1998
Ind. Eng. Chem. Res. 1998, 37, 1998-2004
GENERAL RESEARCH Effect of Surface Properties of Fine Particles on Dynamic Bubble Formation in Gas-Supersaturated Systems Z. A. Zhou,† Zhenghe Xu,*,† and J. A. Finch Department of Mining and Metallurgical Engineering, McGill University, Montreal, Canada
Bubble formation by releasing gas-(CO2-)supersaturated slurry of fine particles (d50 < 5 µm) from a pressure chamber was investigated. Gas holdup, defined as the percent volume of gas phase in the form of bubbles in the suspension released from the pressure chamber, was used to evaluate bubble formation. Compared to the result without solids, gas holdup decreased in the presence of hydrophilic silica; a small quantity of naturally hydrophobic coal increased gas holdup by increasing the number of gas nuclei and nucleation sites, while the presence of a large quantity of coal reduced the gas holdup, due to the coalescence of bubbles induced by the hydrophobic particles. For silica particles of chemically induced hydrophobicity in dodecylamine solutions, gas holdup increased sharply with solid content, reaching a maximum at ∼2% w/w solids. The results suggest that for hydrophobic particles to act as bubble nucleation sites, macroscopic contact angles do not have to be greater than 90° under dynamic conditions. Introduction The formation of bubbles in a gas-supersaturated liquid has been the subject of numerous investigations due to its importance in a variety of chemical, biochemical, mineral, coal, and wastewater treatment processes. The presence of solids in liquid has been found to affect bubble nucleation dramatically, posing a challenge to theoretical prediction. Harvey et al. (1944) suggested that the presence of gas nuclei, stabilized in hydrophobic crevices, caused bubble formation in water at lower threshold pressures than predicted. This theory has been adopted and extended by other investigators (Plesset, 1969; Winterton, 1977; Trevena, 1987; Lubetkin and Akhtar, 1996; Lee and Flumerfelt, 1996). However, it was found that while adding hydrophobic particles enhanced bubble formation only in some cases, in others it did not (Dean, 1944; Yount and Kunkle, 1975; Shafer and Zare, 1991). Recently, Wilt (1986) applied the classical nucleation theory in an integral form to predict nucleation rates in CO2-supersaturated aqueous solutions, and proposed two basic requirements for enhancing bubble nucleation rates: (i) the contact angle of water on a particle must be greater than 90°; (ii) only a cavity of conical shape can act as an active nucleation site while other geometries, such as smooth planar interfaces, conical projections, etc., cannot. Recent investigations by Ryan and Hemmingsen (1993) seem to support these two requirements. With 1-2.5 µm particles, they found that smooth surfaces, whether hydrophilic or hydrophobic, did not facilitate bubble formation by nitrogen dissolved in water at * Corresponding author:
[email protected]. Tel.: (403) 492-7667. Fax: (403) 492-2881. † Current address: Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Canada T6G 2G6.
supersaturation pressures up to 50-120 atm. However, Vinogradova et al. (1995) questioned whether the techniques they used could detect submicron cavities on a molecularly smooth surface. It is recognized that counting and measuring few tiny bubbles are intrinsically difficult, especially under static conditions and in the presence of fine solids, since gas diffusion in a liquid is relatively slow and the growth of a bubble from a nucleus to a detectable size may take a long time. A new technique for measuring the bubble nucleation rate was reported recently by Lubetkin and Akhtar (1996). This method exploits the acoustic event upon the release of bubbles from nucleation sites on solid surfaces. As a result, the technique only applies to the systems in which all the bubbles detach from the hydrophilic surface. Some submicron bubbles formed on hydrophobic solid surfaces may not detach, and the errors involved remain to be established. In addition, much of the previous work was conducted with a small amount of solids under static conditions. Therefore, any resultant macroscopic effect may not be easily detected by conventional methods. A different behavior of bubble nucleation in a mixture of molten polymer and volatile liquid under dynamic conditions, compared with static conditions, has been reported (Han and Han, 1988). To examine the solids’ effect on bubble nucleation and to simulate practical applications such as in dissolved air flotation or vacuum flotation, a study in the presence of a large quantity of solids under dynamic conditions is required. This communication discusses the effect of solid surface hydrophobicity on bubble formation in gas supersaturated systems under dynamic conditions. Experimental Section A pressure chamber made of stainless steel with an inner diameter of 10.5 cm and height of 15 cm was used (Figure 1). Compressed carbon dioxide (Produits de
S0888-5885(97)00489-2 CCC: $15.00 © 1998 American Chemical Society Published on Web 04/14/1998
Ind. Eng. Chem. Res., Vol. 37, No. 5, 1998 1999
Figure 1. Schematic setup used in bubble nucleation experiment.
