1 Introduction

Calcium phosphate-based biomaterials have been used frequently as bone substitutes and osteoconductive scaffolds due to their chemical similarity to the inorganic phase of bone [1]. Currently, they are widely used to promote bone repair and regeneration instead of autogenous (self) bone graft/flab transplantation used for orthopedics, oral and maxillofacial bone defects [2] or implant coatings [3]. Calcium phosphates (CAP) exist in different forms and phases including hydroxyapatite (HA), calcium-deficient HA, tricalcium phosphates (α or β-TCP), calcium monophosphate and biphasic calcium phosphate (BCP) [4]. Carbonated hydroxyapatite (CHA) differs from stoichiometric HA in which carbonate ions substitute for the hydroxyl ion (A type) or phosphate ion (B type) in the apatite structure [5]. Although CHA is more similar to the biological apatites [6], it shows a reduced crystallinity percent, sintering and decomposition temperatures when compared to stoichiometric HA [5, 7].

CHA prepared via sol–gel method offers a distinct advantage of a molecular-level mixing of the calcium and phosphorus precursors, which improves the yield and the chemical homogeneity of the resulting CHA [8]. Furthermore, the high reactivity of the sol–gel powders usually results in the decrease in the heat treatment and the sintering temperatures [9].

β-Tricalcium phosphate (β-TCP) has been used recently as bone and tooth implants due to its unique biocompatibility, osteoconductivity and bioresorbability [10, 11]. β-TCP is mainly prepared via solid-state [12, 13] and wet-chemical methods [14]. In order to synthesize β-TCP, the current preparation methods require the heating of calcium phosphate precursors for very long periods [15] (>24 h) and/or at very high temperatures (up to 1000 °C) [13, 16–18].

Microwave (MW) processing of materials is an advanced processing technology with several advantages [19, 20]. It provides a powerful and significantly different tool to process materials and in most cases improves the performance characteristics of the existing materials. These anticipated benefits include more precise and controlled volumetric heating, faster ramp-up to temperature, lower energy consumption, and enhanced quality and properties of the processed materials [21, 22]. Nano-structured calcium phosphates can be better tailored to obtain desired chemical compositions, surface properties (specifically topography), mechanical properties and distribution of grain size similar to those of physiological bone (which contains 70 % by weight of HA ceramic with grain sizes <100 nm) [23]. MW-assisted synthesized nano-HA was approved for its cell biocompatibility [24] and the ability of bioresorption in stimulated body fluid with a dissolution rate fairly higher than conventional HA and closer to biological apatite due to its nano-structure processing [25]. Others investigated the effects of micro-sized MW synthesized versus nano-sized conventionally precipitated HA (μ/n-HA) on the osteogenic differentiation of rat bone marrow-derived mesenchymal stem cells (rBMSCs) [26]. It was shown that rBMSCs expressed higher levels of osteoblast-related markers with nHA than μ-HA stimulation; thus, the size of calcium phosphates used is an important factor for affecting the osteogenic differentiation of rBMSCs and to obtain more committed differentiated cells. The purpose of the current study was to investigate the possibility to prepare β-TCP from sol–gel-derived CHA via two different routes, the MW treatment and the aging techniques.

2 Materials and methods

2.1 Sol–gel synthesis, aging and MW treatment

For the synthesis of the CHA, calcium nitrate tetrahydrate (Ca(NO3)2·4H2O) and ammonium dihydrogen phosphate ((NH4)H2PO4) were used as the starting materials for Ca and P (Oxford Chemicals, India). Sol–gel-derived CHA samples were reported recently by Bakan et al. [27] and were prepared according to the protocol. In the synthesis process, a solution of Ca(NO3)2·4H2O in ethanol (1.67 M) was prepared, whereas ammonium hydroxide (NH4)OH was used to adjust the pH of the solution (pH = 10) (solution A). On the other hand, a solution of (NH4)H2PO4 (1 M) was prepared and adjusted at pH = 10 (solution B). Solution A was added drop wise to solution B (in equal volumes) at a constant rate (1 mL/min) under vigorous stirring at room temperature, in which the total Ca/P feed ratio = 1.67. Immediately after the solution has transformed into gel, the rate of stirring was adjusted to keep vigorous stirring. A flow chart shown in Fig. 1 outlines the experimental procedure and conditions used to prepare four groups of samples.

