Antibacterial Plasma Polymer Coatings on 3D Materials for Orthopedic Applications

Covalent biofunctionalization of implant surfaces using anti microbial agents is a promising approach to reducing bone infection and implant failure. Radical‐rich, ion‐assisted plasma polymerized (IPP) coatings enable surface covalent biofunctionalization in a simple manner; but until now, they are limited to only 2D surfaces. Here a new technology is demonstrated to create homogenous IPP coatings on 3D materials using a rotating, conductive cage that is negatively biased while immersed in RF plasma. Evidence is provided that under controlled energetic ion bombardment, this technology enables the formation of highly robust and homogenous radical‐rich coatings on 3D objects for subsequent covalent attachment of antimicrobial agents. To functionally apply this technology, the broad‐spectrum antimicrobial CSA‐90 is attached to the surfaces, where it retained potent antibacterial activity against Staphylococcus aureus. CSA‐90 covalent functionalization of stainless‐steel pins used in a murine model of orthopedic infection revealed the highly promising potential of this coating system to reduce S. aureus infection‐related bone loss. This study takes the previous research on plasma‐based covalent functionalization of 2D surfaces a step further, with important implications for ushering in a new dimension in the biofunctionalization of 3D structures for applications in bone implants and beyond.


Introduction
Bone infection is a major clinical problem and can be of exceptionally high risk for open fractures and surgeries. [1]lthough re-operation and debridement combined with systemic antibiotic treatment can often resolve an initial superficial infection, deep infections are often refractory to intervention. [2]Treating the resultant non-unions can be challenging and costly. [3,4]Orthopedic implants can act as a nidus for biofilm formation, making them particularly prone to infection.Joint replacement or total joint arthroplasty is a high-volume, high-cost surgery where preventable post-operative complications are significant cost drivers. [5]nfection rates are particularly high in joint replacement revisions, even when implants are being revised for complications not previously associated with infection.Hence, there is a growing trend toward establishing antimicrobial coatings for medical implants, particularly those implanted in the bone. [6]igure 1.Schematic illustration of a two-step process consisting of ion-assisted plasma polymerization to form radical-rich coatings on 3D materials followed by CSA-90 covalent biofunctionalization.The antibacterial efficacy of the coatings to reduce S. aureus infection-related bone loss was evaluated in a murine model of orthopedic infection.
][11][12][13][14] While bioactive organic agents can have high potency against infection, there remains considerable scope to develop new methods for surface attachment. [15]Simple adsorption methods can result in burst release and a lack of sustained protection.Further, physically attached molecules are susceptible to desorption or competitive replacement by other molecules in vivo. [16]Agents can also be embedded in soluble polymer or hydrogel matrices that can be used for surface coating.This approach can yield challenges with controlling degradation rates and the negative biological effects of breakdown products.19][20][21][22] Covalent attachment of bioactive molecules on surfaces has been traditionally achieved using methods relying on wetchemical steps. [16]Examples of these approaches include linker chemistry methods based on salinization, [23] PEGylation, [24] and heparinization. [25]However, these methods are typically substrate-dependent, meaning they apply to only a particular surface chemistry.For example, salinization, a commonly used chemical linker-mediated immobilization approach to biofunctionalize orthopedic substrates, is limited to only hydroxylated substrates bearing a high concentration of OH groups. [26,27]Further, the wet-chemical methods for covalent biofunctionalization are often time-consuming and complex, requiring multistep processes.For example, the processing time for linkermediated protein immobilization approaches is typically in the order of days. [16]Another undesirable aspect of such wetchemistry methods is the extensive waste produced in the process, making them environmentally questionable.In addition, many side reactions may occur during the multi-step processes, producing by-products that may reduce the overall reproducibility and introduce significant challenges to obtaining regulatory approvals. [16]lasma immersion ion implantation (PIII) is a promising method for covalently attaching a wide range of biomolecules [28][29][30][31] and hydrogels [32] to polymeric surfaces without the limitations mentioned previously.PIII produces radicalrich surfaces, which allow covalent anchorage of biomolecules on contact with no need of chemical linkers. [33]Direct covalent biofunctionalization through this technique is, however, limited to polymeric materials with a carbon-based backbone.
We have recently introduced ion-assisted plasma polymerization (IPP) that combines PIII and plasma polymerization as a versatile, alternative method to biofunctionalize surfaces, including those made of metals.[36][37] Pulse biasing the substrate during the IPP process facilitates the energetic ion bombardment of the coating as it grows, resulting in the formation of a high concentration of reactive radicals buried within the plasma polymer structure.These radicals, stabilized in pi-conjugated nanoclusters, migrate to the surface, where they enable the covalent attachment of a wide range of organic and inorganic bioactive compounds. [33]The covalent attachment in this process is independent of chemical-linker reagents.In the context of bone implants, a range of bioactive compounds has the potential to improve clinical outcomes.Among the bioactive compounds, ceragenins or cationic steroid antimicrobials (CSAs) would be suitable candidates for surface immobilization.CSAs are a class of small molecule agents that disrupt bacterial membranes and are bactericidal against Gram-positive and Gram-negative bacteria. [38]CSAs' mechanism of action is microbial membrane insertion, which leads to membrane depolarization.They act in a manner similar to antimicrobial peptides but have comparatively reduced cytotoxicity and greater in vivo stability. [39]42] IPP coating can be used to immobilize and present a variety of bioactive agents, including proteins such as BMP2, [36] peptides such as MEL4 and caspofungin, [35] and silver nanoparticles. [34]Despite all these intriguing potential applications of the IPP technology in biomimetic surface engineering; to date, it has been limited to only 2D objects such as titanium sheets and silicon wafers.However, almost all implantable biomedical devices, including dental screws, bone grafts, tissue engineering scaffolds, and neurostimulator implants, are 3D in shape.
Ion-assisted plasma polymerization, as well as plasma polymerization in general, are not as straightforward on 3D objects as they are on 2D surfaces.The challenge lies in achieving homogenous and robust coatings on 3D surfaces, as some areas may be shielded from plasma exposure or receive non-uniform coating due to the direction of exposure. [44]Thus, novel technologies with customized reactor designs and geometries are required.As schematically illustrated in Figure 1, here we present a new ion-assisted plasma polymerization process using a rotating cage made of an electrically conductive mesh that is negatively biased and rotates while immersed in RF plasma.We hypothesized that this strategy enables the deposition of radical-rich polymeric coatings onto 3D surfaces for their subsequent covalent biofunctionalization. IPP's efficacy as a surface modification method using CSA-90 was tested for antimicrobial potential using both in vitro assays and an in vivo model of stainless-steel pin infection.

