- Email: [email protected]
Effect of agitation on the interaction of coal and kaolinite in flotation
Powder Technology 313 (2017) 122–128
Contents lists available at ScienceDirect
Powder Technology journal homepage: www.elsevier.com/locate/powtec
Effect of agitation on the interaction of coal and kaolinite in flotation Yuexian Yu a,b, Gan Cheng c,d, Liqiang Ma a,⁎, Gen Huang a, Lun Wu a, Hongxiang Xu a,⁎ a
School of Chemical and Environmental Engineering, China University of Mining and Technology (Beijing), Beijing 100083, China Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta T6G 1H9, Canada c College of Chemistry and Chemical, Henan Polytechnic University, Jiaozuo 454000, Henan, China d Synergism Innovative Center of Coal Safety Production in Henan Province, Jiaozuo 454000, China b
a r t i c l e
i n f o
Article history: Received 10 November 2016 Received in revised form 8 February 2017 Accepted 2 March 2017 Available online 03 March 2017 Keywords: Agitation Slime coating Zeta potential distribution Heterocoagulation
a b s t r a c t In this work, the interaction behavior and underlying mechanism of coal and kaolinite particles in the presence of agitation was investigated by flotation tests, homo and hetero settling tests, turbidity meter, zeta potential distribution, focused beam reflectance measurement (FBRM) and scanning electron microscope (SEM). The results show that, in the presence of kaolinite, the flotation combustible matter recovery of coal decreases from 0 rpm to 1200 rpm and then increases from 1200 rpm to 2000 rpm, indicating that the mild agitation enhances kaolinite-coating and the high intensity agitation mitigates the kaolinite-coating, which is demonstrated by the zeta potential distribution and turbidity measurements. However, the degree of the heterocoagulation between coal and kaolinite is not serious. It is found that the kaolinite particles partially cover on the coal surface by two morphologies includes individuals and aggregates through the size-based scanning electron microscope analysis. In addition, the affinity between fine kaolinite and fine coal particles is more pronounced. The kaolinite-coating enhancement caused by the mild agitation is explained by the Derjaguin–Landau–Verwey–Overbeek (DLVO) theory. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Flotation is widely used in coal industry for treating fine coal. This process greatly relies on the distinct interfacial properties between coal and gangue minerals. Good hydrophobicity of the coal surface is the key to ensure the success of the flotation. Unfortunately, nowadays China coal industry is facing the challenge of processing clayey coal as a result of the depletion of high quality coal deposits. According to the literature, clay particles, usually b2 μm, have a significant effect on coal flotation [1–3]. For example, fine hydrophilic clay particles are easily dragged by the interstitial liquid film between air-bubbles and enter the froth layer with liquids in flotation, resulting in mechanical entrainment and low concentrate quality [1,2]. Clay minerals can also increase reagent consumption and pulp viscosity, which lowers the flotation efficiency [4,5]. One of the most important effects caused by clay minerals is slime coating, defined as a layer of fine or ultrafine colloidal particles coated on the larger value mineral surface, which has been recognized to have a deleterious effect on flotation. It is speculated that these slimes on mineral surface form a hydrophilic “armor” preventing mineral particles from contacting with collectors and/or air bubbles, consequently resulting in a lower flotation recovery [6–8]. Numerous researchers have demonstrated that hydrophilic clay ⁎ Corresponding authors. E-mail addresses: [email protected] (L. Ma), [email protected] (H. Xu).
http://dx.doi.org/10.1016/j.powtec.2017.03.002 0032-5910/© 2017 Elsevier B.V. All rights reserved.
particles may coat the coal surface, making it hydrophilic and preventing the adsorption of collectors and therefore depress coal flotation [4,6,9–12]. Originally, Derjaguin–Landau–Verwey–Overbeek (DLVO) theory was developed to interpret the aggregate stability of colloidal particles, and recently researchers used this theory to explain the interaction behaviors of mineral particles in suspension. The DLVO interaction energy and force between coal and clay particles was calculated by some researchers [3,13]. The results indicate that clay coating is governed by the van der Waals attraction and that the double-layer interaction played a secondary role [3]. It is interesting to find that there is an energy barrier in the DLVO interaction energy curve as shown in Fig. 8, which means the energy barrier should be overcome when clay coating occurs. Therefore, the kinetic energy input (e.g. agitation) may enhance slime coating by overcoming the energy barrier, aggravating the flotation performance. In addition, agitation will improve the particleparticle collision, providing a higher attachment probability of coal and clay particles. However, so far, many works were conducted to investigate the effect of slime coating on coal flotation, which mainly focused on clay types, slurry pH and water hardness [4,11,12,14,15]. The aim of this work is to investigate the effect of agitation on the interaction of coal and kaolinite in flotation. The interaction behavior and underlying mechanism of agitation on slime coating was investigated by flotation tests, homo and hetero settling tests, turbidity meter, zeta potential distribution, focused beam reflectance measurement (FBRM)
Y. Yu et al. / Powder Technology 313 (2017) 122–128
and scanning electron microscope (SEM). The results of this research are expected to provide a more detailed description of the flotation behavior of coal-kaolinite system in the presence of kinetic energy input and give a reference for the coal flotation. 2. Experimental 2.1. Materials The pure coal used in this study was obtained by hand-picking the coarse gravity separation clean coking-coal supplied from Huainan coal preparation plant, China. In order to obtain a purer coal, the original coal samples were separated by the heavy liquid of 1.30 g/cm3 and a light fraction of the coal with the ash content of around 5% was produced. Then obtained coal were crushed, ground and used as the pure coal for the subsequent tests. Some impurities might be re-liberated during crushing and grinding, but their amount should be very low since the ash content of the light fraction is quite low. In addition, most of impurities in coal belong to clay minerals which having a similar property with the kaolinite, therefore, the re-liberated impurities would cause little effect on the tests. Kaolinite, chemical grade (purity N99%), purchased from Sinopharm Chemical Reagent Co., Ltd. was tested as the pure clay mineral in this study. The size distributions of the kaolinite and the pure coal were measured by a laser diffraction particle size analyzer (Mastersizer 2000, UK). As shown in Fig. 1 that D90 of the kaolinite is 10.832 μm, which approximately matches the magnitude of clay minerals in the coal flotation [1]. D90 of the coal is 258.862 μm, which also matches the size of flotation feed in coal industry. 2.2. Methods 2.2.1. Batch flotation Batch flotation tests were conducted using a 1.5 L Denver flotation cell at neutral pulp pH. For the scenario of extra agitation flotation, 40 g of pure coal was first added into a baffle built-in tank and mixed with 1 L tap water by a glass bar until the coal was wetted fully. Then the desired amount of kaolinite was added into the tank and stirred at five different agitation speeds of 400, 800, 1200, 1600, 2000 rpm for 10 min. After that, the suspension was immediately transferred into the flotation cell and additional 200 ml water was added at the same time. 4 μl MIBC was added for each test. No collectors were used during the whole course of flotation since the collector may improve the surface hydrophobicity of coal particles even when kaolinite-coating occurs and cover the changes caused by kaolinite-coating. The impeller speed for flotation was kept constant at 900 rpm and air flow was
Cumulative size distribution (%)
100 Coal Kaolinite 80
60
40
20
0 0.1
1
10
100
1000
Particle size ( µm) Fig. 1. The cumulative size distributions of the kaolinite and coal sample.
