the influence of gas velocity, salt type and concentration on transition concentration for bubble...
TRANSCRIPT
Tta
PS
1
Ascmnat1
tmH1ebwc
0d
chemical engineering research and design 9 0 ( 2 0 1 2 ) 33–39
Contents lists available at ScienceDirect
Chemical Engineering Research and Design
j ourna l ho me page: www.elsev ier .com/ locate /cherd
he influence of gas velocity, salt type and concentration onransition concentration for bubble coalescence inhibitionnd gas holdup
hong T. Nguyen, Marc A. Hampton, Anh V. Nguyen ∗, Greg R. Birkettchool of Chemical Engineering, The University of Queensland, Brisbane, QLD 4072, Australia
a b s t r a c t
The influence of gas velocity (3.5, 10, and 18 mm/s), salt type (NaCl, NaF, NaBr, NaI and CsCl) and salt concentration
(0.001–3 M) on bubble coalescence in a small bubble column were studied. The bubble coalescence was determined
by the relative change in the measured light intensities passing through the salt solutions and clean deionised water.
It was shown that the transition salt concentration for bubble coalescence inhibition (determined at 50% of bubble
coalescence) of all investigated salts decreases with increasing superficial gas velocity. The difference in the transition
concentration between NaCl, NaF, NaBr and CsCl does not significantly change with the gas velocity. However that
difference between NaI and the other salts significantly decreases with increasing the gas velocity. The gas holdup
significantly increases with NaCl, NaF, NaBr and CsCl concentrations but does not significantly change with NaI
concentration. These new results imply that the transition salt concentration for bubble coalescence and gas holdup
depend not only on the salt properties (i.e. the ion type and their combination) as previously reported, but also on
the hydrodynamic conditions.
© 2011 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Keywords: Bubble; Coalescence; Gas holdup; Salt; Ion specificity
with the data obtained with bubbles having a mobile surface
. Introduction
ir bubbles are used in industrial applications to provide highpecific surface areas for mass and/or heat transfer in bubbleolumns and chemical reactors or for flotation separation inineral processing, wastewater treatment and water desali-
ation. Salts have been found to significantly affect the sizend the coalescence of bubbles, and hence the performance ofhese applications (Harvey et al., 2002; Klassen and Mokrousov,963; Paulson and Pugh, 1996).
Numerous studies have been carried out to investigatehe effects of salts on bubble coalescence and the underlying
echanisms (Chesters and Hofman, 1982; Craig et al., 1993;ofmeier et al., 1995; Marrucci, 1969; Marrucci and Nicodemo,967; Prince and Blanch, 1990a). Many studies have focused onxplaining the ion effects specific to the ion types and com-inations. Some salts have been found to inhibit coalescence,hereas many other salts do not. Ctrans is defined as the con-
entration at which bubble coalescence is 50%, where 100%
∗ Corresponding author. Tel.: +61 7 336 53665; fax: +61 7 336 54199.E-mail address: [email protected] (A.V. Nguyen).Received 8 March 2011; Received in revised form 5 August 2011; Accep
263-8762/$ – see front matter © 2011 The Institution of Chemical Engioi:10.1016/j.cherd.2011.08.015
is for pure water. Bubble coalescence has been found to beion-specific, i.e., Ctrans strongly depends on ion types and theircombination (Christenson et al., 2008; Craig et al., 1993). Insome of the studies, Ctrans was identified as a unique parame-ter for each of the salts examined. Some researchers (Chan andTsang, 2005; Marrucci and Nicodemo, 1967; Prince and Blanch,1990b) developed models to predict Ctrans as a function of sur-face tension, bubble size, and van der Waals forces. It is notedthat these theories were developed for the immobile air–waterinterface, which are expected to be different to the mobileair–salt solution interfaces. The mobility of air–salt solutioninterface is hypothesised to be a function of ion partition atthe interface. Researchers (Zahradnik et al., 1995) have usedthe proposed model equations and compared the theorieswith various experimental data (Lessard and Zieminski, 1971),but have not realised that the models developed for bubbleswith the immobile surface might not be physically consistent
ted 8 August 2011
(Zahradnik et al., 1995). Tsang et al. (2004) reportedthe first
neers. Published by Elsevier B.V. All rights reserved.
