Morphological Changes, Antibacterial Activity, and Cytotoxicity Characterization of Hydrothermally Synthesized Metal Ions-Incorporated Nanoapatites for Biomedical Application

The objective of this study was to prepare hydroxyapatite (HA) with potential antibacterial activity against gram-negative and gram-positive bacteria by incorporating different atomic ratios of Cu2+ (0.1–1.0%), Mg2+ (1.0–7.0%), and Zn2+ (1.0–7.0%) to theoretically replace Ca2+ ions during the hydrothermal synthesis of grown precipitated HA nanorods. This study highlights the role of comparing different metal ions on synthetic nanoapatite in regulating the antibacterial properties and toxicity. The comparisons between infrared spectra and between diffractograms have confirmed that metal ions do not affect the formation of HA phases. The results show that after doped Cu2+, Mg2+, and Zn2+ ions replace Ca2+, the ionic radius is almost the same, but significantly smaller than that of the original Ca2+ ions, and the substitution effect causes the lattice distance to change, resulting in crystal structure distortion and reducing crystallinity. The reduction in the length of the nanopatites after the incorporation of Cu2+, Mg2+, and Zn2+ ions confirmed that the metal ions were mainly substituted during the growth of the rod-shape nanoapatite Ca2+ distributed along the longitudinal site. The antibacterial results show that nanoapatite containing Cu2+ (0.1%), Mg2+ (3%), and Zn2+ (5–7%) has obvious and higher antibacterial activity against gram-positive bacteria Staphylococcus aureus within 2 days. The antibacterial effect against the gram-negative bacillus Escherichia coli is not as pronounced as against Staphylococcus aureus. The antibacterial effect of Cu2+ substituted Ca2+ with an atomic ratio of 0.1~1.0% is even better than that of Mg2+- and Zn2+- doped with 1~7% groups. In terms of cytotoxicity, nanoapatites with Cu2+ (~0.2%) exhibit cytotoxicity, whereas Mg2+- (1–5%) and Zn2+- (~1%) doped nanoapatites are biocompatible at low concentrations but become cytotoxic as ionic concentration increases. The results show that the hydrothermally synthesized nanoapatite combined with Cu2+ (0.2%), Mg2+ (3%), and Zn2+ (3%) exhibits low toxicity and high antibacterial activity, which provides a good prospect for bypassing antibiotics for future biomedical applications.


Introduction
Antibiotics are the most common drugs currently used to fight bacterial infections to prevent the emergence of biofilms. However, several types of bacteria can develop antibiotic resistance due to a combinatorial drive of bacteria exposed to antibiotics [1]. The current treatment of bacterial biofilms is complex, expensive and inefficient, leading to an

FT-IR Spectroscopic Analysis
The characteristic apatite absorption bands are shown in Figure 1. Each group showed the librational vibration (ν L ) and asymmetric stretching absorption (ν 3 ) of hydroxyl (OH − ) groups observed at wavenumbers of 632 and 3568 cm −1 , respectively. The apatite PO 4 3-absorption bands at wavenumbers 563 and 603 cm −1 were observed and attributed to the bending vibration of O−P−O (triply degenerate δ4 bending in PO 4 3− ). The symmetric stretching band (ν 1 mode in PO 4 3− ) at 964 cm −1 and asymmetric stretching bands (ν 3 mode in PO 4 3− ) at 1034 and 1097 cm −1 were detected [47][48][49][50]. In addition, the NH 4 + functional group was found at 1385 cm −1 , and the C-H symmetrical and asymmetric stretching vibration bands were observed at 2854 and 2924 cm −1 , respectively, which were attributed to the diammonium hydrogen phosphate ((NH 4 ) 2 HPO 4 ) and residual functional groups of sodium citrate (Na 3 C 6 H 5 O 7 ). The C=O stretching vibration bands of CO 2 were observed at 1448 and 1635 cm −1 [51].
Here, we proposed the co-precipitation of Cu 2+ , Mg 2+ , and Zn 2+ with Ca 2+ for the preparation of nanoapatites with antibacterial properties through hydrothermal wet synthesis and compared their antibacterial properties and biocompatibility. The composition, phase identification, morphological changes, antibacterial activity and cytotoxicity of synthesized nanoapatite doped with Cu 2+ , Mg 2+ , and Zn 2+ ions were investigated.

