Surface hydrophobization of hydrogels via interface dynamics-induced network reconfiguration

Effective and easy regulation of hydrogel surface properties without changing the overall chemical composition is important for their diverse applications but remains challenging to achieve. We report a generalizable strategy to reconfigure hydrogel surface networks based on hydrogel–substrate interface dynamics for manipulation of hydrogel surface wettability and bioadhesion. We show that the grafting of hydrophobic yet flexible polymeric chains on mold substrates can significantly elevate the content of hydrophobic polymer backbones and reduce the presence of polar groups in hydrogel surface networks, thereby transforming the otherwise hydrophilic hydrogel surface into a hydrophobic surface. Experimental results show that the grafted highly dynamic hydrophobic chains achieved with optimal grafting density, chain length, and chain structure are critical for such substantial hydrogel surface network reconfiguration. Molecular dynamics simulations further reveal the atomistic details of the hydrogel network reconfiguration induced by the dynamic interface interactions. The hydrogels prepared using our strategy show substantially enhanced bioadhesion and transdermal delivery compared with the hydrogels of the same chemical composition but fabricated via the conventional method. Our findings provide important insights into the dynamic hydrogel–substrate interactions and are instrumental to the preparation of hydrogels with custom surface properties.


Comparison of silicone oil-coated mold and DNR mold for hydrogel preparation
Silicone oil-coated mold and DNR mold were used to prepare PAA hydrogels to demonstrate the importance of covalently grafting silicone chains onto the mold surface for hydrogel surface wettability regulation (Supplementary Fig. 8a).Before hydrogel preparation, both the silicone oilcoated mold and DNR mold showed hydrophobicity (Supplementary Fig. 8b).However, the WCA value of the hydrogel prepared by the oil-coated mold was much lower and declined much faster than that of the DNR hydrogel (Supplementary Fig. 8c and 8d).Moreover, the silicone oil was entrained by the hydrogel upon extraction from the mold, as confirmed by the emerging signal of siloxane in the ATR-FTIR curve (Supplementary Fig. 8e).This resulted in a dramatic decrease in the WCA of the oil-coated mold after hydrogel preparation (Supplementary Fig. 8b).In contrast, the WCA of the DNR mold did not change even after 20 uses for hydrogel preparation (Supplementary Fig. 8f).

Determination of the grafting density of silicone chains on the DNR mold surface
We determined the density of silicone chains on the DNR mold surface by AFM topographical imaging to quantify the fraction of the mold surface occupied by silicone based on the color difference (Supplementary Fig. 20).To make the color difference more observable and quantifiable, we converted the original AFM images to gray mode and split the color channel to red with a threshold of 80 by ImageJ.The red region represents the mold substrate, while the gray region is the silicone.The fraction of silicone chains was determined by measurement of the proportion of the gray region (Fig. 2g).Discussion: The more significant reduction of WCA for PAAm hydrogel is possibly due to the lower polymer content of PAAm hydrogel (20 wt%) compared to that (25 wt%) of PAA and PMAA hydrogels, which like makes the hydrogel more permeable by the probing water droplet.On the other hand, we reason that the difference in side functional groups, i.e., amino in PAAm and carboxyl in PAA/PMAA, could also result in the varied recovery rate of the hydrogel surface network when contacting with the probing water droplet.Discussion: With higher DMDMS concentrations, a weaker signal assigned to silicon wafer (Si, 98.1 eV) can be observed for the DNR mold, suggesting a reduced substrate area and higher silicone chain density in the mold surface.values less than 0.05 were considered statistically significant differences among the compared groups.
Water contact angles (WCAs) of conventional and DNR hydrogels.The poly(acrylic acid) (PAA) hydrogels were prepared using the conventional method or the DNR strategy.Surface wettability of PAA hydrogels prepared by different molds.a WCA of different mold substrates (blue curve) and corresponding WCA of PAA hydrogels prepared using the molds (gray columns).The substrate of the DNR mold is glass.b WCA of hydrogels prepared by the DNR mold with different substrate materials.Notably, because PMMA and PTFE are chemically inert and the surface modification of the two substrates with organosilanes is highly challenging, we did not prepare DNR molds by PMMA and PTFE substrates.Data are shown as the mean ± SD; n = 3 independent samples.a WCA evolution and b photo images of the PAA-based DNR hydrogel after preparation and standing for 24 h at room temperature and 90% relative humidity.Values in a are shown as the mean ± SD; n = 4 independent samples.Surface wettability of different hydrogels prepared by the conventional method or DNR strategy.a WCA evolution of three kinds of conventional hydrogels (PMAA, PAA and PAAm hydrogels) in 5 min.b WCA evolution of the corresponding DNR hydrogels in 5 min.c WCA difference of the above three hydrogels.WCA difference means the WCA value of the DNR hydrogel minus that of the conventional hydrogel.d WCA evolution of the PAA-based DNR hydrogel in 30 min.Data are shown as the mean ± SD; n = 4 independent samples.