123
kept at 5.23 L/min by an outside independent gas supply system. Froth products were consecutively collected at intervals of 15, 15, 30, 60, 60 s. During the collection, a metronome was placed to obtain a constant scraping speed of 30 times per min. Flotation concentrates and tailings were filtered, dried at 105 °C and weighed for ash analysis and combustible matter recovery calculation. The flotation test was evaluated by the combustible matter recovery, which was calculated by the following formula. ε¼
γc ð100−Ac Þ 100% 100 100−A f
ð1Þ
where ε represents the combustible matter recovery, %; γc represents the yield of the concentrate, %; Ac represents the ash content of the concentrate; Af represents the ash content of the feed. All the flotation tests were repeated for three times and the test results were averaged. 2.2.2. Settling test Settling tests were conducted in 100 ml graduated glass cylinders. Given amounts of solids (5 g for coal and 1.25 g for kaolinite) were accurately weighed and placed in the cylinders separately or mixed together. The cylinders were then filled with 100 ml distilled water and shaken briefly to mix up the slurry. The cylinders were repeatedly inverted for 20 times and left still for 14 h, and then the test photos were taken. This technique was proposed by Xu et al. [11]. 2.2.3. SEM analysis The SEM analysis was conducted by ZEISS MERLIN VP Compact field emission scanning electron microscope. Firstly, the pure coal was wet sieved strictly and a size fraction between 74 μm and 125 μm was obtained. Then 2 g of the screened coal and 0.2 g of kaolinite were mixed in distilled water for 10 min by a magnetic stirrer. After that the suspension was passed through a sifter with pore size of 74 μm and rinsed mildly with distilled water so that the remaining kaolinite particles in the suspension could be separated from the coal-kaolinite aggregation since D90 of the kaolinite was 10.832 μm and the maximum size was 30.200 μm known from the laser size analysis in Section 2.1. The oversized coal-kaolinite aggregates were naturally dried in the air in order to avoid any possible change to the surface structure and then detected by SEM. 2.2.4. Turbidity measurement Mixed solids (5 g for coal and 1.25 g for kaolinite) were accurately weighed and placed into a beaker and agitated by a magnetic stirrer to make sure the solids suspend thoroughly in the water. Ten 10 ml of the stock mineral suspensions was taken and diluted to 200 ml and agitated for 10 min at five different agitation speeds 400, 800, 1200, 1600, 2000 rpm and followed by settling for 1 h, then the turbidity values of supernatants were determined immediately by the turbidity meter, VWR Model 66120-200. 2.2.5. Zeta potential measurement Zeta potential distribution was measured by CAD Zetaphoremetre IV™ (CAD, France). The Smoluchowski model was used to calculate zeta potentials from the measured electrophoretic mobility. Suspensions containing 0.5 g of coal or kaolinite were prepared separately in two jars with 10−3 M KCl solution and conditioned by magnetic stirrer for 20 min and then left to stand for 24 h. For the zeta potential distribution measurement of pure mineral, the suspension was stirred for another 5 min and settled for 5 min, and then the upper portion of dilute fine particle suspension was taken for the measurement. In the case of mixed coal-kaolinite, the two suspensions in the jars were totally transferred into the agitation tank which was used in the flotation stage and diluted into 1 L with 10− 3 M KCl solution and then agitated at each value of desired agitation speeds for 10 min. After settling for 5 min, the upper portion of the suspension was taken for zeta potential
124
Y. Yu et al. / Powder Technology 313 (2017) 122–128
distribution measurement. The pH of the suspension was monitored by an outside pH meter during the measurement. The electrophoretic mobility of about 20–200 particles in the stationary layer were traced and converted into zeta potential distribution. The details of principle were described in the literature [11,16]. In addition, zeta potentials of individual coal and kaolinite under different agitation speeds were measured by a Brookhaven Instruments ZetaPlus Zeta Potential Analyzer. Each reported zeta potential value was the mean of 10 runs, with 10 measurement cycles per run. The pH of suspension for each measurement was recorded by a pH meter.
bar to make sure the sample was wetted fully before the measurement. Then the suspension was placed on a stage and determined by FBRM. 3. Results and discussion 3.1. Batch flotation results Firstly, the effect of agitation on pure coal flotation was studied, and the combustible matter recovery of the pure coal as a function of flotation time at different agitation speeds are shown Fig. 2 (a). As can be seen that the recovery-flotation time curves basically overlapped for the agitation speeds from 0 to 1200 rpm, which indicates that the agitation (when rpm b1200) has very little effect on the pure coal flotation. However, when the agitation speed is over 1200 rpm, the two curves of 1600 and 2000 rpm slightly diverge from the curve of no agitation, showing that the flotation rates of 1600 and 2000 rpm decreased. From Fig. 2 (d), the ultimate combustible matter recovery of the pure coal from 0 to 1200 rpm basically did not change. However, the recovery decreased slightly when the agitation speed continued to increase. This may be probably attributed to the fracture of coal-coal aggregates due to the high intensity agitation as proved in Fig. 9, increasing the amount of fine coal particles which usually having a higher detour probability when collide with bubbles and hence resulting in a lower probability of bubble-particle attachment [19]. However, all in all, the agitation has little effect on the combustible matter recovery of the pure coal flotation in this test. The effect of agitation on the interaction between coal and kaolinite in flotation is shown in Fig. 2 (b) and (c), that the flotation recovery of mixed coal-kaolinite in each moment, no matter in the presence or absence of agitation, is always lower than that of the pure coal, indicating that the kaolinite depressed the flotation of the coal. The ultimate combustible matter recovery of pure coal and mixed coal-kaolinite are shown in Fig. 2 (d), respectively. The combustible matter recovery of 10% and 20% kaolinite content in the absence of agitation are both less than that of the pure coal, same results were obtained by other researchers [11,14]. In addition, it is interesting to find that the flotation
2.2.6. FBRM measurement The measurement of FBRM was carried out by METTLER TOLEDO FBRM® Technology G400. The technique allows real-time, in-process measurement of highly turbid samples. The technique uses an inline probe containing a laser that is focused to a point at the surface of the probe window, which is in contact with the process sample. The focused beam is rotated in a circular path at a controlled scan speed, and pulses of backscattered light are detected by the probe as the focused beam passes over particles and agglomerates. The length of these pulses correlates to the distance across each particle and, by multiplying the length of each pulse by the scan speed, the distance across each particle (chord length) can be calculated. High numbers of chord lengths are measured each second, which enable changes in the particle size distribution and particle count to be monitored [17]. The FBRM probe measures chord lengths, grouped in a chord length distribution, rather than the actual sizes (i.e. diameters) [18]. Chord is supposed to be a fraction of the particle size. Hence, there should be no chord above the maximum particle diameter in the system. The results can be used for qualitative analysis. When flocculation occurs, the chord length value increases. Conversely, when the suspension becomes more dispersed, the chord length decreases. The chord length value is a sensitive indicator of the flocculation or dispersion state of suspensions [1]. For FBRM measurement in this study, 0.5 g of coal and 0.15 g of kaolinite were accurately weighed and added into a beaker with 100 ml distilled water separately or mixed together and first agitated manually with a glass
100
(a)
Combustible matter recovery (%)
Combustible matter recovery (%)
100
80
60
no agitation 400 rpm 800 rpm 1200 rpm 1600 rpm 2000 rpm
40
20
0
60 Pure coal, no agitation 10% Kaolinite, no agitation 10% Kaolinite, 400 rpm 10% Kaolinite, 800 rpm 10% Kaolinite, 1200 rpm 10% Kaolinite,1600 rpm 10% Kaolinite, 2000 rpm
40
20
0 0
50
100
150
200
0
Flotation time (s)
100
(c)
80
60 Pure coal, no agitation 20% Kaolinite, no agitation 20% Kaolinite, 400 rpm 20% Kaolinite, 800 rpm 20% Kaolinite, 1200 rpm 20% Kaolinite,1600 rpm 20% Kaolinite, 2000 rpm
40
20
50
100
150
200
Flotation time (s)
100 Combustible matter recovery (%)
Combustible matter recovery (%)
(b)
80
(d)
95 90 85 80 75 70
Pure coal, no agitation 10% Kaolinite 20% Kaolinite
65 60
0 0
50
100 Flotation time (s)
150
200
0
500
1000
1500
2000
Agitation speed (rpm)
Fig. 2. The combustible matter recovery of pure coal and mixed coal-kaolinite in the absence and presence of agitation. (a) Pure coal; (b) 10% kaolinite added; (c) 20% kaolinite added; (d) the ultimate combustible matter recovery versus the agitation speed.
Y. Yu et al. / Powder Technology 313 (2017) 122–128
recovery presents a further decline in the presence of agitation. In the case of 10% kaolinite addition, the minimum flotation recovery is between 800 and 1200 rpm and the flotation recovery begins to rise again at 1200 rpm. The flotation recovery at the agitation speed of 1600 rpm is higher than the original flotation recovery, and a further increase of agitation speed does not improve recovery a lot, which essentially means the maximum value of flotation recovery obtains. When it comes to the case of 20% kaolinite addition, the recovery curve shows a similar trend compared with that of 10% kaolinite addition. However, there are some differences between them. For example, the minimum combustible matter recovery of 20% kaolinite addition occurs at 1200 rpm, which shows a delay compared with that of the 10% kaolinite addition. Meanwhile, the recovery increases at 1600 rpm and continue to rise obviously at 2000 rpm. Confined by the speed limit of the agitation apparatus, it is speculated that the recovery may continue to increase after 2000 rpm. This means a stronger agitation is needed for a higher kaolinite content to remove the detrimental effect caused by kaolinite. No collectors were added or used during the whole process of flotation, as such, the surface hydrophobicity played a critical role in the floating of the coal in this test. Thus, once the kaolinite-coating occurs, the hydrophobicity of the coal surface should be decreased. Therefore, the lower combustible matter recovery observed in the presence of kaolinite as shown in Fig. 2 (d) should be attributed to the kaolinitecoating. However, clay minerals can affect flotation by affecting froth stability that also lowers flotation recovery of value minerals [20]. Therefore, the depression of the coal flotation in this test should be ascribed to a synergistic effect caused by kaolinite-coating and froth stability changing. In this test, the effect of kaolinite on froth stability should be constant since the kaolinite content for each group is fixed. Therefore, it is indicated that the kaolinite-coating should be responsible for the fluctuation of combustible matter recovery, which indicating the agitation has an important influence on the kaolinite-coating. From the results of this test as shown in Fig. 2 (d), it is expected that the high intensity agitation mitigates slime coating as suggested by many studies [21–24]. However, it is interesting to find that the mild agitation aggravates the kaolinite-coating and its underlying mechanism was investigated in the subsequent tests. 3.2. Settling tests To confirm the presence of slime coating of kaolinite clays on coal surface in the flotation process, the homo and hetero settling tests were conducted at neutral pulp pH same as the flotation solution. The photographs of pure mineral and mixed minerals suspensions after 14 h settlement are presented in Fig. 3. The pictures gave a directly visual observation for the understanding of kaolinite-coating. In the left cylinder that mixed suspension of coal-kaolinite exhibited a clear supernatant. In contrast, the supernatants of pure mineral coal or kaolinite were still turbid after the same period of settling as shown in the middle and right cylinders. By the comparison of those two cases, it is revealed that heterocoagulation between coal particles and kaolinite clays definitely occurred; and this demonstrated that the slime coating of hydrophilic kaolinite is responsible for the reduction of flotation recovery in Section 3.1. 3.3. SEM analysis The directly visual images in Section 3.2 gave a macroscopic observation for the kaolinite-coating on coal surface. Meanwhile, in order to provide a microscopic description of slime coating, SEM analysis was performed as shown in Fig. 4. In the previous works, most of researchers always investigated the phenomenon of slime coating by observing SEM images of the flotation concentrates or tailings directly but failed to realize the important defect during this procedure, i.e. during the filtration or dewatering of flotation products, clay minerals could naturally settle
125
Fig. 3. Photographs of pure mineral and mixed mineral suspension after 14 h settling, coal-kaolinite (left) is clarified; kaolinite (middle) is turbid; coal (right) is turbid.
