Distinct Difference in High-temperature Characteristics between Limonitic Nickel Laterite and Ordinary Limonite

Abstract

In order to achieve the effective utilization of limonitic nickel laterite for stainless steel production at lower cost, the distinct difference in high-temperature characteristics between limonitic nickel laterite and ordinary limonite has been systematically expounded via compact sintering method and the comparative analysis of the relevant sintering performance at the suitable basicity has also been conducted through sinter pot tests. The results indicate that limonitic nickel laterite possesses much poorer assimilability and liquid phase fluidity due to the abundant high-smelting minerals compared with ordinary limonite. The strength of bonding matrix of limonitic nickel laterite is relatively better as the basicity is not exceed 1.4 while that of ordinary limonite is maintained at a higher level with the basicity of no less than 1.8. Meanwhile, the former is far weaker than the latter due to the much less amount of liquid phase, rather higher porosity and looser microstructure of the bonding matrix. Limonitic nickel laterite is identified as a far more refractory ore for sintering compared with ordinary limonite, further supported by sinter pot tests. It is essential to strengthen limonitic nickel laterite sintering from the standpoints of how to promote the liquid phase formation and densification of the sinter.

1. Introduction

The increasing demand for stainless steel worldwide leads to the gradual depletion of nickel sulfide resources as the major nickel source.^1,2) Thus, the abundant nickel laterite resources with about 70% share of the total nickel resources have been given serious emphasis.^3,4) Meanwhile, due to the serious shortage of the domestic nickel laterite resources and strict export limitation of high-grade nickel laterite resources overseas, limonitic nickel laterite accounting for 60% of the total nickel laterite resources has become the major type of imported nickel laterite in China.^5,6) As indicated in the previous studies,^7,8,9) the suitable pyrometallurgical smelting processes of limonitic nickel laterite is confirmed the sintering-blast furnace (BF) process rather than Rotary Kiln-Electric Furnace (RKEF) process owing to its high iron and low nickel grades. However, the exceedingly poorer sintering performance of limonitic nickel laterite has been a huge barrier to the ferronickel alloy smelting at lower cost via sintering-blast furnace process, which is even far worse than that of ordinary limonite.^10,11) Predictably, the sintering behavior of limonitic nickel laterite and ordinary limonite should be very different. Hence, it is of great necessity to reveal the difference of limonitic nickel laterite and ordinary limonite during sintering.

It is generally known that high-temperature characteristics of ores mainly including assimilability, liquid phase fluidity, strength of bonding matrix and formation characteristics of liquid phase are considered as the main evaluation indexes indicating the effect of ore properties on its sintering performance.^12,13,14) The assimilability is characterized as the reactivity of ores with CaO reflecting the difficulty of liquid phase formation during sintering, whereas liquid phase fluidity represents the effective bonding area of liquid phase and is essential to sinter consolidation.^15,16,17) The strength of bonding matrix is used to describe the bonding ability of liquid phase to solid particles and the self-strength of bonding phase.^18,19) Formation characteristics of liquid phase demonstrate the properties of bonding phase including the morphology, component and formation amount.^20,21) All of which has significant influence on sintering process. Therefore, it is a practical approach to reveal the different sintering behavior of limonitic nickel laterite and ordinary limonite by the comparative analysis of their high-temperature characteristics. In the previous study,¹¹⁾ the effect of basicity on sintering performance of limonitic nickel laterite and the optimum sinter indices were reported. However, the high-temperature characteristics such as assimilability, liquid phase fluidity, strength of bonding matrix have not been involved and the comparison of high-temperature characteristics of limonitic nickel laterite and ordinary limonite has also not been investigated yet.

Thus, in this paper, based on the characterization of limonitic nickel laterite and ordinary limonite, the difference of high-temperature characteristics of the two ores such as assimilability, liquid phase fluidity, strength of bonding matrix and formation characteristics of liquid phase was clarified through compact sintering method and the relevant sinter pot tests were also carried out for the mutual authentication, not only contributing to the better understanding of the distinctive sintering characteristics of limonitic nickel laterite but also providing good guidance for improving its sintering performance.

2. Materials and Methodology

2.1. Raw Materials

As shown in Table 1, the adopted ores include limonitic nickel laterite and ordinary limonite imported from Philippines and Australia, respectively. The former contains 0.86% total nickel, 3.45% Cr₂O₃ and 45.09% total iron and its LOI (Loss on ignition), SiO₂, Al₂O₃ and MgO contents reach as high as 12.49%, 5.70%, 4.50% and 5.58%, respectively, 100% of which is used to produce nickel-bearing pig iron by middle BF in stainless steel plant. The latter mainly consists of 56.67% total iron, 5.76% SiO₂, 1.28% Al₂O₃ and 10.65% LOI, which is generally mixed at up to 50% ratio with other iron ores and used to produce pig iron by large BF in steel plants. Compared with ordinary limonite, limonitic nickel laterite possesses higher LOI, lower iron grade and richer high-smelting minerals such as Cr₂O₃, Al₂O₃ and MgO. The primary size distributions of the two ores are presented in Table 2. During compact sintering tests, the two ores are both crushed to below 0.5 mm for eliminating the interference of different size distributions. Similarly, the fluxes include calcium hydroxide and calcium oxide reagents with chemical grade produced from Aladdin company and their fractions passing 0.074 mm are both maintained at 100%. The detailed data of size distributions of the used ores and fluxes are shown in Tables 3 and 4, respectively.

Table 1. Chemical compositions of the two ores (wt-%).

