Effects and Mechanisms of Grinding Media on the Flotation Behavior of Scheelite.

Grinding, an essential procedure for size reduction and fresh surface exposure of mineral particles, plays an important role in mineral flotation. The grinding media are the key factors for effective grinding and thus for successful flotation. In this study, ceramic ball (CB) and cast iron ball (CIB), two representative grinding media, were chosen to investigate the effects and mechanisms of grinding media on the flotation behavior of scheelite. The results of pure scheelite flotation show that scheelite ground by CB has a better floatability than that ground by CIB. Scanning electron microscopy-energy dispersive spectroscopy (SEM-EDS) and X-ray photoelectron spectroscopy (XPS) analyses indicate that there are Fe species, namely, elemental iron (Fe), ferrous oxide (FeO), and iron oxyhydroxide (FeOOH), coated on the surfaces of scheelite ground by CIB but not in the case of scheelite ground by CB. The dissolved oxygen (DO) and inductively coupled plasma-atomic emission spectroscopy (ICP-AES) tests show that Fe ions exist in the CIB grinding slurry but not in the case of CB grinding slurry. Compared with the CB grinding slurry, the CIB grinding slurry has a lower DO content and higher Ca ion concentration. Zeta potential results reveal that the Fe species in the CIB grinding reduce the NaOl adsorption on the scheelite surfaces. Finally, the deleterious effect of CIB grinding on the flotation behavior of scheelite is verified by the actual scheelite ore flotation experiments.

Introduction

Tungsten, a rare metal element, is located at the VIB group of the sixth period of the periodic table. Its atomic number is 74, and its relative atomic mass is 183.85. While having the highest melting point (3410 °C) of all metals, tungsten has a density of 19.35 g/cm3 and is categorized as a heavy metal. Tungsten has already been widely used in many fields such as chemistry, mining, electronics, metallurgy, etc. Scheelite (CaWO4) and wolframite ((Fe, Mn)WO4) are the two main tungsten-bearing minerals that are being mined commercially. However, with the increasing depletion of the easily beneficiated wolframite resources, scheelite resources are being exploited more and more broadly and play an important role in tungsten production.[1]

Flotation is the most widely used beneficiation method in the mining industry. In most scheelite flotation plants, the flotation recoveries are merely 60–85%, such as Cantung mine (75–79%) in Canada,[2] Shizhuyuan mine (65%) in Hunan Province, China,[3] and a skarn scheelite type (71%) in Jiangxi Province, China.[4] A great amount of research has been conducted to improve the flotation recovery of scheelite with the focus on the flotation reagents.[5−9] For example, Han et al.[10] combined the Pb2+ ions with benzohydroxamic acid (BHA) to self-assemble the Pb-BHA complex, which showed superior selective collecting ability for scheelite compared with Pb2+ ions added before BHA. In the Shizhuyuan mine, using Pb-BHA as the collector increased the flotation recovery by 12.57%.[10]

While improved scheelite flotation recovery can be obtained via flotation reagent adjustment, an important factor (grinding media) that could potentially affect the scheelite flotation performance is ignored by researchers. It is well known that cast iron ball (CIB) is the first choice for scheelite grinding because of its cheap price and high grinding efficiency. However, using CIB in grinding inevitably results in Fe contaminations. It has been reported that Fe contamination has deleterious effects on flotation of a great number of minerals, such as sulfide minerals,[11−13] platinum group metal (PGM),[14] carbonate minerals,[15] and so on, whereas using inert grinding media such as ceramic ball (CB) can enhance the sulfide flotation performance.[16−19] Nevertheless, Zhu and Yu[20] and Hu and Sun[21] found that using CIB in grinding could increase the flotation recovery of beryl and spodumene. The adhered iron fillings on the fresh beryl and spodumene surfaces would be oxidized to Fe ions, which activated the flotation of beryl and spodumene. Therefore, it can be concluded that the CIB in grinding can have different effects on flotation depending on the minerals processed. However, the effects of CIB grinding on the flotation behavior of scheelite are still unclear. Herein, this research investigated the effects of grinding media on the flotation behavior of scheelite. CIB and CB were chosen as the grinding media and tested in parallel. The mechanisms were revealed by zeta potential measurements, dissolved oxygen (DO) tests, inductively coupled plasma-atomic emission spectroscopy (ICP-AES) tests, scanning electron microscopy-energy dispersive spectroscopy (SEM-EDS) tests, and X-ray photoelectron spectroscopy (XPS) tests.