ca
WCA of double-network hydrogels prepared by the conventional method or DNR strategy.Data are shown as the mean ± SD; n = 3 independent samples.Supplementary Fig. 7 | Fabrication of the Janus hydrogel with asymmetric wettability using the DNR strategy.a Schematic illustration of the fabrication process by placing the untreated mold and DNR mold on the different sides of the hydrogel.b Image shows the Janus hydrogel with totally different wettability on the two sides of the hydrogel.Water droplets were dyed by rhodamine to improve visibility.c Comparison of the WCA on the two sides of the Janus hydrogel.Values in c are shown as the mean ± SD; n = 4 independent samples.e a te d m o ld s id e D N R m o ld s id e Supplementary Fig. 8 | Comparing silicone oil-coated mold and DNR mold for hydrogel surface wettability regulation.a Schematic of the hydrogel preparation process by using a silicone oil-coated mold or DNR mold.b WCA of the molds before and after hydrogel preparation.c WCA evolution, d Corresponding decline percentage of WCA, and e ATR-FTIR spectra of the PAA hydrogels prepared by silicone oil-coated mold (named silicone oil-coated hydrogel) and DNR mold.f WCA of the DNR mold after 20 uses for hydrogel preparation.Values in b-d and f are shown as the mean ± SD; n = 3 independent samples.Influence of mold surface roughness on hydrogel surface wettability.Average roughness (Ra) of a laser-polished glass mold measured by a roughmeter (W912C, Jenoptik).b WCA of a rough DNR mold, which shows higher hydrophobicity than a smooth DNR mold.c WCA evolution of hydrogels prepared by smooth or rough DNR molds.Values in c are shown as the mean ± SD; n = 4 independent samples.a ATR-FTIR and b XPS spectra of conventional and DNR hydrogels.To make the contrast of XPS curves more distinctive, the overall intensity of DNR hydrogel was adjusted to being comparable with the conventional hydrogel.High-resolution XPS O 1s spectra with deconvolution of polymer structures for conventional and DNR hydrogels.
high-resolution spectra (Binding energy/eV) Supplementary Fig. 12 | a Schematic of surface etching for XPS profiling of the DNR hydrogel surface at various depths.b XPS depth profile of the DNR hydrogel with depth from the top surface to ~60 nm.c Element contents of carbon and oxygen on the DNR hydrogel surface at various depths.The values were determined by integration of the C 1s and O 1s peak areas.d Highresolution XPS spectra of the DNR hydrogel depth profile.e Statistics of the element content distribution of polymer segments in the DNR hydrogel at various depths.

d
Supplementary Fig.13| a XPS depth profile of the conventional (Con.)hydrogel with depth from the top surface to ~60 nm.b Element contents of carbon and oxygen on the conventional hydrogel surface at various depths.The values were determined by integration of the C 1s and O 1s peak areas.c High-resolution XPS spectra of the conventional hydrogel depth profile.d Statistics of the element content distribution of polymer segments in the conventional hydrogel at various depths.
Bulk properties of conventional and DNR hydrogels.a Rheological frequency sweep, b Images showing transparency, c swelling ratio, and d water loss ratio of conventional and DNR hydrogels.Water droplets dyed by rhodamine were dropped on hydrogel surfaces to distinguish DNR hydrogel from conventional hydrogel and improve visibility.e Photo images of conventional and DNR hydrogels over time during the swelling test.Values in c and d are shown as the mean ± SD; n = 3 independent samples.
Regulation of the crosslinking degree of silicone chains using methyltrimethoxysilane (MTMS).The proportion of MTMS to DMDMS in the reactive silane solution is a 0 wt% MTMS, b 50 wt% MTMS, and c 100 wt%.WCA of the DNR molds prepared with different crosslinking degrees of silicone chain, which were regulated by the MTMS content.Data are shown as the mean ± SD; n = 4 independent samples.Statistical analyses were performed by using ordinary one-way analysis of variance (ANOVA) with Tukey's post hoc test.P values less than 0.05 were considered statistically significant differences among the compared groups.Supplementary Fig.17 |a XPS spectra of DNR hydrogels prepared by the DNR molds with different crosslinking degrees of silicone chain.b Element contents of carbon and oxygen on the DNR hydrogel surface.The values were determined by integration of the C 1s and O 1s peak areas in A. c High-resolution XPS spectra of DNR hydrogels prepared by the DNR molds with different crosslinking degrees of silicone chain.d Statistics of the element content distribution of polymer segments in the DNR hydrogels.
WCA of PAA hydrogels prepared by DNR molds with different lengths and densities of grafted silicone chains.The length of silicone chains was regulated by different condensation time of DMDMS from 0.5 h to 5 h, where the density was regulated by various concentration of DMDMS from 1 wt% to 20 wt%.Data are shown as the mean ± SD; n = 4 independent samples.Statistical analyses were performed by using ordinary one-way analysis of variance (ANOVA) with Tukey's post hoc test.P values less than 0.05 were considered statistically significant differences among the compared groups.

XPS spectra of the DNR mold with different chain densities of silicone chains.
The silicone chain was grafted onto a silicon wafer substrate, and the chain density was regulated by the DMDMS concentration (1 wt%, 5 wt%, 10 wt%, and 20 wt%).