Fig. 1
figure 1

Schematic of the methodology to prepare synthetic CHA via sol–gel, classified into four groups

Full size image

The as-prepared group was just dried, neither aged nor calcined. Groups A and B were subjected to an aging process for 24 and 90 h, respectively. Group C was subjected to MW treatment for 10 min using a domestic MW oven working at 2.45 GHz and using 80 % power of the total 1400 W. The three groups (A, B and C) were dried at 80 °C for 12 h and then calcined at 750 °C for 4 h using an electrical furnace at a heating rate of 10 °C/min in air.

2.2 Characterization of the processed CHA samples

All groups were characterized using scanning electron microscopy (SEM). The morphology of the prepared samples (gold coated) was observed using SEM (JEOL 5300-JSM-Japan) operated at 30 keV. Semi-quantitative analyses of the chemical composition of CHA and β-TCP samples were conducted by EDX (energy-dispersive X-ray spectroscopy). Analysis was performed on uncoated powder samples at 15 kV for 60 s. Ca/P mole ratios of the samples were determined as the average of five random runs on each sample. The samples were analyzed using a spectrum 100 FT-IR spectrometer (Perkin Elmer–USA) using KBr pellet technique, wave number region 400–4000 cm−1. XRD was performed using an X’pert diffractometer (Philips–Netherlands), 2θ ranging from 2° to 60°. The phases were identified and compared with the literature reports as well as the Joint Committee on Powder Diffraction Standards (JCPDS) for calcium phosphate samples: CHA, β-TCP and CPP with the PDF numbers: 74-0566, 70-2065 and 45-1016, respectively.

3 Results

In sol–gel synthesis of CHA, the pH of the reactant solutions was found to be very critical. When the pH was lower than 10, a very poor precipitation of amorphous white precipitant with very low yield was observed. On the other hand, solutions adjusted at pH = 10 produced an immediate gel (when added drop wise).

3.1 Characterization of the CHA samples

3.1.1 Morphological and chemical composition analysis (SEM and EDX)

Scanning electron micrographs of the as-prepared sol–gel-derived powder showed the formation of flakes of highly aggregated spherical calcium-deficient carbonated HA in a submicron size (Fig. 2a). Groups A and B after calcinations (750 °C for 4 h) revealed platelet-like morphology with minor percent of rod-like particles (Fig. 2b, c). For group C, the produced particles appeared to be agglomerated spherical particles (diameter about 200–400 nm) and minor rod-like particles (micro-size) (Fig. 2d). Aging was found to significantly increase the particle size; however, MW-treated samples showed minor increase in particle size. Mixed morphologies were also observed in previous studies [28, 29], where several researchers further used other effective chemical (nucleating agents) [30] and physical agents (ultrasonic waves) [7, 31] to control the size and shape of the produced CHA. The EDX analyses performed on the prepared samples are listed in Table 1. Within the accuracy limits of this method, the as-prepared group was identical to the expected Ca/P ratios of CHA [32], while the MW-treated samples (group C) was very close to the expected Ca/P ratios of pure β-TCP [14] (Fig. 3).

