Synthesis and characterization of non-stoichiometric hydroxyapatite nanoparticles using unmodified and modified starches

Non-stoichiometric hydroxyapatite (HAp) presents an additional phase in its structure due to calcium or phosphorus excess, which can influence the material’s mechanical properties, as well as its bioactivity and biodegradability. While stoichiometric HAp, with calcium to phosphorus ratio (Ca/P) of 1.67, has been widely investigated, only a few studies have reported the synthesis of HAp with higher Ca/P ratio. In this work, non-stoichiometric HAp nanoparticles were synthesized using chemical precipitation method followed by a calcination protocol. In order to achieve better process control with chemical precipitation, starch, a natural additive, was applied. Three types of starch were selected for comparison: nonionic starch (NS), soluble starch (SS), and cationic starch (CS). Infrared spectroscopy and chemical analysis results confirmed the non-stoichiometric profile of the synthesized HAp, with a 1.98 Ca/P ratio. X-ray diffraction (XRD) results showed that HAp and calcium oxide (CaO) crystalline phases were obtained and no residual starch was detected. Rietveld refinements confirmed that, for all three types of starch, the content of crystalline HAp was greater than 96.5% and the unit cell volume was not affected. Scanning electron microscopy (SEM) showed agglomeration of particles. Nanoparticle tracking analysis (NTA) results demonstrated that the use of SS produced the smallest particles (approximately 60nm).

Differences in mechanical proprieties, bioactivity and biodegradability are a result of variation in Ca/P ratio in the HAp structure. These differences are related to the presence of a second phase within the material, especially calcium triphosphate (TCP) in HAp with Ca/P ratio below 1.67 and calcium oxide (CaO) when Ca/P ratio is higher than 1.67 (Ramesh et al., 2007;Ravaglioli & Krajewski, 1992), while at stoichiometric ratio these phases aren't observed. Some authors have observed that the presence of CaO in HAp can reduce the material's flexural strength, possibly because stoichiometric HAp presents densification around 99% while non-stoichiometric HAp presents a 97% densification degree after sinterization (Ramesh et al., 2007;Royer et al., 1993). Regardless of the negative effect over its mechanical proprieties, non-stoichiometric HAp can be considered satisfactory for some applications because both CaO and TCP are more soluble in water than pure HAp, thus benefiting the material's biodegradability. The CaO presence stimulates bone formation as the excess of calcium ions promotes HAp nucleation (Kim et al., 2005;Wang et al., 2003). Also, when synthesizing HAp nanoparticles, CaO formed around the particles may be interesting to prevent nanoparticles aggregation (Kakiage et al., 2015;Okada & Furuzono, 2007).
The most commonly used method to prepare HAp is the chemical precipitation method due to its simplicity and low cost (Gomes et al., 2020). A disadvantage of this method, however, is the production of unwanted crystalline phases and low crystallinity of the powders (Yang et al., 2013). Organic additives can be introduced to the synthesis protocol to induce anisotropic growth of the crystal faces after initial nucleation by face-selective adsorption of the additive (Meskinfam, 2011;Sadat-Shojai et al., 2013). An alternative for better process control is the use of additives and/or modifiers in the conventional process, such as cetyl trimethyl ammonium bromide, polyvinylpyrrolidone, and polyethylene glycol.
However, the use of these surfactants has raised concern due to controversial results regarding Research, Society and Development, v. 9, n. 12, e30791210996, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i12.10996 5 toxicity and biocompatibility. In this context, the use of natural additives such as corn starch is promising to produce biocompatible HAp nanoparticles (Masina et al., 2017;Small, 1919;Yang et al., 2013). Water-soluble starch can be used in HAp synthesis because its characteristics facilitate strong adhesion between the HAp and starch (Miculescu et al., 2017).
Starch is a natural biopolymer, biodegradable, biocompatible, water-soluble and inexpensive in comparison to other biodegradable polymers (Jane, 2015;Motta et al., 2020;Sadjadi et al., 2010) and seems to be a good candidate as an additive to produce HAp nanoparticle.
Starch consists of amylose and amylopectin in different percentages, varying according to its botanical source (Jane, 2015;Omori et al., 2014). Starch has several applications in bioceramics synthesis: it can reduce agglomeration, act as a control agent of pore-forming, and improve biological and mechanical properties (Mastalska-Poplawska et al., 2019). Different types of starch (corn, crosslinked, cationic, and calcium-modified) were used by Liu et al. to improve the properties of cement-based calcium phosphates (Liu et al., 2016;Miculescu et al., 2017). Nonionic starch is already widely used to obtain nanocrystalline HAp powders (Meskinfam, 2011;Sadjadi et al., 2010); however, starch modifications can be produced in order to change its physical-chemical and mechanical properties (Jane, 2015;Omori et al., 2014). Cationic starches, for example, are obtained from the reaction with a cationic agent that promotes a physical-chemical transformation responsible for enhanced viscosity and lower gelatinization temperatures (Motta et al., 2020). Soluble starches present high solubility in hot water and can be produced by enzymatic action, acid hydrolysis, or by superheated steam (Small, 1919).
In this work, HAp nanoparticles were synthesized and characterized using starch as an additive. To compare the effect of different types of starch over the synthesis process and the final material, three samples of HAp powders were prepared using nonionic (NS), soluble (SS) and cationic (CS) starches.
The synthesis protocol was based on Lima et al. (Lima et al., 2017) work with Research, Society and Development, v. 9, n. 12, e30791210996, 2020 (CC BY 4. were added dropwise to achieve non-stoichiometric ratio Ca/P (in mol)=1.98. The pH of the mixture was adjusted from 1to10 by NH4OH addition, resulting in a viscous mixture with precipitate. The removal of solvents or supernatant in a drying oven at 100ºC for 72h yielded a raw yellowish material that was pre-calcined at 300ºC for 2h and calcined at 700ºC for 4h in a muffle furnace. After each pre-calcination and calcination process, the particle size was reduced by crushing in an agate mortar. The final product was a white powder. The same preparation protocol used for NS was applied in the syntheses using SS and CS with the difference that for CS the gelatinization temperature applied was 80ºC.
A chemical analysis was performed by inductively coupled plasma with optical emission spectroscopy (ICP-OES) to determine the Ca/P molar ratio. A small amount of each powder sample (0,1g) was digested in concentrated nitric acid, then diluted with distilled water. The samples were analyzed with a Shimadzu atomic emission spectrophotometer (ICP 9000).
An absorption spectrophotometer was used in the infrared region under attenuated total reflectance (ATR) (model FT/IR-4700, Jasco Corporation, Japan) in transmittance, between 400 and 3500 cm −1 , using 32 scans at a resolution of 4 cm −1 .
X-ray diffraction of HAp powders was performed using Panalytical equipment, Nifiltered Cu(kα) radiation (λ=1.5418Å), with an exposure time of 3s/point and step of 0.02°.
Phase identification and Rietveld refinements from XRD data were performed using the The HAp particles were dispersed in isopropyl alcohol (1wt%) and analyzed by Nanoparticle Tracking Analysis (NTA-Malvern Panalytical). 60 seconds videos were recorded for each sample at 25ºC with a red laser application (635 nm), estimating a range of 10-40 particles per image, and the data collected were processed with NTA2.3 Analytical software.