Materials
Silica beads (1 mm diameter) were purchased from Sigma-Aldrich.Double-side polished silicon wafers (10 mm × 10 mm) were ultrasonicated in acetone and ethanol each for 10 min, then dried using a stream of nitrogen gas before film deposition.Stainless-steel (SS) Kirshner-wires (1.1 mm in diameter) cut to 1 cm segments, referred to as SS pins, were obtained from Zimmer, Warsaw, Indiana, USA.For surgery, stainless steel pins (size 000) were purchased from the Australian Entomological Supplies Pty Ltd (South Murwillumbah, Australia).CSA-90 was obtained from Prof Paul Savage (Brigham-Young University, Provo, Utah, United States).Approximately 6 ml of Milli-Q water was added to 9.6 mg of CSA-90 to produce a clear solution.

Ion-Assisted Plasma Polymerization on 3D Substrates
Ion-assisted plasma polymerized films were deposited on silica beads, stainless-steel pins and silicon wafer substrates using a retrofitted plasma polymerization system (Figure 2A) equipped with a rotating, cylindrical cage (Figure 2B).The plasma polymerization system, without the rotating cage, was described in detail previously. [35,45]The rotating cage, with an inner diameter of 4.3 cm and a height of 6.9 cm, was coupled to a pulse generator was installed into the side of the chamber to enable homogenous plasma polymerization of IPP films on the 3D objects, i.e., silica beads and stainless-steel pins.The cage contained a stainless-steel mesh outer lining supported by two inner rods.The plasma polymerization system was equipped with a radio frequency (RF) electrode and a DC-pulsed voltage source connected to the rotating cage.Approximately 450 silica beads, or 20 SS pins were used for each deposition.In each batch, a silicon wafer (10 mm × 10 mm) was attached to the inner rod of the rotating cage.The bias voltage pulses applied to the rotating cage were generated by a RUP-6 pulse generator (GBS-Electronik) at 3 kHz for a pulse duration of 20 μs.The cage rotation motor was set to 2.5 V, providing ten revolutions per minute.Prior to ionassisted plasma polymerization and once the chamber base pressure was below 5.0 × 10 −5 Torr, the silica beads or SS pins, silicon wafer, and the cage were cleaned using argon plasma (Ar flow rate = 40 standard cubic centimeters per minute (sccm)) at 75 W and -500 V pulsed bias applied to the cage for 10 min.Then, a mixture of 5 sccm acetylene and 25 sccm argon was injected into the chamber, and the pressure was adjusted to 110 mTorr.The pulse bias voltage was varied from 0 to 1000 V, while the RF input power was kept unchanged at 50 W.The polymerization time for each coating process was 15 min unless otherwise stated.

X-Ray Photoelectron Spectroscopy (XPS)
The surface chemistry of IPP-coated materials was analyzed using a SPECS FlexMod spectrometer within 24 h after deposition.The spectrometer was equipped with a hemispherical analyzer (PHOIBOS 150), an MCD9 electron detector, and a monochromatic X-ray source, operating at 10 kV and 20 mA (AlKa, hv = 1486.7 eV).The samples were mounted on the holder using double-sided, conductive carbon tape.Each sample was measured at a take-off angle of 90 degrees once the base pressure was below 5.0 × 10 −8 mbar.The survey spectra were obtained at a pass energy of 30 eV (0.5 eV resolution) over an energy range of 0-1000 eV.The carbon high-resolution (C1s) spectra were col-lected at pass energy of 20 and 0.1 eV resolution.Calculations of the atomic concentrations of elements and curve fittings of high-resolution spectra were carried out using CASA XPS software (version 2.3.14).

Time of Flight Secondary Ion Mass Spectrometry (ToF-SIMS)
ToF-SIMS data were obtained using a nanoTOF instrument (PHI TRIFT V, Chanhassen, MN) with a 30 eV pulsed liquid (79Au+) metal primary ion source (LMIG).All measurements were carried out in the positive mode of SIMS at the base pressure below 5 × 10 −6 Pa.Dual charge compensation was achieved by employing an electron flood gun and Ar+ ions at 10 eV.The raster size was recorded for at least six spots with areas of 100 μm × 100 μm per sample.WincadenceN software (version 1.8.1, Physical Electronics) was utilized for all spectral analyses.