Fig. 2
figure 2

SEM of the as-prepared synthetic sol–gel-derived CHA (a) and the processed groups: b group A (aged for 24 h), c group B (aged for 90 h) and d group C (MW treated)

Full size image
Table 1 Mean atomic Ca/P ratio values of the prepared samples using EDX
Full size table
Fig. 3
figure 3

EDX of the MW-treated sol–gel-derived β-TCP powder (group C)

Full size image

3.1.2 Fourier transform infrared spectroscopy (FT-IR)

The FT-IR spectrum of the as-prepared gel group (unaged, un-calcined) showed the formation of a pure CHA powder when compared with the literature [33] (carbonate and phosphate bands are shown in Fig. 4a). The carbonate ion was formed and detected due to the reaction occurred with the atmospheric carbon dioxide in the alkaline pH solutions (pH = 10). The intensity of [CO3]−2 peaks was decreased at high calcination temperature (>400 °C) due to its release as volatile gases [34]. The residual nitrate group (NO3 −) was not detected (peaks 820 and 1380 cm−1) as it was removed through washing [27].

Fig. 4
figure 4

FT-IR spectrum of the sol–gel-derived CHA samples a as-prepared powder (without aging or calcination), b group A: powder aged for 24 h. c Group B: powder aged for 90 h. d Group C: unaged MW-treated powder (10 min). Arrows represent the bands of the phosphate groups related to β-TCP phase

Full size image

The FT-IR spectrum of group A (aged for 24 h and calcined at 750 °C) revealed that the [PO4]−3 band around 1000 cm−1 has appeared with peaks at 1034, 1070 and 1097 cm−1 which was attributed to the asymmetric stretching of (O–P–O). A medium intensity band at about 970 cm−1 which correspond to the symmetric stretching (O–P–O) indicated the degradation of CHA (Fig. 4b) [35]. In addition to the complete dehydration (absence of OH− peak at 1640 cm−1) and the removal of carbonate ions from the apatite structure, the appearance of peaks at 947 and 975 cm−1 (Fig. 4b, arrows) was characteristic of the presence of β-TCP [36]. The peak at 1215 cm−1 was due to the bending mode of the hydrogen belonging to the group (–O3PO–H–O–PO3) [35], while the peaks at 748 and 1165 cm−1 indicated the formation of calcium pyrophosphate (CPP) with P2O7 group [37]. These results are also in accordance with previous reports indicating that the prepared CHA decomposes into β-TCP and CPP and calcium hydrogen phosphate when calcined at temperature above 600 °C [34].

A similar decomposition of CHA was detected for group B samples (aged for 90 h) (Fig. 4c); however, the bands of CPP (748 and 1165 cm−1) were nearly absent. This indicates the formation of β-TCP phase which could be attributed to the positive effect of aging on the properties of sol–gel-derived CHA [27]. On the other hand, group C (Fig. 4d) showed almost the same peaks as in group B with wider (PO4) peak around 1000 cm−1 which is characteristic to the formed β-TCP.

3.1.3 X-ray diffraction (XRD)

The XRD phase analysis pattern of all groups compared with the original β-TCP pattern are shown in Fig. 5. The as-prepared group (Fig. 5b) was identical to the HA/CHA with all its major peaks and in the absence of any other phases (JCPDS-PDF no. 74-0566) [34]. The broad patterns at (002) and (211) indicate that the crystallites are very tiny in nature [33]. The produced CHA was not thermally stable, as it degrades into major amounts of β-TCP and minor amounts of CPP and CaO when calcined at 750 °C (group A) (Fig. 5c) [34]. However, the XRD pattern for group B powder (90-h aging) (Fig. 5d) showed much higher content of the formed β-TCP that appeared as the dominant phase (JCPDS-PDF no. 70-2065). This indicates that the phase type and structure of the sol–gel-derived CHA are dependent on the aging time that affects its phase maturation [9]. On the other hand, group C (unaged, MW-treated for just 10 min) showed the development of pure β-TCP after calcination without the need of any aging time (Fig. 5e). Results also confirmed the absence of any other minor phases (e.g., octacalcium phosphate that appear around 5° (2θ) form the final β-TCP product.