Samples	Fetotal	Nitotal	Cr2O3	SiO2	CaO	Al2O3	MgO	P	S	LOIa)
Limonitic nickel laterite	45.09	0.86	3.45	5.70	0.12	4.50	5.58	0.001	0.011	12.49
Ordinary limonite	56.67	–	–	5.76	0.08	1.28	0.10	0.004	0.012	10.65

^a)  LOI: Loss on ignition of dried samples at 1000°C in air atmosphere.

Table 2. Primary size distributions of the two ores (wt-%).

Size/mm	+8	6.3–8	5–6.3	3–5	1–3	0.5–1	0.25–0.5	–0.25
Limonitic nickel laterite	6.10	4.06	6.00	22.35	44.78	6.43	5.83	4.45
Ordinary limonite	21.96	5.67	11.31	14.76	26.19	8.80	4.39	6.92

Table 3. Size distributions of the two ores after crashing and filtration (wt-%).

Size/mm	0.15–0.5	0.074–0.15	0.045–0.074	0.038–0.045	−0.038
Limonitic nickel laterite	5.92	40.36	18.59	23.95	11.18
Ordinary limonite	6.35	39.88	19.27	23.89	10.61

Table 4. Size distributions of the fluxes after crashing and filtration (wt-%).

Size/mm	0.045–0.074	0.038–0.045	−0.045
Calcium hydroxide reagent	35.80	14.85	49.35
Calcium oxide reagent	36.28	15.19	48.53

Figure 1 compares the mineralogy of the two ores. The XRD analysis was conducted by means of Simens D500 automatic X-ray diffractometer (Siemens AG, Berlin, Germany) with copper target, which was operated at 40 kV and 250 mA in step mode with 0.02° 2θ step and a count time of 0.5 s per step over a 2θ range from 10° to 80°. Subsequently, the software of MDI Jade 6.5 was used to analyze the XRD data. The scanning electron microscopy and energy dispersive spectrum (SEM-EDS) analyses were made by means of environmental scanning electron microscope (ESEM, FEI Quanta-200, FEI, Hillsboro, OR, USA) equipped with an EDAX energy dispersive X-ray spectroscopy (EDS) detector. As described in Figs. 1(a) and 1(b), limonitic nickel laterite primarily contains goethite (FeO(OH)), maghemite (Fe₂O₃), hematite (Fe₂O₃), enstatite (MgSiO₃) and stishovite (SiO₂) whereas ordinary limonite mainly consists of goethite, hematite, kaolinite (Al₂O₃·2SiO₂·2H₂O) and quartz (SiO₂). Combined with Table 5, Figs. 1(c) and 1(d), alumina in limonitic nickel laterite mostly exists in iron-bearing minerals by the partial substitution of Al³⁺ for Fe³⁺ ions and magnesium mainly occurs in enstatite and Cr-spinel. The iron occurrence of limonitic nickel laterite and ordinary limonite is similar, except for silicon. Silicon containing minerals in limonitic nickel laterite are lumpy enstatite and amorphous stishovite, which in ordinary limonite include kaolinite and quartz. The detailed mineral compositions are shown in Table 6. The mineral compositions of ordinary limonite and limonitic nickel laterite are different due to the different ore deposit genesis.^22,23) Compared with ordinary limonite, limonitic nickel laterite is formed under the much wetter and rainier climatic conditions. Therefore, limonitic nickel laterite contains much higher amount of goethite. In addition, Mg-rich peridotite is the material source basis of the formation of nickel laterite and the long-term weathering process contributes to the formation of limonitic nickel laterite with relatively higher iron grade on the upper strata of nickel laterite. Although the MgO content is reduced to some content, it in limonitic nickel laterite is still higher than that in ordinary limonite. Furthermore, the lattice substitution in limonitic nickel laterite is more active than that in ordinary limonite whereas iron and chromium are always associated with nickel, which leads to nickel, chromium and alumina massively occurring in iron-bearing minerals such as alumogoethite and Cr-spinel. Thus, limonitic nickel laterite contains much higher amount of minerals such Cr₂O₃, Al₂O₃ and MgO. The different mineralogy of the two ores is expected to have significant impact on their high-temperature characteristics.

Fig. 1.

Mineralogy of the two ores (a and b are the X-ray diffraction patterns of limonitic nickel laterite and ordinary limonite, respectively; c and d are the micrographs of limonitic nickel laterite and ordinary limonite under scanning electron microscopy (SEM), respectively; G-Goethite, He-Hematite, Ma-Maghemite, E-Enstatite, Si-Stishovite, C-Cr-spinel, Ka-Kaolinite, Q-Quartz). (Online version in color.)

Table 5. EDS analysis results for areas in Fig. 1.

Area No.	Elemental compositions/(atomic conc, %)								Mineral phases
	Fe	Cr	Ni	Mg	Al	Si	Ca	O
1	0.11	–	–	0.12	0.14	33.58	–	66.05	SiO2
2	35.46	0.35	0.09	0.29	0.81	0.15	0.08	62.77	Fe2O3
3	37.27	0.31	0.13	0.48	1.16	0.17	0.11	60.37	Fe2O3
4	30.59	0.62	0.17	0.26	1.02	0.12	0.18	67.04	FeO(OH)
5	0.55	–	–	17.08	0.45	20.19	–	61.73	MgSiO3
6	13.64	20.23	–	6.94	11.75	–	–	47.44	(Fe, Mg)·(Cr, Fe, Al)2O4
7	32.39	0.75	0.18	0.38	1.27	0.21	0.16	64.66	FeO(OH)
8	35.21	–	–	0.52	4.93	0.48	0.08	58.78	FeO(OH)
9	0.60	–	–	0.02	–	38.44	–	60.94	SiO2
10	39.66	–	–	0.05	0.57	0.34	0.01	59.37	Fe2O3
11	37.87	–	–	0.45	3.42	0.67	0.04	57.55	FeO(OH)
12	39.27	–	–	0.20	0.61	0.21	0.03	59.68	Fe2O3
13	0.51	–	–	0.09	17.38	33.60	0.07	48.35	Al2O3·2SiO2·2H2O

Table 6. Mineral compositions of the two ores (%).

