2024 Volume 64 Issue 6 Pages 909-919
Supposing to utilize the slag with high P2O5 content as a new phosphorus resource produced from a novel steelmaking process, a fundamental investigation was carried out with particular focus on spontaneous liquid-liquid phase separation occurring in the system CaO–P2O5–FeO for effective phosphorus recovery. At 1773 K, two types of phase separation, which is a double-layered or a dispersed structure of two liquid phases, were observed according to the bulk composition. The double layers separated vertically and consisted of the phase with higher P2O5 content over 40 mass%, and the phase with higher FeO content over 90 mass%. Such structure corresponds to the liquid-liquid separation in the compositional region which is located in higher P2O5 content (upper side) against the tie line between Ca3P2O8–FeO apex on the isothermal section of CaO–P2O5–FeO ternary system. At 1673 K, various separation was observed such as a double-layer or a dispersed structure of two liquid phases, and a coexistence of solid/liquid phases. The slag with the bulk composition lying in the upper two liquid region showed the double-layered structure, giving a promising result for phosphorus recovery through phosphate concentration. The condition for formation of double-layered structure was considered based on the index originally proposed for co-continuity in polymer blends processing, which consists of viscosity and volume fraction of the co-existing phases. The index was found to represent a specific condition for making co-continuous phases in this study and would be significant in view of effective phosphorus recovery.
Phosphorus is an essential element for life since it composes bone, tooth, or even genes. Phosphorus is also important for industries such as plating, surface treatment of metal, semiconductors, and so on. Global consumption of phosphorus increases year by year. One of the reasons is growing needs of fertilizer for foods due to increment of world population. Another reason would be raising demands of bioethanol. Bioethanol is produced from crops like corn, sugar cane or soy bean, so more fertilizer is required to cultivate these plants. Therefore, phosphorus consumption is getting larger.
On the other hand, phosphorus is mainly produced from phosphorus ore. However, phosphorus ore with high quality is non-renewable and exhaustible. Miki suggested that a promising alternative to natural phosphorus ore would be steelmaking slag which contains phosphorus.1) In the steelmaking process, phosphorus originally comes from iron ore and must be sufficiently removed from molten iron into steelmaking slag as a harmful impurity in order to produce high quality steel. Though phosphorus content in steelmaking slag is normally less than 1 mass%, the generation amount is more than 100 million tons per year globally. With respect to the Japanese situation, the amount of phosphorus which is inevitably contaminated and imported within iron ore is about 10 kilo ton. Therefore steelmaking slag contains a huge amount of phosphorus and would be recognized as a potential phosphorus resource. But steelmaking slag is normally used for civil engineering applications such as road constructing materials, which is not requested in terms of phosphorus utilization.
The major components of steelmaking slag are CaO, SiO2, FeOx and P2O5. Basicity is often used to characterize a chemical feature of multicomponent oxide systems, which the simplest index is (mass%CaO)/(mass%SiO2). In the case of steelmaking slag, the basicity can be normally ranging from 2 to 4, depending on processing temperature as to hot metal pretreatment or decarbulization, in view of the benefit for effective removal of phosphorus from molten iron. Under such basicity, phosphorus in the slag normally forms calcium phosphate (Ca4P2O9, Ca3P2O8) or Ca2SiO4–Ca3P2O8 solid solution. These are stable compounds, however, difficult to separate from slag bulk in terms of phosphorus recovery. Several researches have tried to separate such phosphorus compounds through pyro-process.2,3,4,5) Another method to leach phosphorus from the slag by wet process was also reported by Du et al.6)
A new approach is proposed and under development for phosphorus recovery from steelmaking slag, based on a steelmaking process with utilization of high phosphorus-containing iron ore.7) A schematic of the process flow is given in Fig. 1. At first, P-containing steelmaking slag is reduced at elevated temperature using an electric arc furnace (EAF) or rotary kiln. In this reduction process, iron oxide and phosphorus oxide are reduced simultaneously and high-P iron is generated. This P-reduction process is followed by dephosphorization (De-P) process with oxidation refining of the high-P iron. After De-P treatment, high P2O5 content and almost SiO2-free slag is obtained, since high-P iron is not contaminated by Si through P-reduction process. This slag is further served for final P concentration process, based on spontaneous liquid-liquid phase separation occurring in the system CaO–P2O5–FeO.