Spectroscopic Ellipsometry
To estimate the cross-linking degree of the IPP coatings deposited on silicon wafers, their refractive indices were measured using a J.A. Woollam spectroscopic ellipsometer (EC-400 light source).Measurements were carried out at incidence angles of 65, 70, and 75 degrees and the data were analyzed using the WVASE32 software.All measurements were taken within a wavelength range of 200-1000 nm with 5 nm steps.A Cauchy model was applied to obtain a fit of the data to calculate the thicknesses and refractive indices.

Electron Paramagnetic Resonance (EPR) Spectroscopy
An ADANI SPINSCAN X EPR device was used to assess the concentration of radicals embedded in the coatings on the IPP-coated silica beads.The silica beads were placed in a 3 mm diameter quartz tube and secured with a Teflon holder ≈9.5 cm from the middle of the sample cavity.The central magnetic field was set to 336 mT, and the modulation amplitude was 200 μT.The data from an average of 10 scans were reported.

Stability Evaluation in Tyrode's Simulated Body Fluid (SBF)
Tyrode's solution was used as a biologically relevant medium to evaluate the robustness of the IPP coatings deposited on silica beads.IPP-coated silica beads were placed in a falcon tube and covered with 0.5 ml of Tyrode's solution with the chemical composition listed elsewhere. [45]After one month (at 37 ± 1°C), the SBF solution was removed, and each sample was rinsed with 1 ml of MilliQ water three times, followed by drying with a stream of nitrogen gas.XPS spectra of the washed samples and SEM images were obtained within 24 h.

Scanning Electron Microscopy (SEM)
The physical stability of the IPP-coated silica beads incubated in SBF solution (1 month at 37 ± 1 °C) was evaluated using SEM micrographs.The SEM images were obtained using a Phenom Table-top SEM at a vacuum pressure of 60 Pa, an acceleration voltage of 10 kV and a working distance of 7 mm.

AF488 Antibody Covalent Attachment and Fluorescence Imaging
To confirm attachment and demonstrate the homogenous distribution of attached biomolecules on the IPP-coated 3D materials (silica beads and SS pins), Alexa Fluor 488-conjugated IgG antibody (4 μg ml −1 , Abcam, USA) was used as an example biomolecule.The materials were incubated in the AF488 antibody solution for one hour at room temperature.Prewash images were taken by imaging the surfaces using an upright fluorescence microscope (Zeiss Z1, Oberkochen, Germany) with a 5× objective and an exposure time of 200 ms.Samples were then individually transferred to 50 mL falcon tubes filled with 5% sodium dodecyl sulfate (SDS, Sigma-Aldrich, USA) in sterile MilliQ water and allowed to rotate for 4 h at room temperature to desorb unbound AF488 antibody from the implant surface.The samples were then reimaged using the same imaging protocols to obtain post-wash images.

Bacterial Culture
Patient-derived Staphylococcus aureus (American Type Culture Collection-12600) stored at −80 °C in glycerol stocks was grown overnight on lysogeny broth (LB) agar plates at 37 °C, and single colonies were picked for culture in lysogeny broth (LB) the day prior to surgical inoculation.Bacteria were quantified using a spectrophotometer (Cary 300 UV-vis, Agilent, Las Vegas, NV) at 600 nm with an optical density of 1 representing 1 × 10 9 colonyforming unit (CFU)/milliliter (mL).Colonies were picked the day before surgery and enabled 12 h growth in LB broth to ensure accurate quantification of live/active bacteria from broth culture for inoculation.

Beads and Antimicrobial Coating In Vitro Assay
Silica beads (1 mm) were modified by various coatings: IPP coating only, IPP coating with CSA-90, CSA-90 without IPP coating, and IPP coating with CSA-90.For the CSA-90 coating processes, 60 uncoated or IPP-coated silica beads were covered for 12 h at room temperature with 300 μl of CSA-90 solution (1 mg ml −1 ).The CSA-90 solution was removed, and the samples were rinsed with copious amounts of MilliQ water three times, followed by washing in 5% sodium dodecyl sulphate (SDS) solution at 70 °C for 1 h.The coated silica beads were then placed in either LB with 10 3 CFU of S. aureus (Figure 2C -left) or sterile LB after inoculating 10 4 CFU of S. aureus on the surface and air-dried for 10 min (Figure 2C-right).

Covalent Attachment of CSA-90 on Stainless Steel Pins
For in vitro validation of the coating on metallic implants, stainless steel pins (Zimmer, Warsaw, Indiana, USA; 1 cm in length and 1.1 mm in diameter) were used as substrates.For surgery, stainless-steel pins (Australian Entomological Supplies; size 000) were used.The stainless steel pins were coated by IPP as described in 2.2 prior to incubation in 1 mg mL −1 CSA-90 solution in a petri dish for 12 h on a shaker.The pins were then water-washed for three times and air-dried in an incubator (37 °C) overnight and stored at room temperature.

Animal Ethics and Study Design
C57BL/6 12-week-old female mice (n = 30) were purchased from the Australian BioResources (Moss Vale, NSW, Australia).They were group-housed 5-6 per cage with access to food and water ad libitum.Mice could acclimatize for a week prior to surgery.Animal work was approved by the local Animal Ethics Committee (K339) and carried out in accordance with the Australian Code for the Care and Use of Animals for Scientific Purposes (2013).Prior to surgery, mice were randomly assigned to 3 groups (n = 10 per group) to be surgically implanted with stainless-steel pins that were i) uncoated, ii) coated by ion-assisted plasma polymerization (IPP), or iii) coated by IPP and subsequently incubated in CSA-90.