Results and Discussion

Pure Scheelite Flotation Performance

To investigate the effects of NaOl concentration on the flotation behavior of scheelite ground by CB and CIB in parallel, pure scheelite flotation experiments were conducted as a function of NaOl concentration, and the results are shown in Figure a. It can be seen that when the NaOl concentration is lower than 1.0 × 10–4 mol/L, the flotation recoveries of scheelite increase gradually with the increase of NaOl concentration. The two flotation curves shown in Figure a reach their highest levels at 1.0 × 10–4 mol/L NaOl concentration. After that, the flotation recoveries of scheelite remain at about 84.0% for CB and 69.0% for CIB until 1.5 × 10–4 mol/L NaOl concentration. However, when the NaOl concentration is above 1.5 × 10–4 mol/L, the flotation recoveries of scheelite show a slight decrease with increasing NaOl concentration, which may result from the multilayer adsorption of NaOl on the scheelite surface or the formation of NaOl micelles.[22] Clearly, within the experimental NaOl concentration, the flotation recoveries of scheelite ground by CB are always higher than those of scheelite ground by CIB.

Figure 1

Flotation behavior of scheelite ground by CB and CIB in parallel as a function of (a) NaOl concentration and (b) slurry pH.

Figure b shows the effects of slurry pH, across the pH range of 6–12, on the scheelite flotation ground by CB and CIB in parallel at 1.0 × 10–4 mol/L NaOl concentration. With the increase of slurry pH, the flotation recoveries of scheelite increase very slightly and reached their highest level at pH 9.0–10.0. Above the slurry pH 10.0, the flotation recoveries of scheelite first decrease very slightly (pH < 11.0) and then sharply (pH > 11.0) with increasing slurry pH. This sharp decrease may be induced by the competitive adsorption between NaOl and abundant OH– for the active Ca sites on the scheelite surface.[23,24] However, across the experimental slurry pH range, the flotation recoveries of scheelite ground by CB are higher than those of scheelite ground by CIB.

According to the flotation results shown in Figure , it can be concluded that scheelite ground by CB has a better performance in subsequent flotation than that ground by CIB.

SEM-EDS Analysis

As flotation is a method based on the physicochemical characteristics of mineral surfaces, the surface morphology and elemental composition are the predominant factors in determining mineral flotation behavior. Figure shows the SEM-EDS results of scheelite surfaces subjected to different grinding media. From Figure a, it can be known that the surfaces of scheelite ground by CB are smooth and clean without small particles visible on the surface. Figure b shows the EDS image of scheelite ground by CB. It shows that the elements Ca, W, and O are all detected on the scheelite surfaces, which are the constituent elements of scheelite. No impurity element is found on their surfaces. However, when the scheelite is ground by CIB, the surfaces are rough and uneven as shown in Figure c. It is evident that there are small particles coated on their surfaces. Figure d shows the EDS image of scheelite ground by CIB. Besides the elements Ca, W, and O, Fe is also detected, which originates from the grinding media of CIB. This indicates that there are Fe contaminations from the CIB grinding coated on the scheelite surfaces but not in the case of CB grinding.

Figure 2

(a) SEM and (b) EDS images of scheelite ground by CB, (c) SEM and (d) EDS images of scheelite ground by CIB.