34 chemical engineering research and design 9 0 ( 2 0 1 2 ) 33–39
Fig. 1 – Schematic of bubble column apparatus fordetermining bubble coalescence.
0
10
20
30
545250484644424038
Freq
uenc
y (c
ount
)
Light energy (mW)
Fig. 2 – A histogram of light energy measured by the photodetector. The solid line describes a six-order polynomial fit.
evidence of Ctrans dependence on bubble size. They investi-gated bubble coalescence in aqueous MgSO4 solutions usingthe pair-bubble interaction method. The Ctrans was found toincrease with decreasing the bubble size.
The effect of the salt ion chemistry on bubble coalescencehas also been investigated and provided interesting insight(Craig et al., 1993; Henry and Craig, 2010; Kunz, 2010). Craig(2004) demonstrates the combining rule of the salt ions inbubble coalescence: the �� or �� salts such as sodium chlo-rate do not inhibit bubble coalescence, but �� or �� salts suchas sodium chloride or lithium chloride inhibit bubble coales-cence. Similarly, molecular dynamics simulations (Jungwirthand Tobias, 2006; Marcelja, 2006) show that the �� or �� saltions equally partition at the interface, but the �� or �� saltions unequally distribute at the interface. These results showthe importance of the ion specificity (size and polariability)effect at high salt concentrations on bubbles coalescence inhi-bition. Recently, the empirical Collins’ concept of matchingwater affinities “oppositely charged ions form direct ion pairsspontaneously only when they have equal water affinities” hasbeen applied to explain many experimental observations inbiology (Kunz, 2010) and flotation separation of hydrophobicparticles in brine solutions (Ozdemir et al., 2011). It is con-sidered in the empirical Collins concept that two stronglyhydrated small (hard) ions of opposite charge, such as Na+ andCl-, can strongly attract each another, forming direct ion pairsand expelling the hydration spheres between them. Likewise,weakly hydrated large (soft) ions of opposite charge experiencemuch weaker electrostatic attraction but have loosely boundhydration shells, and can also form direct ion pairs expellingthe hydration water between them. However, the hard and softoppositely charged ions, such as Na+ and I- do not experiencethe similar interactions between themselves and with water,and are always separated by water hydration shell around thehard ion, preventing them from forming strong ion pairs. Thesalt ion affinities towards water appear to play an importantrole in the specific ion effects observed in a number of exper-iments.
The available experimental data and models show thatsalts significantly influence bubble coalescence. However, theavailable data are scattered and the models are not able todescribe the available experimental results. This paper aimsto re-examine the influence of salts on bubble coalescence.Specifically, the paper focuses on the effect of gas superficialvelocity, salt type and concentration on bubble coalescenceinhibition. It is expected that both hydrodynamics and salt
chemistry influence bubble coalescence and hence Ctrans.2. Materials and methods
The bubble column setup (Fig. 1) used for the bubble coa-lescence experiments consisted of a cylindrical glass column(4.5 cm inner diameter and 20 cm height) and an optical sys-tem for measuring the light intensity passing through thecolumn. The change in bubble coalescence in salt solutionswas determined by the change in solution turbidity mea-sured by the light intensity. High purity nitrogen bubbles (BOC,Australia) were created using a glass frit (porosity 11–16 �m)at the bottom of the column. Experiments were conductedat superficial velocities of 3.5, 10, and 18 mm/s. The opticalsystem composed of a cold light source 24 V 150 W (XGY-II(A)Halogen Lamp Unit, ProSciTech, Kirwan, QLD, Australia) onone side of the column, and a condensing lens followed by aphoto sensor (S121C, Thorlabs Inc., Newton, NJ, USA) on theother side, which was connected to a laptop computer via aUSB power and energy meter interface (PM100USB, ThorlabsInc., Newton, NJ, USA). Only the 635 nm wavelength light of thewhite light source emitted from the XGY-II(A) Halogen Lampunit was selected by the built-in software of the PM100USBUSB interface for calibrating and measuring. This experimen-tal method is based on measuring the solution turbidity insidethe bubble column. The measured light intensity was cali-brated against the light intensity measured for clean deionisedwater (100% coalescence) at which the solution turbidity andlight absorbance are minimum. For each measurement, thesensor measured the energy of the transmitted light every 0.5 sfor 100 s, i.e., 200 data points were taken to reliably obtain ahistogram of energy (Fig. 2). The histogram did not changewith time. The histogram of the light energy measured by thephoto detector was then used to calculate the mean value ata probability of 95%, which was used to provide stable andreproducible results.