onto the surface of coal particles. Therefore, the phenomenon of slime coating observed by SEM may not be the true slime coating. In the present work, the test procedure was well-designed and avoided this deficiency successfully. The details were described in Section 2.2.3. As shown in Fig. 4 (a), kaolinite particles were scattered on the coal surface. These hydrophilic kaolinite particles covered on the coal surface and made coal surface less hydrophobic, corresponding to the results of flotation and settling tests. Fig. 4 (b) exhibited a magnified image of partial coal surface which covered by many kaolinite particles and some of these small kaolinite particles aggregated together by edge–edge (E–E), edge– face (E–F), and face–face (F–\\F) structures as expected [25,26]. Meanwhile, some of kaolinite particles covered on the coal surface individually. Therefore, kaolinite may cover on the coal surface by two morphologies includes individuals and aggregates. 3.4. Turbidity measurement In order to investigate the effect of agitation strength on particle flocs of the kaolinite-coal suspension, supernatant turbidity of coal-kaolinite suspension after settling for 1 h at different agitation speeds was tested to monitor the situation of flocs in the suspension. As shown in Fig. 5, the supernatant turbidity after settling for 1 h in the absence of agitation was around 180 NTU and decreased to 104 NTU in the case of 400 rpm, then increased gradually with the rising agitation speed and was higher than the original turbidity at the end. This can be explained as follows: firstly, there are three types of aggregates in the suspension, i.e. coal-coal, coal-kaolinite and kaolinite-kaolinite. At the first descending point of the curve, the quantity of three types of aggregates including the coal-kaolinite raised corresponding to a lower combustible matter recovery at the agitation speed of 400 rpm as shown in Fig. 2 (d). However, when the agitation speed increased to 800 rpm, the turbidity of supernatant exhibited a light increase. This may be attributed to the competitive outcome between the crack of homo coagulations (coal-coal and kaolinite-kaolinite) and the formation of coal-kaolinite aggregate. When the agitation speed increased from 800 rpm to 1200 rpm, an obvious increase in supernatant turbidity appears, which represents the further crack of coal-coal coagulation and the initial crack of coal-kaolinite aggregate, therefore, the combustible matter recovery began to increase. After 1200 rpm, the supernatant turbidity continued to show an increase mainly due to the fractures of
126
Y. Yu et al. / Powder Technology 313 (2017) 122–128
Fig. 4. SEM images of coal–kaolinite aggregates.
coal-kaolinite aggregates corresponding to the further increase in combustible matter recovery as shown in Fig. 2 (d). To summarize, it is indicated that low intensity agitation enhances the particle aggregation; however, high intensity agitation breaks up the aggregates. 3.5. Zeta potential distribution As shown in Fig. 6 (c), the zeta potentials of coal and kaolinite essentially both remained unchanged regardless of the agitation speed. Therefore, the agitation does not change the surface electrical properties of coal and kaolinite. In order to recognize the effect of agitation on the interaction of coal and kaolinite, the zeta potential distribution of mixed coal-kaolinite at different agitation speeds in distilled water with 10−3 M KCl were determined. As proposed by literature that if strong heterocoagulation occurs, the zeta potential distribution of a binary mixture will show single distribution peak. On the contrary, a bimodal zeta potential distribution which is consistent with the peaks of the two individual minerals usually means there is no attraction between the two components [11,27]. Therefore, it is assumed that if the agitation has an effect on the interaction of coal and kaolinite, a change or shift of the zeta potential distribution of the binary system will be expected. As presented in Fig. 6 (a) and (b), the individual zeta potential distributions of coal and kaolinite are close to each other with some parts overlapping from −60 mV to −80 mV. However, most of kaolinite particles sit on the right side of the dash line at −75 mV and on the contrary, most of coal particles sit on the left side. The zeta potential distributions of mixed coal-kaolinite systems are shown in Fig. 7. A single distribution peak appears when there is no 240 Turbidity
Turbidity (NTU)
200 160 120 80 40 0 0
500
1000
1500
2000
Agitation speed (rpm) Fig. 5. Supernatant turbidity of coal-kaolinite blending suspension after settling for 1 h at different agitation speeds in distilled water.
agitation as shown in Fig. 7 (a), however, this does not definitely mean that strong attraction occurs between coal and kaolinite since the zeta potential distributions of the two individual components are too close as shown in Fig. 6. The zeta potential distributions of the mixed system under different agitation speeds look similar as listed in Fig. 7 (b), (c), and (d). However, when drawing a line at −75 mV, it is found that the frequency distributions on the left side of the line firstly decreased and then increased as the increasing agitation speed. As discussed above, most of kaolinite particles sit on the right side of the line and meanwhile, most of coal particles sit on the left side. Therefore, it can be inferred that more and more coal particles were covered by kaolinite clays as the agitation speed increasing from 0 to 1200 rpm. When the agitation speed increases to 2000 rpm, the frequency distribution on the left side of the line increases significantly, which indicates that the kaolinite clays are removed from the coal surfaces at 2000 rpm. It is well consistent with the flotation results obtained in Fig. 2. In addition, as shown from the four plots in Fig. 7, the zeta potential distributions of coal and kaolinite always coexist, suggesting that there are still some remnant coal and kaolinite particles in the mixed suspension, which implies that the aggregation between coal and kaolinite is not strong. In other words, the coal surface is partially covered by the kaolinite particles as shown in Fig. 4. This is also proved by the flotation results, i.e. the combustible matter recovery does not show a dramatic fall in the presence of kaolinite (just from 94% to 72%), indicating that the slime coating between coal and kaolinite is not serious. The interactions between coal and clay particles based on Derjaguin– Landau–Verwey–Overbeek (DLVO) theory of the aggregate stability of colloidal particles are shown in Fig. 8. As can be seen from Fig. 8, the double-layer forces for the two surface charging mechanisms are repulsive and hence the double-layer interaction under the surface charge regulation occurring between the two extreme cases is also repulsive. On the contrary, the van der Waals force is attractive at all separation distances. In particular, at short separation distance, the van der Waals attraction is so strong that the double-layer repulsion is overcome, resulting in strong net DLVO attraction between the particles at close contact. One point should be mentioned is that there is an energy barrier in the DLVO interaction force curve and the energy barrier must be overcome before the slime coating occurs. Therefore, it can be concluded that agitation provides external kinetic energy to the coal-kaolinite suspension and render particles to overcome the energy barrier more easily than that in the absence of agitation, producing a more serious kaolinite-coating phenomenon, corresponding to the further decline of flotation recovery in the presence of agitation. On the other hand, agitation has an important function in the collision and adhesion of fine particles, i.e. agitation increases the collision probability between coal and kaolinite particles [28]. However, Chen et al. [23,24] proposed that high-intensity conditioning can remove the impurities from the mineral surface and decreasing the slime coating by the hydrodynamic forces which overcome the
Y. Yu et al. / Powder Technology 313 (2017) 122–128 60 60
0 Pure kaolinite, pH = 7.34 Pure coal, pH = 7.76
Pure coal, pH =7.88
Pure kaolinite, pH = 7.25 50
-20
50
30
Zeta potential (mV)
Frequency (%)
Frequency (%)
40 40
30
20
20
0
0 0
-40
-60
-80
10
10
-140 -120 -100 -80 -60 -40 -20 Zeta potential (mV)
127
-100 -140 -120 -100 -80 -60 -40 -20 Zeta potential (mV)
0
0
500 1000 1500 2000 Agitation speed (rpm)
Fig. 6. Zeta potential distributions of pure minerals in distilled water with 10−3 M KCl.
45 40
(a)
Mixed, no agitation pH = 7.48
Frequency (%)
35 30 25 20 15
3.6. FBRM measurement results
10 5 0 -140 40
Frequency (%)
35
-120
-100
-80
-60
-40
-20
0
Zeta potential (mV)
(b)
Mixed, 800 rpm pH = 7.69
30 25 20 15 10 5 0 -140
Relative Frequency (%)
50
adhesion resistance force and remove particles from surfaces by rolling and sliding, therefore, as expected that the combustible matter recovery increases after 1600 rpm as shown in Fig. 2. To sum up, it can be inferred that mild agitation aggravates kaolinite-coating by overcoming energy barrier and increasing collision between coal and kaolinite particles. Conversely, the high intensity agitation mitigates the kaolinite-coating.