DO and ICP-AES Analysis

Figure shows the contents of DO, Ca, and Fe ions in the pulp of scheelite ground by CB and CIB in parallel. The DO contents of the scheelite pulp ground by CB and CIB are 8.36 and 5.14 ppm, respectively. The lower content of DO in the CIB grinding is due to the consumption of oxygen in galvanic reactions.[17] Additionally, the Ca ion concentration in the pulp of scheelite particles ground by CIB is significantly higher than that in the pulp of scheelite particles ground by CB. This indicates a considerably increased dissolution of the scheelite surface in the CIB grinding, which in turn will reduce the flotation recovery of scheelite.[25] Regarding the Fe ion concentration, 0.078 mg/L Fe ion concentration is detected in the CIB grinding but not in the case of CB grinding. This indicates that the Fe contaminations exist in the scheelite pulp using CIB as the grinding media, in line with the SEM-EDS results.

Figure 3

Pulp properties of scheelite ground by CB and CIB in parallel.

Zeta Potential Analysis

The zeta potentials of scheelite particles ground by CB and CIB, respectively, in the absence and presence of NaOl were measured, and the results are shown in Figure . In the absence of NaOl, the zeta potential of scheelite is negatively charged and decreases with increasing slurry pH, which is consistent with the published literature.[26,27] However, it should be noted that the zeta potential of scheelite ground by CB is higher than that of scheelite ground by CIB. This may be caused by the coating of positively hydrophilic Fe species generated in the CIB grinding on the scheelite surfaces.[28] In the presence of NaOl, the zeta potential of scheelite shifts to a more negative value, indicating the chemisorption of the anionic carboxylate functional group in NaOl onto the negatively charged scheelite surfaces.[29,30]Figure also shows that the decreases in zeta potential of scheelite particles before and after NaOl adsorption are larger for CB grinding than for CIB grinding, indicating that more NaOl is adsorbed on the surfaces of scheelite ground by CB. This may be caused by the coating of the hydrophilic Fe species, which not only cover the Ca2+ active sites, which are essential for NaOl adsorption, but also increase the hydrophilicity of scheelite surfaces.[17,31,32] This thickens the hydration film, which in turn makes the NaOl adsorption more difficult.[33,34]

Figure 4

Zeta potentials of scheelite ground by CB and CIB in parallel as a function of slurry pH.

XPS Analysis

To have further knowledge of the elemental compositions and states of scheelite surfaces ground by CB and CIB in parallel, XPS tests were conducted, and the results are shown in Figure . From Figure a, it can be seen that the peaks of Ca2p, O1s, W4f, and C1s are all detected on the surfaces of scheelite ground by CIB and CB. Ca, O, and W are the constitution elements of scheelite, while C is the background element for XPS tests. Additionally, it can be seen that a weak Fe2p peak at 711.3 eV is shown in the survey spectra of scheelite ground by CIB, which further proves that there are Fe-containing species coated on the scheelite surfaces in CIB grinding. Figure b shows the atomic concentrations of Ca, O, W, C, and Fe on the surfaces of scheelite ground by CB and CIB in parallel. It can be known that 1.36% Fe is found on the scheelite surfaces ground by CIB but not in the case of those ground by CB. Besides, in the case of CIB grinding, the calcium content on the scheelite surfaces is lower while the oxygen content is slightly higher than those in the case of CB grinding. These may be attributed to the galvanic reactions in CIB grinding, which consume the oxygen to form Fe–O species and cover the active calcium sites on the scheelite surfaces.[31,32,35]

Figure 5

(a) XPS survey spectra and (b) atomic concentrations of scheelite ground by CB and CIB in parallel.