All glassware was thoroughly cleaned by soaking in alka-line cleaning solution prepared from potassium hydroxide,water and ethanol (12.5:16:84 mass ratio), and vigorouslyrinsed many times with pure deionised (DI) water, producedby a Milli-Q system (Millipore, USA). The glassware was thensoaked in a diluted hydrochloric acid solution, and flushedthoroughly with the DI water. The column was cleaned usingthe above cleaning procedure, followed by bubbling with nitro-gen for 10 min. This cleaning method ensured the column andthe glass frit were free from contamination. All salts (NaCl,NaF, NaBr, CsCl and NaI) used were of the highest purity (99.5%A.C.S grade, Sigma–Aldrich, Australia). Further recrystallisa-
tion and calcination did not improve the salt purity but foamfractionation by bubbling the salt solutions for 2 min before
chemical engineering research and design 9 0 ( 2 0 1 2 ) 33–39 35
0
20
40
60
80
100
10.10.010.001
Perc
enta
ge c
oale
scen
ce
Salt concentration (M)
NaI solution
3.5 mm/s10 mm/s18 mm/s
0
20
40
60
80
100
10.10.010.001
perc
enta
ge c
oale
scen
ce
Salt concentration (M)
NaCl solution
3.5 mm/s10 mm/s18 mm/s
Ctrans
0
20
40
60
80
100
10.10.010.001
Perc
enta
ge c
oale
scen
ce
Salt concentration (M)
CsCl solution
3.5 mm/s10 mm/s18 mm/s
Fig. 3 – Bubble coalescence versus salt concentration andsuperficial gas velocity. The dashed lines with arrow showthe decreasing trend of the transition concentration (opencircles) with increasing gas velocity.
siabek
Fv
tarting the experiments removed the trace organic contam-nants as evidenced by surface tension and chromatographynalysis. The three salts were chosen to provide suitable com-ination of halide anions and cations for comparison. The
ffect of the salt ion chemistry on bubble coalescence is alsonown (Craig et al., 1993).0
0.1
0.2
0.3
0.4
150Tran
sitio
n co
ncen
trat
ion,
Ctrans
(M)
Superficial gas
ig. 4 – Transition salt concentration, Ctrans, for bubble coalescenelocity.
The gas holdup, ε, was determined by the bed expansionmethod which gives ε = 100(1 − H0/H)[%], where H is the totalheight of the aerated solution and H0 is the initial height ofthe solutions, which was 8.2 cm (100 mL solution). For the saltsolutions studied, the foam volume formed on the surfacewas significantly small and the bed expansion method gives agood estimation for the gas holdup in the bubble columns. Allexperiments were carried out at controlled room temperature(21.5–23 ◦C).
3. Results
Shown in Fig. 3 are the typical results for the bubble coales-cence versus salt concentration and gas superficial velocity.The bubble coalescence decreases with increasing salt con-centration for each salt. A similar trend was previouslyreported for a low gas velocity (Craig et al., 1993). However,the result in Fig. 3 shows a strong influence of superficial gasvelocity on bubble coalescence. Fig. 4 summarises the experi-mental results for the transition salt concentration, Ctrans, forbubble coalescence inhibition as a function of superficial gasvelocity. The result in Fig. 4 shows that Ctrans of all investi-gated salts decreases with increasing superficial gas velocity.The decrease in Ctrans of NaCl and CsCl with increasing thegas velocity is similar. Therefore, the difference between Ctrans
of these two salts does not significantly change with the gasvelocity. However, with increasing the gas velocity Ctrans ofNaI decreases much faster than Ctrans of NaCl and CsCl. Theinfluence of gas superficial velocity on Ctrans of all investi-gated salts follows the order: NaI > NaBr > CsCl > NaCl > NaF.This result agrees with the combining rule (Craig et al.,1993) in that the combination of the cations (Na and Cs)and anions (I and Cl) in the salts inhibits bubble coales-cence, but the order of the inhibition strength as revealedby our experiments is also important. The strongest influ-ence of gas superficial velocity on transition salt concentrationby NaI indicates that of the cations and anions examined,iodide appears to influence bubble coalescence most signif-icantly.
