All reagents used were analytic grades, and the solutions were prepared with milli-Rho water. The Hydroxyapatites synthesis used: calcium hydroxide, (Ca(OH)₂, E. Merck); phosphoric acid (H₃PO₄); urea ((NH₂)₂CO, 99.3%, SIGMA); calcium nitrate tetrahydrate (Ca(NO₃)₂.4H₂O, 99%, Mallinckrodt); diammonium hydrogen phosphate ((NH₄)₂HPO₄, 99.5%, Mallinckrodt); sodium hydroxide (NaOH, 97%, Reagent SA, Laboratorios Cicarelli); sodium fluoride (NaF, 99.2%, J.T. Baker); strontium nitrate anhydrous (Sr(NO₃)₂, 99.0%, SIGMA); calcium nitrate tetrahydrate (Ca(NO₃)₂.4H₂O, 99%, Mallinckrodt); diammonium hydrogen phosphate ((NH₄)₂HPO_4, 99.5%, Mallinckrodt); sodium fluoride (NaF, 99.2%, J.T. Baker).

Two samples of biological apatites were obtained. One of them consisted of a bone calcification of the shoulder tendon and was obtained by open surgery (performed at CASMU Hospital in Uruguay and with the patient’s consent). The second sample was a dental piece obtained by a dental professional from an unknown patient.

X-ray diffraction Spectroscopy (XRD) was carried out using a Philips PW3710 diffractometer CuK radiation. All the hydroxyapatites were finely ground in an agate mortar. The FTIR spectra were obtained using an IR Prestige-21 Shimadzu (Japan). The sample mixture was pressed using a Pike Crush IRTM Technologies at 10 tons.

Nanostructured hydroxyapatite was synthesized according to Kumar et al. (2004). An aqueous solution was prepared using 0.2 M calcium hydroxide (Ca(OH)₂, E. Merck) in 250 mL of 0.12 M phosphoric acid (H₃PO₄), stirring for 2 h at 52°C and pH 5. The suspension was kept at room temperature for approximately 15 h and centrifuged at 40,000 rpm for 15 min. The resulting precipitate was rinsed with milli-Rho water and dried in the oven overnight at 62°C.

Type B carboxyapatite was synthesized according to Wu et al. (2009). A 50 mL solution of urea ((NH₂)₂CO, 99.3%, SIGMA) 2 M and heated at 80°C for 22 h was added to 50 mL of 0.5 M Ca(NO₃)₂ solution drop by drop. After that, 50 mL of 0.3 M (NH₄)₂HPO₄ solution was added slowly to the previous solution using a buret. The mixture was made, stirring constantly at 80°C, and the final pH was 5. The final solution rested for 12 h. Then, it was centrifuged at 5000 rpm for 15 min, and the precipitate was rinsed with water. This procedure was repeated four times. The precipitate was dried in the oven at 60°C for 24 h.

The synthesis followed that reported by Manjubala et al. (2001). A calcium nitrate tetrahydrate 1 M Ca(NO₃)₂ solution was prepared with 5 mol% NaF. The pH was adjusted to 9 with 2 M NaOH. To 50 mL of the resultant solution, 50 mL of 0.6 M H₃PO₄ was added at room temperature at constant stirring (2 mL per minute). The pH was adjusted to 9 with 2 M NaOH. The resultant solution was rested for 24 h at 80°C. Finally, the solution was centrifuged at 5000 rpm (4 times) for 15 min, and the precipitate was rinsed with milli-Rho water and dried at 90°C for 48 h.

Strontium apatite was synthesized according to Li et al. (2007). A 0.2 M Ca(NO₃)₂ solution with 1.5% Ca/Sr molar ratio with Sr(NO₃)₂ was prepared. In parallel, another 0.2 M (NH₄)₂HPO₄ solution was prepared. Both solutions were adjusted at pH 10 with 25% ammonium solution. The (NH₄)₂HPO₄ solution was added to Ca(NO₃)₂ solution in a constant stirring (1.36 mL/min) at 50°C for 5 h. The final solution was rested for 48 h at room temperature. The solution was centrifuged at 5000 rpm for 10 min, and the precipitate was rinsed three times with milli-Rho water and the last with absolute ethanol. Then, it was dried at 120°C in the oven for 12 h.

Following the synthesis of the four species of hydroxyapatites, about ¼ of each powder was finely ground in an agate mortar and mixed with ¾ of KBr for infrared analysis. The resulting mixture was pressed at 10 tons for 5 min t o form a pellet 1 cm in diameter and 0.5 mm in thickness. The FTIR spectra covered the region 1, ν4 phosphate domain (400–700 cm^-1), and region 2, ν1 and ν3 domain of phosphate (800–1,400 cm^-1). The ASCII data was then analyzed using Origin^® software. The spectra were first normalized between 0 and 1, and the second derivative was calculated in the spectral range between 400 and 700 cm⁻¹ (region 1) and between 800 and 1,300 cm⁻¹ (region 2). Finally, the Levemberg - Marquand algorithm was used to identify the hidden spectral bands. The results were used to construct the spectral library.

A Colombian medical materials company’s commercial titanium (Ti, 99%) hydroxyapatite-coated surgical screws were examined. The hydroxyapatites were studied from commercial powder samples and deposits made on the screws given by the company. The screw was coated using an electrophoretic method, widely used to deposit bioceramic coatings in the industry.

We performed an examination of the spectra bands (400–700 cm⁻¹ and 800–1,300 cm⁻¹ region) to identify vibrational modes corresponding to the non-apatitic environments of phosphate. This involved the second derivative of regions identified in the FTIR spectra. The data obtained from the examination, including the first and second derivative, was then used as the starting point for the deconvolution procedure (Figure 2).

