Chromatographic analysis was performed with the use of a Shimadzu liquid chromatograph equipped with a model LC-2010AHT gradient pump, CBM-20A system controller and Shimadzu CLASS-VP UV-Visible detector (Shimadzu Corporation, Kyoto, Japan). The present study also utilized a Sigma D-37520 high-speed centrifuge (Merck KGaA, Darmstadt, Germany), a Kunshan Shu Mei KQ2200B ultrasonic cleaner (Kunshan Ultrasonic Instruments Co., Ltd., Kunshan, China), a Mettler-Toledo one-hundred-thousandth electronic balance (Mettler-Toledo GmbH, Greifensee, Switzerland), an Elix10 water purification system (Milli-Q Gradient A10; EMD Millipore, Billerica, MA, USA), and a UV-2600 UV-Vis spectrophotometer (Shimadzu Corporation). Sample separation was performed on a Sepax BR-C18 column (250×4.6 mm; Sepax Technologies, Inc., Newark, DE, USA), with a particle diameter of 5 µm.

Levofloxacin was purchased from the National Institutes for Food and Drug Control (cat. no. 130455-201106; Beijing, China). Ciprofloxacin was purchased from Sigma-Aldrich; Merck KGaA (cat no. 17850-5G-F) and applied as the internal standard. Methanol [high-performance liquid chromatography (HPLC)-grade] and tetrabutylammonium bromide (analytically pure) were purchased from Merck KGaA.

The chromatographic separations were performed using a Sepax BR-C18 column (250×4.6 mm; Sepax Technologies, Inc.) with 5-µm particle diameter. The column temperature was set at 40°C, and the mobile phase was prepared by mixing 0.01 mol/l KH₂PO₄, methanol and 0.5 mol/l tetrabutylammonium hydrogen sulphate (proportions, 75:25:4). This was delivered at a flow rate of 1 ml/min. The wavelength of the detector was set at 290 nm, and the injection volume was either 20 µl for associated substances, or 10 µl for assay determination.

The specific production method was described in detail in our previous article (10). Briefly, MSNs were synthesized according to the protocol reported by Argyo et al (17). A total of 0.2 g cetyltrimethylammonium bromide (CTAB) was combined with 0.75 ml 2 mol/l sodium hydroxide (NaOH), and the resultant solution was then added to 100 ml H₂O with constant stirring. Subsequently, 20 ml n-hexane was added into the above solution with stirring (400 rpm). Iron (II, III) oxide (Fe₃O₄) was stabilized with oleic acid. Ethyl acetate (2 ml) and 0.5 ml tetraethyl orthosilicate were added to the solution at 70°C for 3 h. CTAB and Fe₃O₄ separated out as nanocrystals when the pH value of the solution was decreased. The solid product was obtained by filtering, prior to drying under vacuum at room temperature, and these were named as MSNs. Finally, MSNs were resolved in suspension and Levofloxacin (1,500 µg/ml) was loaded by electrostatic attraction.

The n-HA/PU composite porous scaffolds were successfully synthesized using the in situ foaming method (13,18). Initially, 30 g castor oil was combined with 40 g n-HA particles in a nitrogen atmosphere and thorough stirring. Subsequently, 30 g isophorone diisocyanate was added to the suspension at 70°C for 3 h in order to obtain the prepolymer, and 1 ml 1,4-butanediol was used to extend the prepolymer. During this step, the n-HA/PU composite scaffolds were actually obtained. The n-HA/PU was cut into small cuboids (10×6×6 mm) and immersed into the Levofloxacin-MSN suspension at 25°C for 30 min, prior to drying in a vacuum oven at 40°C. Following completion of the above steps, 1 mg Levofloxacin-MSN-n-HA/PU (Lev@ MSN/n-HA/PU) was successfully synthesized.

Levofloxacin (30.00 mg) was weighed precisely and dissolved in simulated body fluid (SBF; Hangzhou Haoxin Biotech, Co., Ltd., Hangzhou, China). Subsequently, the Levofloxacin solution was transferred to a 10 ml volumetric flask and used to obtain a standard solution (3 mg/ml Levofloxacin). The standard solution was diluted into 14 different concentration gradients: 300, 200, 100, 50, 25, 10, 5, 2.5, 1, 0.5, 0.25, 0.1, 0.05 and 0.01 µg/ml. Ciprofloxacin (25.00 mg) was also weighed accurately and dissolved in methanol. The concentration of ciprofloxacin was adjusted to 500 µg/ml, and this was used as the internal standard.

