Considering a coating with antibacterial ability, the first candidate would be Ag-containing coating since Ag is a well-known antibacterial element. For example, the silver nanoclusters-silica composite coating with 0.4 at. %
Ag showed antibacterial ability and improved bacterial anti-adhesion performance [1
]. Another study on a magnetron-sputtered Zr-Cu-Ag composite coating found it showed good mechanical performance, high chemical inertness, and good antimicrobial ability [2
In the literature, one can find some, but not much, researchers have endeavored to study the sputtered coatings against bacteria. Groessner-Schreiber et al. [3
] found that the microflora and Streptococcus
adhesion can be reduced by TiN and ZrN coatings. A study by Huang et al. [5
] proved that the TaN coating possessed antibacterial ability against Staphylococcus aureus
), with a conclusion due to the TaN coating being with high hydrophobic surface. In a study by Julius Andrew et al. [6
], the magnetron-sputtered TiO2
coating on PMMA material was studied, which exhibited antibacterial efficiency reaching 70% and a large decrease in the contact angle from 33–66 degrees to 5–30 degrees.
Few studies have been conducted on applying carbon-based coatings against microbes; those which have were limited to diamond-like carbon (DLC) related coatings. Marciano et al. [7
] evaluated the antibacterial ability of DLC coatings using S. aureus
, Escherichia coli
), Pseudomonas aeruginosa
), and Salmonella
. They found that a DLC coating with an H content of 22 at. % and an intensity ratio of D/G peaks (ID/IG) of 1.23 (detected by Raman spectroscopy) reduced bacterial activity by 25–55% compared with an uncoated sample, depending on the type of bacteria used. Zhou et al. [8
] prepared DLC and hydrogenated DLC coatings on SS316L substrates and evaluated their antibacterial ability using E. coli
. Compared with the uncoated sample, the DLC and hydrogenated DLC coatings reduced the viable bacteria to 15% and 33%, respectively. Studies on carbon-based coatings have demonstrated that surface properties are another factor affecting the adhesion of bacteria [9
]. Ishihara et al. [12
] prepared DLC coatings using a radio frequency magnetron sputtering system and evaluated the effect of the hydrophilicity of coating surfaces on the bacterial retention. The results indicated that the higher the hydrophilicity was, the less the bacteria were retained.
In notable research by Liu and Cohen [13
], carbonitride (CNx) coatings were found to exhibit hardness comparable to diamonds, particularly when involving covalent β-C3
bonding structures. Over the past 30 years, a few studies have been performed on CNx. Some results indicated that CNx coatings exhibited low friction, high adhesion, and high wear and corrosion resistance. Wäsche et al. found that among evaluated samples, a CNx coating showed the highest hardness under deposition conditions with a N/Ar ratio of 0.2 [14
]. Takadoum et al. determined [15
] that a CNx coating displayed higher wear resistance than a DLC coating. The same result has been obtained by Park et al. [16
] and Camero et al. [17
]. The CNx coatings, under deposition conditions with N/Ar ratios of 0.25–0.50, possessed the highest content of sp3
bonding and exhibited the highest wear resistance of their samples.
Other studies have found that the properties of CNx coatings, such as hardness, corrosive resistance, antioxidation, and wear resistance, can be enhanced by metal doping [18
Lai et al. [22
] grew a series of CNx/amorphous C-Zr0.55–0.60
(a-C) coatings on biograde Ti samples. They found that the coatings showed high biocompatibility and good antibacterial performance when the C content was higher than 12.7 at. %. Huang et al. [23
] prepared CNx coatings doped with Ti and Zr and evaluated the bacterial viability by using S. aureus
and Aggrigatebacter Actinomycetemcomitans
; the results revealed that bacterial viability levels of 21% and 36%, respectively, were obtained. Moreover, the coatings showed high biocompatibility with human fibroblasts. Considerably more studies have been conducted on mechanical properties than on antimicrobial performance.