XRD Analysis
To elucidate the phase structure and crystallinity produced by the hydrothermal reaction, the XRD patterns of nanoapatites with different Cu 2+ , Mg 2+ , and Zn 2+ ion doping levels are illustrated in Figure 2 The patterns were compared with those in the JCPDS Figure 1. Fourier transform infrared spectra of rod-shaped nanoapatites synthesized with and without Cu 2+ , Mg 2+ , and Zn 2+ at different ion concentrations.

XRD Analysis
To elucidate the phase structure and crystallinity produced by the hydrothermal reaction, the XRD patterns of nanoapatites with different Cu 2+ , Mg 2+ , and Zn 2+ ion doping levels are illustrated in Figure 2 The patterns were compared with those in the JCPDS database. The diffraction planes (002), (211), (112), and (300) were the main characteristic peaks of HA (JCPD 73−0294, JCPD 74−0565, and JCPD 74−0566). No other characteristic compounds of the diffraction planes containing Cu, Mg, and Zn were detected. The measured increase in the full width at half maximum (FHMW) of most planes observed in Table 1 reflects a decrease in crystallinity as doping metal ions increased in the nanoapatites. This decrease was attributed to the uneven replacement of the apatite bonding sites of Ca 2+ (1.14 Å) by Cu 2+ (0.87 Å), database. The diffraction planes (002), (211), (112), and (300) were the main characteristic peaks of HA (JCPD 73−0294, JCPD 74−0565, and JCPD 74−0566). No other characteristic compounds of the diffraction planes containing Cu, Mg, and Zn were detected. The measured increase in the full width at half maximum (FHMW) of most planes observed in Table 1 reflects a decrease in crystallinity as doping metal ions increased in the nanoapatites. This decrease was attributed to the uneven replacement of the apatite bonding sites of Ca 2+ (1.14 Å) by Cu 2+ (0.87 Å), Mg 2+ (0.86 Å), and Zn 2+ (0.88 Å) ions, resulting in lattice distortion in nanoapatite crystalline structures [52][53][54].     Figure 3 shows the TEM images and SAED analysis results of the original nanoapatite rod and nanoapatite rods doped with different ions (Cu 2+ , Mg 2+ , or Zn 2+ ). Each group was in the form of rod-shaped nanoapatites. In SAED analysis, the diffraction patterns of the characterized (002), (211), and (112) planes and the lattice distances in HRTEM were observed, and HA was observed to correspond to 0.34 and 0.28 nm of the (002) and (300) planes, respectively. This result was verified with the XRD patterns. The doped and undoped rod-shaped nanoapatites had average lengths of 31-49 nm, average widths of 6-12 nm, and average aspect length-to-width ratios of 2-5. No significant differences in the plane ratios between length (002) and width (300) in the lattice calculated from the FWHM value of the XRD pattern were found, but the plane ratio values of Cu 2+ , Mg 2+ , and Zn 2+ were greater than the value of the Ca 2+ -only control group (Table 2). Except in 0.1Cu−nHA, 1Mg−nHA, and 1Zn−nHA groups with low amounts of doped ions, the lengths of the rod-shaped nanoapatites decreased with increasing amounts of doped ions (p < 0.05). Figure 3 shows the TEM images and SAED analysis results of the original nanoapatite rod and nanoapatite rods doped with different ions (Cu 2+ , Mg 2+ , or Zn 2+ ). Each group was in the form of rod-shaped nanoapatites. In SAED analysis, the diffraction patterns of the characterized (002), (211), and (112) planes and the lattice distances in HRTEM were observed, and HA was observed to correspond to 0.34 and 0.28 nm of the (002) and (300) planes, respectively. This result was verified with the XRD patterns. The doped and undoped rod-shaped nanoapatites had average lengths of 31-49 nm, average widths of 6-12 nm, and average aspect length-to-width ratios of 2-5. No significant differences in the plane ratios between length (002) and width (300) in the lattice calculated from the FWHM value of the XRD pattern were found, but the plane ratio values of Cu 2+ , Mg 2+ , and Zn 2+ were greater than the value of the Ca 2+ -only control group (Table 2). Except in 0.1Cu−nHA, 1Mg−nHA, and 1Zn−nHA groups with low amounts of doped ions, the lengths of the rodshaped nanoapatites decreased with increasing amounts of doped ions (p < 0.05).