-120
-100
-80
-60
-40
-20
0
Zeta potential (mV)
(c)
Mixed, 1200 rpm pH = 7.76
40 30
According to the previous studies, it is known that when flocculation occurs, the chord length value increases [29]. Conversely, when the suspension becomes more dispersed, the chord length decreases. The chord length value is a sensitive indicator of the flocculation or dispersion state of suspensions. It is shown in Fig. 9 (a) that with the increase of agitation speed, the D50 of pure coal samples firstly increased and then dropped down but ended up with a higher value than that of the beginning. Evidently, hydrophobic coagulation between coal particles occurred, lower agitation from 100 to 300 rpm enhanced coal-coal coagulation. However, excessive agitation strength destroyed the aggregates and decreased the value of D50 as shown in Fig. 9 (a). For the kaolinite suspension of the bottom curve in Fig. 9 (a), D50 of kaolinite samples slightly increased from 100 to 200 rpm and then decreased as the increasing agitation speed, which indicating that lower agitation from 100 to 200 rpm enhanced the kaolinite-kaolinite coagulation and a higher agitation increased the fracture of kaolinite-kaolinite aggregates, this is similar with the coal-coal coagulation. The observation is consistent with the results shown in Fig. 5. The chord length distributions of
20 10 0 -140
-120
-100
-80
-60
-40
-20
0
Bin Center
Frequency (%)
50
(d)
Mixed, 2000 rpm pH = 7.82
40 30 20 10 0 -140
-120
-100
-80
-60
-40
-20
0
Zeta potential (mV) Fig. 7. Zeta potential distributions of mixed coal-kaolinite in distilled water with 10−3 M KCl under different agitation speeds.
Fig. 8. Colloidal (DLVO) forces between a coal particle and a clay particle versus intersurface separation distance [3].
128
Y. Yu et al. / Powder Technology 313 (2017) 122–128
12 Coal
Chord length (µm)
10 9 Coal+Kaolinite
0 rpm 100 rpm 200 rpm 300 rpm 400 rpm 500 rpm 600 rpm
8
Coal+Kaolinite Coal Kaolinite
(b)
4 Distributions (%)
(a) 11
3
2
7
1
Kaolinite
6
0
5 0
500
1000
1500
2000
2500
Time (s)
1
10
100
1000
Chord Length (µm)
Fig. 9. The results of FBRM measurement (a) D50 of pure and mixed mineral samples at different impeller speeds in distilled water; (b) Distributions of pure and mixed samples at 100 rpm in distilled water.
coal, kaolinite, and mixed coal-kaolinite at 100 rpm are presented in Fig. 9 (b) respectively. It is interesting to find that the curve of mixed coal-kaolinite suspension matches well with that of the pure coal samples. This demonstrated that kaolinite definitely covered on the coal surface and made its characteristic peak disappear. Meanwhile, one point should be noted that the affinity between fine kaolinite and coal particles is more pronounced since the left parts of the two curves match extremely well. In addition, this conclusion was also proved by the settling test in Section 3.2, i.e. hard to settle fine particles of coal and kaolinite remaining in the supernatants at the middle and right cylinders in Fig. 3 aggregated each other and settled successfully, producing a clear supernatant. 4. Conclusions Several conclusions from this work are summarized as follows. Kaolinite clays partially cover on the coal surface by two morphologies includes individuals and aggregates. The affinity between fine kaolinite and fine coal particles is more pronounced. The agitation has little effect on the surface electrical property of coal and kaolinite. However, the mild agitation could enhance kaolinite-coating by overcoming energy barrier and increasing collision between coal and kaolinite particles, and the high intensity agitation could mitigate the kaolinite-coating. This finding is expected to be useful to the process involving heterocoagulation such as tailings disposal, water treatment and flotation separation. Acknowledgements Financial support to the project was provided by National Natural Science Foundation of China (No. 51604280). And we thank Natural Sciences and Engineering Research Council of Canada (NSERC) for the instruments used in this work. Yuexian Yu also appreciates a scholarship from the China Scholarship Council (CSC) to carry out a visiting study at the University of Alberta. References [1] D. Liu, Y. Peng, Reducing the entrainment of clay minerals in flotation using tap and saline water, Powder Technol. 253 (2014) 216–222. [2] L. Wang, et al., A review of entrainment: mechanisms, contributing factors and modelling in flotation, Miner. Eng. 70 (2015) 77–91. [3] W.J. Oats, O. Ozdemir, A.V. Nguyen, Effect of mechanical and chemical clay removals by hydrocyclone and dispersants on coal flotation, Miner. Eng. 23 (5) (2010) 413–419. [4] B.J. Arnold, F.F. Aplan, The effect of clay slimes on coal flotation, part I: the nature of the clay, Int. J. Miner. Process. 17 (3–4) (1986) 225–242.
[5] E. Forbes, K.J. Davey, L. Smith, Decoupling rehology and slime coatings effect on the natural flotability of chalcopyrite in a clay-rich flotation pulp, Miner. Eng. 56 (2014) 136–144. [6] B. Wang, Y. Peng, The interaction of clay minerals and saline water in coarse coal flotation, Fuel 134 (2014) 326–332. [7] M. Polat, H. Polat, S. Chander, Physical and chemical interactions in coal flotation, Int. J. Miner. Process. 72 (1) (2003) 199–213. [8] Y. Yu, et al., Mechanism of entrainment and slime coating on coal flotation, J. China Coal Soc. 40 (3) (2015) 652–658. [9] D. Liu, Y. Peng, Understanding different roles of lignosulfonate in dispersing clay minerals in coal flotation using deionised water and saline water, Fuel 142 (2015) 235–242. [10] B. Wang, Y. Peng, S. Vink, Diagnosis of the surface chemistry effects on fine coal flotation using saline water, Energy Fuel 27 (8) (2013) 4869–4874. [11] Z. Xu, et al., Electrokinetic study of clay interactions with coal in flotation, Int. J. Miner. Process. 68 (1) (2003) 183–196. [12] B.J. Arnold, F.F. Aplan, The effect of clay slimes on coal flotation, part II: the role of water quality, Int. J. Miner. Process. 17 (3) (1986) 243–260. [13] R.Q. Honaker, R.H. Yoon, G.H. Luttrell, Ultrafine coal cleaning using selective hydrophobic coagulation, Coal Prep. 25 (1) (2005) 81–97. [14] Z. Zhijun, et al., Effects of clay and calcium ions on coal flotation, Int. J. Min. Sci. Technol. 23 (5) (2013) 689–692. [15] Y. Xing, X. Gui, Y. Cao, Effect of calcium ion on coal flotation in the presence of kaolinite clay, Energy Fuel 30 (2) (2016) 1517–1523. [16] M. Deng, Q. Liu, Z. Xu, Impact of gypsum supersaturated solution on surface properties of silica and sphalerite minerals, Miner. Eng. 46 (2013) 6–15. [17] J. Moore, E. Cerasoli, Particle light scattering methods and applications, Encyclopedia of Spectroscopy and Spectrometry 2010, pp. 2077–2088. [18] D. Greaves, et al., Measuring the particle size of a known distribution using the focused beam reflectance measurement technique, Chem. Eng. Sci. 63 (22) (2008) 5410–5419. [19] H. Xu, Y. Lu, A study of bubble-particle interactions in a column flotation process, Physicochem. Probl. Miner. Process 53 (1) (2017) 17–33. [20] S. Farrokhpay, D.J. Bradshaw, Effect of Clay Minerals on Froth Stability in Mineral Flotation: A Review, Technowrites, 2012. [21] G. Huang, et al., Flow field simulation of agitating tank and fine coal conditioning, Int. J. Miner. Process. 148 (2016) 116–123. [22] L. MA, et al., Effects of shearing strength in slurry conditioning on coal slime flotation, J. China Coal Soc. 38 (1) (2013) 140–144. [23] G. Chen, et al., The effect of high intensity conditioning on the flotation of a nickel ore, part 2: mechanisms, Miner. Eng. 12 (11) (1999) 1359–1373. [24] G. Chen, et al., The effect of high intensity conditioning on the flotation of a nickel ore. Part 1: size-by-size analysis, Miner. Eng. 12 (10) (1999) 1185–1200. [25] V. Gupta, et al., Particle interactions in kaolinite suspensions and corresponding aggregate structures, J. Colloid Interface Sci. 359 (1) (2011) 95–103. [26] V. Gupta, et al., Crystal lattice imaging of the silica and alumina faces of kaolinite using atomic force microscopy, J. Colloid Interface Sci. 352 (1) (2010) 75–80. [27] T. Chen, Y. Zhao, S. Song, Electrophoretic mobility study for heterocoagulation of montmorillonite with fluorite in aqueous solutions, Powder Technol. 309 (2017) 61–67. [28] H. Schubert, On the turbulence-controlled microprocesses in flotation machines, Int. J. Miner. Process. 56 (1) (1999) 257–276. [29] L. Liang, et al., A review of the modern characterization techniques for flocs in mineral processing, Miner. Eng. 84 (2015) 130–144.