Fig. 5 shows the experimental results for the gas holdup, ε,as a function of salt concentration and superficial gas veloc-ity. In agreement with the literature data (Moshtari et al.,2009; Orvalho et al., 2009; Zahradnik et al., 1995), gas holdupincreases with increasing superficial gas velocity. This depen-dence of ε on increasing salt concentration is opposite tothe dependence of Ctrans (Fig. 4). The results also show thatthe gas holdup increase in NaI solutions is not as rapid
as in the other salt solutions. At the highest salt concen-tration of 3 M and the highest gas flow rate of 18 mm/s,20150
velocity (mm/s)
NaICsClNaCl
ce inhibition for NaCl, CsCl and NaI versus superficial gas
36 chemical engineering research and design 9 0 ( 2 0 1 2 ) 33–39
0
20
40
60
80
10.10.010.001
Gas
hol
dup,
%
Conce ntra�on, M
NaF3.5 mm/s10 mm/s18 mm/s
0
20
40
60
80
10.10.010.001
Gas
hol
dup,
%Concentra�on, M
NaBr3.5 mm/s10 mm/s18 mm/s
0
20
40
60
80
10.10.010.001
Gas
hol
dup,
εε(%
)G
as h
oldu
p, ε
(%)
Gas
hol
dup,
ε(%
)
Salt concentration (M)
NaCl3.5 mm/s
10 mm/s
18 mm/s
0
20
40
60
80
10.10.010.001Salt concentration (M)
CsCl3.5 mm/s
10 mm/s
18 mm/s
0
20
40
60
80
10.10.010.001Salt concentration (M)
NaI3.5 mm/s
10 mm/s
18 mm/s
Fig. 5 – Effect of salt concentration and gas superficial velocity on gas holdup, ε· The dotted lines on the diagram for NaFdescribe the approximate extrapolations to the high concentration region beyond the saturation limit, where the liquid
0
2
4
6
8
10
120100806040200
Bub
ble
term
inal
vel
ocity
, V(m
m/s
)
Bubble diameter, D(μμm)
Bubble coalescence-noninhibiting salt (0.23 M NaClO4)
Bubble coalescence-inhibiting salt (0.11 M HClO4)
Ultra-clean water
Hadamard-Rybczynski rise velocity
Stokes rise velocity
2
12D gV δμ
=
2
18D gV δμ
=
Fig. 6 – Experimental (symbols) (Henry et al., 2008;Parkinson et al., 2008) and model (lines) results for terminalrise velocity of N2 bubbles as a function of bubble size inultra-clean water, bubble coalescence-inhibiting andnoninhibiting salt aqueous solutions. ı, � and g describe
becomes very viscous and crystals are formed.
the highest gas holdup obtained for NaCl and CsCl is 1.9and 2.4 times higher than the highest gas holdup obtainedfor NaI, respectively. The influence of high gas superficialvelocity on ε for all investigated salts follows the increasingorder: NaI < NaBr < NaCl < CsCl, which oppositely mismatchesthe decreasing order of Ctrans: NaI > NaBr > CsCl > NaCl. Fur-thermore, the increase in gas holdup with increasing saltconcentration can be divided into two regimes: the low andhigh salt concentration regimes. In the low salt concentrationregime, the increase in gas holdup with superficial gas veloc-ity is not very sensitive to the salt type. In this regime, theextensive data on gas holdup for a wide range of salts (and atleast two gas rates) were presented by Quinn et al. (2004) whoshowed a weak correlation of gas holdup with ionic strength.In the high salt concentration regime, the gas holdup increaseis rapid for CsCl, NaF, NaBr and NaCl, but not for NaI. In thehigh salt concentration regime the gas holdup appears to besensitive to the salt type.