The standard solution of Levofloxacin was prepared as described above. High (50 µg/ml), medium (25 µg/ml), and low (5 µg/ml) concentrations were selected. After the instrument was calibrated to zero, the maximum absorption wavelength was determined by scanning the standard levofloxacin solutions at 200–400 nm.

A total of 14 different concentrations of the Levofloxacin standard solutions (10 µl) were respectively added into 100 µl blank SBF at room temperature. Subsequently, 10 µl ciprofloxacin was added into the working solution as an internal standard. The working solution was vortex-mixed for 5 min, and 800 µl dichloromethane was subsequently added. The solution was further vortex-mixed for 5 min, and centrifuged at 7,155 × g for 5 min at 25°C. An aliquot (750 µl) of supernatant was extracted, and dried with nitrogen in a 50°C water bath. A total of 100 µl of the mobile phase was added into the tube for redissolution. The supernatant was mixed in an ultrasonicator, vortex-mixed and centrifuged at 16,099 × g for 10 min at 25°C. An aliquot (10 µl) of the supernatant was injected into a Sepax BR-C18 column (250×4.6 mm; Sepax Technologies, Inc., Newark, DE, USA) with a constant flow rate of 1 ml/min. Each concentration of the standard product was analyzed 3 times. The ratio of the sum-of-peak areas of Levofloxacin to ciprofloxacin was set as the vertical axis (y) in the standard curve, while the concentration of Levofloxacin was set as the horizontal axis (x). Linear regression analysis was performed using the weighted least-square method (W=1/C²).

Blank SBF (3 ml) was scanned using UV-Vis, and the instrument was calibrated to zero. Different concentrations (0.5, 2.5, 5, 12.5, 25 and 50 µg/ml) of the standard product were scanned by UV-Vis. The absorbance value was set as the vertical axis (y) in the standard curve, while the concentration of Levofloxacin was set as the horizontal axis (x). Linear regression analysis was performed using the weighted least-square method (W=1/C²).

Three different concentration levels of Levofloxacin (low, 5 µg/ml; middle, 25 µg/ml; high, 50 µg/ml) were prepared, respectively. As quality control samples, these were processed and tested using HPLC in accordance with the method described above. The samples were also investigated by UV-Vis. The ratio between the measured concentration and the theoretical concentration was regarded as the method recovery rate.

All the 1 mg Lev@ MSNs/n-HA/PU composite porous scaffolds (n=10) were sterilized with γ-cobalt 60 radiation. The scaffolds were subsequently immersed in EP tubes containing 3 ml modified SBF and incubated for 24 h at 37°. The scaffolds were subsequently removed. The leach liquor was placed in a refrigerator at −20°C prior to testing. Each sample was determined 3 times. Sample processing for the UV-Vis method was performed as follows: The leach liquor was diluted 100 times with SBF, and 3 ml of the resultant solution was placed in a quartz cell, and scanned at a specific wavelength. The absorbance values were then recorded, and a standard curve was then used in order to calculate the antibiotic concentrations.

SPSS version 19.0 statistical software (IBM Corp., Armonk, NY, USA) was used for statistical analysis. All quantitative data are presented as the mean ± standard deviation. Linear regression was used to establish the standard curve of the HPLC and UV-Vis methods, respectively. Paired Student's t-test was used to compare the mean concentration obtained from the HPLC and UV-Vis methods. Pearson's correlation analysis was used to analyze the concentration detected by HPLC and UV-Vis methods. The relative standard deviation (RSD) was calculated according to the following equation: Standard deviation/calculated arithmetic mean ×100%. P<0.05 was considered to indicate a statistically significant value.

The novel biodegradable composite scaffolds containing a large quantity of MSNs were successfully constructed, and the n-HA/PU was cut into small cuboids (size, 10×6×6 mm). A large number of pores were observed in the material. The appearance of the n-HA/PU composite porous scaffolds is presented in Fig. 1A, and a three-dimensional structural diagram of newly fabricated mesoporous silica microspheres is presented in Fig. 1B. Numerous perforated channels were observed above the mesoporous silica microspheres. A large quantity of Levofloxacin was attached to the surface of the microspheres through electrostatic attraction. Ciprofloxacin was set as the internal standard in the process of determination; the molecular structural formula of Ciprofloxacin is presented in Fig. 1C. Likewise, the molecular structural formula of Levofloxacin is presented in Fig. 1D.

