Because of small interfacial tension of DCPD and OCP compared to HAP and FAP (fluorapatite) in water, suggest their role to be precursors in HAP and FAP biomineralization. Stability of these phases is mentioned to be understood in terms of the DLVO (electrical double-layer repulsion and van der Waals attraction) theory. Obtained values for both OCP and HAP, the interparticle interaction energy was mentioned to be negative. On the other hand, positive value was obtained for DCPD in all studies, which can explain that OCP and HAP habit for rapid flocculation, whereas DCPD was said to remain dispersed because of Lewis acid/base forces. The Lewis base surface tension parameter is mentioned to be the critical parameter for the difference being much smaller to OCP and HAP compared to DCPD. This difference indicates that surfaces of OCP and HAP are hydrophobic while DCPD surface is a hydrophilic, which may result from the crystal structure [19].
10 Characterization of different calcium phosphates by FT-Raman and X-ray diffraction
Karampas and Kontoyannis (2013), established FT-Raman, IR (infrared spectroscopy), and PXRD methods for the identification of different calcium phosphate phases in their mixtures, and performing also quantitative evaluation of the derivative methods. For the semi-quantitative and the characterization analysis mixture of HAP – OCP – DCPD – DCPA was used. According to the results, Raman spectra gave difference at Raman shift in the range of 1200 cm-1 and 800 m
-1, and also between values of 600 cm-1 and 100 cm-1. When results from the infrared spectra were observed, some difference between compounds was seen between 3600 cm-1 and 3200cm-1, and between values of 1600 cm-1 and 400 cm-1. PXRD
analysis showed that HAP has increase and slow decrease pattern in its counts between 0° and 4° (2θ). Its characteristic peaks are at 25.9° and 31.8°. Elsewhere, PXRD patterns in the range of 7°-25° is quite even which makes differentiation of HAP from other phases extremely difficult. Compared to HAP, OCP has its characteristic peak in very early stage, at 4.7°, whereas its other peaks are significantly smaller and situating at 25.9° to 33.4°. OCP and HAP has very similar XRD patterns and overlapped each other, which caused by their poor crystallinity of the compounds. The characteristic peaks of DCPD are at around 11.7° and 20.8°
which can be clearly identified even in the mixture of different phases. Other high intensity peaks of DCPD are at 29.3°, 41.6°, 42.1° and 48.5°. Anhydrous form, DCPA can be detected from the mixture at approximately 26.2-26.3° (2θ) and 30.1°
but it has also peaks at 13.1°, 30.0°, 40.0° and 52.9° [22]. Fig. 6 and Table II show the XRD patters and corresponding data.
Figure 6. X-ray diffraction patterns of different calcium phosphate phases and their mixture [22].
Table II. X-ray diffraction relative intensities and reflection areas for different calcium phosphates, and their mixtures [22].
11 Precipitation of dicalcium phosphate dihydrate
According to Mekmene et al. (2009), a vast number of studies has showed complex process of calcium phosphate formation where depending on the conditions different phases can be obtained. Parameters affecting to calcium phosphates precipitation are pH, calcium and phosphate concentrations, temperature, ionic strength, the presence of other ions and the precipitation duration [29]. According to Karampas and Kontoyannis (2013), the co-precipitation of the other calcium phosphates or the incomplete transformation of precursor/transient phases have been noted in aqueous calcium phosphate solutions, leading to the co-existence of HAP and other calcium phosphate phases [22].
Formation of DCPD is mentioned to take place at 25 °C between pH 4 and 6, whereas OCP, HAP and CDHA formation are mentioned to take place at higher temperatures, either at 37 °C or 60 °C respectively at pH 6.5 (OCP), or between 7 and 9 (HAP and CDHA) [29].
