International Journal of Mineral Processing 124 (2013) 109–116

Contents lists available at ScienceDirect

International Journal of Mineral Processing

j ourna l homepage: www.e lsev ie r .com/ locate / i jminpro

Nanobubble column flotation of fine coal particles andassociated fundamentalsA. Sobhy a,b, D. Tao a,c,⁎a Department of Mining Engineering, University of Kentucky, Lexington, KY 40506, USAb Central Metallurgical Research and Development Institute, Helwan, Cairo 11421, Egyptc College of Chemical Engineering, China University of Mining and Technology, Xuzhou, China

a r t i c l e i n f o

Nanobubble

⁎ Corresponding author at: Department of Mining EnginLexington, KY 40506, USA. Tel.: +1 859 257 2953; fax: +

E-mail address: daniel.tao@uky.edu (D. Tao).

0301-7516/$ – see front matter © 2013 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.minpro.2013.04.016

a b s t r a c t

Article history:Received 29 January 2013Received in revised form 5 April 2013Accepted 20 April 2013Available online 14 May 2013

Keywords:CavitationCoalFroth flotation

Froth flotation is a widely used, cost effective particle separation process. However, its high performanceis limited to a narrow particle size range between approximately 50 to 600 μm for coal and 10 to 100 μmfor minerals. Outside this range, the efficiency of froth flotation decreases significantly, especially fordifficult-to-float particles of weak hydrophobicity (e.g., oxidized coal).This study was aimed at enhancing recovery of an Illinois fine coal sample using a specially designed flotationcolumn featuring a hydrodynamic cavitation nanobubble generator. Nanobubbles that are mostly smaller than1 μm can be formed selectively on hydrophobic coal particles from dissolved air in coal slurry. Results indicatethat the combustible recovery of a−150 μmcoal increased by 5–50% in the presence of nanobubbles, dependingon process operating conditions. Nanobubbles also significantly improved process separation efficiency.Other major advantages of the nanobubble flotation process include lower frother dosage and air consumptionsince nanobubbles are produced from air naturally dissolved in water, thereby resulting in considerably loweroperating costs.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

A 77% of the total global coal production is used by China, USA,India, Russia and Japan (U.S. Energy Information Administration,2009). United States is one of the largest coal producing countrieswith an annual production of more than one billion short tons ofclean coal. There are two main types of coal, low rank coal with 47%of world reserve and high rank coal with 53% of world reserve. Coalranking is determined by degree of transformation of the originalplant materials to carbon. Therefore, low rank coal which can besubdivided into lignite and subbituminous is low in carbon and highin hydrogen and oxygen contents. On the other hand, high rank coalwhich can be subdivided into bituminous and anthracite is high incarbon and therefore energy value but low in hydrogen and oxygencontents. These different types of coal have different uses. For example,lignite is mainly used in power generation. Bituminous and subbitumi-nous are used in power generation, cement manufacture and otherindustrial applications. Anthracite is mainly used as smokeless fuel.A 40% of worldwide electricity is generated from coal and 70% of steelproduced today uses coal (World Coal Association, 2012).

Froth flotation is commonly used in the coal industry to clean−100mesh or −150 μm coal particles from gangue minerals. This processseparates solid particles based on their differences in physical andsurface chemistry properties. It is most efficient and cost effective

eering, University of Kentucky,1 859 323 1962.

rights reserved.

for particles within a narrow size range, nominally from 50 μm to600 μm for coal and from 10 μm to 100 μm for minerals (Feng andAldrich, 1999; King, 1982; Trahar and Warren, 1976). The lower andupper particle size limits are due to the low probability of collisionand the high probability of detachment, respectively (Ralston andDukhin, 1999; Tao, 2004; Yoon, 2000). The previous studies haveshown that flotation recovery of coal particles outside the optimumsize range and/or of poor floatability can be enhanced by use ofnanobubbles (Tao et al., 2008).