4. Discussion
The mechanism of inhibition of bubble coalescence in saltsolutions is yet unknown. However, it is known that saltsinfluence the drainage and stability of liquid films betweenbubbles, which are the driving force for bubble coalescence.The coalescence of bubbles is governed by three consecu-
tive steps: (1) the relative approach of bubbles, which leadsto the formation of a thin liquid film between gas and liquidthe liquid density and viscosity, and gravity acceleration.
interfaces, (2) the drainage of the liquid film under externalforces (due to gravity, inertia and flow) proportional to thebubble volume, and surface forces of molecular origin, which
are proportional to thin liquid film area, and (3) the film rup-ture at a critical thickness under the influence of attractive
chemical engineering research and design 9 0 ( 2 0 1 2 ) 33–39 37
Table 1 – Summary of the key predictions for Ctrans, where Rg is the gas constant, T is the absolute temperature,A = 2.5 × 10−20 J is the Hamaker constant, B = 1.5 × 10−28 m is the retarded van der Waals coefficient, � is the surfacetension, hrup is the film rupture thickness, C is the salt concentration, D is the bubble diameter and v is the number ofions produced upon dissociation.
Author(s) Prediction for Ctrans Molecular forces considered
Marrucci (1969) Ctrans = 0.84vRgT{
2�A2D
}1/3{∂�∂C
}−2Short-range (electromagnetically non-retarded) van der Waals attraction
Prince and Blanch (1990b) Ctrans = 1.18vRgT{
2�BD
}1/2{∂�∂C
}−2Long-range (electromagnetically retarded) van der Waals attraction(Casimir–Lifshitz force)
Chan and Tsang (2005) Ctrans = vRgT�(hrup )2
D
{∂�∂C
}−2Repulsive hydration (decays by a single exponential function ofseparation distance)
Fig. 7 – Ion partition at the air–sodium halide solution (1.2 M) interface as revealed by MD simulation (Jungwirth and Tobias,2006). The ion density, �(z), is normalised by the ion density in the bulk solution, �b.
38 chemical engineering research and design 9 0 ( 2 0 1 2 ) 33–39
surface forces and mechanical disturbances, leading to bubblecoalescence.
The first step of the relative approach of bubbles is governedby the bubble rise velocity. Recently, the effect of salts on thebubble rise velocity was re-examined to determine whetheror not the mobility of the gas–aqueous salt solution interfaceswould influence bubble rise velocity and coalescence inhibi-tion (Henry et al., 2008; Parkinson et al., 2008). The measureddata for terminal rise velocity of small bubbles (with diametersmaller than 110 �m) in both the bubble coalescence inhibit-ing and non-inhibiting salt solutions show that the bubblerise velocity obeys the Hadamard–Rybczynski model (Fig. 6),which corresponds to a mobile interface (slip boundary con-dition) and agrees with observations in ultra-clean water. TheHadamard–Rybczynski theory predicts the bubble rise velocityhigher than the Stokes theory (based on immobile air–liquidinterface) by a factor of 1.5. Only the properties of water(the density ı = 1000 kg/m3 and the bulk viscosity � = 0.001 Pa s)are used in calculating the Hadamard–Rybczynski bubble risevelocity, which excellently compares with the experimentalresults for inhibiting and non-inhibiting salt solutions andultra-clean water. This agreement provides strong evidencethat the bubble coalescence inhibiting and non-inhibiting saltsolutions salts do not influence the rise velocity of small bub-bles and the salts do not influence bubble coalescence throughbubble rise velocity. In fact, the gas holdup remains unchangedalong the bubble column height above the frit surface. Thebubble coalescence should significantly occur during the bub-ble formation on the frit surface, where the film drainage andrupture play an important role. The role of liquid film drainageand rupture, and attractive surface forces is discussed below.