Contents lists available at ScienceDirect
Powder Technology journal homepage: www.elsevier.com/locate/powtec
Effect of agitation on the interaction of coal and kaolinite in flotation Yuexian Yu a,b, Gan Cheng c,d, Liqiang Ma a,⁎, Gen Huang a, Lun Wu a, Hongxiang Xu a,⁎ a
School of Chemical and Environmental Engineering, China University of Mining and Technology (Beijing), Beijing 100083, China Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta T6G 1H9, Canada c College of Chemistry and Chemical, Henan Polytechnic University, Jiaozuo 454000, Henan, China d Synergism Innovative Center of Coal Safety Production in Henan Province, Jiaozuo 454000, China b
a r t i c l e
i n f o
Article history: Received 10 November 2016 Received in revised form 8 February 2017 Accepted 2 March 2017 Available online 03 March 2017 Keywords: Agitation Slime coating Zeta potential distribution Heterocoagulation
a b s t r a c t In this work, the interaction behavior and underlying mechanism of coal and kaolinite particles in the presence of agitation was investigated by flotation tests, homo and hetero settling tests, turbidity meter, zeta potential distribution, focused beam reflectance measurement (FBRM) and scanning electron microscope (SEM). The results show that, in the presence of kaolinite, the flotation combustible matter recovery of coal decreases from 0 rpm to 1200 rpm and then increases from 1200 rpm to 2000 rpm, indicating that the mild agitation enhances kaolinite-coating and the high intensity agitation mitigates the kaolinite-coating, which is demonstrated by the zeta potential distribution and turbidity measurements. However, the degree of the heterocoagulation between coal and kaolinite is not serious. It is found that the kaolinite particles partially cover on the coal surface by two morphologies includes individuals and aggregates through the size-based scanning electron microscope analysis. In addition, the affinity between fine kaolinite and fine coal particles is more pronounced. The kaolinite-coating enhancement caused by the mild agitation is explained by the Derjaguin–Landau–Verwey–Overbeek (DLVO) theory. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Flotation is widely used in coal industry for treating fine coal. This process greatly relies on the distinct interfacial properties between coal and gangue minerals. Good hydrophobicity of the coal surface is the key to ensure the success of the flotation. Unfortunately, nowadays China coal industry is facing the challenge of processing clayey coal as a result of the depletion of high quality coal deposits. According to the literature, clay particles, usually b2 μm, have a significant effect on coal flotation [1–3]. For example, fine hydrophilic clay particles are easily dragged by the interstitial liquid film between air-bubbles and enter the froth layer with liquids in flotation, resulting in mechanical entrainment and low concentrate quality [1,2]. Clay minerals can also increase reagent consumption and pulp viscosity, which lowers the flotation efficiency [4,5]. One of the most important effects caused by clay minerals is slime coating, defined as a layer of fine or ultrafine colloidal particles coated on the larger value mineral surface, which has been recognized to have a deleterious effect on flotation. It is speculated that these slimes on mineral surface form a hydrophilic “armor” preventing mineral particles from contacting with collectors and/or air bubbles, consequently resulting in a lower flotation recovery [6–8]. Numerous researchers have demonstrated that hydrophilic clay ⁎ Corresponding authors. E-mail addresses: [email protected] (L. Ma), [email protected] (H. Xu).
http://dx.doi.org/10.1016/j.powtec.2017.03.002 0032-5910/© 2017 Elsevier B.V. All rights reserved.
particles may coat the coal surface, making it hydrophilic and preventing the adsorption of collectors and therefore depress coal flotation [4,6,9–12]. Originally, Derjaguin–Landau–Verwey–Overbeek (DLVO) theory was developed to interpret the aggregate stability of colloidal particles, and recently researchers used this theory to explain the interaction behaviors of mineral particles in suspension. The DLVO interaction energy and force between coal and clay particles was calculated by some researchers [3,13]. The results indicate that clay coating is governed by the van der Waals attraction and that the double-layer interaction played a secondary role [3]. It is interesting to find that there is an energy barrier in the DLVO interaction energy curve as shown in Fig. 8, which means the energy barrier should be overcome when clay coating occurs. Therefore, the kinetic energy input (e.g. agitation) may enhance slime coating by overcoming the energy barrier, aggravating the flotation performance. In addition, agitation will improve the particleparticle collision, providing a higher attachment probability of coal and clay particles. However, so far, many works were conducted to investigate the effect of slime coating on coal flotation, which mainly focused on clay types, slurry pH and water hardness [4,11,12,14,15]. The aim of this work is to investigate the effect of agitation on the interaction of coal and kaolinite in flotation. The interaction behavior and underlying mechanism of agitation on slime coating was investigated by flotation tests, homo and hetero settling tests, turbidity meter, zeta potential distribution, focused beam reflectance measurement (FBRM)
Y. Yu et al. / Powder Technology 313 (2017) 122–128
and scanning electron microscope (SEM). The results of this research are expected to provide a more detailed description of the flotation behavior of coal-kaolinite system in the presence of kinetic energy input and give a reference for the coal flotation. 2. Experimental 2.1. Materials The pure coal used in this study was obtained by hand-picking the coarse gravity separation clean coking-coal supplied from Huainan coal preparation plant, China. In order to obtain a purer coal, the original coal samples were separated by the heavy liquid of 1.30 g/cm3 and a light fraction of the coal with the ash content of around 5% was produced. Then obtained coal were crushed, ground and used as the pure coal for the subsequent tests. Some impurities might be re-liberated during crushing and grinding, but their amount should be very low since the ash content of the light fraction is quite low. In addition, most of impurities in coal belong to clay minerals which having a similar property with the kaolinite, therefore, the re-liberated impurities would cause little effect on the tests. Kaolinite, chemical grade (purity N99%), purchased from Sinopharm Chemical Reagent Co., Ltd. was tested as the pure clay mineral in this study. The size distributions of the kaolinite and the pure coal were measured by a laser diffraction particle size analyzer (Mastersizer 2000, UK). As shown in Fig. 1 that D90 of the kaolinite is 10.832 μm, which approximately matches the magnitude of clay minerals in the coal flotation [1]. D90 of the coal is 258.862 μm, which also matches the size of flotation feed in coal industry. 2.2. Methods 2.2.1. Batch flotation Batch flotation tests were conducted using a 1.5 L Denver flotation cell at neutral pulp pH. For the scenario of extra agitation flotation, 40 g of pure coal was first added into a baffle built-in tank and mixed with 1 L tap water by a glass bar until the coal was wetted fully. Then the desired amount of kaolinite was added into the tank and stirred at five different agitation speeds of 400, 800, 1200, 1600, 2000 rpm for 10 min. After that, the suspension was immediately transferred into the flotation cell and additional 200 ml water was added at the same time. 4 μl MIBC was added for each test. No collectors were used during the whole course of flotation since the collector may improve the surface hydrophobicity of coal particles even when kaolinite-coating occurs and cover the changes caused by kaolinite-coating. The impeller speed for flotation was kept constant at 900 rpm and air flow was
Cumulative size distribution (%)
100 Coal Kaolinite 80
60
40
20
0 0.1
1
10
100
1000
Particle size ( µm) Fig. 1. The cumulative size distributions of the kaolinite and coal sample.