TEM Analysis
According to Shannon's ionic radii in a previous study [55], which reported the difference in coordination number and high and low spin states of ions, Ca 2+ (1.14 Å) at the longitudinal sites of a nanoapatite were mainly replaced by Cu 2+ (0.87 Å), Mg 2+ (0.86 Å), and Zn 2+ (0.88 Å) with small ionic radii. The average transverse width of 3Mg−nHA (6.80 ± 1.74 nm) was smaller than that of nHA (10.44 ± 2.38 nm; p < 0.05) and more elongated, but the average width of 7Zn−nHA (12.05 ± 2.62 nm) was larger than that of nHA (p < 0.05), making it short and wide.   Table 2. Average longitudinal length, transverse width, length-to-width aspect ratio (n = 50) of rod-shaped nanoapatites and plane ratio of (002) to (300) FWHM values after doping in different Cu 2+ , Mg 2+ , and Zn 2+ ionic reaction concentrations.

Samples Longitudinal Length (nm) Transverse Width (nm) Length-to-Width Aspect Ratio
Plane Ratio of (002) to (300 According to Shannon's ionic radii in a previous study [55], which reported the difference in coordination number and high and low spin states of ions, Ca 2+ (1.14 Å) at the longitudinal sites of a nanoapatite were mainly replaced by Cu 2+ (0.87 Å), Mg 2+ (0.86 Å), and Zn 2+ (0.88 Å) with small ionic radii. The average transverse width of 3Mg−nHA (6.80 ± 1.74 nm) was smaller than that of nHA (10.44 ± 2.38 nm; p < 0.05) and more elongated, but the average width of 7Zn−nHA (12.05 ± 2.62 nm) was larger than that of nHA (p < 0.05), making it short and wide.