Soudure Lte`e, Montreal) was introduced from a gas cylinder to the chamber via a Tygon tube (i.d. ) 3.5 mm) with punched holes of approximately 0.5 mm in diameter. Two release valves were installed, one for liquid (or slurry) and the other for gas. An electromagnetic valve with an inner diameter of 10 mm was used to control slurry release (i.e., on and off, as well as release time). Compared to the nozzle diameter (1.5 mm), the restriction of fluid (i.e., possible cavitation by the electromagnetic valve) is unlikely significant as illustrated in our baseline tests without a cavitation nozzle. A pressure gauge (ranging from 0 to 450 kPa) was used to indicate the pressure inside the chamber. In the center of the lid, there was a glass window which allowed the phenomena within the chamber to be observed. A magnetic stirrer was used to mix the slurry inside the chamber. The solids tested included fine silica (>99% SiO2 with a 50% weight passing size, d50 ) 1.6 µm) from U.S. Silica Co. and bituminous coal chunks (1.4% mineral matter) from Malaysia. The silica was used without further processing. The coal chunks were broken by a hammer and then dry-pulverized. The minus 38 µm (d50 ≈ 5 µm) portion was obtained by dry-screening and stored in a freezer at -18 °C to minimize oxidation. An advancing water contact angle larger than 65° was measured with this coal. Film flotation results showed (Xu et al., 1995) that the coal had a critical surface tension of 29.5 mN/m for complete wetting and 39.5 mN/m for nonwetting. The coal sample used in this work is therefore considered naturally hydrophobic. To stabilize the bubbles formed in the system and to change the particle surface properties, a nonionic surfactant, polypropylene glycol methyl ether, commercially known as Dowfroth 250 or DF250 (Dow Chemical Canada, Inc.), and a cationic surfactant, dodecylamine (DDA) (Prospect Chemicals Ltd., Fort Saskatchewan) were used, respectively. The stock DDA solution of 1.25 × 10-2 M was prepared with the addition of HCl. Both distilled and tap water were used in the experiments. All the experiments were conducted at room temperature which was relatively constant for each set of experiments. In each experiment, 500 mL of water was mixed with surfactant at a given dosage in a 600-mL glass beaker
Figure 2. Gas saturation procedure: (1) step saturation, slowly raising the saturation pressure with bubbling by immersing the gas inlet tubing in the liquid; (2) fast saturation, directly raising the pressure to the final value with bubbling; (3) direct pressurization, directly raising the pressure to the final value without bubbling (i.e., the gas inlet tubing is above the liquid).