123
kept at 5.23 L/min by an outside independent gas supply system. Froth products were consecutively collected at intervals of 15, 15, 30, 60, 60 s. During the collection, a metronome was placed to obtain a constant scraping speed of 30 times per min. Flotation concentrates and tailings were filtered, dried at 105 °C and weighed for ash analysis and combustible matter recovery calculation. The flotation test was evaluated by the combustible matter recovery, which was calculated by the following formula. ε¼
γc ð100−Ac Þ 100% 100 100−A f
ð1Þ
where ε represents the combustible matter recovery, %; γc represents the yield of the concentrate, %; Ac represents the ash content of the concentrate; Af represents the ash content of the feed. All the flotation tests were repeated for three times and the test results were averaged. 2.2.2. Settling test Settling tests were conducted in 100 ml graduated glass cylinders. Given amounts of solids (5 g for coal and 1.25 g for kaolinite) were accurately weighed and placed in the cylinders separately or mixed together. The cylinders were then filled with 100 ml distilled water and shaken briefly to mix up the slurry. The cylinders were repeatedly inverted for 20 times and left still for 14 h, and then the test photos were taken. This technique was proposed by Xu et al. [11]. 2.2.3. SEM analysis The SEM analysis was conducted by ZEISS MERLIN VP Compact field emission scanning electron microscope. Firstly, the pure coal was wet sieved strictly and a size fraction between 74 μm and 125 μm was obtained. Then 2 g of the screened coal and 0.2 g of kaolinite were mixed in distilled water for 10 min by a magnetic stirrer. After that the suspension was passed through a sifter with pore size of 74 μm and rinsed mildly with distilled water so that the remaining kaolinite particles in the suspension could be separated from the coal-kaolinite aggregation since D90 of the kaolinite was 10.832 μm and the maximum size was 30.200 μm known from the laser size analysis in Section 2.1. The oversized coal-kaolinite aggregates were naturally dried in the air in order to avoid any possible change to the surface structure and then detected by SEM. 2.2.4. Turbidity measurement Mixed solids (5 g for coal and 1.25 g for kaolinite) were accurately weighed and placed into a beaker and agitated by a magnetic stirrer to make sure the solids suspend thoroughly in the water. Ten 10 ml of the stock mineral suspensions was taken and diluted to 200 ml and agitated for 10 min at five different agitation speeds 400, 800, 1200, 1600, 2000 rpm and followed by settling for 1 h, then the turbidity values of supernatants were determined immediately by the turbidity meter, VWR Model 66120-200. 2.2.5. Zeta potential measurement Zeta potential distribution was measured by CAD Zetaphoremetre IV™ (CAD, France). The Smoluchowski model was used to calculate zeta potentials from the measured electrophoretic mobility. Suspensions containing 0.5 g of coal or kaolinite were prepared separately in two jars with 10−3 M KCl solution and conditioned by magnetic stirrer for 20 min and then left to stand for 24 h. For the zeta potential distribution measurement of pure mineral, the suspension was stirred for another 5 min and settled for 5 min, and then the upper portion of dilute fine particle suspension was taken for the measurement. In the case of mixed coal-kaolinite, the two suspensions in the jars were totally transferred into the agitation tank which was used in the flotation stage and diluted into 1 L with 10− 3 M KCl solution and then agitated at each value of desired agitation speeds for 10 min. After settling for 5 min, the upper portion of the suspension was taken for zeta potential
124
Y. Yu et al. / Powder Technology 313 (2017) 122–128
distribution measurement. The pH of the suspension was monitored by an outside pH meter during the measurement. The electrophoretic mobility of about 20–200 particles in the stationary layer were traced and converted into zeta potential distribution. The details of principle were described in the literature [11,16]. In addition, zeta potentials of individual coal and kaolinite under different agitation speeds were measured by a Brookhaven Instruments ZetaPlus Zeta Potential Analyzer. Each reported zeta potential value was the mean of 10 runs, with 10 measurement cycles per run. The pH of suspension for each measurement was recorded by a pH meter.
bar to make sure the sample was wetted fully before the measurement. Then the suspension was placed on a stage and determined by FBRM. 3. Results and discussion 3.1. Batch flotation results Firstly, the effect of agitation on pure coal flotation was studied, and the combustible matter recovery of the pure coal as a function of flotation time at different agitation speeds are shown Fig. 2 (a). As can be seen that the recovery-flotation time curves basically overlapped for the agitation speeds from 0 to 1200 rpm, which indicates that the agitation (when rpm b1200) has very little effect on the pure coal flotation. However, when the agitation speed is over 1200 rpm, the two curves of 1600 and 2000 rpm slightly diverge from the curve of no agitation, showing that the flotation rates of 1600 and 2000 rpm decreased. From Fig. 2 (d), the ultimate combustible matter recovery of the pure coal from 0 to 1200 rpm basically did not change. However, the recovery decreased slightly when the agitation speed continued to increase. This may be probably attributed to the fracture of coal-coal aggregates due to the high intensity agitation as proved in Fig. 9, increasing the amount of fine coal particles which usually having a higher detour probability when collide with bubbles and hence resulting in a lower probability of bubble-particle attachment [19]. However, all in all, the agitation has little effect on the combustible matter recovery of the pure coal flotation in this test. The effect of agitation on the interaction between coal and kaolinite in flotation is shown in Fig. 2 (b) and (c), that the flotation recovery of mixed coal-kaolinite in each moment, no matter in the presence or absence of agitation, is always lower than that of the pure coal, indicating that the kaolinite depressed the flotation of the coal. The ultimate combustible matter recovery of pure coal and mixed coal-kaolinite are shown in Fig. 2 (d), respectively. The combustible matter recovery of 10% and 20% kaolinite content in the absence of agitation are both less than that of the pure coal, same results were obtained by other researchers [11,14]. In addition, it is interesting to find that the flotation
2.2.6. FBRM measurement The measurement of FBRM was carried out by METTLER TOLEDO FBRM® Technology G400. The technique allows real-time, in-process measurement of highly turbid samples. The technique uses an inline probe containing a laser that is focused to a point at the surface of the probe window, which is in contact with the process sample. The focused beam is rotated in a circular path at a controlled scan speed, and pulses of backscattered light are detected by the probe as the focused beam passes over particles and agglomerates. The length of these pulses correlates to the distance across each particle and, by multiplying the length of each pulse by the scan speed, the distance across each particle (chord length) can be calculated. High numbers of chord lengths are measured each second, which enable changes in the particle size distribution and particle count to be monitored [17]. The FBRM probe measures chord lengths, grouped in a chord length distribution, rather than the actual sizes (i.e. diameters) [18]. Chord is supposed to be a fraction of the particle size. Hence, there should be no chord above the maximum particle diameter in the system. The results can be used for qualitative analysis. When flocculation occurs, the chord length value increases. Conversely, when the suspension becomes more dispersed, the chord length decreases. The chord length value is a sensitive indicator of the flocculation or dispersion state of suspensions [1]. For FBRM measurement in this study, 0.5 g of coal and 0.15 g of kaolinite were accurately weighed and added into a beaker with 100 ml distilled water separately or mixed together and first agitated manually with a glass
100
(a)
Combustible matter recovery (%)
Combustible matter recovery (%)
100
80
60
no agitation 400 rpm 800 rpm 1200 rpm 1600 rpm 2000 rpm
40
20
0
60 Pure coal, no agitation 10% Kaolinite, no agitation 10% Kaolinite, 400 rpm 10% Kaolinite, 800 rpm 10% Kaolinite, 1200 rpm 10% Kaolinite,1600 rpm 10% Kaolinite, 2000 rpm
40
20
0 0
50
100
150
200
0
Flotation time (s)
100
(c)
80
60 Pure coal, no agitation 20% Kaolinite, no agitation 20% Kaolinite, 400 rpm 20% Kaolinite, 800 rpm 20% Kaolinite, 1200 rpm 20% Kaolinite,1600 rpm 20% Kaolinite, 2000 rpm
40
20
50
100
150
200
Flotation time (s)
100 Combustible matter recovery (%)
Combustible matter recovery (%)
(b)
80
(d)
95 90 85 80 75 70
Pure coal, no agitation 10% Kaolinite 20% Kaolinite
65 60
0 0
50
100 Flotation time (s)
150
200
0
500
1000
1500
2000
Agitation speed (rpm)
Fig. 2. The combustible matter recovery of pure coal and mixed coal-kaolinite in the absence and presence of agitation. (a) Pure coal; (b) 10% kaolinite added; (c) 20% kaolinite added; (d) the ultimate combustible matter recovery versus the agitation speed.
Y. Yu et al. / Powder Technology 313 (2017) 122–128
recovery presents a further decline in the presence of agitation. In the case of 10% kaolinite addition, the minimum flotation recovery is between 800 and 1200 rpm and the flotation recovery begins to rise again at 1200 rpm. The flotation recovery at the agitation speed of 1600 rpm is higher than the original flotation recovery, and a further increase of agitation speed does not improve recovery a lot, which essentially means the maximum value of flotation recovery obtains. When it comes to the case of 20% kaolinite addition, the recovery curve shows a similar trend compared with that of 10% kaolinite addition. However, there are some differences between them. For example, the minimum combustible matter recovery of 20% kaolinite addition occurs at 1200 rpm, which shows a delay compared with that of the 10% kaolinite addition. Meanwhile, the recovery increases at 1600 rpm and continue to rise obviously at 2000 rpm. Confined by the speed limit of the agitation apparatus, it is speculated that the recovery may continue to increase after 2000 rpm. This means a stronger agitation is needed for a higher kaolinite content to remove the detrimental effect caused by kaolinite. No collectors were added or used during the whole process of flotation, as such, the surface hydrophobicity played a critical role in the floating of the coal in this test. Thus, once the kaolinite-coating occurs, the hydrophobicity of the coal surface should be decreased. Therefore, the lower combustible matter recovery observed in the presence of kaolinite as shown in Fig. 2 (d) should be attributed to the kaolinitecoating. However, clay minerals can affect flotation by affecting froth stability that also lowers flotation recovery of value minerals [20]. Therefore, the depression of the coal flotation in this test should be ascribed to a synergistic effect caused by kaolinite-coating and froth stability changing. In this test, the effect of kaolinite on froth stability should be constant since the kaolinite content for each group is fixed. Therefore, it is indicated that the kaolinite-coating should be responsible for the fluctuation of combustible matter recovery, which indicating the agitation has an important influence on the kaolinite-coating. From the results of this test as shown in Fig. 2 (d), it is expected that the high intensity agitation mitigates slime coating as suggested by many studies [21–24]. However, it is interesting to find that the mild agitation aggravates the kaolinite-coating and its underlying mechanism was investigated in the subsequent tests. 3.2. Settling tests To confirm the presence of slime coating of kaolinite clays on coal surface in the flotation process, the homo and hetero settling tests were conducted at neutral pulp pH same as the flotation solution. The photographs of pure mineral and mixed minerals suspensions after 14 h settlement are presented in Fig. 3. The pictures gave a directly visual observation for the understanding of kaolinite-coating. In the left cylinder that mixed suspension of coal-kaolinite exhibited a clear supernatant. In contrast, the supernatants of pure mineral coal or kaolinite were still turbid after the same period of settling as shown in the middle and right cylinders. By the comparison of those two cases, it is revealed that heterocoagulation between coal particles and kaolinite clays definitely occurred; and this demonstrated that the slime coating of hydrophilic kaolinite is responsible for the reduction of flotation recovery in Section 3.1. 3.3. SEM analysis The directly visual images in Section 3.2 gave a macroscopic observation for the kaolinite-coating on coal surface. Meanwhile, in order to provide a microscopic description of slime coating, SEM analysis was performed as shown in Fig. 4. In the previous works, most of researchers always investigated the phenomenon of slime coating by observing SEM images of the flotation concentrates or tailings directly but failed to realize the important defect during this procedure, i.e. during the filtration or dewatering of flotation products, clay minerals could naturally settle
125
Fig. 3. Photographs of pure mineral and mixed mineral suspension after 14 h settling, coal-kaolinite (left) is clarified; kaolinite (middle) is turbid; coal (right) is turbid.