Hosokawa et al. [24
] prepared Ti-doped CNx (CNx-Ti) coatings on drills by using the magnetron sputtering method and conducting drilling tests. They revealed that the life of the CNx-Ti-coated drills was three times that of TiN-coated drills and that the flank wear decreased significantly as well. Kuptsov et al. [25
] studied CNx-Ti coatings and found that corrosive resistance was improved by the formation of a surface TiO2
layer. The same mechanism was also found to achieve improvements in wear resistance at 300 °C [26
In an identical study by Polcara et al. [26
], Cr-doped CNx (CNx-Cr) coatings were prepared and evaluated using an elevated-temperature wear tester. The results showed that at 700 °C, the CNx-Cr coatings displayed a friction coefficient close to that at ambient temperature, due to minor variation in wear behavior between the two conditions.
In a study on a series of Zr-doped CNx (CNx-Zr) coatings, the coatings with 34.3 at. % Zr and a C/N ratio of 0.9 showed a hardness of 22.8 ± 1.1 GPa, low friction, and high adhesion [27
]. Wang et al. [28
] found that deposition temperature influenced the properties of a CNx-Zr coating. As the deposition temperature increased, the mechanical properties improved. At a deposition temperature of 400 °C, the coating showed a hardness of 32.0 ± 0.3 GPa, good adhesion, and a favorable resilience coefficient. Grigore et al. [29
] prepared CNx-Zr coatings using the magnetron sputtering method. Of the samples, the CNx-Zr coating showed the highest hardness (HV 3600) with a Zr content of 52 at. % under deposition of a N/Ar ratio of 0.1125, and that exhibited the highest wear resistance with a Zr content of 54.1 at. % under deposition of a N/Ar ratio of 0.0625. Yao et al. [30
] prepared a series of CNx-Zr coatings with different N content levels and found that the coating with a Zr content of 48.7 at. % under deposition of a N/Ar ratio of 0.2 displayed the best wear resistance and highest hardness (HK 1600).
Ospina et al. [31
] evaluated the mechanical properties and wear performance of W-doped CNx (CNx-W) coatings prepared using the pulse arc deposition method. The results indicated that the critical load (Lc
) and hardness increased, and the friction coefficient decreased to 0.35 from 0.7, as compared with a single W coating. Chen et al. [32
] prepared CNx-W and WNx coatings and compared their wear behavior. The wear rates for the CNx-W and WNx coatings were 3 × 10−8
and 6 × 10−8
/Nm, respectively. A large improvement in wear resistance was produced by the CNx-W coating.
Numerous studies have been conducted on DLC coatings involving various topics. Of these, only a few have discussed antibacterial activity and biocompatibility. Until now, according to our review of the literature, no data have been published concerning CNx coatings. In this study, we prepared CNx and 6 at. % metal-doped CNx coatings on SS316L substrates by using an industrial-scale four-target closed-field unbalanced DC magnetron sputtering system. The metals used were W, Ti, Zr, and Cr. The samples were characterized and evaluated for antibacterial and wear performance. This study was a trial of growing CNx and metal-doped CNx coatings on SS316L steel; their characteristics and possible applications against bacteria and against wear were investigated.