Minerals	Goethite	Hematite	Maghemite	Enstatite	Stishovite	Cr-spinel	Kaolinite	Quartz
Limonitic nickel laterite	71.43	2.05	4.45	18.39	1.25	2.43	–	–
Ordinary limonite	52.88	38.36	–	–	–	–	2.01	6.75

2.2. Experimental Procedure

2.2.1. High-temperature Characteristics

Referring to the relevant investigation,²⁴⁾ compact sintering tests were carried out in the horizontal tube furnace (GSL-1400X, Kejing Material Technology Co., Ltd, Hefei, China) with 60 mm in diameter and 1000 mm in length for revealing the difference of high-temperature characteristics of the two ores. The detailed methods were shown as follows.

(1) Assimilability

For the determination of the assimilability, the ore powder and calcium oxide reagent were compressed into the cylindrical compacts with 8 mm and 25 mm in diameter, respectively, and both 10 mm in height at a set pressure of 200 N. After the compacts were dried, the ore compact was put up on the calcium oxide compact and then put into the horizontal tube furnace and roasted at various temperatures (1235–1320°C) for 5 min in the air atmosphere. As the roasting process was finished, the compacts were slowly removed from the furnace in 5 min and then cooled to room temperature in the air. The beginning of assimilation was characterized by the earliest obvious corrosion on the contact interface between ore and calcium oxide compacts. This roasting temperature was just the lowest assimilation temperature as the valuation index for the assimilability. The increase of the lowest assimilation temperature represented the deterioration of the assimilability.

(2) Liquid phase fluidity

In order to investigate the liquid phase fluidity, the ore powder and calcium hydroxide reagent were mixed with water at 4.0 basicity and compressed into the cylindrical compacts with 10 mm in diameter and 8 mm in height at a set pressure of 200 N. After dried, the ore compact was placed upon the corundum tablet with 40 mm in diameter and 2 mm in height and then put into the horizontal tube furnace roasted at various temperatures (1235–1400°C) for 5 min in the air atmosphere. After followed the same cooling method above, the cooled samples were obtained and the liquid phase fluidity index was ascertained in accordance with the method of the previous studies.^17,24) The higher liquid phase fluidity index indicated the better liquid phase fluidity.

(3) Strength of bonding matrix

To clearly expound the variations of the strength of bonding matrix of the two ores at various basicity (0.6–2.2), the ore powder and calcium hydroxide reagent were mixed with water in different proportions and then compressed into the cylindrical compacts with the same specification as the above. Next, eight dried compacts were put into the porcelain boat with a volume of 8 ml and roasted at 1310°C for 5 min in the air atmosphere in the horizontal tube furnace. After the compacts were cooled to room temperature via the same cooling method, the average compressive strength of the compacts (i.e. the strength of bonding matrix) was mensurated by means of the automatic machine for compression tests (ZQYC, Central South University, Changsha, China).

(4) Formation characteristics of liquid phase

According to the former section, sintered compacts with different basicity were prepared again to clarify the formation characteristics of liquid phase by the mineralogical analysis. Firstly, the sintered compacts were embedded by the acrylic powder and epoxy resin curing agent with the ratio of 1:1.8 in the relevant rubber mold and then made into cylindrical samples with 30 mm in diameter and 15 mm in height at room temperature. After polished by the semi-automatic grinding and polishing machine (Tegramin-25, Struers Ltd., Shanghai, Denmark), the samples were placed in the optical microscopy (Leica DM4500P, Leica Camera AG, Solms, Germany) by using reflected plane polarized light for optical photographs. The mineral compositions and porosity of sintered compacts were determined by the software of Image-Pro Plus 6.0 via the area calculation method due to the different colors of minerals and pores presented into optical micrographs. For each condition, about 50 sheets of optical micrographs with the magnification of 50× were shot evenly on the sintered samples for the measurement of the porosity whereas about 200 sheets of optical micrographs with the magnification of 200× were gained for the determination of mineral compositions in the same way, which can commendably ensure the accuracy and facticity of the statistical results on the conditions of covering all areas of the samples. Each mineral in samples can be more favorable to be distinguished by use of optical micrographs with higher magnification (i.e. 200×). In addition, it should be noted that some pores in the bonding matrix are filled by resin during the preparation process of polished cylindrical samples for optical micrographs. Thus, the marked areas of pore (P) and resin (R) both should be recorded as pores. Then, the SEM-EDS analyses were finished according to the relevant method in section 2.1.

2.2.2. Sinter Pot Tests

Based on the above results, sinter pot tests of 100% limonitic nickel laterite or ordinary limonite were conducted at the respective suitable basicity including proportioning, mixing, granulation, ignition, sintering, cooling, crushing, dropping, sieving and quality testing of sinter. The detailed methods were completely consistent with the descriptions of the previous study.¹¹⁾ Through the optimization of mixture moisture and anthracite dosage, the optimum sintering performance of the two ores was compared to achieve the mutual authentication with the results of high-temperature characteristics.