Nanobubbles can be produced using ultrasonic or hydrodynamiccavitation principle (Farmer et al., 2000; Johnson and Cooke, 1981;Zhou et al., 1997). Nanobubbles preferentially nucleate at the surfaceof hydrophobic particles (Zhou et al., 1997) because work of adhesionbetween a solid particle and water is always smaller than work ofcohesion of water. Furthermore, work of adhesion decreases withincreasing solid surface hydrophobicity measured by the contactangle. Nanobubbles can nucleate on ultrafine particles without theneed for collision, which is often the rate-determining step in frothflotation for ultrafine particles (Weber and Paddock, 1983; Yoon andLuttrell, 1989). Nanobubbles generated on a particle surface also serveas a secondary collector, improving the probability of adhesionand minimizing the need for the hydrophobizing chemical reagents(Luttrell and Yoon, 1992; Zhou et al., 1997). In addition, particles areless likely to detach from tiny bubbles due to their lower ascendingvelocity and centrifugal force associated with the detachment step,reducing the probability of detachment.

The objective of this study was to develop an innovative cavitationnanobubble flotation process based on understanding of nanobubble

110 A. Sobhy, D. Tao / International Journal of Mineral Processing 124 (2013) 109–116

froth flotation fundamentals for enhanced recovery of coal particles byimproving bubble–particle collision and attachment and minimizingdetachment. A 5 cm diameter laboratory flotation column that is char-acterized with dual bubble generators for producing both nanobubblessmaller than 1 μm and regular sized bubbles of about 500 μm wasdesigned and fabricated and investigated to understand effects ofprocess variables on separation performance.

2. Experimental

2.1. Coal sample

A total of four 55-gallon drums of coal slurry were acquired froma mine in Illinois. Upon arrival at the lab, the slurry was thoroughlymixed and then split into 5 gallon buckets and sealed for storageand later usage. The particle size distribution of the coal sample wasmeasured by wet sieve analysis using the following U.S. standardsieves: 300, 150, 75, 45 and 25 μm. The different size fractions werefiltered, dried, and weighed.

Table 1 shows the size distribution data of the sample. The majorityof coal particles (82.97%) were smaller than 150 μm and 40.70% ofcoal particles were smaller than 25 μm.

2.2. Ash analysis

Additional characterization of the coal sample was performed bytaking representative samples, pulverizing them to 150 μm, and thenanalyzing them for ash content. The ash content for each size fractionof the coal sample is shown in Table 1. It can be clearly seen thatash content of the sample decreases as the particle size increases.This results from the concentration of fine clay particles in smallersize ranges. The overall feed ash content was 38.51%.

2.3. Flotation Release Analysis

The flotation release analysis is a procedure used to obtain the bestpossible separation performance achievable by any froth flotationprocess which is analogous to the gravity-based washability analysis.The release analysis was carried out in a conventional laboratoryflotation cell and its data was used as a yardstick for performanceevaluation of the nanobubble flotation technology.

The first stage of the release analysis separates the hydrophobicmaterial away from the hydrophilic material by performing multiplecleaning steps with the original feed. The second stage has the goalof separating particles into fractions of different degrees of surfacehydrophobicity by controlling air flow rate and impeller rotationspeed under starvation reagent conditions.

2.4. Flotation column design, fabrication and testing

A column made of Plexiglas with a 5 cm diameter and 2.4 madjustable height was featured with a Venturi cavitation tube and a

Table 1Feed particle size and ash distribution.

Particle size (μm) Wt. (%) Ash content (%)

Passing Retained

300 1.58 4.62300 150 15.44 5.97150 75 19.96 9.7875 45 16.59 24.1945 25 5.72 33.0525 0 40.70 72.87Total 100.00 38.51

static mixer to generate nanobubbles and conventional sized bubbles(microbubbles), respectively, as shown in Fig. 1.

The typical lengths of collection and froth zones used in the testswere 210 cm and 30 cm, respectively. With a diameter of 5.08 cm,the length-to-diameter ratio of the column was around 41:1, whichprovided near plug-flow conditions. Wash water was added in thefroth zone at a depth of 1/3 of the froth zone height below the overflowlip.

The cavitation tube and the static mixer are compact and have nomoving parts. Frother was pumped into the feed streamwhile air wasinjected into the stream prior to the static mixer. Feed slurry enteredthe column in the upper pulp zone, 45 cm below the overflow lip.After being fed into the column, coal particles collected by risingbubbles ascend to the top. Those that settle to the bottom of the columnare pumped through the static mixer and the cavitation tube to havemore chances for recovery.

