Lubricant skin on diverse biomaterials with complex shapes via polydopamine-mediated surface functionalization for biomedical applications

Background

Implanted and blood-contacting devices are prone to foreign body response, thrombosis and infection, which the authors associate with the adhesion of macrophages, fibroblasts, bacteria and blood proteins. Superhydrophobic coatings rely on air pockets that can disappear by diffusion and require etching or plasma equipment, and the hydration layers of superhydrophilic polymers are thin and vulnerable to abrasion. Lubricant-infused surfaces provide durable repellency, but conventional versions depend on plasma-assisted silane functionalization, which cannot be distributed uniformly inside high-aspect-ratio polymer tubing, and the resulting self-assembled monolayer (SAM) cannot heal once scratched.

Design and approach

Substrates are immersed in dopamine hydrochloride solution containing CuSO4/H2O2 oxidants, which shortened PDA deposition from 24 h to 20 min and gave more even deposition. A carboxyl-terminated PFP is then applied and cured above 100 °C so that it forms amide and hydrogen bonds with the amine and catechol groups of PDA, and the PFP film is immersed in PFPE lubricant, which swells into its free volume. Because the lubricant is held by swelling, the coating retains lubricant without the hierarchical structures of conventional liquid-infused surfaces. Above its glass transition temperature (Tg about 100 °C) the amorphous PFP can reflow, which the authors use for heat-mediated self-healing, while lubricant replenishment restores repellency after abrasion.

How it was evaluated

Swelling was followed by quartz crystal microbalance with dissipation monitoring (QCM-D) and Young's modulus by atomic force microscopy. Contact angles were measured on nine materials: glass, silicone, polytetrafluoroethylene, polypropylene, polyvinyl chloride, stainless steel, polyethylene, polyethylene terephthalate and titanium. Adhesion of albumin, fibrinogen, NIH-3T3 fibroblasts and RAW 264.7 macrophages was compared between bare and L-skin glass, and cytotoxicity was assessed with fibroblasts in an extract assay. Coating adhesion was tested by an ISO 2409 cross-cut test against a commercial particle-based coating and PFP without PDA, heat-mediated self-healing was examined on L-skin, and cotton-swab abrasion and autoclave, ethylene oxide (EO) gas, gamma-ray (25 kGy) and oxygen plasma sterilization were compared with SAM-based lubricated surfaces. Device tests covered pipette tips, friction against an agar tissue phantom, bioprinting of NIH-3T3-laden bioink through 30 G, 32 G and 34 G nozzles, Rhodamine B absorption in PDMS microfluidic channels, 2-m-long tubing for circulation of citrated horse blood, and Foley catheters incubated with Escherichia coli. The study involved no animal or human subjects.

Key results

QCM-D showed that the coating, 35.7 nm thick before lubricant exposure, increased from 90.86 nm to 107.98 nm over 4 h in the lubricant, and its Young's modulus fell from about 100 MPa to about 100 kPa upon swelling and returned to the original value after the lubricant was dried. Static contact angle measurements indicated that the L-skin coating was effective on all nine materials regardless of their original surface characteristics. Albumin, fibrinogen, fibroblast and macrophage adhesion on L-skin differed significantly from that on bare substrates, with low adhesion on L-skin, and extract media from L-skin coated silicone produced no significant difference in fibroblast viability. Contact and sliding angles of L-skin on silicone kept in PBS at 37.5 °C showed negligible changes when measured daily.

In the cross-cut test, more than 60% of the particle-based coating detached after tape peeling and no PFP coating without PDA survived, whereas L-skin with PDA lost none of its blocks. In samples with coating thicknesses of about 235 nm and 400 nm, heating near Tg healed scratches about 1 μm wide, whereas larger scratches left exposed substrate; water droplets slid without pinning over healed scratches about 1.5 μm wide. After at least 20 cotton-swab strokes, droplets were pinned on the SAM surface but continued to slide on L-skin.

Contact angles showed negligible changes after nine autoclave cycles. After EO gas and gamma-ray sterilization, sliding angles increased on SAM surfaces but remained stable on L-skin, and after gamma irradiation the relative F 1s intensity decreased on SAM but not on L-skin. After oxygen plasma (100 W, 2 min), contact angles of both surfaces dropped, but only L-skin recovered to its original values.

L-skin pipette tips showed no nanoparticle fouling, and over 10 aspiration and dispensing cycles the slight mass mismatch seen with bare tips for ethanol and isopropyl alcohol did not occur; L-skin also gave the lowest friction impulse against the agar phantom among hydrophobic, hydrophilic and L-skin coatings. In bioprinting, viability with bare nozzles decreased to 89 ± 2.115% at 32 G and 72 ± 5.15% at 34 G, whereas changes were negligible with L-skin nozzles. Rhodamine B absorption widened the fluorescent band to almost twice the original channel width in bare PDMS, while L-skin channels showed no significant change after an hour. In 2-m-long tubing with an inner diameter of 1 mm, blood left trails in bare tubing but no residue in L-skin tubing, and the meniscus angle was about 41° in bare tubing and 98° in L-skin tubing. After 2 h of closed-loop circulation (3 L/h in the Methods and Fig. 6 legend; 2 L/h in the Results text), blood from bare tubing retained significantly fewer platelets and bare tubing produced more thrombus. Many bacteria adhered to bare catheters after E. coli incubation, whereas adhesion on L-skin catheters was low.

Open questions

The authors state that in vivo examination remains an unmet challenge for L-skin and that further work is required for applications such as chemoports, extracorporeal membrane oxygenation and artificial cardiovascular grafts. The heat-mediated self-healing decreased as scratches became larger, leaving exposed substrate after healing.

How it connects to other research

The slippery liquid-infused porous surfaces of Wong and colleagues (Nature, 2011) lock a lubricant into a textured substrate, whereas L-skin holds the lubricant by swelling a thin PFP film anchored to PDA, which allows coating inside tubing and on materials without surface texturing.