Antibacterial Activity
The results of the quantitative analysis of the antibacterial effect of nanoapatite hydrothermally synthesized with different concentrations of Cu 2+ , Mg 2+ , and Zn 2+ ions on S. aureus are shown in Figure 4. In the hydrothermal synthesis of nanoapatites, the test groups with Cu 2+ , Mg 2+ , and Zn 2+ ions exhibited better bacteriostatic ability than the control nHA group without metal ion addition, especially on day 1 (p < 0.05). As for the individual antibacterial strength of each ionic group, the antibacterial ability of the 0.1Cu−nHA group at a low concentration was comparable to that of 1Cu−nHA at a high concentration until day 3 in terms of strength ( Figure 4a). As shown in Figure 4b, the slender nanoapatite of the 3Mg−nHA group showed a higher inhibitory effect in the group containing Mg 2+ . As for the group with Zn 2+ , as the Zn 2+ ionic addition amount increased, the bacteriostatic ability was enhanced (Figure 4c). ions can bind to thiol bonds in the cytoplasm, disrupting bacterial metabolism and differentiation [60]. Hence, the combined effects of the aspect ratios and surface charges of nanoapatites on the bacteriostatic ability of bacteria should be considered.   Figure 5 shows the quantitative analysis results for the antibacterial effects of the tested nanoapatite groups synthesized with Cu 2+ , Mg 2+ , and Zn 2+ ions on E. coli for 1-4 days, and the concentrations of the groups were compared with the concentration of the control nHA group. Each test group had an antibacterial effect on E. coli on the first day of the analysis. Regarding the antibacterial effect of each ion, although the Cu 2+ ability of 0.3Cu−nHA and 1Cu−nHA groups against E. coli increased with Cu 2+ concentration, the antibacterial ability of 0.1Cu−nHA with a lower concentration of Cu 2+ was maintained for 2 days (Figure 5a). The test groups of 1Mg−nHA and 3Mg−nHA exhibited antibacterial ability against E. coli in 2 days of testing (Figure 5b). Except in the 7Zn−nHA group, the antibacterial ability of Zn 2+ was still statistically significant (p < 0.05) compared with that in the nHA control group on day 2 (Figure 5b), but the difference in antibacterial effect was small. The bacteriostatic ability of the test group with Cu 2+ , Mg 2+ , and Zn 2+ ions for S. aureus was better than that for E. coli.  Figure 6 shows the quantitative and qualitative analysis results of cytotoxicity in L929 cells incubated with nanoapatites with different metal ions for 1 day. Extracts that result in a reduction of control activity below 70% are considered cytotoxic as described in the standard ISO 10993-5 [61]. Figure 6a shows that the cell viability of HDPE was the same as that in the control nHA group, indicating that the groups were effective during the sterilization process. Meanwhile, the control nHA group without Cu 2+ , Mg 2+ , and Zn 2+ showed no toxicity. The nanoapatite groups with Cu 2+ showed cytotoxicity, whereas nanoapatites with Mg 2+ and Zn 2+ and the 7Mg−nHA, 3Zn−nHA, 5Zn−nHA, and 7Zn−nHA groups still showed cytotoxicity [62]. The cell morphology in Figure 6b was qualitatively analyzed, and the results of each group showed the same trend as the quantitative analy- The cell wall structure of E. coli is more complex than that of S. aureus [56], and thus Cu 2+ , Mg 2+ , and Zn 2+ ions can barely disrupt the cell wall of E. coli and hardly exert a bacteriostatic effect [57,58]. In addition, nanoapatites with low concentrations of Cu 2+ ions exhibited better bacteriostatic effects than those with Mg 2+ and Zn 2+ [59]. The particle size of Cu 2+ , Mg 2+ and Zn 2+ ions (~0.86-0.88 Å) is much smaller than that of Ca 2+ (1.14 Å), which causes the possibility of free entry and exit of bacterial cells, destroys cell wall synthesis, and reduces the protective effect of bacteria. For example, literature suggests that Cu 2+ ions can bind to thiol bonds in the cytoplasm, disrupting bacterial metabolism and differentiation [60]. Hence, the combined effects of the aspect ratios and surface charges of nanoapatites on the bacteriostatic ability of bacteria should be considered. Figure 6 shows the quantitative and qualitative analysis results of cytotoxicity in L929 cells incubated with nanoapatites with different metal ions for 1 day. Extracts that result in a reduction of control activity below 70% are considered cytotoxic as described in the standard ISO 10993-5 [61]. Figure 6a shows that the cell viability of HDPE was the same as that in the control nHA group, indicating that the groups were effective during the sterilization process. Meanwhile, the control nHA group without Cu 2+ , Mg 2+ , and Zn 2+ showed no toxicity. The nanoapatite groups with Cu 2+ showed cytotoxicity, whereas nanoapatites with Mg 2+ and Zn 2+ and the 7Mg−nHA, 3Zn−nHA, 5Zn−nHA, and 7Zn−nHA groups still showed cytotoxicity [62]. The cell morphology in Figure 6b was qualitatively analyzed, and the results of each group showed the same trend as the quantitative analysis results of cytotoxicity.