The slurry jet comes out of the neck of the Venturi cavitationtube at a speed of 6 to 10 m/s causing hydrodynamic cavitation inthe stream with nanobubbles formed preferentially on coal particlesurfaces. Nanobubbles formed on hydrophobic coal particle surfacesremain attached while those on hydrophilic particle are detached,which is a selective process that enhances flotation separation efficiency.Hydrophobic particles have a higher collision probability with nano-bubbles, a higher attachment probability, and a lower detachmentprobability, resulting in a greater flotation rate constant and flotationrecovery. The slurry jet enters the column tangentially at the bottom.The total recycling flow rate through the static mixer is 11 L/min, whichsplits at a two-way connector into the cavitation tube and a pipe. As aresult, the flow rate distribution (flow rate in cavitation tube/total flowrate in static mixer) can be adjusted to be 6.6 L/11.0 L = 60% (withnanobubble) or 0 L/11.0 L = 0% (without nanobubble). A microproces-sor (Series 2600 Love Controls) receives signals from a pressure trans-ducer located at the bottom of the column. The signal adjusts a Miniflexpinch valve that controls the underflow flow rate and the desired frothlevel.

Prior to each test, the feed slurrywas conditioned for 5 minwith fueloil used as collector to enhance the hydrophobicity of coal particlesurfaces. Conditioning was conducted in a sump that was equippedwith a mixer and four baffles placed vertically and separated by anequal distance along the circumference of the sump. The slurry wasfed from a feed tank, which utilized a recirculating line to ensure sus-pension of all solids, to the flotation column by a peristaltic pump at apre-determined rate. Unless otherwise specified, all column flotationtests were performed under the following conditions: froth depth of30 cm, superficial gas flow rate of 0.5 cm/s, fuel oil collector dosageof 0.45 kg/ton, MIBC frother concentration of 30 ppm, superficialwash water flow rate of 0.12 cm/s, superficial feed slurry flow rate of0.5 cm/s, feed slurry solids concentration of 7%. Local tap water with6.0 × 10−5 M electrolyte concentration was used to make flotationslurry. A period of time equivalent to three particle retention times wasallowed to achieve steady-state conditions. After reaching the steady-state, samples of feed, product, and tailing streamswere collected simul-taneously. These samples were filtered, dried, weighed, and analyzedfor ash contents.

Major process parameters were examined individually to investi-gate their effects on flotation recovery and concentrate ash contentwith and without the cavitation tube. They included frother concentra-tion, superficial air velocity, superficial feed velocity, etc. Flotationperformancewas evaluated in terms of combustible recovery, ash rejec-tion, and separation efficiency.

3. Results and discussion

Hydrodynamic cavitation integrated into the specially designedcolumn shown in Fig. 1. is a process of creation of nanobubbles in aliquid as a result of the rupture of a liquid–liquid interface (work of

Fig. 1. Specially designed flotation column.

Fig. 2. Size distribution curves for bubbles in water solution with 10 ppm MIBC andF507 (Fan et al., 2010a; Tao et al., 2010, 2008).

111A. Sobhy, D. Tao / International Journal of Mineral Processing 124 (2013) 109–116

cohesion of waterWc) or at a liquid–solid interface due to the ruptureof a liquid–solid interface (work of adhesion Wa between water andsolid). It takes place when the liquid pressure P is reduced to below acritical value with an abrupt increase in flow velocity U (Young,1989), which is well described in Bernoulli's Eq. (1):

P þ 12ρU2 ¼ const: ð1Þ

where ρ is water density.

Wc and Wa can be expressed in Eqs (2) and (3), respectively:

Wc ¼ 2γl ð2Þ

Wa ¼ γl 1þ cosθð Þ ð3Þ

where γl is the liquid surface tension. Eqs. (2) and (3) indicate that thework of adhesion Wa is always smaller than the work of cohesion ofwater Wc, suggesting that cavitation will occur preferentially at thesolid/water interface. Since more hydrophobic particles have a greatercontact angle θ, they will have a smaller value of Wa, indicating thathydrophobic particle surfaces are themore favorable sites for cavitationto take place. Therefore, the generation of nanobubbles by hydro-dynamic cavitation is fundamentally a selective process, which shouldhave a positive effect on flotation efficiency.