The FTIR spectra of synthesized HA, FHA, SrA, and CA are shown in Figure 3. The spectra show strong absorption bands in two characteristic regions at 450 to 700 cm⁻¹ and 800 to 1,300 cm⁻¹ for HA, FHA, CA, and SrA. The bands at 873 cm⁻¹ and the range between 1,420 cm⁻¹ and 1,457 cm⁻¹ represent the characteristic asymmetric stretching of carbonate group type B (CO₃ ^2-) in carbonate apatites (Eichert et al., 2009). The bands corresponding to CO₃ ^2-, both type A and B, and non-apatite were identified in HA, SrA, and FA, indicating carbonate substitution. The peaks were identified at 871 cm⁻¹, 1,429 cm^-1, and 1,470 cm⁻¹ for the HA and SrA, and 872 cm⁻¹, 1,420 years 1,457 cm⁻¹ for FHA. The vibration mode of the free hydroxyl bond bending and stretching band was identified at 630 cm^-1 and the range 3,565–3,570 cm⁻¹, respectively. The bands at 1,600 cm⁻¹ and the broadband in the 2,500–3,700 cm⁻¹ range correspond to the O–H group stretching vibration of absorbed H₂O (Jaafar et al., 2022).

Graphical deconvolution aims to determine the wavenumber of each phosphate band for each region by deconvoluting each curve into n components. This process involves fitting a series of Lorentzian functions using the Levenberg-Marquardt algorithm. The number of bands and the positions of the maximum peaks were identified from the second derivative of each spectral region, resulting in the graphical representation of both the Lorentzian curves of each component and the summed curve of all components (Figure 4).

The deconvolution spectra analysis also allowed us to identify hidden bands in the synthesized apatites (Figure 4). The analysis identified that the vibrational modes for the PO₄ ^3- group are ν₁ (960–964 cm⁻¹), ν₂ (460–474 cm⁻¹), ν₃ (994–1,104 cm⁻¹), and ν₄ (562–604 cm⁻¹) located at the fingerprint region of the spectrum. For example, synthesized HA spectral absorbance showed absorption bands in the range of 1,020 cm⁻¹ and 1,094 cm⁻¹ corresponding to PO₄ ^3- groups and an additional band at 1,109 cm^-1 was identified as having HPO₄ ^2- vibrational modes. FHA showed an absorption band at 864 cm⁻¹, identified as HPO₄ ^2-. CA showed an intense type B CO₃ ^2- band at 873 cm⁻¹ due to substituting PO₄ ^3- for CO₃ ^2- in type B apatite (Eichert et al., 2009).

We constructed a spectral library with the results obtained by the second derivative, to identify the vibrational modes and chemistry environments of apatitic and non-apatitic regions. Table 1 shows the absorption band characteristics of stoichiometric and non-stoichiometric HA. It also includes the bands corresponding to FHA, SrA, and CA. Biological apatites from teeth and bone were also included. We identified apatitic environment bands in stoichiometric and nanocrystalline HA, extra bands corresponding to non-apatitic phosphate environments, and HPO₄ ²⁻.

The analysis of commercial HA in powder and deposited on a surgical screw (Figure 5) allows us to identify changes in the chemical environment of the HPO₄ ²⁻, PO₄ ³⁻, and CO₃ ²⁻ groups.

The spectrum showed distinct absorption bands that can explain the low adhesion on the surface (Figure 6).

The ν₄PO₄ region between 500–700 cm⁻¹ of commercial HA shows absorption bands at 564 cm⁻¹, 572 cm⁻¹, and 602 cm⁻¹, characteristics of PO₄ ^3- apatitic environment, and at 632 cm⁻¹ that correspond to structural OH⁻ group. The same bands are identified in HA deposited on the screw spectrum, except for the band at 572 cm⁻¹ (Figure 7).

From the deconvoluted graph of the commercial HA in the 500–700 cm⁻¹ range, we identified a band at 533 cm⁻¹ and 551 cm⁻¹ corresponding to the non-apatitic and apatitic regions of HPO₄ ^2- vibrational group. However, during the deposition, there was a selective loss of non-apatitic regions, and two peaks were observed at 574 cm⁻¹, and 540 cm⁻¹, corresponding to the apatitic environment of ν₄PO₄ and HPO₄ ^2-, respectively (Figure 8). In the fingerprint of the non-apatitic phosphate environment in the ν₄PO₄ domain (Eichert et al., 2009), characteristic bands are not shown in HA on the screw.

The deconvolution analysis of the ν₁ν₃PO₄ region between 800 cm⁻¹ and 1,400 cm⁻¹ of commercial HA shows a peak at 962 cm⁻¹ and a complex absorption band that extends between 1,030 and 1,082 cm⁻¹ that includes bands of phosphate groups in apatite environment. In addition, hydroxyapatite deposited on the screw shows an overlapping band between 1,004 and 1,031 cm⁻¹ corresponding to the phosphate apatite environment (Figure 8). Between 1,043 cm⁻¹ and 1,106 cm⁻¹, the bands shown from the deconvolution correspond to the PO₄ ³⁻ vibrational band.

Finally, the commercial HA shows a meager absorption band centered at 1,415 cm⁻¹ assigned to the ν₃CO₃ B and a shoulder at 870 cm⁻¹ corresponding to ν₂ CO₃ B; both disappeared in the HA on the screw. The ν₂CO₃ IR domain determines non-apatitic carbonate environments (Eichert et al., 2009). The OH⁻ sharp absorption band at 3,573 cm^-1 is observed in commercial HA as well as deposited on the screw.