for about 10 min, which was then placed in the pressure chamber for gas supersaturation with a prescheduled procedure (Figure 2) while the liquid (slurry) was stirred continuously. All the experiments were conducted using the “step supersaturation” procedure unless otherwise stated. After reaching the dynamic equilibrium (this usually takes 20 min), the pressure chamber was disconnected from the gas cylinder. The amount of suspension released for a given period of time was first calibrated for each supersaturation pressure and electromagnetic valve setting, which allowed an accurate amount of liquid or slurry to be released for each test. About 400 mL of the gas-supersaturated solution was released into a 345-mL glass vessel (i.d. ≈ 7 cm) and placed in a 2.5-L container which was used to collect overflow. It took about 10-18 s to fill the receiving vessel in comparison to 24-43 s to release 400 mL of slurry at a pressure of 313 and 100 kPa, respectively. By placing the releasing tube near the bottom of the 600-mL beaker which still contained about 100 mL of liquid or slurry in the pressure chamber, the entrain-
2000 Ind. Eng. Chem. Res., Vol. 37, No. 5, 1998
ment of any gas phase inside the pressure chamber during slurry release was avoided. The volume of water that overflowed into the 2.5-L container, retained in the 345-mL vessel and left in the 600-mL beaker, were all measured, with which the overall mass balance was checked. For few tiny bubbles formed in gas-supersaturated systems under static conditions, especially in the presence of solids, the measurement of bubble size and population is a difficult task. To examine the effect of different conditions on bubble formation under dynamic conditions and to simplify the analysis, a method used by Jordan and Spears (1990) was adopted to evaluate the macroscopic gas holdup defined as percent volume of gas phase in the form of bubbles in slurry. Recognizing that a large population of visible bubbles are generated by hydrodynamic cavitation as in dissolved air flotation, it is convenient to measure gas holdup. During each test, the gas holdup was measured by releasing the slurry into the 345-mL vessel. After all visible bubbles had risen and burst, and the foam collapsed (usually it takes about 5 min), the volume of the slurry was measured and the content of gas as bubbles per unit volume in the released slurry (i.e., the gas holdup, g) was estimated by a displacement method, which for the current setup is given by
g ) 100(345 - Vt)/345
(%)
Figure 3. Effect of solids on gas holdup. Solids: 5% silica or 2% coal by weight with 30 ppm DF250 and 20 min supersaturation time.
(1)
where Vt is the volume of slurry that remained in the 345-mL vessel after all visible bubbles disappeared. The repeating tests showed a reproducibility of (0.5% in gas holdup at a 95% confidence level. It should be noted that the gas holdup measured as such may underestimate the true gas holdup since the invisible gas nuclei were not considered. However, the error introduced is unlikely significant if there is any. The receiving vessels of different sizes (volume of 345 and 430 mL) were tested, and there was no difference in the measured gas holdup. The measured gas holdup using this technique reflects the combined effect of the number and size of bubbles in the released slurry. At a fixed gas-supersaturation pressure, the higher the gas holdup, the more bubbles are generated. The average size of bubbles formed in the absence of solids was estimated to be less than 200 µm using the Richardson-Zaki equation (Richardson and Zaki, 1954), with the data obtained using a conductivity method (Shen and Finch, 1996). Results Figure 3 shows the effect of coal and silica addition on gas holdup at different saturation pressures. It is noted that in the absence of solid particles gas holdup is similar in both tap water and distilled water, although one may expect more preexisting gas nuclei in tap water than in distilled water (Keller, 1972) due to the presence of impurities. This finding suggests that the original liquid does not contain a sufficient amount of gas nuclei to show any significant impact on overall bubble formation from a gas-supersaturated system under dynamic conditions. Also shown in Figure 3 is that adding solids, whether silica or coal, decreased gas holdup. In the case of coal at 2% solid (w/w), more than a 50% decrease in gas holdup was observed. Visual observation indicated that the behavior of the bubbles and the foam that formed in the presence of coal and silica was different. Com-
Figure 4. Effect of fine coal addition on gas holdup: 30 ppm DF250 and 20 min supersaturation time at 310 kPa.
pared to the case in the absence of solids, there were no obvious changes in bubble size and stability of the foam when silica was present, but much larger bubbles were observed and the foam disappeared within a much shorter period in the presence of coal. This observation suggests that the mechanism causing the decrease in gas holdup by the two solids studied is different. It should be noted that the flotation of hydrophobic coal cannot account for the observed difference as it occurred after hydrodynamic bubble formation and the slurry overflow velocity was greater than the rising velocity of bubble-particle aggregates. To confirm this, the effect of different amounts of coal particles on gas holdup was examined. Figure 4 shows that gas holdup increased initially with the addition of a small quantity of coal, reaching a maximum at a solid concentration of 0.05% (w/w). A further increase in the amount of coal particles decreased the gas holdup sharply, corresponding to the observed formation of large bubbles. This finding indicates that the coalescence of bubbles is induced by coal particles above a critical solid content (or surface area) and suggests that coalescence may be the dominant factor in modifying gas holdup, in contrast to the observations with hydro-
Ind. Eng. Chem. Res., Vol. 37, No. 5, 1998 2001
Figure 6. Setup for testing solids effect on gas holdup: (a) releasing gas-supersaturated slurry into the solution; (b) releasing gas-supersaturated solution into the slurry. Figure 5. Effect of silica addition on gas holdup in the presence of 1.25 × 10-4 M DDA or 1.2 × 10-4 M DF250 with 20 min supersaturation at 310 kPa.