Cytotoxicity
The preparation of metal ion-incorporated nanoapatites with facile synthetic techniques is a significant topic. These nanoapatites inhibit drug resistance and biofilm formation, especially during open wound repair. Functional nanoapatites with anitbacterial ability are clinically useful in preventing the proliferation of bacteria with resistance genes and biofilm-producing bacteria, such as S. aureus resistant to asmethicillin and vancomycin, in implants and related in vitro and in vivo medical devices [31]. The World Health Organization has highlighted the importance of these bacteria, listing them as strains that require the most in-depth microbial research. It mentions all key groups that pose a particular threat to hospitals, nursing homes, and patients in need of medical equipment and can cause serious and often fatal infections, such as bloodstream infections and pneumonia. Nanoparticles exhibit broad antimicrobial activity against gram-positive bacteria (Enterococcus, Staphylococcus, and Streptococcus) and gram-negative bacteria (E. coli and Pseudomonas). However, our study and some other studies have shown that nanoapatites are more active against gram-positive bacteria than against gram-negative bacteria (Figures 4 and 5). This difference may be related to the difference in cell wall structure between the bacterial types [56][57][58].
Arul et al. [63] used a microwave strategy to rapidly synthesize Mg 2+ -incorporated apatite nanorods. They found that the incorporation of Mg 2+ ions did not alter the apatite phase, but significantly reduced the crystallinity and particle size by 48% and 32%. Their results were consistent with the results obtained from the 3Mg-nHA test group (Figure 3, second image of middle row and Table 2), which exhibited biocompatibility and antibacterial activity (Figures 4b and 5b). In the agglomerated apatite shown in the TEM (Figure 3, images in the middle row), the crystallinity and particle sizes of the nanoapatites decreased with increasing Mg 2+ concentration.
Shanmugam and Gopal [39] demonstrated that a small amount of copper ions can induce high activity or antibacterial activity in living cells, HA can be replaced by Cu 2+ ions (Ca 10−x Cu x (PO 4 ) 6 (OH) 2 (x = 0.05-2.0), and the antibacterial ratio against S. aureus is 0.2. Their results were confirmed in the present study (Figures 4a and 5a), although their Cu 2+ -substituted nanoapatite was synthesized by precipitation and sintering at 700 • C.
Zn 2+ ion substitution in nanoapatites has been a subject of particular interest because it is present in all biological tissues and plays diverse roles in many biological functions [64]. Bones and teeth contain large amounts of Zn 2+ , and its absorption or release is strongly mediated by bone depots. Tang et al. found that Zn 2+ is substituted in the apatite structure to form an apatitelike phase and then forms an amorphous structure with parascholzite (CaZn 2 (PO 4 ) 2 ·2H 2 O), which is the most favorable condition for the occurrence of Zn 2+ in tetrahedral coordination and Ca2 sites with obvious local structure distortion. In the coordination of Ca1 and P sites with Zn, all higher shells move to longer and shorter distances and lead to elongated crystallites. These results contradict our results in Figure 3, images in the bottom row. Therefore, the fitting to Zn 2+ at the Ca2 site leads to the overall shrinkage of the local structure, most of the higher shells move toward Zn, and thus the high degree of Zn 2+ incorporation resulted in the shorter morphology of the nanoapatite (Table 2). The preparation of metal ion-incorporated nanoapatites with facile synthetic techniques is a significant topic. These nanoapatites inhibit drug resistance and biofilm formation, especially during open wound repair. Functional nanoapatites with anitbacterial Nanoapatite constitutes the major inorganic components of bones and teeth and can thus improve mechanical properties combined with nanotopographic features that mimic natural bone nanostructures. In this study, we discovered nanoapatite with metal ions, which can be functionalized because of its biocompatibility and effects that inhibit bacterial action. Therefore, the synthesized apatite exhibited antibacterial activity when coprecipitated with Cu 2+ , Mg 2+ , and Zn 2+ , especially in the 0.2Cu−nHA nanoapatite, which showed the highest antibacterial activity and best compatibility in the Cu-incorporated group (Figures 4-6), although still exhibiting cytotoxicity under ISO10993-5 test conditions. The Mg 2+ group had the best biocompatibility. The antibacterial effect of nanoapatite incorporated with Zn 2+ was poor and showed cytotoxicity at high Zn 2+ concentrations [59]. Most of the summarized differences in the antimicrobial properties of Zn 2+ can be attributed to the concentrations studied. Therefore, the relative suggestion of the antimicrobial effect of Cu 2+ over Mg 2+ and Zn 2+ in this study is also concentration-dependent and takes into account the biocompatibility. Another possibility that is different from other literature is that the nanoapatite dry powder does not release metal ions but interacts with the bacterial cell surface, thereby changing the cell permeability, or forming an impermeable layer around the cell, preventing essential solutes from entering the cell.