onto the surface of coal particles. Therefore, the phenomenon of slime coating observed by SEM may not be the true slime coating. In the present work, the test procedure was well-designed and avoided this deficiency successfully. The details were described in Section 2.2.3. As shown in Fig. 4 (a), kaolinite particles were scattered on the coal surface. These hydrophilic kaolinite particles covered on the coal surface and made coal surface less hydrophobic, corresponding to the results of flotation and settling tests. Fig. 4 (b) exhibited a magnified image of partial coal surface which covered by many kaolinite particles and some of these small kaolinite particles aggregated together by edge–edge (E–E), edge– face (E–F), and face–face (F–\\F) structures as expected [25,26]. Meanwhile, some of kaolinite particles covered on the coal surface individually. Therefore, kaolinite may cover on the coal surface by two morphologies includes individuals and aggregates. 3.4. Turbidity measurement In order to investigate the effect of agitation strength on particle flocs of the kaolinite-coal suspension, supernatant turbidity of coal-kaolinite suspension after settling for 1 h at different agitation speeds was tested to monitor the situation of flocs in the suspension. As shown in Fig. 5, the supernatant turbidity after settling for 1 h in the absence of agitation was around 180 NTU and decreased to 104 NTU in the case of 400 rpm, then increased gradually with the rising agitation speed and was higher than the original turbidity at the end. This can be explained as follows: firstly, there are three types of aggregates in the suspension, i.e. coal-coal, coal-kaolinite and kaolinite-kaolinite. At the first descending point of the curve, the quantity of three types of aggregates including the coal-kaolinite raised corresponding to a lower combustible matter recovery at the agitation speed of 400 rpm as shown in Fig. 2 (d). However, when the agitation speed increased to 800 rpm, the turbidity of supernatant exhibited a light increase. This may be attributed to the competitive outcome between the crack of homo coagulations (coal-coal and kaolinite-kaolinite) and the formation of coal-kaolinite aggregate. When the agitation speed increased from 800 rpm to 1200 rpm, an obvious increase in supernatant turbidity appears, which represents the further crack of coal-coal coagulation and the initial crack of coal-kaolinite aggregate, therefore, the combustible matter recovery began to increase. After 1200 rpm, the supernatant turbidity continued to show an increase mainly due to the fractures of
126
Y. Yu et al. / Powder Technology 313 (2017) 122–128
Fig. 4. SEM images of coal–kaolinite aggregates.
coal-kaolinite aggregates corresponding to the further increase in combustible matter recovery as shown in Fig. 2 (d). To summarize, it is indicated that low intensity agitation enhances the particle aggregation; however, high intensity agitation breaks up the aggregates. 3.5. Zeta potential distribution As shown in Fig. 6 (c), the zeta potentials of coal and kaolinite essentially both remained unchanged regardless of the agitation speed. Therefore, the agitation does not change the surface electrical properties of coal and kaolinite. In order to recognize the effect of agitation on the interaction of coal and kaolinite, the zeta potential distribution of mixed coal-kaolinite at different agitation speeds in distilled water with 10−3 M KCl were determined. As proposed by literature that if strong heterocoagulation occurs, the zeta potential distribution of a binary mixture will show single distribution peak. On the contrary, a bimodal zeta potential distribution which is consistent with the peaks of the two individual minerals usually means there is no attraction between the two components [11,27]. Therefore, it is assumed that if the agitation has an effect on the interaction of coal and kaolinite, a change or shift of the zeta potential distribution of the binary system will be expected. As presented in Fig. 6 (a) and (b), the individual zeta potential distributions of coal and kaolinite are close to each other with some parts overlapping from −60 mV to −80 mV. However, most of kaolinite particles sit on the right side of the dash line at −75 mV and on the contrary, most of coal particles sit on the left side. The zeta potential distributions of mixed coal-kaolinite systems are shown in Fig. 7. A single distribution peak appears when there is no 240 Turbidity
Turbidity (NTU)
200 160 120 80 40 0 0
500
1000
1500
2000
Agitation speed (rpm) Fig. 5. Supernatant turbidity of coal-kaolinite blending suspension after settling for 1 h at different agitation speeds in distilled water.
agitation as shown in Fig. 7 (a), however, this does not definitely mean that strong attraction occurs between coal and kaolinite since the zeta potential distributions of the two individual components are too close as shown in Fig. 6. The zeta potential distributions of the mixed system under different agitation speeds look similar as listed in Fig. 7 (b), (c), and (d). However, when drawing a line at −75 mV, it is found that the frequency distributions on the left side of the line firstly decreased and then increased as the increasing agitation speed. As discussed above, most of kaolinite particles sit on the right side of the line and meanwhile, most of coal particles sit on the left side. Therefore, it can be inferred that more and more coal particles were covered by kaolinite clays as the agitation speed increasing from 0 to 1200 rpm. When the agitation speed increases to 2000 rpm, the frequency distribution on the left side of the line increases significantly, which indicates that the kaolinite clays are removed from the coal surfaces at 2000 rpm. It is well consistent with the flotation results obtained in Fig. 2. In addition, as shown from the four plots in Fig. 7, the zeta potential distributions of coal and kaolinite always coexist, suggesting that there are still some remnant coal and kaolinite particles in the mixed suspension, which implies that the aggregation between coal and kaolinite is not strong. In other words, the coal surface is partially covered by the kaolinite particles as shown in Fig. 4. This is also proved by the flotation results, i.e. the combustible matter recovery does not show a dramatic fall in the presence of kaolinite (just from 94% to 72%), indicating that the slime coating between coal and kaolinite is not serious. The interactions between coal and clay particles based on Derjaguin– Landau–Verwey–Overbeek (DLVO) theory of the aggregate stability of colloidal particles are shown in Fig. 8. As can be seen from Fig. 8, the double-layer forces for the two surface charging mechanisms are repulsive and hence the double-layer interaction under the surface charge regulation occurring between the two extreme cases is also repulsive. On the contrary, the van der Waals force is attractive at all separation distances. In particular, at short separation distance, the van der Waals attraction is so strong that the double-layer repulsion is overcome, resulting in strong net DLVO attraction between the particles at close contact. One point should be mentioned is that there is an energy barrier in the DLVO interaction force curve and the energy barrier must be overcome before the slime coating occurs. Therefore, it can be concluded that agitation provides external kinetic energy to the coal-kaolinite suspension and render particles to overcome the energy barrier more easily than that in the absence of agitation, producing a more serious kaolinite-coating phenomenon, corresponding to the further decline of flotation recovery in the presence of agitation. On the other hand, agitation has an important function in the collision and adhesion of fine particles, i.e. agitation increases the collision probability between coal and kaolinite particles [28]. However, Chen et al. [23,24] proposed that high-intensity conditioning can remove the impurities from the mineral surface and decreasing the slime coating by the hydrodynamic forces which overcome the
Y. Yu et al. / Powder Technology 313 (2017) 122–128 60 60
0 Pure kaolinite, pH = 7.34 Pure coal, pH = 7.76
Pure coal, pH =7.88
Pure kaolinite, pH = 7.25 50
-20
50
30
Zeta potential (mV)
Frequency (%)
Frequency (%)
40 40
30
20
20
0
0 0
-40
-60
-80
10
10
-140 -120 -100 -80 -60 -40 -20 Zeta potential (mV)
127
-100 -140 -120 -100 -80 -60 -40 -20 Zeta potential (mV)
0
0
500 1000 1500 2000 Agitation speed (rpm)
Fig. 6. Zeta potential distributions of pure minerals in distilled water with 10−3 M KCl.
45 40
(a)
Mixed, no agitation pH = 7.48
Frequency (%)
35 30 25 20 15
3.6. FBRM measurement results
10 5 0 -140 40
Frequency (%)
35
-120
-100
-80
-60
-40
-20
0
Zeta potential (mV)
(b)
Mixed, 800 rpm pH = 7.69
30 25 20 15 10 5 0 -140
Relative Frequency (%)
50
adhesion resistance force and remove particles from surfaces by rolling and sliding, therefore, as expected that the combustible matter recovery increases after 1600 rpm as shown in Fig. 2. To sum up, it can be inferred that mild agitation aggravates kaolinite-coating by overcoming energy barrier and increasing collision between coal and kaolinite particles. Conversely, the high intensity agitation mitigates the kaolinite-coating.