philic silica, in which case gas holdup decreased continuously with increasing silica contents in the absence of DDA (see Figure 5). Figure 5 shows the effect of silica content on gas holdup in the presence of 1.25 × 10-4 M DDA or 30 ppm DF250. The pH of the released slurry in the receiving vessel was about 3 (natural pH of the suspension after CO2 supersaturation). In the presence of DDA, gas holdup initially increased sharply, from approximately 20% to more than 40% at 2% (w/w) solids. Further increase in solids addition reduced gas holdup, but it was still higher than that without solids. In the presence of DF250, gas holdup decreased progressively to about 10% with increasing solid content above 5% (w/w). It is known (Fuerstenau, 1957) that silica in the presence of DDA at pH ∼3 is moderately hydrophobic (the reported contact angle of about 15-20°). Apparently, preferential formation of gas nuclei on solids and the attachment of the particles to bubbles contributed to the increase in gas holdup, and the mineralized bubbles corresponded to the observed stability of foams on bulk slurry phase. To examine whether a reduced bubble/particle aggregate rising velocity could account for the increased gas holdup, two comparative tests were conducted (see Figure 6): (a) 500 mL of slurry with a given silica content in the pressure chamber was under gas supersaturation for 20 min, and then ∼400 mL was released into a 600-mL beaker containing 400 mL of solution of the same pH and chemical composition as the slurry in the pressure chamber; (b) 500 mL of solution of the same pH and chemical composition as in case (a) was gassupersaturated in the pressure chamber for 20 min and then released (∼400 mL) into a 600-mL beaker containing 400-mL slurry of the same silica content as well as the same pH and chemical composition as in case (a). The results in Figure 7 show that when releasing solution into a slurry (case b), there was essentially no solid effect on gas holdup. This finding suggests that a reduced bubble-rising velocity due to the attached particles was unlikely to be the cause for the observed increase in gas holdup by amine-induced hydrophobic particles. The effect of different gas supersaturation procedures (see Figure 2) on bubble formation is shown in Table 1.
Figure 7. Effect of slurry or solution supersaturation on gas holdup (1.25 × 10-4 M DDA and 20 min supersaturation at 310 kPa). Table 1. Effect of Different Saturation Procedures on Bubble Formation (P ) 313 kPa, 1% silica, and 1.25 × 10-4 M DDA) saturation procedure
saturation time (min)
gas holdup (%)
step fast direct pressurization
20 20 20 90
41.5 17.7 6.4 40.9
At a given pressure (313 kPa) and supersaturation time (20 min), raising the pressure directly from zero to the final value (i.e., the “fast saturation” procedure) produced a lower gas holdup than increasing the pressure slowly to the final level (i.e., “step saturation” procedure), both with bubbling. Also found is that increasing the supersaturation pressure without bubbling (i.e., “direct pressurization” procedure) resulted in a much lower gas holdup. However, this difference can be offset by using a longer saturation time (e.g., 90 min). Discussion Adding solids increases both the liquid density, F, and viscosity, η. If these were all that changed, Banisi et al. (1995) showed that the predicted effect would be a slight increase in gas holdup. The present observations that gas holdup can either decrease or increase, depend-
2002 Ind. Eng. Chem. Res., Vol. 37, No. 5, 1998
ing on the surface properties and the amount of solids added, cannot be explained satisfactorily by changes in these two physical parameters alone. They must, therefore, be also related to the events occurred during gas supersaturation inside the pressure chamber and during slurry release. In gas supersaturated systems, the bubble evolution in the presence of solids may be governed by the nucleation process, by nuclei growth, or by a combination of the two (Carr et al., 1995). Consequently, increasing the dissolved gas molecules and the number of gas nuclei may enhance bubble formation under dynamic conditions and give rise to a higher gas holdup. The factors affecting these two