Effect of Different Crushing Methods on Chalcopyrite Liberation and Heavy Media Preconcentration

Abstract

In order to find a short, economically feasible process for chalcopyrite preconcentration and to provide a reference for the preconcentration of similar copper sulfide ores, the particle size characteristics of the crushed products from a high-pressure grinding roller (HPGR) and jaw crusher (JC) were analyzed, as well as the liberation degree and fracture characteristics of the chalcopyrite. The float–sink test (FST) was carried out on the crushed products, and the effects of the two crushing methods on the FSTs of the crushed products were compared. The research results show that at the same crushing fineness, the chalcopyrite liberation in HPGR products can be enhanced by 14%~18% compared with the JC. The single-particle crushing of the JC tends to produce intergranular fracturing of chalcopyrite, while the lamination crushing of the HPGR produces more transgranular fracturing of chalcopyrite; the chalcopyrite in the −5 + 0.5 mm size fraction mainly produces intergranular fracturing, and the chalcopyrite in the −0.5 mm size fraction mainly produces transgranular fracturing. The FST results show that heavy media preconcentration was suitable for chalcopyrite, and, in the optimal conditions of a size fraction of −3 + 0.5 mm and separation density of 2.55 g/cm³, the grade and distribution rate of Cu in the sinks obtained by HPGR-FST were 0.35% and 89.86%, respectively, and the floats yield was 24.76%, with a better enrichment of sinks and higher floats yields, which was better when compared with that of the JC-FST.

1. Introduction

Copper is a transition metal and one of the earliest metals used by mankind, widely used in electric power, electronics, construction, machinery manufacturing and the defense and military industries. There are various types of copper mines in the world, including the porphyry type, sedimentary rock type, epithermal hydrothermal type, skarn type and IOCG type, etc., among which the porphyry and sedimentary rock types are the most important, and the two together account for more than 80% of the global copper reserves [1,2]. Among the various types of copper ore resources, about 280 types of copper minerals are found, with more than 20 major ones, mostly in the form of copper sulfide ores. Currently, about 90% of the world’s primary copper production comes from copper sulfide ores; the common copper sulfide ores mainly include chalcopyrite, bornite and chalcocite, among which chalcopyrite is the most widely distributed and is one of the main minerals in copper smelting [2,3,4].

Prior to metallurgy, copper sulfide ore is enriched by beneficiation methods to increase the Cu grade in the ore, to reduce the amount of ore in the metallurgical process and lower the production cost. Currently, the main beneficiation method of copper sulfide ore is flotation, followed by bioleaching [5,6,7,8]. Among them, flotation methods mainly include the selective flotation process, the bulk flotation process and the combination process consisting of selective flotation and bulk flotation [9,10,11,12,13]. Through the continuous research on the flotation process of copper sulfide ores by scholars around the world, the quick flotation process and low alkalinity flotation process have been proposed successively [14,15,16]. However, the flotation process requires the use of a large number of flotation reagents and fine grinding of the ore, which increases energy consumption and production costs [17,18]. At the same time, the use of flotation reagents affects the surrounding environment, further increasing the cost of environmental protection treatment. Therefore, reducing the use of flotation reagents and the amount of ore grinding is crucial to the beneficiation of copper sulfide ores.