To identify the chemical states of Fe and O on the scheelite surfaces, the Fe2p2/3 and O1s narrow spectra of scheelite ground by CB and CIB in parallel were analyzed through peak fitting, and the results are shown in Figure . As shown in Figure a, when the scheelite is ground by CIB, there are five different peaks in the spectra of Fe2p3/2. The peak at 706.70 eV is ascribed to elemental iron (Fe),[36,37] and the peaks at 709.45 and 711.30 eV are attributed to ferrous oxide (FeO) and iron oxyhydroxide (FeOOH), respectively.[38,39] Additionally, the peaks at 715.25 and 716.60 eV are the satellites of FeO and FeOOH, respectively.[40]Figure b shows the O1s narrow spectra of scheelite ground by CB and CIB in parallel. When the scheelite is ground by CB, there are only two peaks in the spectra of O1s, which are assigned to the W–O group at 530.49 eV and Ca–O group at 532.03 eV.[41,42] However, when the scheelite is ground by CIB, four different peaks appear in the narrow spectra of O1s. The peaks of the W–O group and Ca–O group shift to 530.55 and 532.45 eV, respectively. Additionally, the peaks at 529.80 and 531.51 eV are assigned to Fe(II)–O and Fe(III)–O, respectively,[36] indicating that there are Fe–O species, namely, Fe(II)–O and Fe(III)–O, on the scheelite surfaces ground by CIB.

Figure 6

Peak fitting of (a) the Fe2p3/2 and (b) O1s narrow spectra of scheelite ground by CB and CIB in parallel.

It can be concluded from Figure that there are elemental iron, iron oxide, and iron oxyhydroxide species generated on the scheelite surfaces ground by CIB but not in the case of scheelite surfaces ground by CB. Elemental iron may originate from the worn parts of CIB, while iron oxide and iron oxyhydroxide can be identified from the galvanic reactions in CIB grinding. The reaction mechanisms can be shown as follows:

As illustrated in Figure , in the presence of oxygen and water, the galvanic reactions can take place between the CIB media. The reactions are as follows:[43,44]

Figure 7

Galvanic reactions between the CIB media in grinding.

The galvanic reactions will result in the worn parts of CIB and the formation of iron oxide and iron oxyhydroxide. In the grinding, these hydrophilic Fe species can coat the scheelite surfaces under the collision and extraction of CIB.

Actual Scheelite Ore Flotation Performance

To verify the effects of grinding media on the flotation performance of actual scheelite ore, the flotation experiments of actual scheelite ore samples ground by CB and CIB in parallel were conducted, and the results are shown in Figure . The flotation recovery was calculated according to the WO3 grade of the scheelite flotation feed. Flotation efficiency was chosen to evaluate the flotation performance. The formula of flotation efficiency is as follows:In the above formula, E is the flotation efficiency; ε and γ are the flotation recovery and flotation yield of rougher concentrates, respectively; α is the WO3 grade of scheelite flotation feed; and βx is the WO3 grade in pure scheelite, which accounts for 80.52%.

Figure 8

Flotation results of actual scheelite ore samples ground by CB and CIB in parallel.

From Figure , it is clear that the WO3 grade and recovery of rougher concentrates in CB grinding are higher than those of rougher concentrates in CIB grinding. Additionally, the flotation efficiency in the CB grinding is 79.18%, while in the CIB grinding, the flotation efficiency is merely 76.63%, illustrating that the flotation performance is better when the actual scheelite ore is ground by CB. Meanwhile, Figure also shows that the slurry color of the actual scheelite ore ground by CB is gray. In contrast, when the actual scheelite ore is ground by CIB, the slurry color is dark green, indicating that the slurry is contaminated when using CIB in grinding.

It is worth noting that in addition to Fe contamination, the difference in mill geometries could contribute to the distinct flotation response when different types of mills are used. Continued work is needed to decouple the effect of Fe contamination and mill geometries on the flotation behavior of mineral particles.

Conclusions

In this study, the effects and mechanisms of grinding media (CB and CIB) on the flotation behavior of scheelite were investigated through pure mineral flotation, actual scheelite ore flotation, SEM-EDS tests, XPS tests, zeta potential measurements, DO, and ICP-AES tests. The results indicate that pure scheelite ground by CIB has a lower flotation recovery than that ground by CB, which is attributed to the coating of Fe species (Fe, FeO, FeOOH) on CIB ground scheelite surfaces, which in turn impede the subsequent adsorption of NaOl. The actual scheelite ore flotation results are in good agreement with the pure scheelite flotation results, which verifies the Fe contaminations in the CIB grinding from a real ore perspective. This research will guide the grinding medium selection in beneficiation of scheelite ore in industry and effectively improve the scheelite flotation performance, which is significant in the efficient use of tungsten resources.