are discussed in the following. Interfacial Properties. Naturally Hydrophobic Particles (Coal). In the presence of naturally hydrophobic particles, the total number of gas nuclei is the sum of free gas nuclei in the water and the gas nuclei entrapped on the particle surface. When dry coal particles are mixed with water, a significant amount of gas nuclei are “entrapped”, due to the low wettability of (unoxidized) coal. In addition, some of the gas nuclei generated by the vigorous mixing during gas supersaturation may attach to coal particles in suspension and thus remain in the slurry inside the pressure chamber. These two contributions account for the observed increase in gas holdup at the low solids content (Figure 4). It is known (Jamialahadi and Muller-Steinhagen, 1991), however, that in gas/liquid/solid systems, hydrophobic particles tend to induce bubble coalescence as bubbles spread across the solid surface to achieve the characteristic contact angle of the system (Dippenaar, 1982). Such coalescence has been observed in “clean” water (i.e., surfactant-free water) (van Weert and Ruizendaal, 1995), or even in the presence of surfactant (Dippenaar, 1982; Cilliers and Bradshaw, 1996). In the later case, the increased bubble coalescence has been attributed to surfactant adsorption on hydrophobic particles by hydrophobic interaction, thus removing the surfactant from the water. One such example is a significant reduction of gas holdup in the presence of coal particles (15% (v/v), d84 ) 53 µm), compared to that without solids (Banisi et al., 1995), which correlated to a reduction in frother (nonionic surfactant) concentration in the liquid phase. Observed in the present study is the reduction in gas holdup by the addition of coal (6.5 ( 0.5%) below the gas holdup of water alone (without DF250 and coal addition), suggesting substantial bubble coalescence induced by hydrophobic particles. Naturally Hydrophilic Particles (Silica). According to Wilt (1986), the heterogeneous nucleation reduces to the homogeneous nucleation when the liquid contact angle on solids equals zero. Since there should be no change in the total amount of gas nuclei in water after addition of hydrophilic silica (in general, there is no entrapped gas nucleus by hydrophilic particles), the bubble formation should remain unaffected. However, the gas holdup reduced in the presence of hydrophilic silica. The thermodynamic analysis by Yount and Kunkle (1975) indicated that the probability of forming a bubble decreases in the presence of hydrophilic solids, which seems to agree with our findings, but the event of bubble formation is a nonequilibrium process and is
generally not regarded as amenable to thermodynamic treatment (Dean, 1944). One contributing factor to the reduction of gas holdup may be related to the formation of a solid-like water film on hydrophilic particles. Because of the strong interaction with silica, water molecules near the surface will exhibit some order, forming a solid-like film (Klassen and Mokrousov, 1963; Derjaguin and Churaev, 1987). Such a film has features distinct from bulk water, with increased viscosity and density, reduced dielectric permeability, and dissolving power (Churaev, 1991). The last point suggests a low gas solubility within this film. The consequence is the reduction of total free water in the system. The quantitative evaluation of this reduction is complicated by the uncertain thickness of solidlike water films (the reported value ranges from 100 Å to 0.1-0.3 µm (Klassen and Mokrousov, 1963; DrostHansen, 1969; Derjaguin and Churaev, 1987)). The exact reason for the reduction of bubble formation by hydrophilic silica remains to be resolved. Amine-Induced Hydrophobic Particle (Silica in DDA Solutions). In this case the stability of the foam that formed depended on the percent solids in the slurry. When the solid content was low (0.2 wt %), the foam was thin (∼1 cm) and unstable, only persisting for a few seconds as in the case of without solid addition. At higher solid concentrations the foam became thicker (up to ∼5 cm, i.e., about half of the receiving vessel was occupied by foam) and stable, persisting for several minutes. This indicates that the