Preconcentration technology is particularly important in the processing of low-grade ore. It can remove part of the waste rock or low-grade ore before coarse grinding or fine grinding, to achieve preconcentration of the raw ore, improve the grade of the feed, and reduce the amount of grinding and the production of fine tailings, thereby improving the utilization rate of resources. At present, the preconcentration of copper ore mainly adopts the gravity concentration process, followed by photoelectric separation. But the photoelectric separation machine has strict requirements for the particle size of the ore, generally requiring a particle size between 10~50 mm, and the processing capacity of the photoelectric separation machine is relatively low [19]. And the gravity preconcentration of copper ore benefits from the advanced gravity concentration equipment that has emerged in recent years. Hu Genhua used a “jig-spiral chute-shaking table” gravity concentration process on natural copper sulfide copper ore preconcentration, finding that the natural copper can be effectively enriched, while the copper sulfide minerals mainly enter the preconcentration tailings, and the enrichment effect of copper sulfide minerals is poor, which is related to the insufficient dispersion particle size and liberation degree of the minerals [20]. Christian Ndolwa Katwika et al. used a Knelson separator to treat fine-particle copper–cobalt flotation tailings and increased the Cu distribution rate by 21.09%. This also indicates that copper sulfide minerals can be enriched by gravity concentration as long as mineral liberation is ensured [21]. Similarly, copper sulfide ore preconcentration by Mozley Mineral Separator can obtain a concentrate with a Cu grade of 5.33%, but the Cu distribution rate was only 26.14% [22]. In addition, many scholars applied fluidized beds to the preconcentration of copper sulfide ores, but the distribution rate of Cu concentrate obtained from fluidized beds was low and required a coarse feed size, which was not effective for the treatment of fine-particle ores [23,24]. Different types of ores react differently to different crushing methods, and choosing suitable crushing methods can effectively improve the mineral liberation and preconcentration effect; for example, an impact crusher can best liberate copper minerals in polymetallic ores [25]. Meanwhile, selective crushing or optimized crushing processes can effectively improve the mineral liberation effect and preconcentration efficiency of ores, which has important application value and promotion prospects [26,27]. In summary, the particle size and liberation degree of copper sulfide ore have a great influence on its gravity preconcentration process [25]. Therefore, it is necessary to explore the influence of the feed particle size and liberation degree of copper sulfide ore on gravity preconcentration and find a gravity preconcentration technology for copper sulfide ore with a high distribution rate and wide particle size.

The high-pressure grinding roller (HPGR) was developed by SCHÖNERT based on the theory of lamination comminution, which is a kind of high-efficiency and energy-saving comminution equipment. The crushed product shows a high content of fine particles and microcracks, which can be favorable for the subsequent grinding, separation or leaching operations [28,29]. Nghipulile et al. compared the difference in performance between the HPGR and cone crusher in processing UG2 platinum ores and found that the HPGR is able to generate microfractures more efficiently during crushing, which contributes to the liberation of sulfide particles [30]. Similarly, Kodali et al. found that HPGR crushing can produce more microfractures and a higher exposure of copper minerals compared to the jaw crusher, resulting in higher reactivity and leaching efficiency during leaching [31]. Therefore, the HPGR has significant advantages in mineral liberation and microfracture generation. In this study, the liberation characteristics of chalcopyrite in the crushed products of the HPGR and jaw crusher (JC) and the effects of two different crushing methods on chalcopyrite heavy media preconcentration were analyzed with low-grade porphyry copper ore in Tibet, China, as the object of study. To achieve this goal, the crushed products of the two crushing methods at different crushing finenesses were characterized in terms of particle size characteristics, liberation degree and fracture characteristics to evaluate the effect of the crushing methods on chalcopyrite liberation. At the same time, a float–sink test (FST) of the crushed products was carried out to study the effects of feed size and separation density on the heavy media preconcentration of chalcopyrite, and the optimal parameters were determined. On this basis, the effects of the two different crushing methods on the FST of the crushed products were evaluated.

2. Materials and Methods

2.1. Materials

The representative ore samples used in this study are from the Julong copper mine in the Tibet region of China, which is a low-grade porphyry copper ore. The samples were initially crushed to −40 mm, and 6 t of samples were randomly obtained from the transportation belt for subsequent experiments, with the sample particle size distribution shown in Figure 1.

2.2. Methods

2.2.1. Sample Property Analysis

Sample property analysis includes mineral composition analysis and chemical com-position analysis. The chemical composition of the sample was measured using inductively coupled plasma–atomic emission spectroscopy (Intrepid II XSP, Thermo Electron, Waltham, MA, USA). The mineral composition and content of the samples were determined by optical microscopy (ZEISS Axioskop 40, Zeiss, Oberkochen, Germany) and mineral liberation analysis (MLA 650F, FEI Company, Hillsboro, OR, USA).