The blank SBF containing no antibiotics was regarded as the blank control. Following its examination via HPLC, no clear indication of impurity interference could be identified on the chromatograms (Fig. 2A). The standard solution of Levofloxacin and ciprofloxacin was examined by HPLC, and the results demonstrated that the retention times of Levofloxacin and ciprofloxacin were ~4.84 and 6.74 min, respectively (Fig. 2B). These results demonstrated that the adjusted mobile phase was optimized for the detection of these composite scaffolds, as it presented the shortest retention time, an improved shape to the peaks, and good resolution of Levofloxacin and ciprofloxacin.

SBF was used as the blank control group, and this was established as the experimental group following addition of the standard solution of Levofloxacin. The performance of the blank control group and the experimental group are presented in Fig. 2C and D. Subsequently, the samples were detected by UV-Vis. It was revealed that Levofloxacin at different concentrations had a maximum absorption wavelength of 288.4 nm. Therefore, 288.4 nm was selected as the wavelength for UV-Vis detection of the samples.

The ratio of the sum-of-peak areas of each of Levofloxacin and ciprofloxacin was set as the vertical axis (y) in the standard curve according to the HPLC method. The concentration of Levofloxacin was used as the horizontal axis (x). Linear regression was calculated using the weighted least square method (W=1/C²). The linear concentration range was identified to be 0.05–300 µg/ml for Levofloxacin. The regression equation of HPLC yielded values of y=0.033x+0.010, with R²=0.9991. The chromatograms of Levofloxacin at different concentrations, and ciprofloxacin are presented in Fig. 3A. The standard curve derived from the HPLC method is presented in Fig. 3B. The absorbance value recorded by UV-Vis was set as the vertical axis (y) in the standard curve according to the UV-Vis method. The concentration of Levofloxacin was set along the horizontal axis (x). The regression equation for the UV-Vis method was identified to be y=0.065x+0.017, with R²=0.9999. The linear concentration range was identified to be 2.5–50 µg/ml for Levofloxacin. The wavelength of Levofloxacin with different concentrations is presented in Fig. 3C, and the representative calibration curve for UV-Vis is presented in Fig. 3D.

The low, medium, and high (5, 25 and 50 µg/ml) concentrations of the three standard solutions were determined by HPLC. The recovery rates were 96.37±0.50, 110.96±0.23 and 104.79±0.06%, respectively. The RSD values were revealed to be 0.52, 0.21 and 0.08%, respectively. The same three concentrations of low, medium, and high (5, 25 and 50 µg/ml) standard solutions were also checked by UV-Vis. The recovery rates were demonstrated to be 96.00±2.00, 99.50±0.00, and 98.67±0.06%, respectively. The associated RSD values were 2.08, 0.00 and 0.06%, respectively. All the results of the recovery analysis met the requirement of the Chinese pharmacopoeia, i.e. 80–120%. The measured concentrations and recovery of the two methods are presented in Table I.

The mean concentration of 10 samples measured according to the UV-Vis method was 187.93±33.52 µg/ml, whereas the mean concentration of 10 samples measured using the HPLC method was 30.43±10.27 µg/ml. Paired t-tests were used to compare the results of the two methods. A statistically significant difference was identified between the two methods (t=20.17, P<0.001; Fig. 4A). The concentration of Levofloxacin measured by UV-Vis was set as the vertical axis (y), while the concentration of Levofloxacin measured by HPLC was set as the vertical axis (x). These two results were used for correlation analysis. The linear equation was determined to be y=2.94x+98.57, and the correlation coefficient (r) was identified to be 0.90 (P<0.001; Fig. 4B). Characteristic parameters according to Lambert-Beer's law were examined. The absorbance value (A) has a logarithmic function with the reciprocal of transmittance (1/T) number (Fig. 4C). The peak wave of Levofloxacin could not be seen when the original sample was detected (Fig. 4D), however, after the sample was diluted the peak wave of Levofloxacin was observed (Fig. 4E).