bubbles formed were stabilized by amine-induced hydrophobic silica particles, which was opposite to the case with naturally hydrophobic coal particles. Hydrophobic particles provide gas nucleation sites. The number of sites depends on the number of particles and the degree of hydrophobicity. In the absence of coalescence, increasing gas holdup with solids addition is expected. This is the situation here up to 2% solid (w/w), but with a further increase in solid content, gas holdup decreased. It should be noted that although the gas holdup reduces with increasing solid concentration above a given level in both systems (naturally hydrophobic coal and amine-induced hydrophobic silica), the visual observation was different for the two systems, suggesting the different mechanisms of the effect. In the case of amine-induced hydrophobic silica, there were no obvious changes in the stability of the foam at silica content >2% (w/w). For coal, large bubbles and fast foam collapse were observed, suggesting a coalescence control mechanism. An increased rate of gas evolution in water in the presence of surfactant-induced hydrophobic particles, compared to that without solids, was demonstrated by Klassen and Mokrousov (1963). This phenomenon appears to be related to the low wettability of these particles, resulting in the accumulation of gas molecules and nuclei at the interface. The ability of hydrophobic particles to “trap” gas nuclei is also evident in the following experiment with a slurry containing amineinduced hydrophobic silica. After step supersaturation and release of about 400 mL of slurry, the remaining slurry (∼100 mL) was kept in the beaker within the pressure chamber. The beaker was removed after releasing the pressure by opening the gas release valve (Figure 1), and virtually no bubbles were observed. However, shaking the beaker produced an abundance of bubbles, which formed a thick layer of foam. In
Ind. Eng. Chem. Res., Vol. 37, No. 5, 1998 2003
contrast, only a few bubbles formed in the absence of solids. A question arises as to how gas nucleates on amine-induced hydrophobic particles, with contact angles being only about 15-20° (Fuerstenau, 1957). According to heterogeneous nucleation theory (Wilt, 1986), among other requirements, a contact angle of greater than 90° is required for a surface to act as a nucleation site. Although the macroscopic contact angle of particles under the present conditions is low, it is possible that DDA molecules adsorb as localized clusters (hemimicelles) which exhibit higher local microscopic contact angles (Kitchener, 1984), reflecting the microscopic heterogeneity of almost all real mineral surfaces. Recent investigations by Carr et al. (1995) revealed a distribution of active nucleation sites on a surface. All these examples (also shown in our experiments) suggest that, for hydrophobic particles to act as nucleation sites, the macroscopic contact angles can be less than 90° especially under dynamic conditions. Dean (1944) proposed that although an adsorbed monolayer of gas on a particle surface would not be expected to promote bubble formation, there is considerable evidence that a thin gas film accumulated on the surface in supersaturated solutions may act as nuclei. Theoretical analysis (Yount and Kunkle, 1975) indicated that, under static conditions, the accumulation of gas molecules on hydrophobic particles to form a gas nucleus requires a high degree of supersaturation (∼1000 atm), which does not correspond to the situation here. Therefore, other mechanisms must exist whereby energy can be concentrated at a point to an extent that the attractive forces between neighboring liquid molecules can be overcome and a cavity be generated (Hayward, 1970). One possibility is the dynamic conditions used in this study, which is considered next. Stirring and Dissolved Gas. Under dynamic conditions, the turbulence generated by stirring causes pressure fluctuations so that at some locations the pressure drops to the water vapor pressure, and thus a water vapor cavity forms. The formation of a cavity would be facilitated in the presence of solid particles due to the pronounced changes in pressure around a particle. Therefore, some cavities