2.2.2. Crushing Methods

The crushing method and process of the samples are shown in Figure 2. Six representative −40 mm samples, each weighing 500 kg, were prepared and crushed by HPGR (GM80/25, Sinosteel Anhui Tianyuan Technology Co., Ltd., Maanshan, China) and JC (XPC60×100, Wuhan Exploration Machinery Co., Ltd., Wuhan, China), respectively, and the crushed product was screened with a screen size of 5/3/2 mm. The oversized products were returned to the feeder to form a closed circuit, and the undersized products were qualified products for subsequent experiments. The final samples of −5 mm, −3 mm and −2 mm after crushing by HPGR and −5 mm, −3 mm and −2 mm after crushing by JC were obtained.

2.2.3. Analysis of Particle Size Distribution and Liberation Characteristics

The six samples obtained from the two crushing methods were screened using a screen set to obtain the particle size distribution of each sample. The average particle size (D_w) could reflect the average size of the sample particle size; the deviation factor (K_d) was used to indicate the degree of uniformity of the sample size; the larger the value of K_d, the more uneven the distribution of the sample particle size [32]. The calculation formulas for both are shown in Equations (1) and (2):

$$D_{\mathrm{w}}={\displaystyle \frac{\sum r_{\mathrm{i}}{d}_{\mathrm{i}}}{\sum r_{\mathrm{i}}}}={\displaystyle \frac{\sum r_{\mathrm{i}}{d}_{\mathrm{i}}}{100}}$$

(1)

$$K_{\mathrm{d}}={\displaystyle \frac{\sqrt{\sum {\left(d_{\mathrm{i}}-D_{\mathrm{w}}\right)}^2{r}_{\mathrm{i}}}}{D_{\mathrm{w}}}}$$

(2)

where D_w was the weighted average particle size, (mm); K_d was the deviation factor; d_i was the arithmetic average particle size of individual particles, (mm); and r_i was the yield of individual particles, (%).

The six samples obtained from the two crushing methods were processed separately, and each sample was made into five polished blocks to ensure that a sufficient number of mineral particles were analyzed for each sample to keep the experimental data representative. The extended back-scattered electron (XBSE) measurement mode was used to measure the polished blocks in the Mineral Liberation Analyzer (MLA 650F, FEI Company, Hillsboro, OR, USA). This mode used backscattered electron images to delineate particles and mineral particles, followed by characteristic X-ray analysis for phase identification. Finally, the mineral particle size distribution, mineral liberation degree and mineral liberation characteristics with different crushing methods were obtained. For the liberation characteristics of the minerals, the main focus was on the fracture produced by mineral liberation. The fracture modes of the minerals, as shown in Figure 3, were classified into intergranular fracturing and transgranular fracturing [33,34]. Intergranular fracturing ensures the liberation of the mineral and at the same time makes the mineral crystals more complete, which is favorable for subsequent separation. Transgranular fracturing results in a finer particle size and incomplete liberation of the minerals. Chalcopyrite particles could be accurately localized by MLA to observe and count the fractures. In addition, the ambient temperature of the scanning electron microscope was 18°C, and the accelerating voltage was 15 kV.

2.2.4. Float–Sink Tests

The feasibility of heavy media separation of samples could be evaluated by float–sink tests [35]. Based on the density difference between the valued minerals and the gangue minerals, potassium iodide, mercuric iodide and water were used to prepare different densities of heavy liquid (SG) as the separation medium for the float–sink test (FST). The density of the heavy liquid was measured by the gravimetric method. Firstly, the HPGR and JC crushed products were sieved separately to obtain several different particle size products, the particles of −0.3 mm size were removed, and then the FSTs were carried out on the other particle size products; the test schematic is shown in Figure 4. In the FST, each particle size was placed separately in a large beaker filled with a heavy liquid, allowed to stand and stratified. Afterwards, light and heavy minerals were removed separately and, after washing, drying and weighing, the copper was analyzed, and the metal distribution rate was calculated. Finally, based on these test results, the suitability of gravity concentration for chalcopyrite was evaluated.

