °C, green squares). While an almost proportional increase of water uptake for Gels 1-3 is found, a pronounced deviation toward larger swelling ratios is seen for Gel 4. This is in accordance to a pronouncedly decreased gel fraction found for Gel 4 (Table 1).