Visible Light Irradiation-Mediated Drug Elution Activity of Nitrogen-Doped TiO 2 Nanotubes

We have developed nitrogen-doped TiO2 nanotubes showing photocatalytic activity in the visible light region and have investigated the triggered release of antibiotics from these nanotubes in response to remote visible light irradiation. Scanning electron microscopy (SEM) observations indicated that the structure of TiO2 nanotubes was not destroyed on the conditions of 0.05 and 0.1M diethanolamine treatment. e results of �-ray photoelectron spectroscopy (�PS) con�rmed that nitrogen, in the forms of nitrite (NO2 ) and nitrogen monoxide (NO), had been incorporated into the TiO2 nanotube surface. A drug-release test revealed that the antibiotic-loaded TiO2 nanotubes showed sustained and prolonged drug elution with the help of polylactic acid. Visible light irradiation tests showed that the antibiotic release from nitrogen-doped TiO2 nanotubes was signi�cantly higher than that from pure TiO2 nanotubes (PP P PPPP).


Introduction
Development of TiO 2 nanostructures has been a focus area in the �elds of photocatalysis [1][2][3], solar cells [4], and biomedical applications [5][6][7].TiO 2 is well known as a highly efficient photocatalyst and has been widely used for degrading organic pollutants for air puri�cation and sterilization [8,9].However, the band gap of TiO 2 allows limited photocatalytic activity, which occurs solely in the narrow ranges of ultraviolet light.Many researchers have tried to develop visible light photocatalysts by modifying the structure of titanium dioxide.Metal doping is one of the typical modi�cations of TiO 2 , and several kinds of metal elements, such as Cr, Co, Mo, Mn, and V, have been used to adjust the band gap of TiO 2 and promote its photocatalytic activity in the visible light range.
Bacterial infection is one of the major reasons for the failure of orthopedic implants.An ideal solution to reduce bacterial infection is antibiotic drug therapy around 2 months aer implant surgery.e delivery methods of antibiotics are generally systemic, intravenous, intramuscular, and topical.However, systemic antibiotic delivery is typically associated with certain side effects, including unwanted cytotoxicity.In recent years, various drug-delivery systems have been developed to facilitate drug effectiveness at the site of implantation [22,23].ese drug-delivery devices release a proper dose of the drug at the site of action and thereby avoid undesirable side effects.However, these devices lack the ability to exert an on-off control over drug release.Consideration of release periods as well as the elution point of drugs is very important to reduce bacterial infection.Unwanted high doses of the drug lead to unavoidable toxic effects.Being able to trigger the antibiotic release, for example, by using light stimulation, would be highly desirable to minimize complications and side effects for the orthopaedic implants.However, ultra violet (UV) light triggering TiO 2 photocatalytic activity is well known to be used for sterilization and to be harmful to mucous membrane in mouth.erefore, the development of new TiO 2 showing photocatalytic activity at visible light region is required to minimize the side effect of UV light irradiation and promote the effect of remote-controlled drug release.In terms of the photocatalytic activity of TiO 2 nanotubes at visible light region, several studies reported that nitrogen or Fe-doped TiO 2 nanotubes showed excellent photocatalytic activity and degree of dye degradation compared to pure TiO 2 nanotubes by visible light irradiation [24][25][26][27].
In this study, we developed nitrogen-doped 100 nm TiO 2 nanotubes on Ti, performed surface analysis to examine the amounts and structure of nitrogen incorporated into Ti, and investigated the triggered release of antibiotic drug from nitrogen-doped TiO 2 nanotubes in response to remote visible light irradiation.

Fabrication of Nitrogen-Doped TiO 2
Nanotubes.As reported previously [28], a machined Ti sheet (0.2 mm thick, 99.5%; Hyundai Titanium Co., Republic of Korea) was electropolished by an electrochemical etching process and cleaned with acetone and deionized water.To prepare nitrogen-doped TiO 2 nanotube arrays on a Ti sheet, 0.05, 0.1, or 0.2 M diethanolamine (DEA; Sigma, MO, USA) was added into 0.5 w/v% hydro�uoric acid (48 w/v%; Merck, NJ, USA) in a mixture of water and acetic acid (98 w/v%; JT Baker, NJ, USA) in the volumetric ratio of 7 : 1. Anodization voltage and time were 20 V and 30 min, respectively.
Samples were then rinsed with deionized water, dried at 60 ∘ C for 24 h, and heat treated at 500 ∘ C for 2 h in an atmosphere of N 2 .Morphological and surface analyses of nitrogen-doped TiO 2 nanotube arrays were performed by �eld emission scanning electron microscopy (FE-SEM, S4800; Hitachi/Horiba, Japan), transmission electron microscopy (TEM, Tecnai G2; FEI Co., USA, power: 300 kV), and X-ray photoelectron spectroscopy (XPS, K-Alpha ESKA system; ermo, USA), respectively.Also, contact angle of experimental specimen was measured by contact angle meter (eta Optical Tensiometer, KSV, Finland).e solvent of contact angle measurement was D.I. water.