Ferreira et al. (2003), investigated DCPD precipitation at 25 °C in a batch system using different initial concentrations (0.05 M – 0.3 M) of orthophosphoric acid and calcium hydroxide as equimolar quantities. In their experiments, the formation of DCPD can be divided to five stages depending on the precipitated phase. In the first stage, HAP precipitation takes place immediately when two reagent solutions are
mixed, stabilizing the pH while Ca(OH)2 dissolves. In the second stage, when Ca(OH)2 is completely dissolved HAP growth and pH decrease. The pH decrease depends on the initial reagent concentration. Lower pH value is obtained when initial concentrations are higher. It is because the phosphate ion was consumed to support the formation and growth of HAP, and thus released hydrogen ions from the H2PO4- dissociation. According to the authors, these two stages occurred in very short time, approximately in 150 seconds, whereas HAP crystallizes is mentioned very slowly. In the beginning of third stage, because of pH decrease the first nuclei of DCPD was formed on the surface of the HAP causing slight increase of pH. That is because acid was consumed because of transformation of HAP into DCPD. In addition, during this stage calcium concentration in the solution was also decreasing because of DCPD formation. According to Ferreira et al. (2003), the equation for hydroxyapatite transformation to DCPD can be written as [35]:
𝐶𝑎5𝑂𝐻(𝑃𝑂4)3+ 2𝐻3𝑃𝑂4+ 9𝐻2𝑂 → 5𝐶𝑎𝐻𝑃𝑂4∙ 2𝐻2𝑂 (5)
In the fourth stage, these two calcium phosphates were mentioned to co-exist in the solution which was later confirmed with SEM analysis. In addition, during this stage the pH value and calcium concentration were constant. In the last stage, because of DCPD became more stable than HAP, leading to HAP transformation into DCPD. It results in pH increase, and again decrease of the calcium concentration. According to results obtained from SEM and XRD analysis, gradual decrease of HAP crystals was seen while simultaneous DCPD crystal growth was observed before completely HAP disappearance. It is mentioned that transformation of HAP to DCPD happens because HAP has low degree of crystallinity and is also in metastable equilibrium with DCPD and thus enhance the transformation. The pH increase lead to increase of HAP supersaturation and on the other hand, slowly decrease the DCPD supersaturation. This may lead probably to DCPD dissolution and HAP formation if the process continues for long time such as more than three hours [35]. However, according to Schmidt and Both (1987), DCPD was the first phase to precipitate when pH was in the range of 5.3-6.8 at 25
°C and 50 °C, ionic strength was around 0.1 [29].
Mekmene et al. (2009) investigated influence of different pH and concentrations on the crystalline structure and quantity of calcium phosphate precipitates. In their experiments 0.02 M phosphate concentration and three different Ca/P molar ratios 1.0, 1.5 and 2.0 were used at constant temperature of 20 °C. The initial pH values 5.5, 6.7, 7.5, 8.5 or 9.5 were either kept constant or allowed to change during the reaction time of 3 hours. Reagents used in the experiments were calcium chloride (CaCl2 ∙ 2H2O), and sodium hydrogen phosphate which was neutralized from phosphoric acid with 1 M NaOH, whereas 2 M NaOH was used in constant pH experiments which resulted to produce calcium-deficient apatite (CDHA) with different levels of crystallinity [29].
When pH value was not controlled, spontaneous precipitation of calcium phosphate followed by pH decrease was noted when calcium solution was added. The pH was decreased rapidly and soon slowed down before reach stable level after two hours.
When pH was not controlled during the experiments, final pH was in range of 4.48-5.96, and the main crystalline phase formed was DCPD. Also, XRD-analysis showed traces of OCP in the precipitate in all Ca/P molar ratios when initial pH was 9.5. The precipitation efficiencies of phosphate which depend on the initial pH and the Ca/P molar ratio were between 21 %-85 % [29].
According to Mekmene et al. (2009), calcium phosphate precipitation with all molar ratios and pH values except when initial pH was 5.50 showed that supersaturation level was not achieved. Also, with pH 6.7 low amount of calcium phosphate was obtained when compared to higher pH values of 7.5, 8.5 and 9.5. It only gave low precipitation efficiency which between 21 %-35 %, whereas with higher pH values from 7.5 to 9.5 precipitation efficiencies were significantly higher ranging from 60 %-85%. The highest phosphate precipitation efficiency, 85.3 % was obtained with pH 8.5 with the Ca/P molar ratio of 2.0. With the Ca/P molar ratio of 1.5 the phosphate precipitation efficiency was almost the same 85.15 %.
But remarkable lower efficiency 59.35 % was observed with equimolar concentrations with equimolar concentrations. In other experiments with equal Ca/P molar ratio, the phosphate precipitation efficiency was 26.65 %, 68.15 % and 64.65 % for the initial pH when 6.7, 7.5 and 9.5, respectively [29].
In all cases, higher phosphate recovery efficiency was obtained when pH was kept constant compared to pH drift during the whole process. Phosphate precipitation was 59.65%, with the equimolar solution at pH 6.7, and 99.95 % at pH 9.5 with the Ca/P molar ratio of 2.0. Both XRD and FTIR analysis showed that precipitated calcium phosphate was calcium-deficient apatite with different levels of crystallinity, whereas DCPD was produced with pH drift experiments except pH 9.5. In addition, traces of OCP was also mentioned. Its low intensity peak was mentioned to be approximately in the 2θ angle of 4.73 º, which mainly crystallized when at temperature 50 °C [29].