Orthopedic Surgical Model
Surgical anesthesia was induced with intraperitoneal ketamine (75 mg kg −1 ) and xylazine (10 mg kg −1 ) and maintained with  inhaled isoflurane (2-3% per 1.5-2L oxygen) as required.The right leg of each animal was shaved and treated with a topical povidone-iodine solution before surgery.
A medial parapatellar approach was used to access the right proximal tibia.A hole (0.5 mm in diameter) was made at the right tibial metaphysis (below the growth plate) using a surgical drill (Stryker 5100-15-250 Straight, Kalamazoo, USA), exposing the medullary canal adjacent to the drilled hole for bacterial infection.A stainless-steel pin was inserted through the subchondral bone at the knee, adjacent to the hole defect.After that, 1 × 10 6 CFU S. aureus in 5 μL was injected directly into the drilled hole with a Hamilton syringe and needles (Hamilton Company, Nevada, USA) immediately after the pin insertion.The incision was closed with 5-0 Vicryl (Ethicon LLC, Puerto Rico, USA), and no dressings were applied to the wound.Baseline radiographs were taken at the time of surgery.Animals recovered on a heated pad after surgery and were given normal saline (200 μL) subcutaneously for rehydration.Buprenorphine (0.1 mg kg −1 ) was given subcutaneously one hour prior to surgery and then every 12 h for three days for post-operative analgesia.
Animals were monitored daily by experienced staff and had twice-weekly radiographs performed under anesthesia (inhaled isoflurane) using digital X-ray (Faxitron Bioptics, Tuscan, AZ) at 25 kV (5 s) with ×2 magnification.X-ray images were assessed by orthopedic surgeons and veterinarians blinded to treatment.To minimize animal pain and distress, animals showing overt physiological and/or radiological evidence of infection judged by declining overall health (loss of body weight, lethargy, pyrexia, poor coat condition, non-weight bearing, and inflammation of the surgical site) or radiological evidence of worsening infection (localized osteolysis at the tibia and joint) were prematurely euthanized to avoid sepsis.The remaining mice were euthanized by carbon dioxide inhalation at three weeks postoperatively.

Sample Collection and Analysis
A biopsy of the soft tissue adjacent to the bone defect was taken after the incision under aseptic conditions.The surgical pin was pulled out from the joint by a sterile needle holder.All specimens were placed in Luria-Bertani (LB) broth (1 mL) for bacterial culture.A bone swab was collected at the defect site and resuspended in the LB broth for 15 s.Pus samples (if present) were also collected by swab for bacterial culture.The right femora and tibiae were harvested.The bacterial culture was incubated overnight at 37 °C.Positive and negative results were determined by the media turbidity and quantified using a plate spectrophotometer (SpectraMax iD3, Molecular Devices, San Jose, USA) at 600 nm.
For micro-CT, the right tibiae were fixed in 10% formalin for 24 h and transferred to 70% ethanol before being scanned with a SkyScan 1272 micro-computed tomography (micro-CT) scanner (SkyScan, Kontich, Belgium).All samples were scanned in 70% ethanol-soaked kimwipe at 50 kV and 200 μA using a 0.5 mm aluminium filter.Images were scanned at a pixel resolution of 9 μm, reconstructed with NRecon, straightened using DataViewer and analyzed with CTAn software (SkyScan).A global threshold to define bone tissue was set at 0.4 g cm −3 calcium hydroxyapatite, calibrated using two phantom samples of a known density.Bone morphometric outcomes included bone volume (mm 3 ), tissue volume (mm 3 ), and bone tissue mineral density (g cm −3 ).3D reconstructions were generated using CTVox software (Skyscan).

Decalcification and Paraffin Histology
The tibiae were decalcified in 0.34 m EDTA (pH 8.0) solution at 4 °C on a shaker for two weeks with solution changes twice a week.Samples were next embedded in paraffin and sectioned coronally through the tibial drilled hole at a thickness of five microns.Mounted sections were stained with hematoxylin and eosin (H&E) to differentiate bone and show the bone defect region.

Statistical Analysis
No data pre-processing techniques such as normalization or outlier exclusion were employed.All data in our charts were presented as mean ± SEM.The sample size (n) for each statistical analysis is specified in the figure legends.All statistical tests performed were two-sided.Statistical power calculations and analyses were performed using GraphPad Prism (La Jolla, California), and the cut-off for significance was set at p < 0.05.The micro-CT data from the in vivo study were analyzed using a Kruskal-Wallis test for multiple groups, followed by a nonparametric post-hoc Mann-Whitney U test to compare individual groups.