[CaO–P2O5–FeO] ternary phase diagram in contact with metallic iron, after Muan and Osborn, is shown in Fig. 2.8) This diagram is characterized by a large liquid-liquid separation region presented roughly over 1573 K, which is known as “lens”. The isothermal sections of the diagram at 1873 K, 1773 K and 1673 K are given in Fig. 3 for the compositional range lower than 50 mass% P2O5. One can see two-liquid regions (hatched area) located in the upper side (hereafter referred as the upper liquid) and lower side (hereafter referred the lower liquid) against the tie line between 3CaO·P2O5 (hereafter denoted as C3P) and FeO in the diagram, respectively. In these regions, one of the liquid phases corresponds to high P2O5 content and low FeO content, and the other liquid phase has high FeO content and low P2O5 content. Figure 3 also shows that these two-liquid regions gradually shrink with lowering of temperature. Systematic works by Tromel’s group and other researchers show detailed information on this system and the effect of another component such as Al2O3, SiO2, MgO on the change of said region.9,10,11,12,13,14,15,16,17,18,19,20,21,22)
In order to apply such spontaneous liquid-liquid phase separation for phosphorus recovery, the key information would be separating behaviour how the coexisting phases distribute in the mixture, as well as phase relation from the phase diagram. However, the previous research did not go far to such concept and there is a critical lack about the separation behaviour in this oxide system.
This study aims to utilize steelmaking slag as a candidate for phosphorus resources in terms of phosphorus recovery based on an enrichment technique of phosphate and a following liquid-liquid phase separation in the slag. In the course of the study, a fundamental research was carried out for obtaining basic information on spontaneous liquid-liquid phase separation occurred in the system CaO–FeO–P2O5 in this paper.
The starting materials for the experiment were calcium carbonate, dicalcium phosphate, iron oxide of reagent grade. Those were prepared by weighing and mixing raw materials at a predetermined composition in advance. The starting materials were mixed, ground and pressed into pellet shape. The amount of sample was about 1 g. The aim compositions of the experimental samples are listed in Table 1. These compositions are plotted on the isothermal section of the ternary phase diagram at each temperature in Fig. 4. It is worth mentioning that the lower liquid disappears and is degenerate in the diagram calculated with FactSage, under the condition at Fe activity being unity.
| Sample Code | Exp. Temp.(K) | CaO | P2O5 | FeO | CaO/P2O5 | Note* |
|---|---|---|---|---|---|---|
| 15-1 | 1773 | 30.0 | 30.0 | 40.0 | 1.00 | Upper 2 Liquid |
| 15-2 | 1773 | 40.0 | 20.0 | 40.0 | 2.00 | Lower 2 Liquid |
| 15-3 | 1773 | 45.0 | 20.0 | 35.0 | 2.25 | Lower 2 Liquid |
| 15-4 | 1773 | 48.0 | 32.0 | 20.0 | 1.50 | Lower 2 Liquid |
| 14-1 | 1673 | 28.0 | 30.0 | 42.0 | 0.93 | Upper 2 Liquid |
| 14-2 | 1673 | 14.5 | 15.5 | 70.0 | 0.93 | Upper 2 Liquid |
| 14-3 | 1673 | 33.7 | 36.3 | 30.0 | 0.93 | Upper 2 Liquid |
| 14-4 | 1673 | 38.5 | 41.4 | 20.0 | 0.93 | Upper 2 Liquid |
| 14-5 | 1673 | 20.0 | 30.0 | 50.0 | 0.67 | Upper 2 Liquid |
| 14-6 | 1673 | 35.0 | 30.0 | 35.0 | 1.17 | C3P + Liquid |
| 14-7 | 1673 | 45.0 | 35.0 | 20.0 | 1.29 | C3P + Liquid |
| 14-8 | 1673 | 30.0 | 20.0 | 50.0 | 1.50 | C3P + Liquid |
| 14-9 | 1673 | 48.0 | 32.0 | 20.0 | 1.44 | Solid + Liquid |
The pellet sample was hold in an iron crucible to ensure iron oxide as FeO. The composition of the crucible material (JIS C2504) is listed in Table 2. Though chemical analysis on Fe2+ in the slag was not carried out, chemical analysis on total Fe made for several slags has revealed that the Fe content agreed well with the aimed one. Furthermore, the compositions of the observed phases were reasonably consistent with the phase diagram as described later. So iron oxide in the slag is treated as FeO throughout this paper.
| Composition (mass%) | C | Si | Mn | P | S | Cu | Ni | Cr |
| tr. | tr. | 0.26 | 0.005 | 0.005 | 0.01 | 0.02 | 0.01 |
Figure 5 shows a schematic illustration of the furnace. The furnace was equipped with 12 SiC heaters (Siliconit co., A8-2) and mullite tube (Nikkato co., HB, od50×id42×1000 mm). The iron crucibles containing pellet specimen are mounted on a refractory boat and set in the uniform temperature zone of the furnace which was determined in preliminary heating. The sample was heated at 100 K/min to the experimental temperature in dehydrated high purity argon stream at 100 mL/min, and kept for 2 hr. The temperature was measured with Pt-PtRh13 thermocouple placed within 10 mm of the samples, and controlled within ±3 K.