Fig. 5
figure 5

XRD of the sol–gel-derived powder: a β-TCP (JCPDS no. 70-2065) representing the base pattern, b CHA (as-prepared group): unaged and un-calcined with indexing. c Group A: aged for 24 h and calcined; CPP indicated as (black circle). d Group B: aged for 90 h and calcined. e Group C: unaged MW-treated for 10 min and then calcined

Full size image

4 Discussion

Previous studies have demonstrated that pure carbonated HA could be rapidly synthesized after exposing a freshly prepared solution of biphasic HA/β-TCP (under reflux) to sufficient MW radiation power [38]. This was attributed to the assisting of the deaquation necessary to the formation of apatite in aqueous solution via the MW energy that diminishes the bond between the Ca and its hydration sphere [7, 38, 39].

On the other hand, the thermal degradation of HA into α/β-TCP, tetra-calcium phosphates (TTCP), CPP and/or CaO has also been widely discussed [40, 41]. The normally precipitated CHA has two types of water: adsorbed water (reversibly lost at low temperature up to 120 °C) and lattice water (irreversibly lost between 400 and 500 °C) [42]. Further dehydroxylation (OH− loss) should begin when HA is heated at temperatures more than 900 °C leading to further decomposition of HA into oxyapatite (unstable) which upon further heating up to 1200 °C decomposes into β-TCP or TTCP. However, α-TCP and/or CaO are formed only above 1200 °C [40, 43].

Previous studies have also showed that other irradiation-induced phenomena were observed during HA examination by high-resolution electron microscopy [44]. The high-energy electron–crystal interaction was believed to cause a local temperature raise (100–500 °C) of the irradiated particles followed by hydroxyl, Ca and oxygen ions displacement and thus the destruction of HA crystal into CaO [45] or α-TCP [41]. Comparable results were mentioned in a previous study, in which Mg+-stabilized CHA was treated with domestic MW, in which the produced β-TCP was the dominant phase but not as a pure phase [5]. Similarly, other studies have also reported that β-TCP could be detected as an unstable minor phase when CHA was heated between 600 and 800 °C [46, 47]. This low-temperature phase transition could be related to the fact that CHA was less crystalline than the stoichiometric HA [48].

Furthermore, a single-phase β-TCP powders were previously produced by using a MW-assisted “combustion synthesis (auto ignition)/molten salt synthesis” hybrid route, with Ca/P feed ratio equal 3:2 [18, 49]. Other studies have demonstrated that MW irradiation could reduce the maturation time required for HA phase formation through enhancing the biomimetic process that takes place at long aging time [50].

In the present study, the MW treatment had reduced the decomposition time of the unaged sol–gel-derived CHA into a stable major β-TCP phase in just 10 min, which had occurred at longer aging times (90 h). This could be attributed to the effect and the interaction of the MW with the CHA particles on the molecular level, especially with water molecules exist on the CHA structure. It is well known and established that the MW energy interacts and heats the water (dipole) molecules very efficiently due to its high dielectric properties [20, 21]. Hence, this interaction resulted in local temperature raise within the CHA structure and reduced the maturation time required for CHA phase, and consequently, a rapid transformation and destruction of CHA crystal into β-TCP phase took place [44, 45, 47].

5 Conclusions

Sol–gel method was used to prepare CHA powder. The sol–gel-derived CHA powder was then used to produce β-TCP after calcination via two different methods, aging and MW treatment. Aging of the sol–gel-derived CHA for 24 h caused its degradation into β-TCP with minor amounts of CPP and CaO. Meanwhile, longer aging for 90 h caused the formation of a β-TCP as the major phase with mixed morphologies. This longer aging of the sol–gel-derived CHA samples was having a positive effect on the maturation of the CHA structure so it could be easily converted into β-TCP phase after calcination. On the other hand, the MW treatment (for only 10 min) of the freshly prepared (unaged) CHA showed also the formation of β-TCP as a major phase with highly agglomerated morphology. This rapid conversion could be attributed to the MW irradiation-induced effect on the CHA structure and hence enhance its conversion into β-TCP. The preparation of β-TCP via long-term aging process or via a fast and short MW treatment of the sol–gel-derived CHA was successfully achieved.