Nanobubbles normally refer to tiny bubbles smaller than 1 μm asshown in Fig. 2 (Fan et al., 2010a; Tao et al., 2008). The nanobubblesare about two orders of magnitude smaller than microbubbles asshown in Fig. 2. Frother F507 produces finer bubbles than frotherMIBC. This is because the surface tension reduction by F507 is moresignificant than by MIBC.

Recent AFM studies have confirmed that nanobubbles are stableand can exist on a hydrophobic surface for several hours withoutdiscernible changes although the conventional Laplace equationsuggests that the capillary pressure of a nanobubble is too great forthe nanobubble to be stable. This is partly because nanoscopic contact

angle is much larger than the macroscopic contact angle. Nanoscopiccontact angle of nanobubbles with a hydrophobic surface is typicallyin the range of 150–160° as shown in Fig. 3 and Fig. 4 (Borkentet al., 2010; Johnson et al., 2012). The nanoscopic contact angle ismore than twice the measured macroscopic contact angle of a waterdroplet deposited on the same surface (Borkent et al., 2010).

An extensive evaluation of effects of nanobubbles on flotationperformance of coal sample was carried out in a specially designedlaboratory-scale flotation column at different operation parameterssuch as frother concentration, superficial air velocity and superficialfeed rate velocity.

Particle recovery by bubbles in froth flotation starts with the collisionand adhesion of hydrophobic particles to the air bubbles followedby transportation of hydrophobic particle–bubble aggregates from thecollection zone to the froth zone, drainage and enrichment of the frothand finally by their overflow from the cell top, whereas hydrophilicparticles remain in the pulp and are discharged as tailings. The successof effective particle separation by froth flotation relies on the efficientcapture of hydrophobic particles by air bubbles in three steps, i.e., colli-sion, attachment anddetachment. To investigate the role of nanobubbles

Fig. 3. AFM image of surface nanobubbles existing on a hydrophobic surface insidea large water droplet deposited by a syringe on the surface. The size of the image is2000 × 2000 × 40 nm3 (Borkent et al., 2010).

Fig. 5. Effect of nanobubbles at varying frother concentrations on combustible recoveryand clean coal ash (A) and separation efficiency (B).

112 A. Sobhy, D. Tao / International Journal of Mineral Processing 124 (2013) 109–116

in flotation recovery, a series of flotation tests have been performedunder different conditions.

Fig. 5(A) shows the combustible recovery and clean coal ash asa function of frother concentration with and without nanobubbles.Nanobubbles reduced required frother concentration by about one-third. For example, without nanobubbles, maximum combustiblerecovery was about 89% achieved at 55 ppm frother concentration;but with nanobubbles 92.5% combustible recovery was obtained at35 ppm frother concentration. Fig. 5(A) shows that the product ashcontent was higher in the presence of nanobubbles than in its absence.This is primarily a result of higher combustible recovery when nano-bubbles were utilized. The separation efficiency, defined as the dif-ference between combustible recovery and ash recovery, vs. frotherconcentration curve shown in Fig. 5(B) indicates that use of nano-bubbles significantly improved the flotation separation efficiency ofcoal particles.

The probability of collision (Pc) between a particle and a bubble isdefined as the fraction of particles in the path of the rising bubble that ac-tually colloids with it. It can be calculated from stream functions for qui-escent conditions (Weber and Paddock, 1983; Yoon and Luttrell, 1989)and micro-turbulence models for well mixed conditions (Schubert andBischofberger, 1979; Yoon, 2000). One of the mathematical models forPc is shown in this Eq. (4):

Pc ¼32þ 4Re0:72

15

" #Dp

Db

� �2

ð4Þ

where Db is the bubble size, Dp is the particle size and Re is the Reynoldsnumber. Fig. 6 (A) and (B) show the simulation results from Eq. (4).The results indicate that Pc increases with increasing particle size anddecreasing bubble size. Fine particles have a low probability of collision

Fig. 4. AFM image of a nanobubble that exists on a hydrophobic

with bubbles and are thus difficult to catch by bubbles, particularly bylarge size bubbles. This is the main reason for low flotation rate of fineparticles. For a nanobubble generated in the liquid, its collision proba-bility with particles is very high because of its tiny size, which is inpart responsible for the higher recovery at a given frother concentrationshown in Fig. 5(A). The improved separation efficiency by nanobubblesobserved in Fig. 5(B) is a result of selectivity of nanobubble generationdiscussed earlier.

surface with a contact angle of 165° (Johnson et al., 2012).