Preparation of Nano-Hydroxyapatite with Metal Ions (Cu 2+ , Mg 2+ , and Zn 2+ )
The hydrothermal synthesis of nanoapatite (nHA), Cu 2+ , Mg 2+ and Zn 2+ ion-doped nanoapatite followed the previously reported research procedure [65]. The concentrations of Cu 2+ , Mg 2+ , and Zn 2+ ions were set at molar levels different from the total molar content of metal cations, as shown in Table 3. Briefly, the prepared cation solutions with different ratios of Ca(NO 3 ) 2 ·4H 2 O and different amounts of Cu 2+ , Mg 2+ , and Zn 2+ ions (as indicated in Table 3) were dissolved and mixed into 90 mL of double-distilled water (ddH 2 O). The pH of the resulting solution was adjusted to 7. Then, 0.713 g of (NH 4 ) 2 HPO 4 and 2.647 g of Na 3 C 6 H 5 O 7 were stirred with 70 mL of ddH 2 O for 30 min to form an anion solution. The aforementioned two cation and anion solutions were mixed and stirred for an additional 30 min. The nanoapatites were synthesized through hydrothermal synthetic reaction at 180 • C and 220 psi (l b/in 2 ) for 12 h. After the reaction, the nanoapatites were washed three times with ethanol and ddH 2 O at 3000 rpm for 5 min each time and then dried at 60 • C for 12 h. The yield of the resulting precipitate in this study was about 30% (w/w).

Characterization of Nanoapatites through Fourier Transform Infrared (FTIR) Spectroscopy, X-ray Diffraction (XRD) Analysis, and Transmission Electron Microscopy (TEM) Morphological Measurements
The functional groups of the nanoapatites were analyzed using a Fourier transform infrared (FTIR) spectroscopy (Nicolet 6700, Thermo Fisher Scientific, Waltham, MA, USA) to confirm the structural composition [66].
The crystalline phases, crystallinity, and preferred orientation changes in the nanoapatites were investigated through X-ray diffraction (XRD) pattern analysis. Diffraction analysis was performed using an X-ray diffractometer (D2 Phaser, Bruker, Billerica, MA, USA), and the diffraction conditions were as follows: Kα diffraction using a Ni-filtered Cu target at a voltage of 30 kV, current of 20 mA, and scan speed of 2 • /min. The scanning range of 2θ was 20-60 • , and the phases were identified by comparing the diffraction patterns with the patterns in the Joint Committee on Powder Diffraction Standards (JCPDS).
The sample was mixed with 99.8% alcohol (Avantor, Inc., Radnor, PA, USA) and dispersed uniformly with an ultrasonic oscillator (DC400H, DELTA ULTRASONIC Co., Ltd., New Taipei, Taiwan). The dispersion solution was dropped onto a copper mesh (TED PELLA, INC., Redding, CA, USA). The morphology of HA was analyzed using transmission electron microscopy (TEM; JEM-2100F, JEOL, Tokyo, Japan) at 200 kV and bright-field images. The microstructure of HA was observed using a selected area electron diffraction (SAED) system with an aperture size of 160 nm, and the crystal structure was analyzed through high-resolution transmission electron microscopy (HRTEM). Digital Micrograph (Gatan) software was used to measure the diffraction plane distance (d). Image-J program (software version 1.53e) was used to analyze the TEM images of each group of samples and calculate the lengths, diameters, and aspect ratios of the nanoapatites.

Antibacterial Abilities
Staphylococcus aureus (ATCC No. 25923) and Escherichia coli (ATCC No. 10798) were cultured in Luria-Bertani broth [67][68][69]. The bacterial suspension was diluted to achieve an optical density at 595 nm (OD 595 ) of 0.2 (equivalent to~1.0 × 10 7 cells/mL on average) and this value was confirmed with an enzyme-linked immunosorbent assay (ELISA) reader (EZ Read-400, Biochrom, Holliston, MA, USA). The bacterial suspensions of S. aureus and E. coli were subsequently diluted to achieve an OD 595 value of 0.2. For quantitative antibacterial testing, 0.03 g of a sample was immersed in 2 mL of bacterial suspension and incubated at 37 • C for 1-4 days. Approximately 100 µL of the bacterial solution was collected from each group, and its OD 595 value was obtained with the ELISA reader (n = 3).