3. Results and Discussion

3.1. Mineral Composition and Chemistry of the Sample Particle Size Characteristics

The chemical composition of the sample is shown in

Table 1. Chemical analysis of the ore sample.

Component	Cu	Mo	S	SiO2	Al2O3	CaO
Content wt.%	0.36	0.01	1.52	65.71	13.59	2.23
Component	MgO	As	K2O	Na2O	P	TiO2
Content wt.%	0.04	0.01	3.88	2.58	0.058	0.47

. The results showed that the Cu content of the sample was 0.36%, the S content was 1.52%, and the Mo contents were low. The mineral composition and content results of the samples are shown in

Table 2. Mineral composition and content of the sample.

Mineral	Formula	Content wt.%	Mineral	Formula	Content wt.%
Chalcopyrite	CuFeS2	0.61	Plagioclase	Na[AlSi3O8]-Ca[Al2Si2O8]	25.50
Chalcocite	Cu2S	0.08	Orthoclase	K[AlSi3O8]	17.23
Bornite	Cu5FeS4	0.02	Mica	/	16.01
Covellite	CuS	0.01	Chlorite	Y3[Z4O10](OH)2·Y3(OH)6	3.02
Enargite	Cu3AsS4	0.01	Dolomite	CaMg(CO3)2	0.11
Malachite	Cu2(OH)2CO3	0.01	Kaolinite	Al4[Si4O10](OH)8	0.85
Molybdenite	MoS2	0.01	Rutile	TiO2	0.32
Pyrite	FeS2	2.43	Apatite	Ca5(PO4)3(F,Cl,OH)	0.29
Magnetite	Fe3O4	0.08	Zircon	ZrSiO₄	0.07
Hematite	Fe2O3	0.01	Others	/	0.24
Quartz	SiO2	30.69	/	/	/
Calcite	CaCO3	2.40	Total	/	100.00

. The results showed that the main copper minerals in the ore were chalcopyrite (0.61%), chalcocite (0.08%) and bornite (0.02%), and the main gangue minerals were quartz (30.69%), plagioclase (25.50%), orthoclase (17.23%) and mica (16.01%).

3.2. Particle Size Characteristics

The particle size distribution characteristics of the crushed products of the two crushing methods with different crushing finenesses are shown in Figure 5, where Figure 5d shows the results of the average particle size (D_w) and deviation factor (K_d) of the different crushed products. The results showed that the P₈₀ (P₈₀ means that the proportion of particles with a size larger than a certain value in the size distribution was 80%) of the HPGR-crushed product was obviously smaller than that of JC-crushed product with the same crushing fineness, indicating that the degree of HPGR crushing was higher, and the product size was finer, which could also be verified by comparing D_w. Meanwhile, the K_d value of the HPGR-crushed product was large, and the particle size distribution was more non-uniform relative to the JC-crushed product, indicating that the slow crushing method of the HPGR was prone to the phenomenon of high content of fine particles and large specific surface area of the product. In the same crushing mode, as the crushing fineness became smaller, the D_w of the crushed product gradually became smaller, the K_d of the HPGR product gradually becomes smaller, and the change of the K_d value of the JC product tended to be flat. It showed that, as the crushing fineness became smaller, it would not improve the particle size uniformity of the JC product, while the particle size distribution of the HPGR product became more uniform. In addition, it was found through Figure 5d that the D_w of the HPGR (−5 mm)-crushed product was close to that of the JC (−3 mm)-crushed product, and the D_w of the HPGR (−3 mm)-crushed product was close to that of JC (−2 mm)-crushed product, which indicated that the HPGR could effectively relax the crushing fineness.