3. Results and Discussion

3.1. Comparison of High-temperature Characteristics

3.1.1. Assimilability

As illustrated in Fig. 2, the lowest assimilation temperatures of limonitic nickel laterite and ordinary limonite are 1290°C and 1250°C, respectively. Furthermore, although the reaction temperature of limonitic nickel laterite is higher, the erodibility between limonitic nickel laterite and calcium oxide is far less than that of ordinary limonite. Compared with ordinary limonite, limonitic nickel laterite possesses much poorer assimilability. This is mainly due to that the excessive high-smelting minerals including Cr₂O₃, Al₂O₃ and MgO dramatically increase the formation temperature of the primary melt during sintering.^11,25) Consequently, limonitic nickel laterite is harder to react with calcium oxide to form liquid phase even at a higher roasting temperature. Thus, the insufficient formation of liquid phase and more solid fuel consumption are apt to be obtained in limonitic nickel laterite sintering, eventually leading to poorer sinter strength and higher sintering cost.

Fig. 2.

Assimilation temperature measurement of the two ores. (Online version in color.)

3.1.2. Liquid Phase Fluidity

According to the description of Fig. 3, the liquid phase fluidity indexes of the two ores are both improved with the roasting temperature, but they have different sensitivity to temperature. As the roasting temperature rises from 1235°C to 1310°C, the liquid phase fluidity index of ordinary limonite is sharply increased from 0.04 to 11.45. However, that of limonitic nickel laterite is slowly elevated from 0.21 to 3.61 with the roasting temperature raised from 1280°C to 1370°C and then rapidly improved to 11.27 when the roasting temperature is further increased to 1400°C. Meanwhile, compared with ordinary limonite, limonitic nickel laterite possesses much lower liquid phase fluidity index at the same or even higher roasting temperature. Combined with Fig. 4, it is fairly clear that the liquid phase fluidity of limonitic nickel laterite is far below than that of ordinary limonite. This is mainly because that the contents of Cr₂O₃, Al₂O₃ and MgO of limonitic nickel laterite are much higher whereas the SiO₂ content of the two ores are very close. The extensive high-smelting minerals have poor reactivity during sintering and significantly increase the viscosity of liquid phase.^16,26) The considerably poorer liquid phase fluidity is extremely harmful to the bonding of solid phase by liquid phase and the homogenization of liquid phase compositions during limonitic nickel laterite sintering, which has great adverse impact on sinter consolidation.

Fig. 3.

Liquid phase fluidity indexes of the two ores at various roasting temperatures. (Online version in color.)

Fig. 4.

Schematic diagram of liquid phase fluidity of the two ores. (Online version in color.)

3.1.3. Strength of Bonding Matrix

Figure 5 demonstrates that the variations of the strength of bonding matrix of the two ores with the basicity are absolutely different. As the basicity is elevated from 0.6 to 1.0, the strength of bonding matrix of ordinary limonite is rapidly worsened from 59.96 MPa to 55.99 MPa, while it is substantially improved to 62.29 MPa with the basicity further increased to 2.2. In addition, as the basicity is increased from 0.6 to 1.4, the strength of bonding matrix of limonitic nickel laterite is decreased slightly from 19.11 MPa to 18.64 MPa. However, this strength is deteriorated sharply to 13.94 MPa with the basicity further rising to 2.2. At the basicity of 1.0–1.4, the strength of bonding matrix of ordinary limonite is rather lower whereas that of limonitic nickel laterite is relatively better. The different suitable basicity for the two ores during sintering indicates the distinctive sinter consolidation characteristics of limonitic nickel laterite. Besides, the strength of bonding matrix of limonitic nickel laterite is far lower than that of ordinary limonite even at the suitable basicity. It can be expected that limonitic nickel laterite should possess very poorer sinter strength compared with ordinary limonite.

Fig. 5.

Strength of bonding matrix of the two ores at various basicity. (Online version in color.)

3.1.4. Formation Characteristics of Liquid Phase

Combined with Figs. 6, 7 and Tables 7, 8, the microstructure and mineral compositions of the bonding matrix of the two ores are very different with the variation of the basicity. Solid phase in the bonding matrix of ordinary limonite is only hematite (Fe₂O₃) whereas that of limonitic nickel laterite mainly contains hercynite ((Fe, Mg)·(Fe, Al)₂O₄), a small amount of hematite (Fe₂O₃) and Cr-spinel ((Fe, Mg)·(Cr, Fe, Al)₂O₄). This is caused by the partial substitution of extensive Al³⁺ and Mg²⁺ ions in limonitic nickel laterite for Fe³⁺ and Fe²⁺ ions, respectively.²⁷⁾ In addition, liquid phase in the bonding matrix of ordinary limonite is composed of fayalite (2FeO·SiO₂) and kirschsteinite (CaO·FeO·SiO₂) at the lower basicity of 0.6. As the basicity rises to 1.4, fayalite is gradually converted into kirschsteinite and a slight of dicalcium silicate (2CaO·SiO₂) and silico-ferrite of calcium and alumina (SFCA) are also formed. With the basicity further increased to 2.2, liquid phase primarily consists of SFCA and tricalcium silicate (3CaO·SiO₂). As to limonitic nickel laterite, liquid phase in its bonding matrix is a type of eutectic spinel olivine (CaO·(Fe, Mg)Al₂O₄·SiO₂) when the basicity is as low as 0.6, formed by the eutectic reaction of hercynite with CaO and SiO₂. The increase of the basicity to 1.4 contributes to the formation of a certain amount of SFCA. As the basicity reaches 2.2, SFCA and tricalcium silicate are the major components of liquid phase. The different formation characteristics of liquid phase of the two ores would lead to the great difference of their sintering properties.