Ion-Assisted Plasma Polymerization on 3D Objects
Plasma polymerization has been predominantly applied to coat planar substrates with negligible to moderate curvatures.
However, almost all structures of interest for biomedical engineering applications, such as implantable medical devices, are 3D.To deposit robust and homogeneous IPP coatings on 3D surfaces, we exploited a new strategy using an electrically conductive mesh connected to a high-voltage power supply that rotates while immersed in the RF plasma of a polymerizable gas.We hypothesized that this novel design would permit homogenous deposition of radical-rich IPP coatings on 3D objects while also accelerating plasma ions toward the 3D objects, thus achieving strong substrate-coating adhesion and significant concentrations of radicals embedded in the coating.
To verify this hypothesis, we initially used silica beads as simple, model 3D substrates and evaluated the role of the applied bias voltage (V b ) in producing the IPP coatings for subsequent covalent attachment of biomolecules.Figure 3A shows the XPS surface chemical composition of IPP films as a function of pulsed bias voltage (V b ).The uncoated silica beads show surface atomic concentrations of ≈25%, 48%, and 26% for silicon, oxygen, and carbon, respectively.XPS results showed that by the deposition of IPP coatings on the beads, the atomic concentration of carbon increased to (80 ± 3)%, while those of silicon and oxygen decreased to (5 ± 1)% and (13 ± 2)%, respectively.Such changes in the surface chemistry indicate the presence of IPP coatings on the surfaces, resulting from successful ion-assisted plasma polymerization of the acetylene and argon mixture on the beads.The deposited IPP coating reduces the silicon and oxygen signals originating from the underlying substrate, thus resulting in a drop in their atomic concentrations.
To further evaluate the chemistry of IPP coatings, we curvefitted the C1s high-resolution spectra, as shown in Figure 3B.Three peaks associated with C─C/C─H at binding energy (BE)∼ = 284.6 eV, C─O at BE∼ = 286.5 eV, C═O at BE∼ = 287.5 eV, COOH at BE∼ = 289 eV were fitted in the C1s high-resolution spectra with their area percentage values plotted in Figure 3C.As the bias voltage increases, the C1s peak becomes narrower.The highest concentration of C1 compounds and the lowest concentration of C2 and C3 groups are observed for the coating prepared using the bias voltage of 0 V.These results suggest that a higher concentration of oxygenated carbon-containing moieties is formed on the surfaces for higher applied bias voltages.Such an increase in the concentration of oxygenated groups at higher bias voltages is explained by the greater ion bombardment that occurred on the growing film, resulting in the formation of a high concentration of radicals susceptible to post-deposition oxidation.The post-deposition oxidation of plasma polymer films, also referred to as auto-oxidation, [46] is an inevitable process that has also been previously observed for other hydrocarbon precursor monomers such as 1,7-octadiene [47] and thiophene. [48,49]e used electron paramagnetic resonance (EPR) spectroscopy to evaluate the differences in the number of electron spins, i.e., the concentration of radicals created in the IPP coatings.The EPR spectra of the IPP coatings deposited using the Vb values of 0, −500, and −1000 V are plotted in Figure 4A.The EPR spectra, measured ten days post deposition, are broad and show single resonance peaks.The resonance peak intensity increases by increasing the applied bias voltage, indicating that higher concentrations of radicals are embedded within the IPP structure.During the growth of plasma polymers, the film-forming compounds on the surface are continuously bombarded by photons, electrons, as well as positively charged ions. [50]These interactions between the growing coating and the reactive species present in the plasma phase produce radicals that can be trapped within the coating structure.In the case of IPP coatings polymerized using a rotating cage, higher bias voltages applied to the cage result in greater degrees of energetic ion bombardment onto the 3D materials.This strategy, in turn, increases the fragmenting chance of acetylene molecules, bond cleavage, and chain scission, resulting in the formation of higher concentrations of reactive radicals within the growing IPP coating.Using 2D substrates, such as titanium and stainless steel, we have previously shown that such radicals are mobile and migrate from sub-surface reservoirs to the surface in a thermally activated manner, where they facilitate the covalent attachment of bioactive molecules. [51]The EPR results shown here underpin the importance of bias voltage applied to the rotating cage in producing high concentrations of radicals for subsequent covalent biomolecule functionalization onto 3D objects.
The chemical stability of the IPP coatings is an essential factor for their application in the surface engineering of biomedical devices, particularly for orthopedic implants.Extensive oxidation of a polymeric coating once exposed to aqueous media can result in the deterioration of its integrity. [52]The degradation rate of a polymeric coating exposed to biological media depends on the diffusion rate of oxygen and water molecules throughout the film thickness and the availability of reactive sites for the formation of metastable compounds during the early stages of oxygen uptake. [50]To investigate the role of ion bombardment on the stability of the IPP films deposited on silica beads, we examined their XPS surface chemistry after incubation in Tyrode's Simulated Body Fluid (SBF) at 37°C for one month.This duration is sufficiently long in the context of bone implant applications, as the aim of the CSA90-IPP coating system is to prevent complications related to surgical site infection, which arise immediately after an orthopaedic surgery.