2020 Volume 61 Issue 8 Pages 1679-1683
The ability of a variety of hydroxyapatite (HAp) sources, including animal bone and calcium phosphate glasses, is demonstrated to immobilize fluorine in slag through calcining processes. The slowly cooled slags were calcined in a furnace along with hydroxyapatite, animal bone, and calcium meta-phosphate glasses. As a result, fluoro-hydroxyapatite was formed, confining the fluorine ions in the formed crystal, leading to the suppression of the dissolution of fluorine ion. Each of the samples demonstrated a low elution of fluorine ions (less than 0.8 ppm), measured by leaching method. The quantity evaluation of formed apatite-like crystal and other subsidiaries were conducted for the treatments by adding recycled animals’ bones as the source of calcium phosphate, as well as reagent HAp and meta calcium phosphate glasses.
Fig. 9 Schematic of fluorine confinement in HAp crystals allowing for the alternation of stable fluoro-apatite (FAp) crystal.
The slag produced as a byproduct of iron smelting contains a high concentration of iron and thus requires further processing to recover this iron. Several treatments have been practiced in many steel refining works.1–3) Among the several processing treatments that have been used to depolymerize and reduce the viscosity of the high-viscous slag consisting of SiO2- and Al2O3-based melts,1,2) one involves the use of CaF2 to reduce the liquidus temperature and reduce the sulfur concentration of the slag.3) However, this process has a low recyclability for source of cement use, as fluorine can elute into the underground water via soil during this process. The concentration of fluorine must be less than 3 wt.% in slag4) and less than 0.8 ppm in the eluted water by dissolution test.5) Confining mobile fluorine ions (F−) into a stable crystalline may prevent their dissolution into water and allow for improved recyclability as secondary materials.
Fluoro-hydroxyapatite (FAp, Ca10(PO4)6F2), which contains alternating layers of CaO and fluorine, and isolated PO4 tetrahedral units with six-membered rings,6) is a stable crystal. It is an advantage that the reduced slag possesses abundant CaO. Therefore, if P2O5 content is additionally provided, the FAp can be formed by confining fluorine from isolated CaF2. As reduced slag does not contain phosphorous pentoxide (P2O5), it must be added to the slag for hydroxyapatite (HAp, Ca10(PO4)6(OH)2) to be created.
As the global population continues to increase, food shortages can be expected.7) Chemically synthesized fertilizers are commonly utilized in grain agriculture and often require phosphorus. Animal bones contain an abundant calcium phosphate matrix and HAp, which partially provides alternation of P2O5 content.7)
This work therefore aims to move elemental fluorine from CaF2 into crystalline FAp obtained by isothermal heating of the reagent HAp with pig bones and a meta CaO–P2O5 glass.
The reagent-grade monoclinic Ca10(PO4)6(OH)2 crystal (99.99%, KISHIDA CHEMICAL Co., Ltd.), pig bones, and 50CaO–50P2O5 glasses in molar percent, which were respectively mixed with conventional reduced slag with 50–60 wt.% of CaO, 10–20 wt.% of SiO2, ∼30 wt.% of Al2O3 (see Fig. 1), and rest of CaF2 with more than 1.5 wt.%, were pulverized into powders of particle size ≤35 µm in diameter using an agate mortar and pestle and then independently isothermally heated. For the purpose of investigating the mixing effect of P2O5 source, we made a sample with a ratio of P2O5 source/slag in wt.% from 0.4 to 2.0 with an insertion step of 0.2–0.6, depending on a kind of P2O5 source.
Ternary phase diagram of the CaO–Al2O3–SiO2 system, including the studied reduced slag.9)
Each powder (mixed) sample of approximately 5 g were put in alumina crucibles and heated in an electrical furnace for 1 h at 1450°C. The samples were then cooled to room temperature for 8 h in air in the furnace. X-ray diffraction (XRD) spectra were taken of the pulverized samples (1–10 mg in weight) at 45 keV and with a 40 mA tube. The results were analyzed using the Rietveld simulation8) to quantitatively identify the molar fractions of the Ca2Al2O7, Ca10(PO4)6F2, Ca3(PO4)2, Ca12Al14O33, and CaF2 crystals in the calcined samples.