After holding at the predetermined temperature for the required time, the sample was withdrawn from the reaction tube with the aid of an mullite rod for rapid cooling. Finally the sample was cut together with the Fe crucible, embedded in curing resin and studied using SEM-EDX analysis.
SEM images obtained for the samples heated at 1773 K are given in Fig. 6. Figure 6(a) shows a SEM image of sample 15-1. Two phases in bright and dark contrast are observed, clearly separated into an upper layer and a lower layer. Since the bulk composition of this sample lies in the upper liquid region as shown in Fig. 7, this observation corresponds to the miscibility gap of those liquidus phase.
Figure 6(b) shows a SEM image of sample 15-2. Different phases in bright and dark contrast are intermixed with each other, showing that both phases were liquid phases. The bulk composition of this sample is located within the solid phase (C3P) + two liquid region with respect to the diagram after Muan et al., and within the lower liquid region for the diagram by FactSage as shown in Fig. 8. The present observation may support the latter diagram.
SEM image of sample 15-3 also shows different phases in bright and dark contrast are intermixed with each other as in Fig. 6(c). The bulk composition of this sample is located within the lower liquid region on both aforementioned diagrams.
Figure 6(d) shows a SEM image of sample 15-4. The bright phase with higher FeO content is dispersed in the dark matrix phase, showing that both phases were liquid phases. The bulk composition of this sample is located within the same manner as sample 15-2. The images may also support the diagram from FactSage.
Quantitative spot analysis of the observed phases with SEM-EDX was carried out. Figure 7 shows the compositions of the phases in sample 15-1 plotted on the isothermal section at 1773 K of [CaO–FeO–P2O5] ternary phase diagram after Muan and from FactSage respectively. The bulk composition of the sample is given together in the figure with X mark. The upper layer has a higher P2O5 content, as high as 40 mass%, and a lower FeO content than the lower layer. This is a promising result suggesting an effective phosphorus recovery from such slag by liquid-liquid separation. The composition of the upper phase is rather consistent with the phase diagram drawn from FactSage compared to that after Muan.
In Fig. 8 the composition of the phases determined by SEM-EDX for sample 15-2 through 15-4 are plotted on the isothermal sections of [CaO–FeO–P2O5] ternary phase diagram after Muan and from FactSage. The different compositions of the phases are shown as relatively higher P2O5 with lower FeO content, and lower P2O5 with higher FeO content. These compositions correspond to the darker and the brighter phase in the SEM images respectively. One can find that the compositions of the phases with higher FeO content for these samples are close one another, and is a similar way for the phases with higher P2O5 content. Despite some mismatch in detail, the compositions of both phases are not inconsistent with the diagram from FactSage, in which the bulk compositions lie in lower two-liquid region. On the other hand, with respect to the diagram after Muan, there seem some discrepancies with the experimental results. Although the bulk compositions of sample 15-2 and 15-4 lie within the solid phase (C3P) + two liquid region in the diagram, the compositional analysis gave no more than two types of the phases. Furthermore, the bulk compositions of sample 15-3 lies within the two liquid region, while the compositions of the phases with higher P2O5 content are located far away from the concerned area. Therefore, it should be mentioned, here at this point, that the diagram from FactSage would better reflect the phase relation under the condition in the present study. So the isothermal section from FactSage is employed in the following compositional representation of the observed phases.
3.2. Phase Separation at 1673 KSEM images obtained for the samples heated at 1673 K are given in Fig. 9. Please note that the black area in the figures correspond to holes filled with resin. Figure 9(a) shows a SEM image of sample 14-1, taken at low magnification by choice. Two layers in bright and dark contrast are clearly separated in vertical direction. The bulk composition of the sample is the same as that of 15-1 and lies in the upper liquid region on the diagram from FactSage. This observation shows that such double layered structure of the liquid phase still holds at 1673 K.