[55][56] The XPS elemental composition of the coatings after SBF incubation is shown in Figure 5A.These results indicate that the coating deposited at the highest voltage bias of -1000 V shows the lowest chemical stability as indicated by the highest variations in its surface chemistry after SBF incubation, where the most significant increase in oxygen and decrease in carbon atomic concentrations are observed.These changes in surface chemistry are consistent with the peak-fitted XPS high-resolution C1s spectra (Figure 5B) and the changes in the calculated area percentage of various carbon-containing components presented in Figure 5C.The XPS data show that the most significant changes in the concentrations of C═O and COOH groups are observed for the coating deposited at -1000 V.The higher oxidation rate observed for the IPP coating deposited at -1000 V is explained by its greater content of embedded radicals, as indicated by EPR spectra (Figure 4A).This observation is also in good agreement with other works in which it has been shown that plasma polymer films containing higher concentrations of radicals suffer from a greater degree of oxidation and hydrolysis. [50,57,58]n addition to the initial density of radicals embedded within the IPP structure, the cross-linking degree of the coating is also an essential factor that can regulate oxidation kinetics.To evaluate the cross-linking degree of the IPP coatings deposited at varied bias voltages, we used spectroscopic ellipsometry and measured their refractive indices at 630 nm, plotted in Figure 4B.The refractive index of a polymer is correlated with its degree of crosslinking and density. [59,60]The coatings deposited using bias voltages of 0 and -500 V showed refractive indices of ≈1.6, whereas a refractive index of ≈1.7 was achieved for the coating deposited using a bias voltage of ─1000 V.The higher refractive index measured for the ─1000 V IPP coating is attributed to the higher fluxes of energetic ions arriving at the substrate, thus, greater fragmentation and recombination of deposited species, yielding highly cross-linked structures.In previous work carried out on flat titanium and silicon wafer surfaces, we observed that the oxidation kinetics could be moderated in an extremely dense and highly cross-linked IPP structure with an n value of as high as 1.8. [61]Such highly cross-linked structures limit the oxygen diffusion into the surface and the mobility of structural elements, including the secondary radicals, e.g., C─O─O., generated from oxidation.In the current work on 3D substrates, however, a maximum refractive index of 1.7 was achieved, leaving the high concentration of reactive radicals as the primary factor for regulating the oxidation kinetics.
All the IPP coatings deposited using negative bias voltages of 0 -1000 V showed excellent physical stability upon incubation in the SBF solution, as indicated by SEM micrographs obtained before and after incubation, shown in Figure 6A,B.From the SEM images, no evidence of physical failure in delamination, cracking or buckling was witnessed, indicating that sufficient adhesion between the IPP coatings and the underlying substrate has been achieved.Informed by these surface characterization results and based on the criteria of chemical stability and concentration of embedded reactive radicals, we used the IPP coating deposited at V b = −500 V for all subsequent experiments.
To evaluate the influence of ion-assisted plasma polymerization time on the surface chemistry, and homogeneity of the coatings, we varied the deposition time from 0 to 20 min while the bias voltage was kept constant at -500 V.The XPS atomic concentrations of uncoated and IPP-coated silica beads as a function of polymerization time are shown in Figure 7A.An increase in the polymerization time resulted in an increase in carbon atomic concentration and a decrease in the concentrations of silicon and oxygen.The increase of carbon atomic concentration versus deposition time is due to the polymerization and deposition of more hydrocarbon fragments from the acetylene precursor onto the silica bead surfaces.The decrease of silicon and oxygen concentrations, on the other hand, is due to the reduction of the signals originating from the underlying substrate.For a polymerization time of 25 min, no contribution of silicon in the XPS surface chemistry was detected, indicative of the formation of IPP coating with a thickness larger than the sampling depth of XPS that is 8 -10 nm. [62]The absence of silicon signals also suggests that the IPP coating deposited for 25 min is conformal and continuous, fully concealing the silica bead substrate.While no oxygen was present in the precursor gas mixture (acetylene + Ar), ≈8% oxygen was measured on the surface of this sample due to autooxidation, as previously explained.
The changes in XPS surface chemistry correlate well with ToF-SIMS normalized positive counts obtained for IPP coatings deposited at various deposition times, as shown in Figure 7B.The ToF-SIMS data demonstrate that by increasing the polymeriza-tion time, the relative counts of hydrocarbon species increase, whereas those of Si decrease.The major changes in ToF-SIMS positive counts are recorded for deposition times of up to 15 min, with no marked changes observed by increasing the deposition time to 25 min.These changes in surface chemistry indicate that at polymerization times longer than 15 min, the IPP coatings formed on the silica beads are thicker than the sampling depth of ToF-SIMS, which is in the range of 1-2 nm. [63]hese changes in surface chemistry can also be visually observed from the ToF-SIMS imaging data shown in Figure 7C.The increase of C2H3 + counts, shown here as representative hydrocarbon species present on the IPP-coated surfaces, is evident from these ion distribution maps.The uniform distribution of C2H3 + counts on these surfaces indicates that ion-assisted plasma polymerization using the negatively biased rotating cage is an effective approach to achieve homogenous coatings on 3D objects.
Figure 7D shows thicknesses of IPP coatings polymerized on silicon wafers as a function of deposition time obtained for asdeposited and incubated in SBF (at 37 °C for 1 month) conditions.These results provide further evidence on the stability of the coatings regardless of the deposition time, as no significant differences are observed in the thicknesses of the coatings measured before and after incubation is SBF.Altogether from the XPS, ToF-SIMS and thickness results, we chose the ion-assisted plasma polymerization time of 25 min for all the subsequent experiments to ensure sufficiently thick coatings are formed on the 3D substrates including stainless-steel pins.