When bubbles are in close contact, a small amount of liq-uid is entrapped between them, forming a small circular lensor liquid film with an initial thickness of a few 100 nm. Capil-lary and molecular forces on the opposite sides of the liquidfilms cause the film to grow thinner and rupture at a crit-ical film thickness. The film drainage equation establishedby the lubrication approximation has been solved employ-ing both short- and long-range van der Waals attraction, andrepulsive hydration (Chan and Tsang, 2005; Marrucci, 1969;Prince and Blanch, 1990b). The change in surface tension dueto increasing salt concentration (the Gibbs–Marangoni effect)was assumed to play a significant role in immobilising theliquid flow inside the film and preventing the film from coa-lescence. The available predictions for Ctrans are summarisedin Table 1. All the three available theoretical models basedon the Gibbs–Marangoni effect predict that Ctrans is propor-tional to the square of surface tension gradient with respect tosalt concentration. Although the surface tension gradient hasbeen found to correlate with the entropies of ion hydration,the Gibbs–Marangoni effect does not provide a satisfactoryexplanation for the bubble coalescence inhibition of salts(Weissenborn and Pugh, 1996). Furthermore, the Hamakerconstants and coefficient (A and B) and many other parame-ters contained in the three model equations do not specificallydepend on specific properties of salt ions such as ion sizeor charge/polarisability. The electrical double-layer forces arenot dependent on the ion-specific properties and normallybecome vanishingly small at high salt concentration. There-fore, the classical DLVO (Derjaguin-Landau-Verwey-Overbeek)theory of colloid stability does not explain the bubble coales-cence and the transition concentration shown in Figs. 3 and 4,
respectively. Several additional hypotheses suggested includeincrease in repulsion due to hydration (Pashley, 1981) anddecrease in hydrophobic attraction (Craig et al., 1993; Pugh andYoon, 1994). The hydration force due to the change in solventstructure and hydration shell is short-ranged and cannot sta-bilise saline water films with thickness >50 nm. The origin ofthe measured hydrophobic attraction is still debated. There-fore, the suggested extensions of the DLVO theory remainunsatisfactory.
Interestingly, the available theories as summarised inTable 1 do not show the influence of the gas superficial velocityon the salt transient concentration. This contradicts the exper-imental results shown in Fig. 4. The theoretical deficiencyrequires further theoretical investigation.
It is further noted that increasing gas rate increases thebubble size and decreases the salt transition concentration(Fig. 4). There could be a correlation between the bubble sizeand the salt concentration, but this correlation while useful inmany cases is of secondary importance here. The bubble sizecannot be the primary factor causing the change in the bub-ble coalescence in the experiments. Indeed, the high-speedmovies show that the bubble coalescence mainly occurs atthe frit surface, where the momentum of the gas stream ishigh and can overcome the resistance of liquid film betweenthe gas-liquid interfaces, causing the film rupture and coa-lescence. Evidently, the gas velocity is the primary factor (i.e.,more important than the bubble size) causing the bubble coa-lescence in different salt solutions. The experiments with fritsof big and small pore sizes show the same decreasing trendsof the salt transition with increasing the gas superficial veloc-ity. The salt transition concentration can change from one toanother experimental system (with the different gas veloci-ties, frit pore sizes, hydrostatic pressures, salts, etc.) and itsaccurate prediction remains a challenge.