Figures 9(b) to 9(d) show the SEM images of sample 14-2, 14-3 and 14-4 respectively. The bulk composition of this sample is located within the upper liquid region. The ratio of mass%CaO/mass%P2O5 are the same of 0.97 for these samples as well as 14-1. In samples 14-2 and 14-3, different phases in bright and dark contrast are separated into an upper layer and lower layer, confirming the phase relation in the diagram. However, the upper layer of sample 14-2 and the lower layer in sample 14-3 formed a discontinuous layer rather than a continuous layer, probably due to the insufficient volume of said phase. In sample 14-4, which composition lies almost on the boundary of the upper liquid region, the brighter phase and darker phase are observed, but the brighter phase is small and is put aside to the inner corner of the crucible.
Figures 9(e) and 9(f) are SEM images of samples 14-5 and 14-6 respectively. The P2O5 content of these samples are the same of 30 mass% as well as 14-1. The bulk compositions lie in the upper liquid region for sample 14-5 and in [C3P + liquid] region for sample 14-6 according to the diagram. In sample 14-5, different phases in bright and dark contrast are clearly separated into an upper layer and lower layer. In sample 14-6, the brighter phase is dispersed in the dark matrix phase.
Figures 9(g) to 9(i) are SEM images of sample 14-7 thorough 14-9. The bulk compositions of these samples are all located outside the upper liquid region. For samples 14-7 and 14-8, the bulk compositions lie on the tie line between C3P and the bulk compositions themselves. Two phases with bright and dark contrast are observed for these samples, those ratio depending on the bulk composition. The darker phase corresponds to higher P2O5 and lower FeO content and this phase forms mostly polygonal shape. Since the bulk compositions of these samples are located within [C3P + liquid] region, this would suggest that solid C3P coexists in the specimen. The bulk composition of sample 14-9 is located on the boundary line, which is originally the lower liquid region at higher temperature. The bright phase in sinuous shape is dispersed in the dark matrix phase, and the dark phases form almost polygonal shape.
The compositions of the observed phases for samples 14-1 to 14-4 analysed with SEM-EDX are shown on the isothermal section at 1673 K in Fig. 10. The compositions of the higher P2O5 phases are close one another among the employed samples, and the higher FeO phases also show similar composition. Both compositions agreed well with the phase boundary on the phase diagram from FactSage. Some scattering of the composition may be due to the difference in tie line between two liquid for the individual slag composition. The P2O5 content of the higher P2O5 phases are larger than 40 mass%, suggesting a promising result for phosphorus recovery by liquid-liquid separation at as low as 1673 K.
Figure 11 shows the results of the spot analysis for samples 14-5 and 14-6. Sample 14-5 gives the composition corresponding to the phase boundary of the upper liquid region as well as sample 14-1. For sample 14-6, the phase with higher P2O5 content, which composition is almost same as C3P, and the phase with higher FeO and lower P2O5 content were determined. This reflects that the phase relation in sample 14-6 consists of solid C3P and liquid, in agreement with the diagram from FactSage.
In Fig. 12, the analysed composition of samples 14-7 to 14-9 are plotted on the diagram. For samples 14-7 and 14-8, two phases are determined at higher P2O5 content and higher FeO content respectively. Though the compositions of the higher P2O5 phases appeared as slightly higher P2O5 content than stoichiochemical C3P, the compositions of the higher FeO phases lie on extension of the tie line between C3P and the bulk compositions. This supports that these samples consist of solid C3P and liquid phase, in accordance with the phase relation on the diagram calculated with FactSage.
In sample 14-9, the brighter phase with higher FeO content is dispersed in the darker solid matrix phase with higher P2O5 content. The bulk composition of this sample almost lies on the phase boundary connecting the high-P2O5 phase and high-FeO phase in the calculated diagram, which boundary might come from degeneration of the lower two liquid at higher temperature. The corresponding high-P2O5 phase gives the composition of 52.1%CaO-37.6%P2O5-10.3%FeO (in mass%) and is output as a liquid phase, while the observed composition of the matrix solid phase of this sample is identified as 50.6%CaO-44.7%P2O5-5.0%FeO. The assignation of the phase is discussed in the following chapter. The composition of the higher FeO phase was also consistent with the diagram.