12 Morphology of dicalcium phosphate dihydrate
Toshima et al. (2015), investigated morphology of DCPD crystals by changing mixing methods. In the experiments, equimolar concentrations, 0.4 M of calcium nitrate [Ca(NO3)2] and ammonium dihydrogen phosphate [NH4H2PO4] were used, both of which initial pH was 5.5. Based on their results, DCPD tended to crystallize in the form of single crystal when initial concentration was higher, whereas it crystallized as agglomerate-like structures which was not change during the crystal growth process at lower initial concentration. Agglomeration at lower concentrations was not typical phenomenon in crystallography because usually it takes place under higher concentration conditions. In one experiment both solutions were added simultaneously, whereas two other experiments were carried out by addition of another solution into another one after which they were let to age for an hour before filtering and drying [32].
According to Toshima et al. (2015), plate-like crystals of DCPD were formed when calcium nitrate solution was added to ammonium dihydrogen phosphate with 1 mL min-1 flow rate. However, some petal-like crystals were also noticed from SEM-images. Petal-like crystals of DCPD together with few plate-like crystals were identified when ammonium dihydrogen phosphate was added to calcium nitrate with 1 mL min-1 flow rate. Only plate-like crystals were obtained when the both solutions were added simultaneously. Therefore, the morphology of DCPD crystals can be affected by changing mixing methods [32]. In Fig. 7 different crystal morphologies are showed when mixing method is changed.
Figure 7. Different morphologies of DCPD crystals obtained when different mixing processes were used. (A) Mixed solutions at once. (B) Calcium ion solution added into phosphate ion solution, (C) phosphate ion solution added into calcium ion solution [32].
Toshima et al. (2014) studied the effects of initial pH and calcium and phosphate concentrations on DCPD morphology. The initial concentrations of calcium nitrate [Ca(NO3)2] and ammonium dihydrogen phosphate [NH4H2PO4] was in the range of 0.02 M to 0.1 M, whereas the initial pH of both solutions was in the range of 4.0 to 7.0. One hour ageing was performed if any solids could be observed. Otherwise, solution was continuously stirred for 24 h before filtration. The results showed that no crystals were observed when the initial concentration of both solutions was 0.01 M even after 24 hours stirring at initial pH under 5.5. With the same concentration, crystals were observed after 24 hours when initial pH was 6.0, and after an hour when initial pH was 7.0. Same kind of behaviour in crystallization was noted in greater concentrations values where crystallization took place more easily at lower initial pH. In all experiments, the pH drop took place whether crystals were produced or not. The difference between the initial pH and terminal pH increased with the increase of the concentrations or initial pH. SEM-analysis showed that formed crystals were either plate or petal-like crystals which depend on the initial conditions. The difference could also be seen with XRD-analysis. As shown in Fig.
8, when as initial concentrations of the solutions were 0.1 M, plate-like morphology
was obtained as initial pH above 5.5, whereas petal-like morphology was obtained with pH 4.5 and mixture was obtained with pH 5.0. Toshima et al. (2014), showed SEM-pictures taken from produced DCPD crystals with different conditions.
Where the difference can be clearly seen, describing that low concentration and low pH formed petal-like DCPD crystals which were agglomerated and formed nest-like structure at the middle of crystal. The plate-nest-like crystals were in parallelogram shapes which were stacked in layers. The length of the crystals was ranged from a few to 100 µm, and the thickness approximately 1-2 µm. The size of petal-like crystals were smaller than plate-like. According to authors, controlling of the early crystallization process, especially the nucleation was the most important step to affect DCPD morphology [36].
Figure 8. Formed DCPD morphology affected by initial pH under various growth conditions [36].
EXPERIMENTAL PART
13 Continuous and semi-batch precipitation of calcium phosphate in a stirred tank and plug-flow crystallizer
Calcium phosphate, dicalcium phosphate dihydrate, DCPD was aimed to crystallize by reaction precipitation in semi-batch, continuous plug-flow and stirred tank crystallizers at room temperature. The initial pH and flow rate were set as variables in semi-batch and plug-flow crystallizer. In addition, effect of ageing to the precipitates of calcium phosphate was also under investigation. In continuous stirred tank effects of residence time and mixing rate on crystallization were studied. Another aim of this study was to compare precipitation of DCPD in plug-flow and continuous stirred tank crystallizers. Attention was also paid to other possible calcium phosphate phases present in the final product.
The precipitated calcium phosphate products were analysed and identified by using powder X-ray diffraction (XRPD). In addition, Scanning Electron Microscope (SEM) was used to identify not only the calcium phosphate phases present in the sample but also the morphology of crystals and presence of by-products.
The changes in morphology could also be seen with microscope used while analysing crystal size distributions with Morphologi 3. In continuous stirred tank experiments, liquid samples were also taken from filtrates and analysed later with Ion Chromatography (IC) to determine phosphate concentration of mother liquor.