The failure of the DLVO theory in explaining the inhi-bition of bubble coalescence in salt solutions implies thation-specific non-DLVO factors may be responsible for bub-ble coalescence. Recently, molecular dynamic (MD) simulation(Jungwirth and Tobias, 2006) and sum-frequency-generationspectroscopy (Liu et al., 2004) have provided useful informa-tion about ion partition at the gas–liquid interfaces, whichmay influence the salt inhibition of bubble coalescence.The MD results show (Fig. 7) that the ions of NaCl andNaI do not equally concentrate at the interfaces, but forNaF the salt ions equally distribute at the gas–salt solutioninterfaces. Specifically, the big anions like iodide are sig-nificantly expelled to the gas phase by the interface whilethe smaller anions like fluoride stay within the interfacelayer, where they strongly associate with the sodium cations.Experiments in this study show that NaI weakly inhibits thebubble coalescence and Craig et al. (1993) found that othersalts with large anion size, for example, NaClO3 do not. Itis possible that the difference in salt ion partition at theinterface can generate different interfacial properties capa-ble of causing or inhibiting bubble coalescence (Henry andCraig, 2010). In particular, the affinity of salt ions to watermolecules appears significant in determining the ion par-tition at the interface and the salt capability of inhibitingbubble coalescence. The Collins’ concept of matching wateraffinities of salt ions (Collins, 2004; Kunz, 2010; Ozdemiret al., 2011) can be applied here. The small (and hard)halide ions like Li+, Na+ and F− are strongly hydrated (kos-motropic), while the big (and soft) halide ions like K+, Rb+,Cs+, Cl−, Br− and I− are weakly hydrated (chaotropic). Ifthe constituent ions of a salt are matched in water affin-
ity (kosmotrope–kosmotrope and chaotrope–chaotrope), such
chemical engineering research and design 9 0 ( 2 0 1 2 ) 33–39 39
atdlicicw
5
TcicwdTastch
A
Tc
R
C
C
C
C
C
C
H
H
H
s in the case of NaF and CsCl, the ion partitions withinhe water–gas interface layer should not be significantlyifferent and the salt capability of inhibiting bubble coa-
escence is strong. Conversely, if salt ions are not matchedn water affinity (kosmotrope–chaotrope), such as in thease of NaI, the salt ion partitions within the water–gasnterface layer are not significantly different and the saltapability of inhibiting bubble coalescence in water becomeseak.
. Conclusion
he influence of gas superficial velocity, salt type and saltoncentration on bubble coalescence and gas holdup wasnvestigated. The transition concentration for bubble coales-ence inhibition was shown to be ion-specific and decreaseith increasing gas velocity, which was hypothesised to beue to the difference in salt ion partitions at the interface.he gas velocity had the strongest influence on Ctrans for NaInd weakest influence on the gas holdup for NaI. Our resultshow that both hydrodynamics and salt properties (the ionype and their combination) influence the transition salt con-entration for inhibition of bubble coalescence and the gasoldup.
cknowledgment
his research is supported under Australian Research Coun-il’s Discovery Projects funding scheme (Grant DP0985079).
eferences
han, B.S., Tsang, Y.H., 2005. A theory on bubble-size dependenceof the critical electrolyte concentration for inhibition ofcoalescence. J. Colloid Interface Sci. 286, 410–413.
hesters, A.K., Hofman, G., 1982. Bubble coalescence in pureliquids. Appl. Sci. Res. 38, 353–361.
hristenson, H.K., Bowen, R.E., Carlton, J.A., Denne, J.R.M., Lu, Y.,2008. Electrolytes that Show a Transition to BubbleCoalescence Inhibition at High Concentrations. J. Phys. Chem.C 112 (3), 794–796.
ollins, K.D., 2004. Ions from the Hofmeister series andosmolytes: eVects on proteins in solution and in thecrystallization process. Methods 34, 300–311.
raig, V.S.J., 2004. Bubble coalescence and specific-ion effects.Curr. Opin. Colloid Interface Sci. 9, 178–184.
raig, V.S.J., Ninham, B.W., Pashley, R.M., 1993. The effect ofelectrolytes on bubble coalescence in water. J. Phys. Chem. 97,10192–10197.
arvey, P.A., Nguyen, A.V., Evans, G.M., 2002. Influence ofelectrical double-layer interaction on coal flotation. J. ColloidInterface Sci. 250, 337–343.
enry, C.L., Craig, V.S.J., 2010. The link between ion specificbubble coalescence and Hofmeister effects is the partitioningof ions within the interface. Langmuir 26, 6478–6483.
enry, C.L., Parkinson, L., Ralston, J.R., Craig, V.S.J., 2008. A mobile
gas–water interface in electrolyte solutions. J. Phys. Chem. C112, 15094–15097.Hofmeier, U., Yaminsky, V.V., Christenson, H.K., 1995.Observations of solute effects on bubble formation. J. ColloidInterface Sci. 174, 199–210.