In Vitro Cytotoxicity Tests
The L929 cell line from newborn mouse fibroblasts was provided by the National Institute of Health in Taiwan and used for the cytotoxicity tests. The testing procedures were conducted by ISO 10993-5:2009. L929 cells were grown with media in an incubator at 37 • C and 5% CO 2 and subcultured when the cell concentration was between 0.8 × 10 6 and 1.0 × 10 6 cells/mL. The medium used was a minimal essential medium alpha medium (Gibco, Thermo Fisher Scientific Inc., Waltham, MA, USA) containing 10% horse serum and was changed every 2 days of culture.
The samples prepared for cell culture were sterilized by autoclaving at 121 • C and 1.05 kg/cm 2 (15-20 psi; TOMIN, TM-328, Taipei, Taiwan). The extraction solution was prepared by placing the sterilized nanoapatites, a negative control group of high-density polyethylene (HDPE) and a positive control group of 15 vol.% dimethyl sulfoxide (DMSO; Sigma-Aldrich, St Louis, MO, USA) in the culture medium. Owing to the hydrophilicity of the nanoapatites [70], the extraction weight ratio (g/mL) of the sample to the medium was set at 1:5, and the samples were placed in a 37 • C incubator for 24 h, and the extracted medium was transferred for cytotoxicity testing.
For the quantitative cytotoxicity test, 100 µL of suspension with an L929 cell concentration of 1 × 10 4 cells was transferred to a 96-well microliter plate and cultured at 37 • C for 24 h in an incubator. The medium was removed, and then 100 µL of the sample extract was added. L929 cells were incubated in a 37 • C incubator for 24 h and then mixed with 50 µL of a XTT cell proliferation assay kit (Biological Industries, Kibbutz Beit Haemek, Israel) for a 4 h extension reaction before treatment with another ELISA reader (SPECTROstar Nano, BMG LABTECH, Offenburg, Germany). The measured OD 490 absorbance was proportional to cell viability.
In the qualitative cytotoxicity test, the preparation of the extracts and the culture medium of the control group were the same as those in the quantitative test. Approximately 1000 µL of cell suspension was collected, and L929 cells were transplanted into a 48-well microliter plate at a cell concentration of 1 × 10 5 cells/well. The original medium was cultured in an incubator at 37 • C and 5% CO 2 for 1 day, and then the medium was removed. Sample extract (100 µL) was added and cultured for 24 h, and cell morphology was observed under an inverted microscope (IVM-3AFL, SAGE VISION Co., Ltd., New Taipei City, Taiwan).

Statistical Analysis
Analysis of variance (ANOVA) was performed using IBM SPSS Statistics version 20 (SPSS Inc., Chicago, IL, USA) for the measurement of average longitudinal length, transverse width, length-to-width aspect ratio and antibacterial activity. ANOVA was used in determining whether differences among the means of multiple groups were significant. The estimates of two different variables were used in comparing the differences.

Conclusions
In summary, almost all nanoapatites with metal ions (Cu 2+ , Mg 2+ , and Zn 2+ ) exhibited initial antibacterial activity, the magnitude of which was Cu 2+ , Mg 2+ , and Zn 2+ in descending order. After the addition of metal ions, the main characteristic peaks were slightly shifted due to the smaller size of the HA crystal structure. In particular, the length shrinkage of the nanorods. Cu 2+ , Mg 2+ , and Zn 2+ replaced Ca 2+ , resulting in changes in the lattice structure, deformation of the crystal structure, and a decrease in crystallinity. In this formulation, Cu 2+ , Mg 2+ , and Zn 2+ ions replaced Ca 2+ in the apatite mainly at the longitudinal positions of the nanorods. Antibacterial test results showed that adding Cu 2+ ions with an atomic ratio of 0.1% was more effective than doping Mg 2+ (3%) and Zn 2+ (5-7%) ions, although they both showed gram-positive antibacterial activity. Nanoapatite with all Cu 2+ ion concentrations was toxic, but 0.2% Cu 2+ showed relatively little cytotoxicity in the group. After adding Mg 2+ ions to the nanoapatite, only the 7% Mg 2+ group with the highest doping amount showed slight cytotoxicity and was the test group with the highest biocompatibility. For Zn 2+ ionic groups that exhibit moderate cytotoxicity, doping at 1% of the Zn 2+ groups is expected to be biocompatible, while 3% of the Zn 2+ groups have little cytotoxicity. Therefore, in terms of biocompatibility and antibacterial synergistic function of nanoapatite, the best group is 3% Mg 2+ -doped nanorods. Future studies should consider expanding the testing range; for example, if the 0.2% Cu 2+ doped nanorods with the strongest antibacterial activity in this study were incorporated into other biocompatible medical devices, it might reduce cytotoxicity but maintain bacteriostatic properties.