Fig. 6.

Microstructure of the bonding matrix of the iron ores under optical microscopy (A-Ordinary limonite, B-Limonitic nickel laterite, H-Hercynite, He-Hematite, C-Cr-spinel, SFCA-Silico-ferrite of calcium and alumina, Ki-Kirschsteinite, F-Fayalite, K-Eutectic spinel olivine, CS-Tricalcium silicate, P-Pore, R-Resin). (Online version in color.)

Fig. 7.

Microstructure of the bonding matrix of the two ores under SEM (a–f are the selected areas in Fig. 6). (Online version in color.)

Table 7. EDS analysis results for areas in Fig. 7.

Area No.	Elemental compositions/(atomic conc,%)								Mineral phases
	Fe	Cr	Ni	Mg	Al	Si	Ca	O
a1	39.48	–	–	0.02	0.26	0.08	0.09	60.07	Fe2O3
a2	16.31	–	–	0.11	0.16	15.93	0.32	67.13	2FeO·SiO2
a3	9.33	–	–	0.84	0.73	14.36	18.61	56.13	CaO·FeO·SiO2
b1	30.86	–	–	0.23	3.47	3.79	7.88	53.77	Dendritic SFCA
b2	0.82	–	–	0.31	0.27	16.58	31.96	50.06	2CaO·SiO2
b3	39.32	–	–	0.24	0.52	0.11	0.07	59.74	Fe2O3
b4	7.27	–	–	0.20	0.62	15.11	19.51	57.29	CaO·FeO·SiO2
b5	27.63	–	–	0.39	4.58	4.30	8.69	54.41	Lamellar SFCA
c1	1.31	–	–	0.13	1.79	10.43	29.87	56.47	3CaO·SiO2
c2	31.16	–	–	0.19	2.96	3.68	8.49	53.52	Dendritic SFCA
c3	28.22	–	–	0.27	4.28	4.46	8.98	53.79	Lamellar SFCA
c4	39.26	–	–	0.07	0.77	0.10	0.19	59.61	Fe2O3
c5	25.09	–	–	0.34	5.57	4.79	9.27	54.94	Tabular SFCA
d1	39.06	0.38	0.13	0.58	0.79	0.3	0.16	58.6	Fe2O3
d2	16.51	0.37	0.32	3.96	5.32	16.5	7.26	49.76	CaO·(Fe, Mg)Al2O4·SiO2
d3	25.42	15.63	0.06	4.74	6.29	0.12	0.25	47.49	(Fe, Mg)·(Cr, Fe, Al)2O4
d4	34.83	0.74	0.91	2.98	2.68	0.54	0.28	57.04	(Fe, Mg)·(Fe, Al)2O4
e1	27.86	16.56	0.1	3.4	4.24	0.14	0.31	47.39	(Fe, Mg)·(Cr, Fe, Al)2O4
e2	17.58	0.25	0.11	3.09	6.72	13.06	10.98	48.21	CaO·(Fe, Mg)Al2O4·SiO2
e3	34.39	0.91	0.33	3.45	2.95	0.26	0.43	57.28	(Fe, Mg)·(Fe, Al)2O4
e4	32.44	0.36	0.21	0.18	1.78	3.72	8.44	52.87	Acicular SFCA
e5	30.93	0.42	0.32	0.24	2.34	4.16	8.35	53.24	Dendritic SFCA
e6	38.63	0.32	0.15	0.26	0.46	0.54	0.47	59.17	Fe2O3
f1	38.86	0.56	0.21	0.4	0.24	0.27	0.57	58.89	Fe2O3
f2	35.31	0.88	0.57	3.08	1.89	0.05	0.33	57.89	(Fe, Mg)·(Fe, Al)2O4
f3	11.91	21.84	0.09	8.98	7.86	0.13	0.33	48.86	(Fe, Mg)·(Cr, Fe, Al)2O4
f4	25.18	0.61	0.19	0.79	5.21	4.82	9.05	54.15	Tabular SFCA
f5	30.83	0.91	0.41	0.28	2.48	4.57	6.89	53.63	Dendritic SFCA
f6	28.39	0.48	0.13	0.43	3.36	4.57	8.86	53.78	Lamellar SFCA
f7	1.91	0.24	0.19	0.13	0.33	10.98	30.9	55.32	3CaO·SiO2

Table 8. Mineral compositions of the bonding matrix of the two ores at diffierent basicity (%).

NO.a)	Solid phases			Liquid phases						Porosity
	Hercynite	Hematite	Cr-spinel	Fayalite	Kirschsteinite	Eutectic spinel olivine	SFCA	Dicalcium silicate	Tricalcium silicate
A-0.6	–	74.13	–	11.27	14.60	–	–	–	–	40.27
A-1.4	–	64.58	–	–	24.46	–	6.76	4.20	–	46.03
A-2.2	–	55.54	–	–	–	–	34.27	–	10.19	29.38
B-0.6	7.54	74.72	2.29	–	–	15.45	–	–	–	43.47
B-1.4	6.78	63.45	2.42	–	–	17.43	9.92	–	–	46.85
B-2.2	5.53	54.98	2.38	–	–	–	23.82	–	13.29	53.06

^a)  A and B represents ordinary limonite and limonitic nickel laterite, respectively.