2. Experimental Details
2.1. Analysis Equipment
A scanning electron microscope (E-SEM Quanta 400F, FEI, Hillsboro, OR, USA) equipped with an electron dispersive X-ray system was used. A glow discharge spectrometer (GDS-750 QDP, LECO, Saint Joseph, MI, USA) was applied to determine the coating thickness. A glancing-angle X-ray diffractometer (D/MAX2500, Rigaku, Tokyo, Japan) was used, setting Cu Kα radiation at 40 kV and 100 mA with a glancing angle of 4°. An X-ray photoelectron spectroscope (PHI 1600 ESCA, Perkin-Elmer Co., Waltham, MA, USA) was used, employing non-monochromatized Mg Kα radiation. An X-ray photon spectroscopy (XPS) system was applied with 3 kV Ar ions to sputter the surface oxide layer and examine the chemical composition of the coatings. Spectra ranging from 0 to 1000 eV were recorded for each sample, followed by high-resolution spectra over different elemental peaks, from which the composition was calculated. The spectral ranges at 287 ± 7, 400 ± 9, 184 ± 10, 582 ± 13, 460 ± 10, and 36 ± 10 eV corresponded to the binding energies of C1s, N1s, Zr3d, Cr2p, Ti2p, and W4f, respectively. Curve fitting was performed after a Shirley background subtraction by a Gaussian fitting [33
]. Energy calibration was conducted with reference to the Au 4f7/2 peak at 83.8 eV. A Raman spectrometer (HR 800 Micro PL, Horiba, Kyoto, Japan) employing a He-Ne laser and Peltier-cooled Princeton CCD camera was used. Two peaks were observed in the Raman analyses result, indicating carbon structures corresponding to a D peak at approximately 1363 cm−1
and a G peak at approximately 1560 cm−1
. A nanoindentation system (LBI Nanoindenter, UNAT-M, Dresden, Germany) was used at an applied load of 10 mN. Adhesion was evaluated using a scratch tester with an applied load that increased from 0 to 100 N at a rate of 10 N/mm. A critical load (Lc
) was obtained.
The coatings were prepared using a closed-field unbalanced DC magnetron sputtering system (CFUBMS KD-550U, MIRDC, Kaohsiung, Taiwan), which features four vertical orthogonally-mounted magnetron cathodes surrounding a rotational substrate holder. The dimensions of a target were 437.3 mm × 178.2 mm × 10.2 mm.
Biograde SS316L was used as the substrate material. Two types of samples, block and plate, were used. The block had dimensions of 24 mm × 8 mm (diameter × thickness) and the plate of 16 mm × 16 mm × 1 mm. The plate sample was used in the antimicrobial test. The substrate surface to be evaluated was polished to a surface roughness Ra < 60 nm.
The standard sample preparation procedure for physical vapor deposition (PVD) was used. After the samples were loaded, the chamber was pumped down to 2 × 10−5 Torr. The main Ar gas and minor H2 gas were introduced. The H ions were used to maintain the chamber temperature. The Ar ions were used to perform the bombardment conditioning (30 min). The Ar gas flow rate was increased at a steady rate until it reached 25 sccm (corresponding to 3 × 10−3 Torr) at the end of the bombardment process and was then maintained for the subsequent coating.
In this study, CNx, CNx-W, CNx-Ti, CNx-Zr, and CNx-Cr coatings were prepared. For growth of metal-doped CNx coatings, four targets were used, with two carbon targets next to each other and two metal targets in the chamber. The metal targets were the doping material and were also used to prepare an effective interface. For example, in the case of the CNx-Ti coating, two Ti and two carbon targets were used. The Ti interlayer was prepared first under a condition of 1 A current for one Ti target for 10 min. The TiC gradual layer was then prepared under conditions of 1 A and 3 A currents for one Ti and two C targets, respectively, for 6 min. Finally, the main CNx-Ti coating was prepared under conditions of 1 A and 3 A currents for two Ti and two C targets, respectively, and a N2
gas flow rate of 6 sccm for 120 min. Identical processes were used to grow the CNx-Zr and CNx-Cr coatings by using Zr and Cr targets, respectively. The coating procedure for the CNx-W coating involved some modification. From long-term experience, we know that the use of a W interface layer leads to poor quality of adhesion for the subsequent main coating. Thus, to grow the CNx-W coating, two C targets, one Ti target, and one W target were used. The Ti interface was prepared first at a Ti target using a 1 A current for 10 min., and then the TiC gradual layer was prepared under conditions of 1 A and 2 A currents for one Ti and two C targets, respectively, for 6 min. Finally, the main CNx-W coating was prepared under conditions of 0.6 A and 3 A currents for one W and two C targets, respectively, and a N2
gas flow rate of 6 sccm for 120 min. The aforementioned conditions were set on the basis of repeated tests and trials for the different types of coatings. With these procedures, the overall thickness of the coatings was approximately 2 μm, as determined using glow discharge optical spectroscopy. The amount of doped metal was well controlled at 6–7 at. %, as determined using XPS. The data are listed in Table 1
. For growth of the CNx coating, one Ti and two carbon targets were used. The Ti interface was prepared first, then the TiC gradual layer, and finally the main CNx coating, using the same condition for the CNx-Ti coating, with 0 A for the Ti target.