The leaching test was demonstrated using the samples with sizes of 2 mm and weighed 20 g. The sample was immersed in 200 mL of ultra-pure water solution contained in a plastic bottle. The solution with calcined slag was shaken with a swing width of 45 mm and at a rotating speed of 200 rpm for 6 h. Then, the remaining solution was filtrated by a membrane filter with a pore diameter of 0.45 µm.
The filtrated liquid sample was adequately diluted to 5–400 times concentrations by adding 3% HNO3 aqueous solution. The amounts of F− ions in the elution liquid were evaluated at least three times for each sample, which were recorded by ion chromatograph method.
A ternary phase diagram of the CaO–Al2O3–SiO2 system is shown in Fig. 1,7) where the studied reduced slag contains, in wt.%, 50–60% CaO, 10–20% SiO2, and ∼30% Al2O3 is highlighted by the hatched ellipsoidal area. Thus, although the composition of the slag resembles cement, which is a possible use for a recycled slag, the large amount of CaO enables to use roadbed material after treatment of confinement of fluorine.
The resulting XRD patterns from the heat-treated powdered slag at HAp-to-slag weight ratios of 0.4, 1.0, 1.5, and 2.0 are shown in Figs. 2(a)–(d), respectively. The gehlenite (Ca2Al2O7), C12A7 (Ca12Al14O33), and FAp (Ca10(PO4)6F2) phases were observed in the samples.
Powder XRD patterns of the heat-treated slag at hydroxyapatite (HAp)-to-slag weight ratios of (a) 0.4, (b) 1.0, (c) 1.5, and (d) 2.0.
The XRD patterns from the heat-treated powdered slag at bone-to-slag weight ratios of 0.4, 1.0, 1.5, and 2.0 are shown in Figs. 3(a)–(d), respectively. The gehlenite (Ca2Al2O7), C12A7 (Ca12Al14O33), and FAp (Ca10(PO4)6F2) phases were observed in the samples. We here note that it is impossible to exclude the process during that OH in the reagent of HAp (Ca10(PO4)6(OH)2) or unexpected and feasible H2O molecule in phosphate glass might react fluorine in CaF2, leading to a formation of HF gas, at high temperature.
Powder X-ray diffraction (XRD) patterns of the heat-treated slag at bone-to-slag weight ratios of (a) 0.4, (b) 1.0, (c) 1.5, and (d) 2.0.
The XRD patterns from the heat-treated powdered slag containing a meta CaO–P2O5 glass-to-slag weight ratio of 0.4, 1.0, and 2.0 are shown in Figs. 4(a)–4(c), respectively. The gehlenite (Ca2Al2O7), C12A7 (Ca12Al14O33), FAp (Ca10(PO4)6F2), TCP (Ca3(PO4)2) phases were observed in the samples.
Powder XRD patterns of the heat-treated slag at meta CaO–P2O5 glass-to-slag weight ratios of (a) 0.4, (b) 1.0, and (c) 2.0, where TCP means tricalcium phosphate, Ca3(PO4)2.
The ion chromatographs of the calcined samples obtained from the dissolution of the mixed HAp and reduced slag at HAp-to-slag weight ratios of 0.8 and 1.5 are shown in Figs. 5(a) and 5(b), respectively, where each peak represents anions in the ultra-pure water solution. The first peak, located near a retention time of 4 min, is due to the presence of F− ions. The concentration of dissolved F− ions in both cases was less than 0.8 ppm, which is within the threshold set by the regulation,10) as is shown in Fig. 5(b). Also, in Fig. 6(a) and (b) the chromatographs of the samples for the pig bone and slag are presented. Here, approximately two times of amount of bone was necessary because the bones contained approximately 50 wt.% HAp.7) As a result, more bone was necessary to form a similar amount of FAp (see Fig. 8(a)). The effect of immobilization of F− are elucidated in the samples of the mixed meta 50CaO–50P2O5 glass and reduced slag at weight ratios of 1.0 and 2.0 are shown in Figs. 7(a) and (b), respectively.