The phase diagram of the ternary system CaO–FeO–P2O5 after Muan is referred in the previous chapters, as well as the calculated diagram from FactSage. At 1773 K, the upper liquid region and the lower liquid region appear in both diagrams, but the lower liquid region is smaller in the Muan’s diagram than the calculated one. The experimental results corresponded to, if anything, the calculated wider area for said region.
At 1673 K, the upper liquid region and the lower liquid region get shrunk compared to 1773 K. Since lower temperature leads to energy saving in an actual process, the phase diagram at 1673 k is examined in more detail. With respect to the upper liquid region, the phase boundaries of different diagrams are drawn in Fig. 13. One can identify the discrepancy among the diagrams. The area is the largest at the diagram from FactSage, and the experimental results are almost consistent with the phase boundary of the largest area as above.
The lower liquid region becomes small but still exists in the diagram after Muan and Tromel. After Muan, the lower liquid remains in 6–20 mass%P2O5. After Tromel, the lower liquid moves to less than 5 mass%P2O5 (not shown here). On the other hand, in the diagram calculated with FactSage, the lower liquid almost disappears and is degenerates into the tie line between two liquid phases. Both the end compositions of this line agree with the present results for sample 14-9 to some extent, however, the high-P2O5 phase in this sample is recognized to be a solid phase with SEM observation.
Fix15) and Tromel et al.17) examined the variation of coexisting phases in the present system in contact with metallic iron in detail. They reported the precipitation of iron oxide apatite in the system under 1843 K, which coexists with 4CaO·P2O5, C3P, and high FeO liquid slag, and determined its composition as 55.0%CaO-40.9%P2O5-4.1%FeO. Similar result was also reported by Riboud.18) In the calculated diagram, the corresponding phase appears in coexistence with 4CaO·P2O5, C3P, and high FeO liquid slag, though it is provided as a liquid phase. In this study, the high P2O5 phase is recognized to be a solid phase and its composition is close to that of Tromel’s work and also the calculated one to some extent. Therefore, the authors consider this phase in sample 14-9 as solid iron oxide apatite, even if some discrepancy against the calculated diagram occurs.
Although the research may not be sufficient, the authors deem that, through the observation so far, the calculated diagram would be reliable at least for the upper liquid region. With respect to the lower region, especially around 1673 K, consideration on iron oxide apatite might be necessary.
4.2. Co-continuity of Two PhasesSome of the samples employed in this study give P2O5 concentrated phases higher than 40 mass% P2O5. In order to utilize such phases for resource of phosphorus, those should be separated clearly from other mineral phases. In this point of view, the samples with the bulk compositions lying in the upper liquid region would be preferable, since those can be separated in a double-layered manner. However, even if the bulk composition is located in said region, some samples did not form the layered structure. Therefore, the specific condition on formation of layered structure should be clarified for better separation of P-rich phase in this system.
Separation or dispersion behaviour of two liquid phases can be interpreted in several aspects. One is related to a formation of modulated structure owing to spinodal decomposition in semi-miscible system.23) If one considers a dispersed oxide system as emulsion, its criteria between stability and instability may be another viewpoint. Though most of the theories for emulsion are based on aqueous system, the phenomena such as sedimentation/flotation balance or electrostatic repulsion between droplets could be taken into account.
Another key may come from processing of polymer materials such as polymer blends or polymer alloys. On mixing of different viscous polymers, formation of the structure in continuous phase and dispersed phase are controlled for mixed and melted state.24) With respect to molten slag system, polymerized oxide may be involved depending on the composition. In the slag samples employed in this study, P2O5 is known as a typical network former. Therefore, in terms of such analogy, a tentative analysis is made by applying an index on controlling of polymer structure to the layered separation in the present slag systems.
A simple equation was proposed for guiding principle about structure control of polymer blends.25) This equation is a function of viscosity of two coexisting phases.
| (1) |
here η and ϕ is a is a viscosity (in Pa·s) and a volume fraction of each component, and subscripts 1 and 2 correspond to a dispersing component and a continuous component respectively. In case of the value α around unity, both components would form co-continuous phase. The α value larger than unity corresponds to relatively high viscosity for dispersing component, so that dispersing phase would rotate under matrix movement and is difficult to form continuous phase. On the other hand, the α value smaller than unity means high viscosity for matrix component and leads to prevention of coalescence of dispersing component.