3.3. Liberation Characteristics

The results of the liberation degree of chalcopyrite in different crushing finenesses with two crushing methods are shown in

Table 3. Liberation degree of chalcopyrite in crushed products.

Crushing Methods	Particle Size (mm)	Single (%)	Aggregation (%)	Total (%)
HPGR	−5	24.15	75.85	100.00
	−3	31.18	68.82	100.00
	−2	40.22	59.78	100.00
JC	−5	21.17	78.83	100.00
	−3	27.26	72.74	100.00
	−2	34.09	65.91	100.00

. The results showed that the liberation degree of chalcopyrite in the crushed product gradually increased with the smaller crushing fineness with the same crushing method. For the −5 mm, −3 mm and −2 mm crushing finenesses, the liberation degree of chalcopyrite in the HPGR was 24.15%, 31.18% and 40.22%, respectively, and that of JC chalcopyrite was 21.17%, 27.26% and 34.09%, respectively. For the same crushing fineness, the liberation degree of chalcopyrite in the HPGR-crushed products was higher, and, relative to the JC-crushed products, the liberation degree could be increased by 14%~18%.

3.4. Crystal Fracture Characteristics

Due to the −3 mm and −2 mm crushing fineness, the liberation degree of chalcopyrite in the crushed products of the two crushing methods was relatively high. In order to better analyze the chalcopyrite generated fractures, only the HPGR (−5 mm)- and JC (−5 mm)-crushed products were analyzed to observe the unliberated chalcopyrite crystals generating fractures. The results of the chalcopyrite fracturing type distribution in the two crushing modes are shown in

Table 4. Distribution of chalcopyrite fracture modes for two crushing methods.

Crushing Methods	Size Fraction (mm)	Intergranular Fracture	Transgranular Fracture	Total (%)
HPGR	−5 + 3	83.34	16.66	100.00
	−3 + 2	67.74	32.26	100.00
	−2 + 1	59.26	40.74	100.00
	−1 + 0.5	61.54	38.46	100.00
	−0.5	23.53	76.47	100.00
5~0 mm		47.97	52.03	100.00
JC	−5 + 3	90.91	9.09	100.00
	−3 + 2	70.00	30.00	100.00
	−2 + 1	81.25	18.75	100.00
	−1 + 0.5	66.67	33.33	100.00
	−0.5	46.15	53.85	100.00
5~0 mm		70.27	29.73	100.00

, and the BSE images of chalcopyrite fracturing are shown in Figure 6. The results showed that HPGR (−5 mm) chalcopyrite produces 47.97% and 52.03% of intergranular fracturing and transgranular fracturing, respectively, and JC (−5 mm) chalcopyrite produces 70.27% and 29.73% of intergranular fracturing and transgranular fracturing, respectively. It indicated that the JC was more inclined to make chalcopyrite produce intergranular fracturing with the same crushing fineness, while the HPGR would make chalcopyrite produce more transgranular fracturing. This phenomenon was mainly related to the crushing mode of the two crushing methods, with the JC mainly focusing on single-particle crushing and the HPGR on lamination crushing. The compression of lamination crushing was much stronger and able to penetrate into the interior of the mineral particles, leading to more transgranular fracturing. At the same time, single-particle crushing was different; lamination crushing could form a number of high-density particle layers within the effective crushing stroke, which would have had enough crushing power to act on the particle group, and, while giving full play to lamination crushing, it could use the powerful fragmentation kinetic energy generated in the process of particle crushing to re-crush the neighboring particles, which had a high crushing rate and increased the fine-particles content in the crushed products easily [36].

In addition, it was found that the generation of fractures was related to the size of the mineral particles through the test results. In the two crushing modes, chalcopyrite in the −5 + 0.5 mm particle size mainly produces intergranular fracturing, while chalcopyrite in the −0.5 mm particle size mainly produces transgranular fracturing. The main reason was that, when the particle size became smaller, the chalcopyrite crystal size and particle size gradually closed, and the extrusion pressure acting inside the chalcopyrite crystal gradually became larger, more easily producing fractures along the intracrystalline cracks caused by the chalcopyrite particle size becoming fine [37]. It should be noted that there were significant differences in the lattice defects of different minerals, which altered the internal structure of the mineral crystals and the bonding forces between particles, thus affecting the hardness of the minerals, which may have limited the applicability of the above conclusions.