Chromatographs obtained from the dissolution of the reduced calcined slag with an HAp-to-slag weight ratios of (a) 0.8 and (b) 1.5 in ultra-pure water, where the inset photos show the calcined slag before dissolution.
Chromatographs obtained from the dissolution of the reduced calcined slag with a bone-to-slag ratios of (a) 1.5 and (b) 2.0 in ultra-pure water, where the inset photos show the calcined slag before dissolution.
Chromatographs obtained from the dissolution of the reduced calcined slag with meta CaO–P2O5 glass-to-slag weight ratios of (a) 1.0 and (b) 2.0 in ultra-pure water, where the inset photos show the calcined slag before dissolution.
The correlation between the P2O5, which means net weight of HAp, bone, and intriguingly P2O5 in 50CaO–50P2O5 glass (taken from 4 g to 20 g), to slag weigh ratio and resulting fraction of FAp and C12A7 after calcination (before elution) and the concentration of eluted F− are shown in Figs. 8(a) and 8(b), respectively. The calculation was done by estimation of weight percentages in heat-treated slag by the Rietveld simulation for collecting XRD patterns (Fig. 8(a)). Table 1 lists analyzed weight ratio of compound by the Rietveld simulation. Then, the weight ratio between weight of FAp, bone, and CaO–P2O5 (taken 4–20 g) and examined slag (10 g). The addition of compounds such as HAp reagent, pig bone, and CaO–P2O5 glass altered the FAp content without increasing the C12A7, which remained unchanged or decreased slightly with an increased weight ratio of additive. The more moderate curve caused by the addition of bones was due to the reduced net amount of HAp component in bones when compared with pure HAp reagent or CaO–P2O5 glass.
The calculated fractions of (a) FAp and C12A7 and (b) eluted F− with a varied P2O5 to slag weigh ratio. The fractions were calculated of Rietveld analysis method on the obtained XRD patterns of the heat-treated slag mixed with HAp, bone, and calcium phosphate glass.
The amount of F− eluted into the water was dependent on both the weight ratio and type of P2O5 additive, as shown in Fig. 8(b). When FAp is formed by calcination mentioned above, fluorine ions originating from those in the isolated CaF2 are confined in the 2a site in the FAp.11) As a result, less than 0.8 ppm of F− was eluted to the water at HAp additive-to-slag weight ratios of 1.0–1.5.
The scheme of movement of F− from CaF2 to FAp during the calcining process, via adding HAp, is shown in Fig. 9. Overall, including CaO–P2O5 compounds not dependent on crystalline and glassy substances nor regent or recycled HAp was demonstrated to be suppress the elution of F− from slag. While when no material is added to stabilize fluorine in the reduced slag, eluted F− are quantified more than 10 ppm as shown in Fig. 8(b). Furthermore, the use of recycled, dried animal bones as a source of calcium phosphate in the absence of wet process.7) Application of drying processes with unused and natural resources enables fluorine confinement in the stable FAp crystals.
Schematic of fluorine confinement in HAp crystals allowing for the alternation of stable fluoro-apatite (FAp) crystal.
Quantitative structural information is presented for the immobilization of F− from CaF2 to FAp. The addition of HAp (via bone or reagent) or meta Ca-phosphate glasses was demonstrated to facilitate the formation of stable FAp, which resulted in the confinement of F− in the FAp. Powder XRD data and a subsequent Rietveld analysis provided qualitative and quantitative information about the dry conversion process of powdered bones to FAp using isothermal heat treatments. The use of dry processes with unused and natural resources, such as bones, enables calcium phosphate to be obtained sustainably, without depending on the limited and declining resources of phosphorus ores.
The authors would like to thank Prof. Dr. H. Aono (Ehime University) for granting them access to the ion chromatography apparatus.
