Design of the dual-functional smart-coating foil for orthopedic implants. Schematics (A) and optical image (B) showing the smart-coating foil with integrated biomimetic mechano-bactericidal and multiplexed strain-sensing functionalities for orthopedic implants (e.g., spinal rods for lumbar fusion). Scale bar, 20 mm.

Biomimetic mechano-bactericidal nanopillar arrays on the outer surface of the dual-functional smart-coating foil protecting orthopedic implants against bacterial infections. A) Schematics depicting the process to prepare polyimide foils featuring nanopillar arrays with tunable structures. (B to F) Scanning electron microscopy (SEM) images showing the critical steps: Top-view SEMs of the colloidal-crystal mask before (B) and after (C) oxygen RIE, and the metal hard mask formed after lift-off (D); cross-sectional SEM micrograph of the template after deep-silicon RIE (E); and tilted-view SEM showing the representative nanopillar arrays on the prepared polyimide foil (F). Scale bars, 400 nm. (G to J) Counts of viable E. coli (MG1655) after being incubated with 1-cm by 1-cm nanopillar arrays with varying nanopillar diameter (d), pitch (p), and height (h) and planar controls for 90 min. Asterisk indicates the number of colony-forming units (CFUs) was less than 1000. (K to N) Counts of viable MRSA (USA300) after being incubated with 1-cm2 nanopillar arrays and planar controls. p = 500 nm and d = 250 nm for (G) and (K); p = 500 nm and d = 100 nm for (H) and (L); p = 240 nm and d = 100 nm for (I) and (M); and p = 240 nm and d = 50 nm for (J) and (N). Zero height represents the planar foils as internal controls included in each experiment. N ≥ 4 in all experiments.

The optimum biomimetic nanopillar-array design ensuring high antimicrobial activities, structural robustness, and biocompatibility with mammalian cells. (A) AFM image of freshly prepared mechano-bactericidal polymer nanopillar arrays. (B and C) AFM scans of 800-nm-tall and 100-nm-diameter polymer nanopillar arrays immersed in water for the first time [(B) p = 500 nm as in part A; (C) p = 240 nm]. (D) AFM image of the nanopillar arrays shown in (A) and (B) after drying. (E to G) SEM micrographs showing the nanopillar arrays with different geometries after drying [(E) as in (C); (F) p = 240 nm, d = 100 nm, and h = 400 nm; (G) p = 240 nm, d = 50 nm, and h = 400 nm]. Scale bars, 1 μm. (H and I) The number of CFUs in the suspension after incubation with the planar controls (orange) and nanopillar arrays with the optimum geometry (purple) for E. coli strain MG1655 [(H) variable incubation time up to 3 hours, N = 3], S. aureus strain 29213, and P. aeruginosa strain 27853 [(I) incubation for 3 hours, N = 6]. (J to M) SEM showing the bacterial biofilms formed by S. aureus on planar (J) and nanostructured (K) foils and P. aeruginosa on planar (L) and nanostructured (M) foils, all after 48-hour incubation. The clustering of nanopillars was caused by the cell fixation protocol. Scale bars, 2 μm. (N) Mammalian cell viability on polymer nanopillar arrays (hatched) and planar controls (solid) after 48-hour incubation as determined in WST-1 (dark) and TBE (light) assays (N = 3). P values for unpaired t test between control and mechano-bactericidal foils are 0.5, 0.7, 0.08, and 0.1 in WST-1 assay and 0.7, 0.9, 0.5, and 0.5 in TBE test for human osteosarcoma cell 143B, MG63, melanoma cell C2C12, and mouse myoblast cell A2058, respectively.

Fig. 4. Multiplexed strain-sensing array on the inner surface of the dual-functional smart-coating foil in contact with the orthopedic implants. (A) Schematic of the multiplexed strain-sensing array. VDD, supply voltage; VSS, ground voltage; V+, plus-node output-sensing voltage; V−, negative-node output-sensing voltage; Vcol, column-selection voltage; Vrow, row-selection voltage. (B and C) Image [(B) scale bar, 100 μm] and circuit diagram (C) of a single strain-sensing pixel. (D) Output voltage difference (∆Vout) of a pixel under applied tensile strain of 0.05%, measured with different bias conditions of the selector transistors. Applied VDD = 1 V. (E) Optical image of a spinal rod coated with the smart-coating foil under four-point bending test. Scale bar, 50 mm. (F) Strain distribution along the spinal rod under 2-kN load as recorded by a 3 × 3 sensor array. (G) ∆Vout recorded by each pixel inside the 3 × 3 array with increasing load applied. Applied VDD = 5 V. (H) ∆Vout recorded by the three strain-sensing pixels in the central column during five consecutive loading cycles. (I and J) Images showing the spine cadaver mounted on the custom platform, without (I) or with (J) the bending moment applied. Scale bars, 50 mm. (K to M) Images showing the spinal specimen with the intact joint facets (K), after destabilization (L), and after applying bone cement (M). Scale bars, 10 mm. (N) Comparing the strain recorded by the strain-sensing pixels located on the dorsal side of the spinal rod for the spinal specimen with intact (orange), destabilized (green), and cemented (purple) facet joint and after pedicle screw loosening (yellow). (O) Strain on the spinal rod as measured by the pixels in the medial column of the array, located on the ventral (orange), lateral (blue), and dorsal (red) side of the implant, for the destabilized (solid) and then cemented (dashed lines) spinal specimen.