Moreover, the porosity of the bonding matrix of ordinary limonite is increased from 40.27% to 46.03% with the basicity elevated from 0.6 to 1.4 and then decreased to 29.38% as the basicity further reaches 2.2 (Fig. 6 and Table 8). This is due to that the former increase of the basicity favors the formation of kirschsteinite (CaO·FeO·SiO₂) and dicalcium silicate (2CaO·SiO₂). The crystal transformation from β-2CaO·SiO₂ to γ-2CaO·SiO₂ and large cooling shrinkage stress of different types of liquid phases lead to the generation of micro-cracking and the increase of the porosity.²⁸⁾ At the higher basicity of 2.2, the formation and propagation of SFCA contribute to the lower porosity. However, the porosity of the bonding matrix of limonitic nickel laterite is increased along with the basicity. This is mainly because that hercynite as the major solid phase has high corrosion resistence to alkali slags and is doubly difficult to be wetted by liquid bonding phases with the increase of the basicity, eventually leading to the loose microstructure and higher porosity of the bonding matrix.²⁹⁾ Besides, the porosity of the bonding matrix of limonitic nickel laterite is always higher than that of ordinary limonite at the same basicity and the texture of pores is more irregular due to the more abundant crystal water and high-smelting minerals. This is a large part of the reason why the strength of bonding matrix of limonitic nickel laterite is far below than that of ordinary limonite.

During iron ore sintering, SFCA is considered the most desirable bonding phase due to its higher mechanical strength, better reducibility and lower reduction degradation.^30,31,32) The formation characteristics of SFCA is mainly dominated by the factors such as the chemical compositions of iron ores, sinter basicity, gas atmosphere and cooling rate.^{33,34,35,36,37)} Normally, the higher sinter basicity, iron grade and SiO₂ content of iron ores and oxygen partial pressure contribute to the formation of SFCA while the excessive Al₂O₃ and MgO contents of iron ores and overquick cooling rate during sintering are adverse to the formation of SFCA. In general, SFCA can be divided into two types on the basis of the composition, morphology and crystal structure, i.e. high-iron SFCA and low-iron SFCA.^{38,39,40,41,42)} The former possesses relatively higher mechanical strength and lower formation temperature compared with the latter. Thus, the more formation of high-iron SFCA is more beneficial to the improvement of sinter strength.

Based on the above analysis, the amounts of kirschsteinite (CaO·FeO·SiO₂) and dicalcium silicate (2CaO·SiO₂) of the bonding matrix of ordinary limonite as well as the porosity are increased with the basicity increased from 0.6 to 1.4. As indicated in the previous literatures,^28,43) the self-strength of kirschsteinite (CaO·FeO·SiO₂) is weaker than that of fayalite (2FeO·SiO₂) and the micro-cracking is formed due to the crystal transformation of dicalcium silicate (2CaO·SiO₂) and the existence of large cooling shrinkage stress. Thus, although a small amount of SFCA is formed, the strength of bonding matrix of ordinary limonite is still greatly weakened. However, as the basicity is continuously increased to 2.2, SFCA amount rises from 6.76% to 34.27% and the porosity is reduced from 46.03% to 29.38%, contributing to the substantial increase of the strength of bonding matrix of ordinary limonite. In addition, although the porosity of the bonding matrix of limonitic nickel laterite is increased from 43.47% to 46.85% with the basicity elevated from 0.6 to 1.4, the increased amount of liquid phases and the formation of acicular and dendritic SFCA with higher strength are helpful to maintain the relatively good strength of bonding matrix. As the basicity further reaches 2.2, the porosity is sharply increased to 53.06% and the major solid phase (i.e. hercynite) cannot be commendably bonded by liquid phases due to the high corrosion resistence to alkali slags. Meanwhile, SFCA mainly exhibits tabular or lamellar texture with coarse size and its strength is lower than that of acicular or dendritic SFCA due to the lower iron content. Despite SFCA amount is increased, the strength of bonding matrix of limonitic nickel laterite suffers a huge deterioration at the higher basicity of 2.2. At the respective suitable basicity, the grains of solid phase in the bonding matrix of limonitic nickel laterite are pretty smaller with weaker connectivity and the microstructure of its bonding matrix is bitterly looseer compared with that of ordinary limonite. Besides, the amount of liquid phase is much less while the porosity presents the opposite rule due to the lower iron grade and poorer assimilability of limonitic nickel laterite. Thus, the strength of bonding matrix of limonitic nickel laterite is dramatically lower than that of ordinary limonite. All of which is exactly consistent with the results of Fig. 5.

Overall, limonitic nickel laterite possesses much poorer assimilability and liquid phase fluidity compared with ordinary limonite. Moreover, the variations of the strength of bonding matrix and formation characteristics of liquid phase of the two ores with the basicity are exceedingly different. In view of the yield and quality of product sinter comprehensively, the suitable basicity of limonitic nickel laterite sintering should be not exceed 1.4 while that of ordinary limonite sintering ought to be not below 1.8. Furthermore, the strength of bonding matrix of limonitic nickel laterite is far weaker than that of ordinary limonite due to the limited formation amount of liquid phase and looser microstructure. It is expected that sintering performance of limonitic nickel laterite would be much worse than that of ordinary limonite.