Results and Discussion
Hydroxyapatite is a ceramic material which applicability can be defined by its chemical stoichiometry. The Ca/P molar ratio obtained for each sample is presented in Table 1. For all samples, the HAp obtained was non-stochiometric (with Ca/P ratio higher than the stoichiometric value, 1.67), although there was no relevant difference among the different starches used. The obtained infrared spectra, presented in Figure 1, were similar for all synthesized samples and indicate incorporation of HPO4 2groups in the crystalline lattice, rather than PO4 3groups. Table 2 presents the most significant bands observed in the FTIR spectra with their respective assigned groups. Bands located around 470, 570, 600 and 1050 cm -1 were assigned to PO4 3group, while the band located around 875 cm -1 is attributed to HPO4 2- (Fowler, 1974;Meskinfam, 2011;Omori et al., 2014;Rey et al., 2017), and the band at 1420 cm -1 can be attributed to CO3 2substituted phosphate positions in the HAp lattice (Omori et al., 2014). These observed FTIR peaks suggest the chemical formula Ca10-x(PO4)6x(HPO4)x(OH)2-x for the synthesized HAp (Ishikawa et al., 1993).   Figure 2 shows XRD patterns of synthesized HAp and Table 3 presents Rietveld refinements results obtained about the three studied powders. These results indicate the presence of HAp and residual CaO as nanocrystalline phases. Under all synthetic conditions used in this work, the content of crystalline HAp evaluated by Rietveld refinement was Research, Society and Development, v. 9, n. 12, e30791210996, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i12.10996 9 greater than 96.5%. Furthermore, no residual starch appeared in the powders, which indicates that after calcination all organic content was eliminated (Lima et al., 2017).  The refined crystalline parameters from the diffractogram (Table 3)  Usually, studies about synthetic HAp apply a 1.67 ratio between Ca and P, corresponding to the formation reaction of stoichiometric hydroxyapatite Ca10(PO4)6(OH)2. In this study, a 1.98 Ca/P ratio was used and the excess of calcium justifies the formation of residual CaO phase. The amount of CaO obtained was about 2.3 -3.5 %, corresponding to the 16 % excess of calcium added to the formation reaction. observed. Note that the agglomeration degree doesn't allow a detailed size and morphology analysis of individualized particles and therefore does not allow the identification of nanometric characteristics in the materials via SEM. Thus, these three powders were analyzed using a particle analyzer associated with crystallographic characterizations. represents the number of particles identified for each scanned size in terms of particle concentration. Curves for all studied systems presented one peak (monomodal), regardless of the type of starch used, indicating a monodisperse material. The synthesis using SS produced the smallest particles, with an average diameter of 60nm, and a narrow size distribution curve.
The use of CS and NS yields bigger HAp particles, with 87nm and 105nm, respectively, and broader particle size distributions.  Nonmodified starches (NS and SS) present the highest gelatinization temperature (90ºC) (Motta et al., 2020) and the amylopectin in its composition is associated with the crystallinity of the granules (Jane, 2015). The main limitation of using soluble starch is its high degree of purity, increasing production cost (Nawaz et al., 2020). Usually, modified starches containing cationic and anionic groups are less sensitive for retrogradation and syneresis (Kaur & Singh, 2016) which makes them useful in the industry. Thus, one can achieve gelation temperature more easily and decrease retrogradation and syneresis process when using modified starches.

Conclusions
The high percentage of HAp crystalline phase observed in the synthesized material demonstrated that the method investigated, which used starch as an additive, was a valid synthetic route. CaO phase in the non-stoichiometric HAp makes it more susceptible to degradation, benefiting its application as a biomaterial.
Although SEM micrographs exhibited agglomerated particles with irregular-shape morphology, different crystallite and particle sizes, within the nanometer scale, were obtained for HAp powders synthesized with the three types of starches.
All three types of starch added in synthesis led to nanosized HAp particles, with no difference relevance when comparing the results and considering the standard error.
In order to continue the work, we suggest homogenization and deglomeration of particles within the hydroxyapatite powders using the ball mill and characterization regarding the particles' size and morphology.