2.3. Hydrophilicity Test
The hydrophilicity of the surface of the samples was evaluated by measuring the static contact angles using a contact-angle analyzer (FTA-1000B, Portsmouth, VA, USA). Each specimen was alternately washed in ethanol and deionized water in an ultrasonic cleaner for 30 min and then dried at 55 °C for 4 h. A drop of 2 mL distilled water was place on the sample surface using a micrometric pipette at room temperature. The sample was immediately imaged, and the contact angle was measured automatically using a spherical fitting approach. Each contact angle reported herein is the mean of five independent measurements.
2.4. Antimicrobial Test
The antimicrobial performance of a coating was evaluated according to the antimicrobial activity of a given coating surface against bacteria in contact with it. The antimicrobial test procedures were performed following the Japan Industrial Standard JIS Z 2801:2010 Antimicrobial products—Test for antimicrobial activity and efficacy, with minor modifications. Prior to the test, the plate samples were sterilized in a high-pressure steam autoclave operated at 121 °C, 103 kPa for 15 min, and then placed in sterile Petri dishes. The test bacteria used were S. aureus Gram-positive bacteria in suspension, adjusted at a cell concentration of 5 × 108 colony-forming units per milliliter (cfu/mL). A 50 μL suspension of the test bacteria was instilled onto the surface of a sample, and was then covered with a sterilized polyethylene film measuring 10 mm × 10 mm. The film was pressed so that the inoculum (test bacteria) spread evenly over the film, to avoid drop evaporation. An uncoated SS316L sample was used as a reference. Finally, the Petri dishes containing the inoculated test pieces were incubated at a temperature of 37 °C and a relative humidity of not less than 90% for 24 h.
After the inoculation, the test bacteria were washed out using 10 mL of phosphate-buffered saline. To conduct a viable cell count of bacteria, 10-fold serial dilutions and the agar plate culture method were used. Commercial Lysogeny broth agar was used (Difco Laboratories, Detroit, MI, USA). The incubation was performed at 37 °C for 24 h. After incubation, the number of visible bacterial colonies was counted using an agar plate in which 30 to 300 colonies appeared. To determine the antimicrobial performance, the following formula was used:
Antibacterial Efficiency (η) = (N0 − N)/N0 × 100%
where N (expressed in cfu) represents the number of viable cells of bacteria on the sample and N0
represents the number for the uncoated SS316L reference sample. The N value was calculated using C*D, where C is an average of the number of colonies in three test pieces, and D is the corresponding dilution ratio of the diluted solution. The data presented herein were obtained from seven independent experiments.
The final condition of the bacterial colonies was imaged using a traditional camera. The statistical correlation of the antibacterial efficiency between the coated samples was determined using Student’s t
]. Differences were considered to be statistically significant when p
was less than 0.01.
2.5. Wear Test
The wear performance of the coatings was evaluated using a pin-on-disk rotating-sliding wear tester (Jun-Yan Precision Machine Co., Kaoshiung, Taiwan). The material of ball attached at the pin end was 100% Al2O3. The testing conditions were as follows: applied load, 10 N; line speed, 0.2 m/s; rotating radius, 6 mm; and sliding distance, 350 m. The resulting wear track was measured by scanning four cross-sections at an interval of 90° using a white-light interferometer (BMT, Nuremberg, Germany). The wear rate was calculated by the formula 2πRA/WL, where R is the rotating radius, A is an average of four measurements of a wear cross area, W is the load, and L is the sliding distance.