3.2. Contrastive Analysis of Sintering Performance

After the optimization of mixture moisture and anthracite dosage, the optimum sintering performance of the two ores are obtained at the suitable basicity of 1.40 and 1.80, respectively (Table 9). The suitable mixture moisture and anthracite dosage during limonitic nickel laterite sintering are 18% and 7.5%, respectively, whereas they during ordinary limonite sintering are 8.5% and 4.6%, respectively. Compared with ordinary limonite, limonitic nickel laterite possesses much poorer sinter indices indeed with tumble index and productivity of only 45.87% and 0.97 t·m⁻²·h⁻¹, respectively, and solid fuel rate of as high as 140.52 kg·t⁻¹. Meanwhile, the vertical sintering speed of limonitic nickel laterite is considerably faster than that of ordinary limonite due to its richer crystal water. This is extremely unfavorable to the sinter consolidation and decrease of sinter porosity. Besides, the poorer assimilability and liquid phase fluidity seriously limits the liquid phase formation and the bonding between solid phase and liquid phase. All of which eventually results in the rather lower tumble index and productivity. And the extensive existence of high-smelting minerals promotes much more solid fuel consumption during sintering. Thus, sintering performance of limonitic nickel laterite is confirmed far worse than that of ordinary limonite, perfectly agreeing with the results of Section 3.1. According to the above investigations, it is an effective method to strengthen limonitic nickel laterite sintering by promoting the liquid phase formation and densification of the sinter.

Table 9. Comparision of sintering performance of the two ores.

Samples	Vertical sintering speed/mm/min	Tumble index/%	Productivity/t·m−2·h−1	Solid fuel rate/kg·t−1	Optimum basicity
Limonitic nickel laterite	32.77	45.87	0.97	140.52	1.40
Ordinary limonite	28.04	63.60	1.51	74.24	1.80

4. Conclusions

(1) The assimilability and liquid phase fluidity of limonitic nickel laterite are much weaker than that of ordinary limonite due to the massive existence of high-smelting minerals. This is extremely adverse to the formation and propagation of liquid phase during limonitic nickel laterite sintering and then apt to readuce its sinter strength.

(2) The variations of the strength of bonding matrix of limonitic nickel laterite and ordinary limonite with the basicity are quite different. During sintering, the suitable basicity of limonitic nickel laterite should be no more than 1.4 while that of ordinary limonite ought to be not below 1.8. Besides, the strength of bonding matrix of limonitic nickel laterite is far lower than that of ordinary limonite.

(3) Compared limonitic nickel laterite with ordinary limonite during sintering, the formation amount of liquid phase of the former is much less at the suitable basicity, especially for high-strength liquid phase such as SFCA. Meanwhile, the porosity of its bonding matrix is exceedingly higher than that of ordinary limonite and the solid phase cannot be tightly bonded by liquid phase. All of which leads to the generation of rather looser microstructure and then the poorer strength of the bonding matrix.

(4) Through sinter pot tests, sintering performance of limonitic nickel laterite is indeed much poorer than that of ordinary limonite, characterized by much higher solid fuel rate, lower tumble index and productivity, which is nicely consitent with the investigations of high-temperature characteristics. Limonitic nickel laterite sintering can be effectively improved by promoting the liquid phase formation and densification of the sinter.

Acknowledgements

Financial support from the Major Project of Master Alloy Manufacture for Heat Resistant Stainless Steel Production No. (AA18242003) funded by the Provincial Government of Guangxi Zhuang Autonomous District, is sincerely acknowledged.