-120
-100
-80
-60
-40
-20
0
Zeta potential (mV)
(c)
Mixed, 1200 rpm pH = 7.76
40 30
According to the previous studies, it is known that when flocculation occurs, the chord length value increases [29]. Conversely, when the suspension becomes more dispersed, the chord length decreases. The chord length value is a sensitive indicator of the flocculation or dispersion state of suspensions. It is shown in Fig. 9 (a) that with the increase of agitation speed, the D50 of pure coal samples firstly increased and then dropped down but ended up with a higher value than that of the beginning. Evidently, hydrophobic coagulation between coal particles occurred, lower agitation from 100 to 300 rpm enhanced coal-coal coagulation. However, excessive agitation strength destroyed the aggregates and decreased the value of D50 as shown in Fig. 9 (a). For the kaolinite suspension of the bottom curve in Fig. 9 (a), D50 of kaolinite samples slightly increased from 100 to 200 rpm and then decreased as the increasing agitation speed, which indicating that lower agitation from 100 to 200 rpm enhanced the kaolinite-kaolinite coagulation and a higher agitation increased the fracture of kaolinite-kaolinite aggregates, this is similar with the coal-coal coagulation. The observation is consistent with the results shown in Fig. 5. The chord length distributions of
20 10 0 -140
-120
-100
-80
-60
-40
-20
0
Bin Center
Frequency (%)
50
(d)
Mixed, 2000 rpm pH = 7.82
40 30 20 10 0 -140
-120
-100
-80
-60
-40
-20
0
Zeta potential (mV) Fig. 7. Zeta potential distributions of mixed coal-kaolinite in distilled water with 10−3 M KCl under different agitation speeds.
Fig. 8. Colloidal (DLVO) forces between a coal particle and a clay particle versus intersurface separation distance [3].
128
Y. Yu et al. / Powder Technology 313 (2017) 122–128
12 Coal
Chord length (µm)
10 9 Coal+Kaolinite
0 rpm 100 rpm 200 rpm 300 rpm 400 rpm 500 rpm 600 rpm
8
Coal+Kaolinite Coal Kaolinite
(b)
4 Distributions (%)
(a) 11
3
2
7
1
Kaolinite
6
0
5 0
500
1000
1500
2000
2500
Time (s)
1
10
100
1000
Chord Length (µm)
Fig. 9. The results of FBRM measurement (a) D50 of pure and mixed mineral samples at different impeller speeds in distilled water; (b) Distributions of pure and mixed samples at 100 rpm in distilled water.
coal, kaolinite, and mixed coal-kaolinite at 100 rpm are presented in Fig. 9 (b) respectively. It is interesting to find that the curve of mixed coal-kaolinite suspension matches well with that of the pure coal samples. This demonstrated that kaolinite definitely covered on the coal surface and made its characteristic peak disappear. Meanwhile, one point should be noted that the affinity between fine kaolinite and coal particles is more pronounced since the left parts of the two curves match extremely well. In addition, this conclusion was also proved by the settling test in Section 3.2, i.e. hard to settle fine particles of coal and kaolinite remaining in the supernatants at the middle and right cylinders in Fig. 3 aggregated each other and settled successfully, producing a clear supernatant. 4. Conclusions Several conclusions from this work are summarized as follows. Kaolinite clays partially cover on the coal surface by two morphologies includes individuals and aggregates. The affinity between fine kaolinite and fine coal particles is more pronounced. The agitation has little effect on the surface electrical property of coal and kaolinite. However, the mild agitation could enhance kaolinite-coating by overcoming energy barrier and increasing collision between coal and kaolinite particles, and the high intensity agitation could mitigate the kaolinite-coating. This finding is expected to be useful to the process involving heterocoagulation such as tailings disposal, water treatment and flotation separation. Acknowledgements Financial support to the project was provided by National Natural Science Foundation of China (No. 51604280). And we thank Natural Sciences and Engineering Research Council of Canada (NSERC) for the instruments used in this work. Yuexian Yu also appreciates a scholarship from the China Scholarship Council (CSC) to carry out a visiting study at the University of Alberta. References [1] D. Liu, Y. Peng, Reducing the entrainment of clay minerals in flotation using tap and saline water, Powder Technol. 253 (2014) 216–222. [2] L. Wang, et al., A review of entrainment: mechanisms, contributing factors and modelling in flotation, Miner. Eng. 70 (2015) 77–91. [3] W.J. Oats, O. Ozdemir, A.V. Nguyen, Effect of mechanical and chemical clay removals by hydrocyclone and dispersants on coal flotation, Miner. Eng. 23 (5) (2010) 413–419. [4] B.J. Arnold, F.F. Aplan, The effect of clay slimes on coal flotation, part I: the nature of the clay, Int. J. Miner. Process. 17 (3–4) (1986) 225–242.
[5] E. Forbes, K.J. Davey, L. Smith, Decoupling rehology and slime coatings effect on the natural flotability of chalcopyrite in a clay-rich flotation pulp, Miner. Eng. 56 (2014) 136–144. [6] B. Wang, Y. Peng, The interaction of clay minerals and saline water in coarse coal flotation, Fuel 134 (2014) 326–332. [7] M. Polat, H. Polat, S. Chander, Physical and chemical interactions in coal flotation, Int. J. Miner. Process. 72 (1) (2003) 199–213. [8] Y. Yu, et al., Mechanism of entrainment and slime coating on coal flotation, J. China Coal Soc. 40 (3) (2015) 652–658. [9] D. Liu, Y. Peng, Understanding different roles of lignosulfonate in dispersing clay minerals in coal flotation using deionised water and saline water, Fuel 142 (2015) 235–242. [10] B. Wang, Y. Peng, S. Vink, Diagnosis of the surface chemistry effects on fine coal flotation using saline water, Energy Fuel 27 (8) (2013) 4869–4874. [11] Z. Xu, et al., Electrokinetic study of clay interactions with coal in flotation, Int. J. Miner. Process. 68 (1) (2003) 183–196. [12] B.J. Arnold, F.F. Aplan, The effect of clay slimes on coal flotation, part II: the role of water quality, Int. J. Miner. Process. 17 (3) (1986) 243–260. [13] R.Q. Honaker, R.H. Yoon, G.H. Luttrell, Ultrafine coal cleaning using selective hydrophobic coagulation, Coal Prep. 25 (1) (2005) 81–97. [14] Z. Zhijun, et al., Effects of clay and calcium ions on coal flotation, Int. J. Min. Sci. Technol. 23 (5) (2013) 689–692. [15] Y. Xing, X. Gui, Y. Cao, Effect of calcium ion on coal flotation in the presence of kaolinite clay, Energy Fuel 30 (2) (2016) 1517–1523. [16] M. Deng, Q. Liu, Z. Xu, Impact of gypsum supersaturated solution on surface properties of silica and sphalerite minerals, Miner. Eng. 46 (2013) 6–15. [17] J. Moore, E. Cerasoli, Particle light scattering methods and applications, Encyclopedia of Spectroscopy and Spectrometry 2010, pp. 2077–2088. [18] D. Greaves, et al., Measuring the particle size of a known distribution using the focused beam reflectance measurement technique, Chem. Eng. Sci. 63 (22) (2008) 5410–5419. [19] H. Xu, Y. Lu, A study of bubble-particle interactions in a column flotation process, Physicochem. Probl. Miner. Process 53 (1) (2017) 17–33. [20] S. Farrokhpay, D.J. Bradshaw, Effect of Clay Minerals on Froth Stability in Mineral Flotation: A Review, Technowrites, 2012. [21] G. Huang, et al., Flow field simulation of agitating tank and fine coal conditioning, Int. J. Miner. Process. 148 (2016) 116–123. [22] L. MA, et al., Effects of shearing strength in slurry conditioning on coal slime flotation, J. China Coal Soc. 38 (1) (2013) 140–144. [23] G. Chen, et al., The effect of high intensity conditioning on the flotation of a nickel ore, part 2: mechanisms, Miner. Eng. 12 (11) (1999) 1359–1373. [24] G. Chen, et al., The effect of high intensity conditioning on the flotation of a nickel ore. Part 1: size-by-size analysis, Miner. Eng. 12 (10) (1999) 1185–1200. [25] V. Gupta, et al., Particle interactions in kaolinite suspensions and corresponding aggregate structures, J. Colloid Interface Sci. 359 (1) (2011) 95–103. [26] V. Gupta, et al., Crystal lattice imaging of the silica and alumina faces of kaolinite using atomic force microscopy, J. Colloid Interface Sci. 352 (1) (2010) 75–80. [27] T. Chen, Y. Zhao, S. Song, Electrophoretic mobility study for heterocoagulation of montmorillonite with fluorite in aqueous solutions, Powder Technol. 309 (2017) 61–67. [28] H. Schubert, On the turbulence-controlled microprocesses in flotation machines, Int. J. Miner. Process. 56 (1) (1999) 257–276. [29] L. Liang, et al., A review of the modern characterization techniques for flocs in mineral processing, Miner. Eng. 84 (2015) 130–144.