3.5. Chalcopyrite Heavy Media Preconcentration Feasibility Evaluation

3.5.1. Narrow Range Particle Size Tests

In order to evaluate the effect of different crushing methods on the gravity concentration of copper minerals, FSTs were carried out on different size fractions of HPGR (−5 mm) products at a separation density of 2.60 g/cm³, and the optimum range of particle size for separation was determined. The enrichment ratio (E) refers to the ratio of the grade of valuable elements in the concentrate to the grade of valuable elements in the feed. The results of the FST for each size fraction are presented in

Table 5. Results of FST on different size fractions in separation density of 2.60 g/cm³.

Size Fraction (mm)	Products	Yield (%)	Cu Grade (%)	Cu Distribution Rate (%)	E (Cu)
−5 + 3	Sinks	28.66	0.56	47.65	1.69
	Floats	71.34	0.25	52.35
	Feed	100.00	0.33	100.00
−3 + 2	Sinks	36.78	0.58	59.32	1.61
	Floats	63.22	0.23	40.68
	Feed	100.00	0.36	100.00
−2 + 1	Sinks	36.09	0.47	60.83	1.67
	Floats	63.91	0.17	39.17
	Feed	100.00	0.28	100.00
−1 + 0.5	Sinks	34.59	0.40	56.47	1.66
	Floats	65.41	0.16	43.53
	Feed	100.00	0.24	100.00
−0.5 + 0.3	Sinks	29.74	0.33	45.11	1.50
	Floats	70.26	0.17	54.89
	Feed	100.00	0.22	100.00

. From the results, it could be seen that the sinks Cu grade was 0.56%, the distribution rate was 47.65%, and the E of Cu was 1.69 in the −5 + 3 mm size fraction. Theoretically, the poor separation performance of the −5 + 3 mm size fractions was due to insufficient mineral liberation of copper minerals. When the size fraction was reduced from −5 + 3 mm to −3 + 2 mm, there was no significant increase in Cu grade in the sinks, Cu distribution rate increased from 47.65% to 59.32%, and the E value of Cu was 1.61. With the reduction of the size fraction to −2 + 1 mm and −1 + 0.5 mm, the Cu results in the sinks were similar to those of the −3 + 2 mm size fraction. When the size fraction was further reduced to −0.5 + 0.3 mm, the Cu grade in the sinks decreased to 0.33%, the Cu distribution rate decreased to 45.11%, the E value of Cu was 1.50, and the separation efficiency of fine particles was deteriorating. The main reason was that the separation performance of the fine particles was more sensitive to changes in media rheology, and the settling velocity became slower, resulting in the separation efficiency of fine particles being usually lower than that of coarse particles [38]. Therefore, the optimal size fraction for the FST should be −3 + 0.5 mm.

3.5.2. Separation Density Tests

FSTs with different separation densities were carried out with the −3 + 0.5 mm size fraction in the HPGR (−3 mm) product, and the results are shown in

Table 6. FST results of −3 + 0.5 mm size fraction at different separation density.