References

• 1)   X. T.  Ma,  D. L.  Yang,  Y. J.  Zhai,  X. X.  Shen,  R. R.  Zhang and  J. L.  Hong: Int. J. Life Cycle Assess., 24 (2019), 1840.
• 2)   M.  Landers,  R. J.  Gilkes and  M.  Wells: Appl. Clay Sci., 42 (2009), 615.
• 3)   L.  Panda,  D. S.  Rao,  B. K.  Mishra and  B.  Das: Min. Metall. Explor., 31 (2014), 57.
• 4)   H. T. B. M.  Petrus,  A. D. P.  Putera,  E.  Sugiarto,  I.  Perdana,  I. W.  Warmada,  F.  Nurjaman,  W.  Astuti and  A. T.  Mursito: Miner. Eng., 132 (2019), 126.
• 5)   A. E. M.  Warner,  C. M.  Díaz,  A. D.  Dalvi,  P. J.  Mackey and  A. V.  Tarasov: JOM, 58 (2006), 11.
• 6)   G.  Mongelli,  B.  Taghipour,  R.  Sinisi and  S.  Khadivar: Ore Geol. Rev., 111 (2019), 102990.
• 7)   S.  Sarbishei and  L. T.  Khajavi: Fuel, 280 (2020), 118648.
• 8)   M. J.  Eckelman: Resour. Conserv. Recycl., 54 (2010), 256.
• 9)   H.  Tsuji: ISIJ Int., 52 (2012), 1000.
• 10)   L. T.  Pan,  J.  Yang and  Y. B.  Xu: Sinter. Pellet., 38 (2013), 25 (in Chinese).
• 11)   D. Q.  Zhu,  Y. X.  Xue,  J.  Pan,  C. C.  Yang,  Z. Q.  Guo,  H. Y.  Tian,  H.  Liao,  L. T.  Pan and  X. Z.  Huang: Powder Technol., 367 (2020), 616.
• 12)   D. H.  Liu,  J. H.  Li,  Y.  Peng,  J. L.  Zhang,  G. W.  Wang and  X.  Xue: J. Iron Steel Res. Int., 26 (2019), 691.
• 13)   S. L.  Wu and  G. L.  Zhang: Steel Res. Int., 86 (2015), 1014.
• 14)   G. S.  Feng,  S. L.  Wu,  H. L.  Han,  L. W.  Ma,  W. Z.  Jiang and  X. Q.  Liu: Int. J. Miner. Metall. Mater., 18 (2011), 270.
• 15)   S. L.  Wu,  G. L.  Zhang,  S. G.  Chen and  B.  Su: ISIJ Int., 54 (2014), 582.
• 16)   S. L.  Wu and  X. B.  Zhai: Metall. Res. Technol., 115 (2018), 505.
• 17)   S. L.  Wu,  H. P.  Li,  W. L.  Zhang and  B.  Su: Metals, 9 (2019), 404.
• 18)   G. L.  Zhang,  S. L.  Wu,  B.  Su,  Z. G.  Que,  C. G.  Hou and  Y.  Jiang: Int. J. Miner. Metall. Mater., 22 (2015), 553.
• 19)   X. B.  Zhai,  S. L.  Wu,  H.  Zhou,  L. X.  Su and  X.  Ma: Ironmaking Steelmaking, 47 (2020), 405.
• 20)   K. J.  Li,  J. L.  Zhang,  Z. J.  Liu,  B. X.  Su,  L. T.  Kong and  T. J.  Yang: Int. J. Miner. Metall. Mater., 21 (2014), 216.
• 21)   S. L.  Wu,  D.  Oliveira,  Y. M.  Dai and  J.  Xu: Int. J. Miner. Metall. Mater., 19 (2012), 217.
• 22)   R. L.  Thorne,  S.  Roberts and  R.  Herrington: Geology, 40 (2012), 331.
• 23)   C. R. M.  Butt and  D.  Cluzel: Elements, 9 (2013), 123.
• 24)   Y. X.  Xue,  J.  Pan,  D. Q.  Zhu,  Z. Q.  Guo,  C. C.  Yang,  L. M.  Lu and  H. Y.  Tian: Minerals, 10 (2020), 802.
• 25)   H.  Guo and  X. M.  Guo: Steel Res. Int., 90 (2019), 1900138.
• 26)   T.  Umadevi,  A. V.  Deodar,  P. C.  Mahapatra,  M.  Prabhu and  M.  Ranjan: Steel Res. Int., 80 (2009), 686.
• 27)   P.  Jiang,  J. H.  Chen,  M. W.  Yan,  B.  Li,  J. D.  Su and  X. M.  Hou: Int. J. Miner. Metall. Mater., 22 (2015), 1219.
• 28)   S. C.  Panigrahy,  P.  Verstraeten and  J.  Dilewijns: Metall. Trans. B, 15 (1984), 23.
• 29)   H. S.  Park,  Y.  Kim,  S.  Kim,  T.  Yoon,  Y.  Kim and  Y.  Chung: Ceram. Int., 44 (2018), 17585.
• 30)   X.  Jiang,  J. X.  Yu,  L.  Wang,  T. Y.  He,  H. Y.  Zheng,  Q. J.  Gao and  F. M.  Shen: ISIJ Int., 60 (2020), 2191.
• 31)   S.  Cheng,  M.  Shevchenko,  P. C.  Hayes and  E.  Jak: Metall. Mater. Trans. B, 52 (2021), 2416.
• 32)   J.  Chen,  S.  Cheng,  M.  Shevchenko,  P. C.  Hayes and  E.  Jak: Metall. Mater. Trans. B, 52 (2021), 517.
• 33)   T. J.  Park,  J. S.  Choi and  D. J.  Min: Steel Res. Int., 90 (2019), 1900001.
• 34)   X.  Jiang,  J. D.  Zhao,  L.  Wang,  H. W.  An,  Q. J.  Gao,  H. Y.  Zheng and  F. M.  Shen: ISIJ Int., 61 (2021), 86.
• 35)   S.  Nicol,  S.  Cheng,  E.  Jak and  P. C.  Hayes: ISIJ Int., 60 (2020), 2659.
• 36)   Y. F.  Chai,  W. T.  Yu,  J. L.  Zhang,  S. L.  An,  J.  Peng and  Y. Z.  Wang: Ironmaking Steelmaking, 46 (2019), 424.
• 37)   J.  Park,  R.  Rajavaram,  I. K.  Suh,  J.  Jeon,  S.  Son and  J.  Lee: Metall. Mater. Trans. B, 51 (2020), 3016.
• 38)   A.  Koryttseva,  N. A. S.  Webster,  M. I.  Pownceby and  A.  Navrotsky: J. Am. Ceram. Soc., 100 (2017), 3646.
• 39)   N. A. S.  Webster,  M. I.  Pownceby,  R.  Pattel,  J. R.  Manuel and  J. A.  Kimpton: ISIJ Int., 59 (2019), 263.
• 40)   E. P.  Heikkinen,  M.  Iljana and  T.  Fabritius: ISIJ Int., 60 (2020), 2633.
• 41)   D. W.  Yang,  W.  Wang,  J. X.  Li,  R. S.  Xu,  X. Z.  Wang and  G.  Wang: Metall. Mater. Trans. B, 51 (2020), 776.
• 42)   W.  Wang,  D. W.  Yang,  Z. L.  Ou-Yang,  R. S.  Xu and  M. M.  Song: Metall. Mater. Trans. B, 50 (2019), 678.
• 43)   X. R.  Tang,  D. Y.  Wang and  Q. C.  Zhang: Sintering Theory and Technology, Central South University of Technology Press, Changsha, (1992), 188 (in Chinese).