Fig. 6. Bubble–particle collision probability as a function of particle diameter (A) andair bubble diameter (B).

Fig. 7. Effect of nanobubbles at varying superficial air velocities (A) on the combustiblerecovery and clean coal ash (B) on separation efficiency.

113A. Sobhy, D. Tao / International Journal of Mineral Processing 124 (2013) 109–116

Higher frother concentration decreased the microbubble andnanobubble sizes and increased their concentrations in liquid whichincreased the probability of collision (Pc). Also, atomic forcemicroscopy(AFM) images show that the coalescence of nanobubbles forms gaseouscapillary bridges as hydrophobic surfaces approach, and thus a capillaryforce (Hampton and Nguyen, 2010). The resulting concave capillarybridge produces an attractive force that forces the two surfaces intocontact. Thus, nanobubble coated surface of very fine particles canlead to particle aggregation and thus more easily recovered due to anincreased collision probability.

Different superficial air velocities of 0.4, 0.8, 1.2, and 1.6 cm/swere used to investigate the influence of the superficial air velocityon flotation performance in the presence and absence of nanobubblesand the results are shown in Fig. 7(A) and (B). The flotation recoveryremained essentially unchanged at a high level of approximately95% in the presence of nanobubbles when the superficial air velocitydecreased from 1.6 to 0.4 cm/s. Product ash increased from 6.0% to9.1% as superficial air velocity increased from 0.4 to 0.8 cm/s, remainedunchanged as superficial air velocity increased to 1.2 cm/s, and thenincreased another 3% as superficial air velocity increased to 1.6 cm/s.The greater separation associated with high recovery and low productash observed at lower superficial air velocities was believed to be theresult of good selectivity of cavitation generated nanobubbles attachingto coal particles.

Combustible recovery was much higher in the presence ofnanobubbles at all superficial air velocities examined, as shown inFig. 7(A). For example, use of nanobubbles increased combustiblerecovery by about 50% at a superficial air velocity of 0.8 or 1.2 cm/s.This result is in agreement with the data shown in Fig. 5, confirmingthat use of nanobubbles increased flotation recovery. The separationefficiency vs. superficial air velocity curve shown in Fig. 7(B) clearlyindicates that use of nanobubbles greatly improved separation perfor-mance of flotation, which is also consistent with the result shown inFig. 5(B).

Fig. 7(B) shows that the separation efficiency in presence ofnanobubbles decreased from 80% to about 65% as superficial air velocityincreased from 0.4 to 1.6 cm/s. This is because nanobubbles increasedthe attachment probability of the weak hydrophobic particles by in-creasing their hydrophobicity. This increased the clean coal ash contentfrom 5% to about 12% whereas combustible recovery remained essen-tially unchanged. The data clearly indicates that a lower superficial airvelocity is preferred in the presence of nanobubbles, which reducesthe air consumption and thus operating cost. The lower requirementfor external air consumption in the presence of nanobubbles is a directresult of the fact that nanobubbles are produced from air naturallydissolved in water.

Many studies have confirmed that nanobubbles generated byhydrodynamic cavitation change the surface characteristics of minerals(Hampton and Nguyen, 2010), increase contact angle of solids andhence attachment force (Fan et al., 2010b), bridge fine particles toform aggregates, and in consequence reduce reagent consumption(Fan and Tao, 2008; Zhou et al., 1997).

The probability of attachment (Pa) is related to the energy barrierfor the bubble–particle adhesion Ei and the kinetic energy of collisionEk in Eq. (5) (Mao and Yoon, 1997; Yoon and Luttrell, 1989):

Pa ¼ exp − EiEk

� �: ð5Þ

Fig. 9. Effect of nanobubbles at varying superficial feed velocities (A) on combustiblerecovery and clean coal ash (B) on separation efficiency.