Jungwirth, P., Tobias, D.J., 2006. Specific ion effects at theair/water interface. Chem. Rev. 106, 1259–1281.
Klassen, V.I., Mokrousov, V.A., 1963. An introduction to the theoryof flotation. Butterworth.
Kunz, W., 2010. Specific ion effects in colloidal and biologicalsystems. Curr. Opin. Colloid Interface Sci. 15, 34–39.
Lessard, R.R., Zieminski, S.A., 1971. Bubble coalescence and gastransfer in aqueous electrolytic solutions. Ind. Eng. Chem.Fundam. 10, 260–269.
Liu, D., Ma, G., Levering, L.M., Allen, H.C., 2004. Vibrationalspectroscopy of aqueous sodium halide solutions andair–liquid interfaces: observation of increased interfacialdepth. J. Phys. Chem. B 108, 2252–2260.
Marcelja, S., 2006. Selective coalescence of bubbles in simpleelectrolytes. J. Phys. Chem. B 110, 13062–13067.
Marrucci, G., 1969. A theory of coalescence. Chem. Eng. Sci. 24,975–985.
Marrucci, G., Nicodemo, L., 1967. Coalescence of gas bubbles inaqueous solutions of electrolytes. Chem. Eng. Sci. 22,1257–1265.
Moshtari, B., Babakhani, E., Moghaddas, J., 2009. Experimentalstudy of gas hold-up and bubble behavior in gas–liquid bubblecolumn. Petroleum & Coal 51, 27–32.
Orvalho, S., Ruzicka, M.C., Drahos, J., 2009. Bubble column withelectrolytes: gas holdup and flow regimes. Ind. Eng. Chem.Fundam. 48, 8237–8243.
Ozdemir, O., Du, H., Karakashev, S.I., Nguyen, A.V., Celik, M.S.,Miller, J.D., 2011. Understanding the role of ion interactions insoluble salt flotation with alkylammonium and alkylsulfatecollectors. Adv. Colloid Interface Sci. 163, 1–22.
Parkinson, L., Sedev, R., Fornasiero, D., Ralston, J., 2008. Theterminal rise velocity of 10–100 mm diameter bubbles inwater. J. Colloid Interface Sci. 322, 168–172.
Pashley, R.M., 1981. Hydration forces between mica surfaces inaqueous electrolyte solutions. J. Colloid Interface Sci. 80,153–162.
Paulson, O., Pugh, R.J., 1996. Flotation of inherently hydrophobicparticles in aqueous solutions of inorganic electrolytes.Langmuir 12, 4808–4813.
Prince, M.J., Blanch, H.W., 1990a. Bubble coalescence andbreak-up in air-sparged bubble columns. AIChE J. 36,1485–1499.
Prince, M.J., Blanch, H.W., 1990b. Transition electrolyteconcentrations for bubble coalescence. AIChE J. 36, 1425–1429.
Pugh, R.J., Yoon, R.H., 1994. Hydrophobicity and rupture of thinaqueous films. J. Colloid Interface Sci. 163, 169–176.
Quinn, J.J., Kracht, W., Gomez, C.O., Gagnon, C., Finch, J.A., 2004.Comparing the effect of salts and frother (MIBC) on gasdispersion and froth properties. Miner. Eng. 20, 1296–1302.
Tsang, Y.H., Koh, Y.-H., Koch, D.L., 2004. Bubble-size dependenceof the critical electrolyte concentration for inhibition ofcoalescence. J. Colloid Interface Sci. 275, 290–297.
Weissenborn, P.K., Pugh, R.J., 1996. Surface tension of aqueoussolutions of electrolytes: relationship with ion hydration,oxygen solubility, and bubble coalescence. J. Colloid InterfaceSci. 184, 550–563.
Zahradnik, J., Fialova, M., Kastanek, F., Green, K.D., Thomas, N.H.,1995. The effect of electrolytes on bubble coalescence and gas
holdup in bubble-column reactors. Chem. Eng. Res. Des. 73,341–346.