Although inorganic oxides are involved in this study, an attempt was made to evaluate a critical condition for formation of continuous phase in slag based on the above equation, assuming a primitive physical phenomena. As for the present slag system, subscripts 1 and 2 go to phases with relatively lower and higher volume fraction, respectively. The viscosity of the slag phase was calculated using Urbain’s model,26) and the calculated viscosity was checked comparing with the literature data27) for CaO–P2O5 melts, FeO melts, and other related melts. The volume fraction of each phase was calculated using the weight fraction read from the phase diagram and the density referred from the literature.27)
In Fig. 14, the α values thus obtained are shown in the isothermal section of the [CaO–P2O5–FeO] system at 1773 K and 1673 K respectively. The α values are given adjacently in the bulk composition of the corresponding samples. The black circle and the white circle represent a double-layered structure and a partially-layered structure, and the X mark denotes a non-layered structure. At 1773 K, the sample 15-1 with its bulk composition lying in the upper liquid region formed fully double-layered structure and gives the α value of 1.67, which is slightly larger compared to the condition for polymer blends. The Eq. (1) was derived for processing of polymer blends under forced flow of the raw materials, so it is not unreasonable if the present results, obtained under gravitational flow, could diverge against the typical value for polymer blends. On the other hand, the samples 15-2 to 15-4 with its bulk composition lying in the lower liquid region give much smaller α values than unity. This difference is mainly due to lower viscosity of the FeO rich phase for the latter samples.
At 1673 K, the calculation of α was done for the samples with those bulk compositions lying in the upper liquid region, since the other samples did not reveal liquid-liquid separation at all and then was regarded out of scope in view of effective phosphorus separation. The α values of the samples formed fully double-layered structure are 1.17 (sample 14-1) and 0.98 (sample 14-5), which are consistent with the typical value for polymer blends. For the samples with a partially-layered or a non-layered structure, the α values are larger than unity. Such difference should be considered in more detail, however, the α values in this study may represent some criterion for the formation of co-continuous phases, here at this moment, even if it occurs rather widely around unity.
Based on the above evaluation, an attempt was made to show an aimed compositional region to obtain a double-layered structure in the present system. According to the present results, the α value from 1.5 to 0.67 (=1/1.5, in case of reversion of phase fraction) is applied to the preferable condition. Under a quite rough estimation, a tentative target may be given as a gray-hatched area in Fig. 14. Such macroscopic criterion would become more worthful in view of effective phosphorus recovery from slag.
Supposing to utilize the slag with high P2O5 content as a new phosphorus resource produced from a novel steelmaking process, spontaneous liquid-liquid phase separation occurring in the system CaO–P2O5–FeO. a fundamental investigation was carried out with particular focus on spontaneous liquid-liquid phase separation occurring in the system CaO–P2O5–FeO for effective phosphorus recovery.
- CaO–P2O5–FeO slag prepared from chemical reagent was heated in an iron crucible at 1673 and 1773 K and was followed by quenching to obtain the specimen for investigating phase separation behaviour.
- At 1773 K, two types of phase separation, which is a double-layered or a dispersed structure of two liquid phases, were observed according to the bulk composition. The double layers separated vertically and consist of the phase with higher P2O5 content over 40 mass%, and the phase with higher FeO content over 90 mass%. Such structure corresponds to the liquid-liquid separation in the region which is located in higher P2O5 content (upper side) against the tie line between Ca3P2O8–FeO apex on the isothermal section of CaO–P2O5–FeO ternary system.
- At 1673 K, various separation was observed such as a double-layer or a dispersed structure of two liquid phases, and a coexistence of solid/liquid phases. The slag with the bulk composition lying in the upper liquid region showed the double-layered structure even at 1673 K, giving a promising result for phosphorus recovery through phosphate concentration.
- The phase diagram of the system reported by the previous researchers, as well as the calculated diagram with FactSage 8.1, were examined in comparison with the experimental results. The calculated diagram was rather in better agreement with the phase separation observed in this study.
- The condition for formation of double-layered structure was considered based on the index originally proposed for co-continuity in polymer blends processing, which consists of viscosity and volume fraction of the co-existing phases. Despite some discrepancy in the criterion in original application, the index in this study was found to represent a specific condition for making co-continuous phases and would be macroscopically significant in view of effective phosphorus recovery.
The authors would like to express their sincere appreciation to Professor Masayuki Yamaguchi, Japan Advanced Institute of Science and Technology, for his helpful comments and suggestions, and a former undergraduate student at Nippon Institute of Technology, Mr. Takuya Naito, for his assistance in the experimental works. This paper is based on the results obtained from a project, JPNP12004, subsidized by the New Energy and Industrial Technology Development Organization (NEDO).