114 A. Sobhy, D. Tao / International Journal of Mineral Processing 124 (2013) 109–116

Pa can be calculated using Eq. (6) (Yoon and Luttrell, 1989):

Pa ¼ sin2 2 tan−1 exp −45þ 8Re0:72ubti15Db

DbDp

þ 1� �

0@

1A

24

35 Dp

Db

� �2

: ð6Þ

Eq. (6) indicates that Pa increases with increasing particlehydrophobicity or decreasing induction time ti. Pa increases with de-creasing bubble rising velocity ub anddecreasing bubble sizeDb, meaningsmaller bubble size ismore favorable for increasing probability of attach-ment (Ralston and Dukhin, 1999; Yoon, 2000; Yoon and Luttrell, 1989).Pa also decreases with increasing Dp, suggesting that coarse particlesare more difficult to attach to air bubbles.

Simonsen et al. (2004) and Stockelhuber et al. (2004) found thatnanobubbles can cause the rupture of the wetting films betweenmineral particles and conventional-sized bubble which is a basic stepin the flotation process. In the process of drainage of the wetting film,the largest nanobubble is almost as thick as the wetting film. Fig. 8(A)shows that nanobubbles play no roles in rupture process when thethickness of thick film hw is greater that the height of bubbles hrupture.As the film thickness is close to the bubble height, the surface forcesbegin to act between the biggest nanobubble and film surface asshown in Fig. 8(B). It has been observed that the rupture alwayshappens at the biggest nanobubble, because the thinnest place is theweakest spot to break at the same interaction force (Stockelhuberet al., 2004).

To investigate the influence of the feed flow rate on flotationperformance, different feed flow rates of 0.25, 0.50, 0.75 and 1.00 cm/swere examined. Nanobubbles increased flotation recovery, particularlyat higher superficial feed velocities, as shown in Fig. 9(A). For example,an increase of approximately 10%, i.e., from 60% to almost 70%, incombustible recovery was obtained at a superficial feed velocity of0.75 cm/s in the presence of nanobubbles, while an increase of approxi-mately 22%, i.e., from 33% to almost 55%, in combustible recovery wasobtained at a superficial feed velocity of 1.0 cm/s in the presence ofnanobubbles. Although combustible recovery decreased from 94% to55% when superficial feed velocity increased from 0.25 to 1.0 cm/s,it was still much higher in the presence of nanobubbles than in theirabsence.

The product ash content was lower or almost the same in thepresence of nanobubbles than in their absence. The separation efficiencycurve shown in Fig. 9(B) indicates that the use of nanobubbles signifi-cantly improved the separation performance at superficial feed velo-cities larger than 0.5 cm/s. This is because nanobubbles reduced theprobability of detachment of the attached hydrophobic particles from

Fig. 8. AFM image: Wetting film with gas bubbles adhered to the solid substrate (A) for abehavior (B) Film rupture occurs at a film thickness on the order of the biggest bubble heig

bubbles. Furthermore, after collision and attachment, particles attachedto air bubbles do not all report to the froth phase and some of themdetach from bubble surface and drop back into the pulp phase whensuperficial feed rate was very high.

large film thickness hw, where no rupture occurs, nanobubbles play no roles in filmht (hw = hrupture) (Stockelhuber et al., 2004).

115A. Sobhy, D. Tao / International Journal of Mineral Processing 124 (2013) 109–116

The probability of detachment Pd is related to the energy barrierfor the bubble–particle detachment Ei′, work of adhesion betweenbubble and particle Wa and kinetic energy of collision Ek in Eq. (7)(Mao and Yoon, 1997):

Pd ¼ exp − E0

i þWa

Ek

!ð7Þ

Particle detachment occurs when detachment forces exceedthe maximum adhesive forces. Pd can be calculated using Eq. (8)(Tao, 2004):

Pd ¼ 1þ 3 1− cosθdð Þγg ρb−ρw

12 þ 3

4 cos θd2

� �� �0@

1A 1þ Dp

Db

D2p

0@

1A

24

35−1

: ð8Þ

Eq. (8) shows that Pd increases with increasing particle size Dp andincreasing bubble sizeDb. Therefore, coarse heavy and less hydrophobicparticles are more likely to detach from large air bubbles and use ofsmall bubbles by increasing the frother concentration will increase theflotation recovery of coarse particles, as shown in Fig. 10(A) and (B).The effects of bubble size and particle size on the particle detachmentprobability determined from Eq. (8) for coal and ash particles with acontact angle of 60° and 10° and a specific gravity of 1.3 and 2.6, respec-tively, are shown in Fig. 10(A) and (B). The results clearly show that ashparticles have substantially higher detachment probabilities than coalparticles and smaller air bubbles reduce the detachment probability.Obviously, the coal particle has a substantially lower probability ofdetachment than the ash particle for a given particle and bubble size.