Separation Density (g/cm3)	Products	Yield (%)	Cu Grade (%)	Cu Distribution Rate (%)
2.50	Sinks	90.43	0.31	96.06
	Floats	9.57	0.12	3.94
	Feed	100.00	0.29	100.00
2.55	Sinks	74.43	0.36	88.21
	Floats	25.57	0.14	11.79
	Feed	100.00	0.30	100.00
2.60	Sinks	36.24	0.48	60.25
	Floats	63.76	0.18	39.75
	Feed	100.00	0.29	100.00
2.70	Sinks	16.32	0.73	41.20
	Floats	83.68	0.20	58.80
	Feed	100.00	0.29	100.00
2.80	Sinks	11.15	0.83	32.13
	Floats	88.85	0.22	67.87
	Feed	100.00	0.29	100.00
2.85	Sinks	3.03	0.87	9.01
	Floats	96.97	0.27	90.99
	Feed	100.00	0.29	100.00

, which describes the trend of Cu grade and distribution rate in the sinks with the separation density. It could be seen that, with the increase in separation density, the Cu grade in the sinks gradually increased and the Cu distribution rate gradually decreased. When the separation density was 2.55 g/cm³, the low-grade floats with a yield of 25.57% could be removed, the grade and distribution rate of Cu in the sinks were 0.36% and 88.21%, respectively, and Cu was well enriched and recovered. When the separation density was increased from 2.55 g/cm³ to 2.60 g/cm³, the distribution rate of Cu in the sinks decreased greatly. Therefore, it can be concluded that the heavy media preconcentration of chalcopyrite was feasible, and the separation density could be determined as 2.55 g/cm³.

3.5.3. Comparison of Product FST for Different Crushing Methods

The separation density of the comparison test was 2.55 g/cm³, and the FST was performed on the −3 + 0.5 mm size fraction in the HPGR (−3 mm) product and the −3 + 0.5 mm size fraction in the JC (−3 mm) product, respectively. The results are shown in

Table 7. Comparison results of FST of products with HPGR and JC crushing.

Crushing Methods	Products	Yield (%)	Cu Grade (%)	Cu Distribution Rate (%)
HPGR	Sinks	75.24	0.35	89.86
	Floats	24.76	0.12	10.14
	Feed	100.00	0.29	100.00
JC	Sinks	93.35	0.29	96.62
	Floats	6.65	0.14	3.38
	Feed	100.00	0.28	100.00

. It could be seen that the grade and distribution rate of Cu in the sinks obtained by HPGR-FST were 0.35% and 89.86%, respectively, and the floats yield was 24.76%, while the grade and distribution rate of Cu in the sinks obtained by JC-FST were 0.29% and 96.62%, respectively, and the floats yield was 6.65%. For the same size fraction, the sinks enrichment effect obtained by HPGR-FST was better, and the floats removal rate was higher, compared with JC-FST. The main reason was that, even for the same size fraction, the HPGR crushed product was finer in particle size, and the chalcopyrite liberation degree was higher, compared with JC crushed product.

4. Conclusions

The effects of HPGR crushing and JC crushing on chalcopyrite liberation and chalcopyrite heavy media preconcentration were studied. Compared with JC crushing, the HPGR was prone to the phenomenon of high fine-particles content and uneven particle size distribution of the product, which also led to higher chalcopyrite liberation in the HPGR product, and the liberation could be increased by 14%~18%.

At the same crushing fineness, the single-particle crushing of the JC tended to produce intergranular fracturing of chalcopyrite, while the lamination crushing of the HPGR produced more transgranular fracturing of chalcopyrite. This phenomenon was mainly related to the crushing mode of the two crushing methods, with the JC mainly focusing on single-particle crushing and the HPGR on lamination crushing. The compression of lamination crushing was much stronger and able to penetrate into the interior of the mineral particles, leading to more transgranular fracturing. The chalcopyrite in the −5 + 0.5 mm particle size of both crushing methods mainly produced intergranular fracturing, and the chalcopyrite in the −0.5 mm particle size mainly produced transgranular fracturing.

The FST results showed that heavy media preconcentration was suitable for chalcopyrite, and the optimum test parameters were determined with the size fraction of −3 + 0.5 mm and the separation density of 2.55 g/cm³. At the same size fraction, the sinks enrichment obtained by HPGC-FST was better, and the floats removal rate was higher, compared to JC-FST. The grade and distribution rate of Cu in the sinks obtained by HPGR-FST were 0.35% and 89.86%, respectively, and the floats yield was 24.76%.