Fig. 5. In vivo authentication of antimicrobial performance of the smart-coating foils. (A) Schematic showing the mouse subcutaneous implantation model. (B to E) Images of the hematoxylin and eosin (H&E)–stained histologic sections showing the tissues surrounding the planar foils subject to S. aureus (B) or P. aeruginosa (D) challenges (scale bars, 50 μm), with the magnified views [(C) for S. aureus and (E) for P. aeruginosa; scale bars, 20 μm) highlighting the formation of microcolonies. (F to J) Tissues surrounding the smart-coating foils challenged with S. aureus under low [(F) scale bar, 500 μm] and high [(G) scale bar, 50 μm] magnifications or P. aeruginosa under low [(H) scale bar, 500 μm], medium [(I) scale bar, 200 μm], and high [(J) scale bar, 50 μm] magnifications, showing only mild neutrophilic inflammation and tissue damage but no evidence of intralesional bacterial colonies. (K to N) Comparisons of the S. aureus (K) and (M) and P. aeruginosa (L) and (N) burdens on the implanted planar controls (green, solid bars), the smart-coating foils (diagonal hatched bars), and their surrounding skin (orange) and muscle (blue) tissues after 3 days [(K) and (L) N = 4] and 2 weeks [(M) and (N) N = 5] in vivo, respectively. P values for unpaired t test with unequal variance between planar controls and smart-coating foils as well as their associated surrounding tissues are all less than 0.008.

Fig. 6. In vivo authentication of biocompatibility of the smart-coating foils. (A to D) Time evolution of the abundance of neutrophils (A), dendritic cells (B), macrophages (C), and T cells (D) in the tissues surrounding the implanted smart-coating foils (black), with the tissues collected from mice without receiving the surgery as control (red). N = 5. P values are determined from unpaired t test with unequal variance between controls and implants. (E to G) Histopathology of H&E-stained tissues surrounding the smart-coating foils 2 weeks (E), 4 weeks (F), and 8 weeks (G) (note that the foil was retrieved beforehand for the antimicrobial assays displayed in Fig. 7, D and E) after implantation. (H) Masson’s trichrome staining demonstrated a mild degree of peri-implant fibrosis, as evidenced by a thin layer of blue-stained collagen fibers surrounding the smart-coating foil 8 weeks after implantation. Scale bars, 50 μm.

Fig. 7. Authentication of long-term stability of the smart-coating foils in vivo. (A) ∆Vout of the silicon-nanomembrane Wheatstone bridge gauges subject to tensile strain of 0.1%, before (purple) and after 2 weeks (yellow), 4 weeks (blue), and 8 weeks (pink) in vivo. P value determined by one-way analysis of variance (ANOVA) is 0.2. (B) Current-voltage characteristics of the selector transistors before (black solid lines) and after 8 weeks in vivo (red dashed lines), measured with the whole device submerged in PBS. VDS, source-drain bias; IDS, source-drain current. (C) SEM micrographs showing the in vitro bacterial biofilm formation on the control planar films (top frames) and the mechano-bactericidal smart-coating foils retrieved after 8 weeks in vivo (bottom frames). Films were incubated with 105 to 106 CFUs of S. aureus (left frames) or P. aeruginosa (right frames) for 48 hours in a nutrient-rich medium. Scale bar, 5 μm. (D and E) The number of CFUs of S. aureus (D) or P. aeruginosa (E) after in vitro incubation with the planar controls (orange) or the smart-coating foils that were freshly prepared (0 weeks; purple) or those that were retrieved after subcutaneous implantation in mice for different periods of time, showing their quantitatively similar bactericidal efficacy. N ≥ 3. P values for unpaired t test with unequal variance between planar controls and smart-coating foils are all less than 0.0004.