Fig. 10. Effect of bubble size and particle size on coal (a) and ash (b) particles detachmentprobability.

A high froth flotation efficiency is limited to the narrow particlesize range between 50 μm and 600 μm for coal, as shown in Fig. 11(Jowett, 1980). Nanobubbles expand the froth flotation particle sizelimit by enhancing the recovery of fine and relatively coarse particlesas a result of elevated probabilities of particle–bubble collision andattachment and reduced probability of detachment (Fan et al., 2010b).The improvement by nanobubbles on the difficult-to-float particleswas more significant than that on the easy-to-float particles, especiallyat lower collector dosages (Fan et al., 2010c). Tao (2004) found thatnanobubbles can extend the lower size limit for effective flotation to afewmicrons, even submicrons, and the upper limit to 1–2 mm, increasingthe process efficiency for ultrafine and relatively coarse particles andexpanding applications of flotation.

Fig. 12 shows the effect of nanobubbles on combustible recoveryas a function of ash rejection at a collector dosage of 0.49 kg/ton,frother concentration of 30 ppm, superficial wash water flow rate of0.41 cm/s, and feed slurry solids concentration of 7%. The results inFig. 12(A) were generated by changing superficial slurry feed velocitiesfrom 0.25 cm/s to 1.23 cm/s while data for Fig. 12(B) were generatedby changing superficial gas velocities from 0.4 cm/s to 1.6 cm/s. Bycomparing these combustible recovery data generated in the presenceand absence of nanobubbles with the release analysis curve, it canbe clearly seen that the flotation combustible recovery produced withnanobubbleswasmuch higher thanwithout nanobubbles and in generalthe data points generated in the presence of nanobubbles were closerto the release analysis curve, indicating that the use of nanobubblesimproved flotation separation performance.

4. Conclusions

Nanobubble application to laboratory column coal flotation hasproven very successful. It not only significantly improved combustiblerecovery, but also reduced reagent and air consumption. Laboratoryflotation results have shown that nanobubbles significantly enhancedthe coal flotation process efficiency with higher recovery and/orlower product ash. The flotation recovery of fine coal was increased by5 to 50 absolute percentage points, depending on process operatingconditions. The frother dosage was reduced by one-third becausenanobubbles are mostly smaller than 1 μm and they are formed fromprecipitation air naturally dissolved in water. The improved flotationperformance by nanobubbles can be attributed to increased probabilitiesof collision and attachment and reduced probability of detachment.

Fig. 11. Froth flotation particle size limitations (Jowett, 1980).

Fig. 12. Effect of nanobubbles on flotation combustible recovery with and withoutnanobubbles (A) at different superficial feed slurry velocities (B) at different superficialair velocities.

116 A. Sobhy, D. Tao / International Journal of Mineral Processing 124 (2013) 109–116

Acknowledgment

This research project is supported by National Natural ScienceFoundation of China (Grant No. 50921002 and 51274200) and theDepartment of Commerce and Economic Opportunity of the state ofIllinois through the Office of Coal Development and the IllinoisClean Coal Institute under the project Number 10/4B-3.

References

Borkent, B.M., Beer, S.D., Mugele, F., Lohse, D., 2010. On the shape of surface nanobubbles.Langmuir 26 (1), 260–268.

Fan, M., Tao, D., 2008. A study on picobubble enhanced coarse phosphate froth flotation.Sep. Sci. Technol. 43 (1), 1–10.

Fan, M., Tao, D., Honaker, R., Luo, Z., 2010. Nanobubble generation and its applicationin froth flotation (part I): nanobubble generation and its effects on the propertiesof microbubble and millimeter scale bubble solutions. Min. Sci. Technol. (China) 20(1), 1–19.

Fan, M., Tao, D., Honaker, R., Luo, Z.F., 2010. Nanobubble generation and its application infroth flotation (part II): fundamental study and theoretical analysis. Min. Sci. Technol.20 (2), 159–177.