Drug Release
Test. ree antibacterial drugs, tetracycline (Sigma, MO, USA), cetylpyridinium chloride (CPC; Sigma, MO, USA), and chlorhexidine (Sigma, MO, USA), were mixed with polylactic acid (PLA; Sigma, MO, USA) and loaded into TiO 2 nanotubes.Tetracycline is an antibiotic drug that serves as a protein synthesis inhibitor.Chlorhexidine is a chemical antiseptic, while cetylpyridinium chloride is a strong bactericide.e amount of antibiotics released as a function of incubation time was measured by a microplate ELISA reader (Spectra Max 250; ermo Electron Co., USA).e amount of antibiotics released in response to visible right irradiation was also measured by the microplate ELISA reader.e source of visible light was a dental light curing unit (intensity, 1000 mW/cm 2 ; wavenumber of irradiated light, 470 nm; Elipar Free-Light 2; 3 M ESPE Co., USA).

Data Analysis.
All data were expressed as mean ± standard deviation values and analyzed statistically by oneway ANOVA (SPSS 12.0; SPSS GmbH, Germany) and post hoc Duncan�s multiple range test.Signi�cant differences were considered if  values were less than 0.05.

Results and Discussion
As shown in Figures 1(a), 1(b), and 1(c), SEM images show the differences in appearance among N-doped 100 nm TiO 2 nanotubes with 0.05 M, 0.1 M, and 0.2 M of DEA.e micrographs of N-doped TiO 2 nanotubes show somewhat randomly organized nanotube geometry with different concentrations of DEA, in contrast to the SEM image of undoped TiO 2 nanotubes.However, the nanotubular structure was not formed on the Ti surface at a DEA concentration of 0.2 M. erefore, we examined the characteristics of N-doped TiO 2 nanotubes at a DEA concentration of 0.1 M. TEM image (Figure 1(d)) of N-doped 100 nm TiO 2 nanotubes treated by 0.1 M DEA indicates that nanosized (100 nm thickness) porous layer was formed at top surface of TiO 2 nanotubes.High-magni�cation TEM image (Figure 1(e)) illustrates that the lattice spacing of newly formed layer is 0.35 nm, and this spacing is corresponding to the (101) planes of anatase TiO 2 as previously reported [29].
Figure 2 indicates X-ray diffraction (XRD) patterns of undoped and N-doped TiO 2 nanotubes.As shown, XRD mainly detected anatase TiO 2 and Ti crystalline phases.ere was no dramatic difference in crystallinity between undoped and N-doped TiO 2 nanotubes aer heat treatment.erefore, we expect that DEA treatment had essentially no effect on the crystallinity of TiO 2 nanotubes in this study.
e XPS spectra of TiO 2 nanotubes and N-doped TiO 2 nanotubes are shown in Figure 3(a).In terms of pure TiO 2 nanotubes, Ti, O, and C elements were detected at 459.6, 531.2, and 285.5 eV, respectively.Among these elements, carbon is supposed to be contaminant deposited at the surface of TiO 2 nanotubes.e surfaces of N-doped TiO 2 nanotubes were composed of Ti, O, N, and C contaminants, and a very weak N signal was detected at the surface of Ndoped TiO 2 nanotubes.e XPS analysis also resulted that the N amounts in undoped TiO 2 nanotubes and N-doped TiO 2 nanotubes were 0.57 and 3.39 atomic%, respectively.e atomic ratio of Ti to N of N-doped TiO 2 nanotubes was 5.13.As previously reported, the photocatalytic effect of N dopant was affected by both the N content of N-doped TiO 2 and the degree of nitrogen atoms reacting with TiO 2 precursor [30].erefore, N-doped TiO 2 nanotubes having high N amounts and the atomic ratio of Ti to N are supposed to result in enhanced photocatalytic activity by visible light irradiation.Also, the N amount doped in TiO 2 structure is related to the intensity of photocatalytic activity at visible light region, and N amount is affected by the reaction temperature of dopant and TiO 2 .Previous studies have reported 5-8 atomic% of nitrogen incorporation into the TiO 2 surface by chemical treatment and excellent photocatalytic activity in the visible light region [30][31][32].ese studies involved heat treatment of N-doped TiO 2 nanoparticles at temperatures of 800-900 ∘ C to maximize the concentration of nitrogen doping.However, we could not heat treat TiO 2 nanotubes above 500 ∘ C because heat treatment above 500 ∘ C resulted in the formation of rutile structure destroying the nanotubular structure of TiO 2 .is limitation provides lower N amount of TiO 2 nanotubes compared to that of other TiO 2 nanoparticles.ere are several researches obtaining highly N-doped TiO 2 nanomaterials without conventional sintering process.Xiang et al. developed nitrogen-, sulfuror carbon-doped TiO 2 nanosheets with exposed (001) facets showing excellent photocatalytic activity at visible region by solvothermal process [29,33,34].Also, solvothermal process is seemed to be one of the techniques enhancing N-doping into the structure of TiO 2 nanotubes without the destruction of nanotubular structure as previously reported [24,26].Further experiment is required to investigate the comparison of the photocatalytic activity between sintering process and solvothermal process for doping nitrogen into TiO 2 nanotubes.
e N 1s peaks were detected at the surface of N-doped TiO 2 nanotubes at 406.1 and 402.5 eV of binding energy, respectively (see Figure 3(b)).In terms of the location of nitrogen dopant in TiO 2 structure, many researches have reported that nitrogen species are doped into TiO 2 in different forms due to doping process, reaction technique, and nitrogen sources [24-27, 29, 33-38].From the results of previous researches [35][36][37], typical N 1s peak of TiN species mainly in substitutional N was less than 397.5 eV, whilst interstitial N in TiN species showed above 400 eV of typical N 1s binding energy.From the results of XPS analysis, typical binding energies of 402. on the basis of previous studies [35][36][37]39].us, it is con�rmed that nitrogen from �EA is effectively doped into TiO 2 nanotubes, and the N 1s peaks obtained from this study are assigned to interstitial N-doped TiO 2 nanotubes.
Figure 4(a) shows the cross-sectional views of water droplets on machined Ti, electropolished Ti, and undoped and N-doped TiO 2 nanotubes.Electropolishing did not change the hydrophilicity of Ti surface dramatically, but nitrogen doping did change the wettability of TiO 2 nanotubes by changing the hydrophilic surface to a super hydrophobic (>120 ∘ ) surface.erefore, we are investigating the effect of the super hydrophobicity of N-doped TiO 2 nanotubes on the behavior and functionality of human mesenchymal stem cells.
Presented in Figure 5 is the effect of PLA on the release behavior of antibacterial drugs such as tetracycline, CPC, and chlorhexidine.As shown in Figure 5(a), the experimental group with 10% tetracycline shows only an initial burst of drug release in the incubation period.However, all experimental groups with 1% PLA show sustained release of drugs as a function of incubation time.erefore, we con�rmed that 1% PLA could alter the elution behavior of all drugs and allow sustained and prolonged drug release regardless of the drug type.
Figure 6 shows the elution concentrations of the three antibacterial drugs loaded on the surface of undoped and N-doped TiO 2 nanotubes, respectively, aer 30 seconds of visible light irradiation with the dental curing unit.In  the tetracycline and chlorhexidine elution tests, the release concentrations of drugs from N-doped TiO 2 nanotubes were signi�cantly higher than those from undoped TiO 2 nanotubes (  ).However, the total amount of CPC released was much lower than the amounts of tetracycline and chlorhexidine, and there was no signi�cant difference between CPC release from undoped and N-doped TiO 2 nanotubes.
On the basis of these results, we can summarize that nitrogen doping into the TiO 2 nanotubular structure was performed successfully by DEA treatment, even though the amount of nitrogen doping was lower than reported in other studies because of the lower heat treatment temperature used in this study.Moreover, drugs stored in N-doped TiO 2 nanotubes were released effectively by the visible light irradiation with the dental light curing unit.