may form at lowpressure regions on hydrophobic particles, into which dissolved gas would migrate. If the liquid is not saturated with gas, or the concentration of gas molecules in the liquid is low, the mechanical equilibrium may not be reached between the cavity and the surrounding liquid, and the incipient cavity will collapse. Therefore, the formation of bubbles assisted by mechanical agitation in gas-supersaturated systems is a function of local turbulence in a slurry and the concentration of dissolved gas. In addition to inducing pressure fluctuation, the continuous stirring employed during the saturation period also served to promote the mass transfer of gas into the slurry (dispersion and dissolution). This in turn increased the diffusion of gas molecules into any newly generated cavities, thus preserving them and resulting in a higher gas holdup for the same gas-supersaturation pressure, as compared to that without stirring. The results in Table 1 support this contention: a high gas holdup was obtained by either bubbling slurry (indicating efficient dispersion and dissolution of gas) or using a long saturation time. This suggests that stirring and bubbling enhanced the generation of cavities and that
a long saturation time allowed more gas molecules to dissolve and migrate into the cavities. More importantly, the fast release of slurry from the pressure chamber may have further enhanced cavity formation and gas diffusion, due to pressure fluctuations in the stream as predicted from Bernoulli’s equation. It was observed by Jackson (1994) and Zhou et al. (1995) that although the saturation pressures were similar, faster liquid release velocities produced more bubbles. Since gas holdup is a measure of the amount of gas bubbles in the system, the increased gas holdup in the liquid released from the pressure chamber suggests that gas nucleation was facilitated by hydrodynamic conditions. Formation of nuclei/small bubbles is the desired situation for most gas-supersaturation-based solid/liquid and mineral-separation processes such as dissolved air (gas) flotation and vacuum flotation. Conclusions Bubble formation in gas-supersaturated systems under dynamic conditions is affected by different gassaturation procedures, the number of gas nuclei, the total amount of dissolved gas, and the presence of solid particles. Gas holdup decreased upon the addition of hydrophilic silica, and the exact reason remains to be explored. In the presence of a small quantity (0.1% (w/w) coal, gas holdup reduced by up to 50%, which was attributed to the coalescence of bubbles by the hydrophobic solids. In the silica-DDA system, gas holdup increased initially with increasing solid content, more than double at 2% (w/w), compared with no solid addition. A further increase in solids content reduced gas holdup as in the case of naturally hydrophobic coal. Acknowledgment Financial support for this work was provided by the Natural Sciences & Engineering Research Council of Canada under Strategic Grant NSERC-STR0149414 and is gratefully acknowledged. The support of Inco, Cominco, and Noranda in obtaining the grant is gratefully acknowledged, as is the supply of silica by the U.S. Silica Co. Literature Cited Banisi, S.; Finch, J. A.; Laplante, A. R.; Weber, M. E. Effect of Solid Particles on Gas Holdup in Flotation Columns. Chem. Eng. Sci. 1995, 50 (14), 2329-2342. Carr, M. W.; Hillman, A. R.; Lubetkin, S. D. Nucleation Rate Dispersion in Bubble Evolution Kinetics. J. Colloid Interface Sci. 1995, 169, 135-142. Churaev, N. V. Surface Forces and Their Role in Mineral Processing. XVII International Mineral Processing Congress; Polygraphischer Bereich: Dresden, Germany, 1991; Vol. 2, pp 1-15. Cilliers, J. J.; Bradshaw, D. J. The Flotation of Fine Pyrite Using Colloidal Gas Aphrons. Miner. Eng. 1996, 9 (2), 235-241. Dean, R. B. The Formation of Bubbles. J. Appl. Sci. 1944, 15, 446451. Derjaguin, B. V.; Churaev, N. Structure of Water in Thin Layers. Langmuir 1987, 3 (5), 607-612. Dippenaar, A. The Destabilization of Froth by Solids. 1sThe Mechanism of Film Rupture. Inter. J. Miner. Process. 1982, 9, 1-14. Drost-Hansen, W. Structure of Water near Solid Interfaces Ind. Eng. Chem. 1969, 61 (11), 10-47.