Covalent Bio-Functionalization of IPP-Coated 3D Objects
Fluorescently labelled IgG was used as an example molecule to image the homogeneity of covalent biofunctionalization achieved on silica beads and stainless-steel pins.Detergents such as SDS remove physically attached molecules but cannot remove covalently attached molecules. [64,65]Washing the surfaces with SDS is a well-known technique used to test for the covalent nature of macromolecular attachment. [66]The images shown in Figure 8, therefor suggest that homogenous covalent attachment of fluorescently labelled IgG onto the 3D surfaces has been achieved.The data show that the molecules were resistant to SDS detergent washing in the IPP-coated group, whereas most of the physically adsorbed IgG was removed from the uncoated surfaces.The results also indicate homogenous deposition of IPP coatings on the 3D objects, as in the absence of homogenous coatings on the surfaces, homogenous distribution of covalently attached molecules could not be observed in the images.

CSA-90 Functionalization
To functionally assess the antimicrobial effects of antimicrobial treatments, coated silica beads were exposed to S. aureus.The inoculated bacteria were allowed to air dry on the surface, and then beads were placed in a solution of nutrient media.As controls, plasma-polymer (PP)-coated beads alone and unwashed PP + CSA-90 coated beads were included, giving positive and negative signals for bacterial growth.Beads where CSA-90 was adsorbed to the surface without plasma polymerization and then washed with a sodium dodecyl sulfate (SDS) solution did not impair bacterial growth, suggesting that CSA-90 was rapidly lost when washed.In contrast, PP + CSA-90 beads washed with SDS still showed potent antimicrobial activity in this assay (Figure 9A).These results agree with the SDS washing of fluorescently labelled IgG (Figure 8) and indicate that covalent attachment of CSA-90 onto IPP-coated surfaces is achieved as facilitated by surface-embedded radicals (Figure 4A).We note that the incubation times and detergent washing protocol, including the temperature and agitation conditions, for these two experiments were different as they needed to be adjusted according to the structure and geometry of the objects to ensure a thorough wash is achieved.
Next, 1 cm stainless-steel pins (analogous to those used in later mouse surgery) were tested for antimicrobial activity with and without PP + CSA-90 coating.Even with a higher inoculation dose of S. aureus, PP + CSA-90 coated pins showed a potent antimicrobial activity when grown in nutrient broth (Figure 9B).

Antibacterial Activity in an Infected Fracture Model
A preclinical murine orthopedic infection model was performed, where a localized bone defect was made in the proximal tibia and a stainless-steel pin was inserted into the intramedullary space as a nidus for biofilm formation and infection.Prior model development showed that a foreign surface adjacent to the bone defect (i.e., a metal pin) is required for reliable and progressive infection. [67]n the control group without any infection control measures, 8/10 developed pyogenic infections (pus present) and 8/10 were positive for bacteria by swab test at the study endpoint.Those with PP-coating only (no CSA-90) had 5/10 with pus, but 9/10 were positive for bacteria by swab test.The PP + CSA-90 group showed 5/10 with pus (p = 0.35) and 4/10 with bacteria by swab test (p = 0.16).While this did not reach statistical significance, bone loss imaged by radiography (Figure 10) and micro-CT (Figure 11A) suggested worse infection in the group featuring uncoated pins.Indeed, the quantification of bone loss in the drilled hole defects showed a more significant amount of regen-erated bone with PP + CAS-90 coating compared to no coating, and it is higher than PP-coating alone (Figure 11B).
Semi-quantitation measurements of bacterial load from the infection site grown in nutrient broth were compared at the bone defect (swab), the cultured removed pin, and soft tissue excised adjacent to the defect.Across the specimens, there was a trend toward reduced bacteria densities with no coating > PP > PP + CSA-90.From the local bone swabs, PP + CSA-90 showed significantly fewer bacteria than uncoated (p < 0.05).From the soft tissue sample cultures, PP and PP + CSA-90 both showed significantly fewer bacteria than uncoated (p < 0.01) (Figure 12).
To complement the micro-CT analysis, descriptive histology was performed on specimens from each group.In the nocoating group, the infected drilled hole persisted (Figure 13A), and the defect in the marrow was infiltrated with inflammatory macrophages and lymphocytes (Figure 13B; yellow arrows).All groups showed evidence of new woven bone formation in the metaphyseal bone at or adjacent to the defect site (Figure 13A,C,D; black arrows).Specimens that featured CSA-90 delivered by PP-coating showed no histological evidence of persistent infection and showed superior bone repair (Figure 13D).
CSAs are a class of small molecules with antimicrobial effects.The mechanism is associated with direct disruption of the negatively charged membrane molecules of bacteria, including phosphatidylglycerol and lipid A lipopolysaccharides, which then lead to bacterial cell membrane permeabilization and depolarization. [38]The retention of CSA activity after attachment to IPP-coated surfaces implies that this mode of action is retained; suggesting a surface presentation that allows interaction with the bacterial membrane.However, the precise nature of the presentation and mechanistic biology has scope for further research.Taken together, the in vitro and in vivo results suggest that CSA-90 molecules retain such activity after presentation on the IPP-coated surfaces.
Here we demonstrated the efficacy of the newly developed ion-assisted plasma polymerization technology in surface engineering of small-scale 3D objects used as implantable medical devices.Examples of such devices include cardiovascular stents, dental screws, and bone grafts.The size of the rotating cage could, however, be scaled up to accommodate larger-scale   implantable devices such as total hip and knee replacement implants.The economical and feasibility considerations for a scaledup process based on plasma polymerization are described by Yasuda et al. [68] In summary, our findings suggest that radical-rich IPP coatings on 3D objects, fabricated through a new technology presented here, hold great potential to open new avenues in simple surface biofunctionalization of antibacterial implantable medical devices, in particular bone implants.