Summary
e visible light irradiation-mediated drug elution activity of nitrogen-doped TiO 2 nanotubes has been investigated in this study.We found that nitrogen was effectively doped into the TiO 2 nanotubular structure, with the existence of NO 2 − (406.1 eV) detected by the XPS analysis playing an important role in the photocatalytic activity of TiO 2 in the visible light region.e results of the drug release test showed that PLA facilitated sustained and prolonged elution of drugs.We conclude that N-doped TiO 2 nanotubes are expected to overcome the limited usage of TiO 2 , which shows photocatalytic activity only within the UV region, thereby allows the development of novel fusion technologies in the �eld of implant materials.

�on��ct of �nterests
e authors declare having no con�ict of interests about all materials in this paper.

F 1 :F 2 :
SEM images of 100 nm TiO 2 nanotubes with (a) 0.05 M, (b) 0.1 M, and (c) 0.2 M of DEA treatment.(d) TEM and (e) high-magni�cation TEM images of 100 nm TiO 2 nanotubes with 0.1 M DEA treatment.X-ray diffraction patterns of undoped and nitrogendoped TiO 2 nanotubes treated by 0.1 M DEA.

F 3 :F 4 :
5 and 406.1 eV are assigned to NO and NO 2 − generating highest localized state for interstitial species, which are characteristics of interstitial N-doped TiO 2 (a) XPS spectra of TiO 2 nanotubes and N-doped TiO 2 nanotubes.(b) High resolution XPS spectra of N 1s for TiO 2 nanotubes and N-doped TiO 2 nanotubes.(a) Cross-sectional views of water droplets on Ti, electropolished Ti, 100 nm TiO 2 nanotubes, and N-doped 100 nm TiO 2 nanotubes.(b) Graph showing water droplet contact angles at the surface of the nanotubes.

F 5 :
e release amounts of antibacterial drugs as a function of incubation time.(a) 10% tetracycline mixed with/without 1% PLA.(b) 10% CPC and chlorhexidine mixed with 1% PLA.