Fan, M., Tao, D., Honaker, R., Luo, Z.F., 2010. Nanobubble generation and its applicationin froth flotation (part III): specially designed laboratory scale column flotationof phosphate. Min. Sci. Technol. 20 (3), 317–338.

Farmer, A.D., Collings, A.F., Jameson, G.J., 2000. Effect of ultrasound on surface cleaningof silica particles. Int. J. Miner. Process. 60, 101–113.

Feng, D., Aldrich, C., 1999. Effect of particle size on flotation performance of complexsulfide ores. Miner. Eng. 12, 721–731.

Hampton, M.A., Nguyen, A.V., 2010. Nanobubbles and the nanobubble bridging capillaryforce. Adv. Colloid Interf. Sci. 154 (1–2), 30–55.

Johnson, B., Cooke, R., 1981. Generation of stabilized microbubbles in seawater. Science213, 209–211.

Johnson, D.J., Al Malek, S.A., Al-Rashdi, B.A.M., Hilal, N., 2012. Atomic force microscopyof nanofiltration membranes: effect of imaging mode and environment. J. Membr.Sci. 389, 486–498.

Jowett, A., 1980. Formation and disruption of particle–bubble aggregates in flotation.In: Somasundaran, P. (Ed.), Fine Particles Processing, 1, pp. 720–754.

King, R.P., 1982. Flotation of fine particles. In: King, R.P. (Ed.), Principles of Flotation.South African Institute of Mining and Metallurgy, Johannesburg, pp. 217–219.

Luttrell, G.H., Yoon, R.-H., 1992. A hydrodynamic model for bubble–particle attachment.J. Colloid Interface Sci. 154, 129–137.

Mao, L., Yoon, R.-H., 1997. Predicting flotation rates using a rate equation derived fromfirst principles. Int. J. Miner. Process. 51, 171–181.

Ralston, J., Dukhin, S.S., 1999. Interaction between particles and bubbles. Colloids Surf.,A Physicochem. Eng. Asp. 151, 3–14.

Schubert, H., Bischofberger, C., 1979. On the optimization of hydrodynamics in flotationprocesses. Proceedings of 13th International Mineral Processing Congress, Warszawa,2, pp. 1261–1287.

Simonsen, A.C., Hansen, P.L., Klosgen, B., 2004. Nanobubbles give evidence of incompletewetting at a hydrophobic interface. J. Colloid Interface Sci. 273, 291–299.

Stockelhuber, K.W., Radoev, B., Wenger, A., Schulze, H.J., 2004. Rupture of wetting filmscaused by nanobubbles. Langmuir 20, 164–168.

Tao, D., 2004. Role of bubble size in flotation of coarse and fine particles — a review.Sep. Sci. Technol. 39 (4), 741–760.

Tao, D., Yu, S., Zhou, X., Honaker, R.Q., Parekh, B.K., 2008. Picobubble column flotationof fine coal. Int. J. Coal Prep. Util. 28 (1), 1–14.

Tao, D., Fan, M., Honaker, R.Q., 2010. Enhanced Fine Coal Column Flotation usingCavitation Concept. International Coal Preparation Congress, Conference Proceedings,pp. 413–420.

Trahar, W.J., Warren, L.J., 1976. The floatability of very fine particles— a review. Int. J. Miner.Process. 3, 103–131.

U.S.Energy Information Administration, 2009. http://www.eia.gov/cneaf/coal/page/acr/tables2.html.

Weber, M.E., Paddock, D., 1983. Interceptional and gravitational collision efficienciesfor single collectors at intermediate Reynolds numbers. J. Colloid Interface Sci.94, 328–335.

World Coal Association, 2012. http://www.worldcoal.org/coal/uses-of-coal/.Yoon, R.-H., 2000. The role of hydrodynamic and surface forces in bubble–particle

interaction. Int. J. Miner. Process. 58, 128–143.Yoon, R.-H., Luttrell, G.H., 1989. The effect of bubble size on fine particle flotation.

Miner. Process. Ext. Metall. 5, 101–122 Revised.Young, F.R., 1989. Cavitation. McGraw-Hill Book Company, London.Zhou, Z.A., Xu, Z., Finch, J.A., Hu, H., Rao, S.R., 1997. Role of hydrodynamic cavitation

in fine particle flotation. Int. J. Miner. Process. 51, 139–149.