Experimental Section

Materials and Reagents

The pure scheelite samples were collected from Jiangxi Province, China. The massive crystals were hand-ground by a hammer to −1.7 mm and hand-selected for the high-purity parts. The chemical multielement analysis results of the pure scheelite samples (Table ) show that the WO3 grade is 76.98%, which indicates that the purity of pure scheelite samples is as high as 95.56%. The X-ray diffraction (XRD) patterns of the pure scheelite samples (Figure ) further show the high purity of the samples.

Table 1

Chemical Multielement Analysis Results of the Pure Scheelite Samples

element	CaO	WO3	SiO2	MgO	Al2O3	Fe	purity
content/%	18.54	76.98	1.16	0.98	1.22		95.56

Figure 9

XRD patterns of pure scheelite samples.

The actual scheelite ore samples used in batch flotation experiments were collected from a scheelite beneficiation plant in Jiangxi Province, China. Table shows the chemical multielement analysis results of the actual scheelite ore samples. The WO3 grade of the samples is 0.45%. The main gangue minerals are quartz, calcite, and fluorite with contents of 51.53, 13.14, and 8.51%, respectively. Table shows the phase analysis of tungsten in the actual scheelite ore samples. It shows that the main tungsten-bearing mineral is scheelite, which accounts for 95.11%, with only 3.11% wolframite and 1.78% tungstite.

Table 2

Chemical Multielement Analysis Results of the Actual Scheelite Ore Samples

element	WO3	Cu	Bi	Mo	Pb	Zn
content/%	0.45	0.158	0.049	0.0044	0.029	0.076

Table 3

Phases of Tungsten in the Actual Scheelite Ore Samples

phase	scheelite	wolframite	tungstite	total tungsten
WO3/%	0.428	0.014	0.008	0.450
distribution rate/%	95.11	3.11	1.78	100.00

Sodium oleate (NaOl, C18H33O2Na, analytical grade) used as the collector was purchased from Shanghai Maikun Chemical Co., Ltd., China. Hydrochloric acid (HCl, analytical grade) and sodium hydroxide (NaOH, analytical grade) bought from Sinopharm Chemical Reagent Co., Ltd., China, were used as the pH modifiers in pure scheelite flotation experiments. Sodium carbonate (Na2CO3, analytical grade) bought from Sinopharm Chemical Reagent Co., Ltd., China, and water glass (Na2SiO3, technical grade) obtained from the scheelite beneficiation plant in Jiangxi Province, China, were used as the pH modifier and depressant, respectively, in the actual scheelite ore flotation experiments. Potassium butyl xanthate (KBX, C4H9OCSSK, technical grade) and ammonium dibutyl dithiophosphate (ADD, (C4H9O)2PSSNH4, technical grade) obtained from a sulfide flotation plant were used as the collectors to remove sulfides in the actual scheelite ore before scheelite flotation. Deionized (DI) water was used in all the pure scheelite experiments and tests. Tap water was used in the actual scheelite ore flotation experiments.