2004 Ind. Eng. Chem. Res., Vol. 37, No. 5, 1998 Fuerstenau, D. W. Correlation of Contact Angles, Adsorption Density, Zeta Potentials, and Flotation Rate. Trans. AIME 1957, 208, 1365-1367. Han, J. H.; Han, C. D. A Study of Bubble Nucleation in a Mixture of Molten Polymer and Volatile Liquid in a Shear Flow Field. Polym. Eng. Sci. 1988, 28 (24), 1616-1627. Harvey, E. N.; Whiteley, A. H.; McElroy, W. D.; Pease, D. C.; Barnes, D. K. Bubble Formation in Animals. II. Gas Nuclei and Their Distribution in Blood and Tissues. J. Cell. Comput. Physiol. 1944, 24, 23-34. Hayward, A. T. J. The Role of Stabilized Gas Nuclei in Hydrodynamic Cavitation Inception. J. Phys. D: Appl. Phys. 1970, 3, 574-579. Jackson, M. L. Energy Effects in Bubble Nucleation. Ind. Eng. Chem. Res. 1994, 33 (4), 929-933. Jamialahadi, M.; Muller-Steinhagen, H. Effect of Solid Particles on Gas Holdup in Bubble Columns. Can. J. Chem. Eng. 1991, 69, 390-393. Jordan, C. E.; Spears, D. R. Evaluation of a Turbulent Flow Model for Fine-Bubble and Fine-Particle Flotation. Miner. Metall. Process. 1990, May, 65-73. Keller, A. The Influence of the Cavitation Nucleus Spectrum on Cavitation Inception, Investigated with a Scattered Light Counting Method. J. Basic Eng. 1972, 94, 917-925. Kitchener, J. A. Surface Forces in FlotationsA Critique. In Principles of Mineral Flotation; Jones, M. H., Woodcock, J. T., Eds.; The Australasian IMM: Australia, 1984; pp 65-72. Klassen, V. I.; Mokrousov, V. A. An Introduction to the Theory of Flotation; Butterworths: London, 1963. Lee, J. G.; Flumerfelt, R. W. A Refined Approach to Bubble Nucleation and Polymer Foaming Process: Dissolved Gas and Cluster Size Effects. J. Colloid Interface Sci. 1996, 184, 335348. Lubetkin, S. D.; Akhtar, M. The Variation of Surface Tension and Contact Angle under Applied Pressure of Dissolved Gases, and the Effects of These Changes on the Rate of Bubble Nucleation. J. Colloid Interface Sci. 1996, 180, 43-60. Plesset, M. S. The Tensile Strength of Liquids. In Cavitation State of Knowledge; Robertson, J. M., Wislicenus, G. F., Eds.; The American Society of Mechanical Engineers: New York, 1969; pp 15-25.
Richardson, J. F.; Zaki, W. N. Sedimentation and Fluidization: Part 1. Trans. Inst. Chem. Eng. 1954, 32, 35-52. Ryan, W. L.; Hemmingsen, E. A. Bubble Formation in Water at Smooth Hydrophobic Surfaces. J. Colloid Interface Sci. 1993, 157, 312-317. Shafer, N.; Zare, R. N. Through a Beer Glass Darkly. Phys. Today 1991, Oct, 48-52. Shen, G.; Finch, J. A. Bubble Swarm Velocity in a Column. Chem. Eng. Sci. 1996, 51, 3665-3674. Trevena, D. H. Cavitation and Tension in Liquids; Bristol: Adam Hilger, 1987; p 12. van Weert, G.; Ruizendaal, A. E. Effect of Hydrophobic Solids on Bubble Size and Oxygen Transfer in an Air-Sparged Column. In Processing of Hydrophobic Minerals and Fine Coal; Laskowski, J. S., Poling, G. W., Eds.; CIM: Canada, 1995; pp 387-412. Vinogradova, O. I.; Bunkin, N. F.; Churaev, N. Y.; Kiseleva, O. A.; Lobeyev, A. V.; Ninham, B. W. Submicrocavity Structure of Water between Hydrophobic and Hydrophilic Walls as Revealed by Optical Cavitation. J. Colloid Interface Sci. 1995, 173, 443447. Wilt, P. M. Nucleation Rates and Bubble Stability in WaterCarbon Dioxide Solutions. J. Colloid Interface Sci. 1986, 112 (2), 530-538. Winterton, R. H. S. Nucleation of Boiling and Cavitation. J. Phys. D: Appl. Phys. 1977, 10, 2041-2056. Xu, Z.; Zhou, Z. A.; Liu, Q.; Yordan, J. L. Role of Interfacial Phenomena in Fine Coal Dewatering. In Processing of Hydrophobic Minerals and Fine Coal; Laskowski, J. S., Poling, G. W., Eds.; CIM: Canada, 1995; pp 513-525. Yount, D. E.; Kunkle, T. D. Gas Nucleation in the Vicinity of Solid Hydrophobic Spheres; J. Appl. Phys. 1975, 46 (10), 4484-4486. Zhou, Z. A.; Xu, Z.; Finch, J. A. Fundamental Study of Cavitation in Flotation. XIX International Mineral Processing Congress; SME: Littleton, CO, 1995; Vol. 3, pp 93-97.
Received for review July 16, 1997 Revised manuscript received February 19, 1998 Accepted February 19, 1998 IE970489D