Conclusion
We have developed an ion-assisted plasma polymerization (IPP) technology to generate radical-rich, organic coatings for sing-step covalent functionalization, which has numerous potential applications for developing biomimetic and antimicrobial interfaces on real-world 3D objects.We demonstrated that this technology, previously adapted for 2D surfaces, [35,61] can be re-engineered to modify 3D materials.Notably, using a combination of spectroscopic data and fluorescent labelled antibody attachment and imaging, we provided evidence that the IPP coatings are uniformly formed on 3D materials and are robust, with no evidence of failure after incubation in a simulated body fluid at 37 °C for up to one month.The reactivity, uniformity and robustness of these coatings make them ideal for the surface engineering of biomaterials in various applications.In this study, we showcased one application and demonstrated its excellent potential in creating antibacterial 3D interfaces using CSA-90 antimicrobial agent as a test molecule.The CSA-90-functionalized pins were permissive for healing in an infected bone defect model.These findings support the potential of this dry and environmentally friendly plasma technology to affix and retain antimicrobials on implants, mainly where antibacterial prophylaxis is advisable.Considering the substrate-independent nature of the IPP process, it can be applied to modify other 3D objects used in tissue engineering and regenerative medicine applications, including implantable medical devices.
(DP190103507).The authors acknowledge the facilities, scientific and technical assistance of Microscopy Australia at the University of South Australia, a facility that is funded by the University of South Australia, the State and Federal Governments.
Open access publishing facilitated by The University of Newcastle, as part of the Wiley -The University of Newcastle agreement via the Council of Australian University Librarians.

Figure 2 .
Figure 2. A) Schematic illustration of a retrofitted plasma polymerization system equipped with B) a negatively biased rotating cage to create IPP coatings on 3D materials.C) Schematic illustration showing the in vitro assay used for testing the antimicrobial properties of coated glass beads.The coated silica beads were placed in either lysogeny broth (LB) with 10 3 CFU of S. aureus (left) or sterile LB after inoculating 10 4 CFU of S. aureus on the surface and air-dried for ten minutes (right).

Figure 3 .
Figure 3. XPS Surface chemistry of IPP coatings deposited on silica beads as model 3D substrates A) XPS elemental composition of uncoated and IPP-coated silica beads as a function of bias voltage (N = 3 per group).B) XPS C 1s high-resolution spectra of IPP films.The spectra are curve-fitted using three components: C1: C─C/C─H, C2: C─O, C3: C═O, and C4: COOH.C) Area percentage for carbon-containing components (C1 -C4) fitted in C1s high-resolution spectra as a function of bias voltage (N = 3 per group).

Figure 4 .
Figure 4. A) EPR spectra obtained for IPP coatings deposited on silica beads using bias voltages of 0, −500, and −1000 V. B) Refractive indices of IPP coatings (at 630 nm) obtained using spectroscopic ellipsometry (N = 3 per group.The coatings were deposited on silicon wafers using various bias voltages as indicated.

Figure 5 .
Figure 5. Surface chemistry of IPP coatings after incubation in SBF at 37 °C for 1 month.A) XPS elemental composition of IPP-coated silica beads as a function of bias voltage after incubation in SBF (N = 3 per group.B) XPS C 1s high-resolution spectra of IPP films after incubation in SBF solution.The spectra are curve-fitted using three components: C1: C─C/C─H, C2: C─O, C3: C═O, and C4: COOH.C) Area percentage for carbon-containing components (C1 -C4) fitted in C1s high-resolution spectra as a function of bias voltage.(N = 3 per group).

Figure 6 .
Figure 6.SEM images of IPP coatings deposited on silica beads using applied bias voltages of 0, −500, and −1000 V before and after 1 month incubation in SBF.Scale bar = 200 μm in (A) and 30 μm in (B).

Figure 7 .
Figure 7. A) Surface chemistry of the IPP-coated silicon beads deposited at −500 V as a function of deposition time (N = 3 per group).B) Surface chemistry results from ToF-SIMS showing normalized Si and hydrocarbon positive counts for various deposition times (N = 6 per group).C) ToF-SIMS distribution maps of C2H3+ obtained for IPP coatings prepared using various plasma polymerization times.D) Thickness of the IPP coatings measured using spectroscopic ellipsometry before and after incubation in SBF at 37C.The coatings were fabricated on silicon wafers using various deposition times as indicated.(N = 3 per group).

Figure 8 .
Figure 8. Alexa Fluor 488-conjugated IgG antibody on uncoated and IPP-coated silica beads and pins before and after SDS wash.Scale bar = 500 μm.

Figure 10 .
Figure 10.Radiographs (X-ray images) A-C) showing osteolytic bone loss in infected tibiae with A) uncoated pins and bone healing in select tibiae from groups with B) PP coated pins and C) PP + CSA-90 coated pins.Radiographs show images before (left) and after (right) pin removal.The white arrows point to the osteolytic lesions of the tibial drilled hole due to infections.Scale bar = 6 mm.

Figure 11 .
Figure 11.A) Micro-CT reconstructions.Scale bar = 1 mm.B) Micro-CT data showing bone volume (mm 3 ) in the tibial drilled hole; N = 10 per group with Kruskal-Wallis test with post-hoc multiple comparisons (Mann-Whitney U).There was significantly improved defect repair in the PP + CSA-90 group (p < 0.05*).BV was increased in the PP group, but this did not reach statistical significance (p = 0.319, no coating versus PP; and p = 0.4313, PP vs PP + CSA-90).The white arrows point to the non-union osteolytic lesions of the tibial drilled holes.

Figure 12 .
Figure 12.Semi-quantitative analysis of relative bacterial load (absorbance at 600 nm) from A) bone swabs, B) intramedullary steel pins, and C) soft tissue adjacent to the defect site.N = 10 per group with Kruskal-Wallis test with post-hoc multiple comparisons (Mann-Whitney U).Error bars show mean and SEM; (* p-value ≤ 0.05; ** p-value ≤ 0.01).