Grinding and Flotation

The pure scheelite samples were ground in a tumbling mill equipped with a 1.5 L ceramic pot. CIB (Mohs hardness 7.0) and CB (Mohs hardness 9.0) were chosen as the grinding media in parallel with a 50.0% media filling rate (volume ratio of grinding media to the mill cavity) and 50.0% grinding concentration (mass ratio of grinding materials to the water and grinding materials). After grinding, the ground products were sieved to obtain the −38 + 75 μm fractions for pure mineral flotation experiments, SEM-EDS, and XPS tests. The −38 μm fractions were further ground to −5 μm for zeta potential tests. The flotation experiments of pure scheelite samples were conducted in an XFGC flotation machine manufactured by Jilin Exploration Machinery Plant, China. The speed of the impeller was set at 1600 rpm. For each experiment, 2 g of pure scheelite samples and 35 mL of DI water were mixed in a 40 mL flotation cell. Then, the pH modifier and NaOl were added sequentially and stirred for 2 and 3 min, respectively. After that, the concentrates were collected for about 4 min, and both the concentrates and tailings were filtered, dried, and weighed to calculate the flotation recovery. The flotation flowsheet of pure scheelite samples is shown in Figure .

Figure 10

Flotation flowsheet of pure scheelite samples.

The actual scheelite ore samples were ground using a laboratory conical ball mill (HLXMQ-Φ240 × 90, Wuhan Hengle Mineral Engineering Equipment Co., Ltd.) and a tumbling mill (5-II, Haoqiang Machinery Factory, Yixing town, Jiangsu Province, China) equipped with a 5.0 L ceramic pot. The laboratory conical ball mill was filled with CIB, and the mill shell composition was cast iron; meanwhile, the tumbling mill was filled with CB, and the mill shell composition was ceramic. For each grinding, the medium filling rate and grinding concentration were 50.0 and 50.0%, respectively. The −75 μm fractions of the ground products accounted for about 62.0%, which was consistent with the scheelite beneficiation plant. The flotation experiments of actual scheelite ore samples were conducted in a 1.5 L HLXFD flotation machine manufactured by Wuhan Hengle Mineral Engineering Equipment Co., Ltd., China. For each experiment, 800 g of the ground products and 1200 g of tap water were mixed. Then the reagents were added sequentially and stirred according to the flotation flowsheet of actual scheelite ore samples shown in Figure . After flotation, the rougher concentrates and tailings were filtered, dried, and weighed. The WO3 grades of the rougher concentrates and tailings were analyzed to calculate the flotation recovery.

Figure 11

Flotation flowsheet of actual scheelite ore samples.

Zeta Potential Measurements

Zeta potential measurements were carried out using a Zetasizer Nano ZS90 analyzer manufactured by Malvern Panalytical Ltd., England. A dilute suspension was prepared by adding 50 mg of pure scheelite samples to 35 mL of KCl solution (10–2 mol/L) and stirred with a magnetic stirrer for 2 min. Then, the pH modifier and NaOl were added sequentially and stirred for 2 and 3 min, respectively. After that, the suspension was allowed to settle naturally for 10 min. The upper layer of the suspension was sucked out and injected into the testing sample cell for measurement. Each measurement was repeated three times at 25 °C, and the average was calculated.

DO and ICP-AES Tests

After grinding, the slurry of pure scheelite samples ground by CIB and CB in parallel was transferred into sealed beakers for settlement. After settlement, the upper clarified liquid was decanted for DO tests (AZ8403, Hengxin Technology Co., Ltd., China) and ICP-AES tests (ICPE-9800, Shimadzu Corporation, Japan).

SEM-EDS Tests

Scanning electron microscopy (SEM, JSM-6610, JEOL Ltd., Japan) coupled with energy dispersive spectroscopy (EDS, QUANTAX200-30, Bruker Corporation, Germany) was applied to observe the surface topography and elemental composition of scheelite ground by CIB and CB in parallel.

XPS Tests

To analyze the elemental composition and elemental states of scheelite surfaces ground by CIB and CB in parallel, XPS tests were conducted using an X-ray photoelectron spectrometer (ESCALAB 250 XI, Thermo Fisher Scientific, USA). The testing parameters were as follows: Al Kα (hυ = 1486.6 eV) energy resource, 8 × 10–10 Pa of the analytical chamber vacuum, 12.5 kV of the testing voltage, and 16 mA of the testing current. All the binding energies were referred to the C1s (284.80 eV).