Using sound to study bubble coalescence

https://doi.org/10.1016/j.jcis.2008.12.025Get rights and content

Abstract

Frothers are surfactants used in flotation to aid generation of small bubbles, an effect attributed to coalescence prevention. Studying coalescence at the moment of bubble creation is a challenge because events occur over a time frame of milliseconds. This communication introduces a novel acoustic technique to study coalescence as bubbles are generated at a capillary. The sound signal was linked to bubble formation and coalescence events using high-speed cinematography. The technique has the resolution to detect events that occur within 1–2 ms. The results show that for common flotation frothers and n-alcohols (C4–C8) coalescence prevention is not simply related to surface activity. A total stress model is used to give a qualitative explanation to the action observed. Results for salt (sodium chloride) are included for comparison.

Graphical abstract

A novel acoustic technique to study coalescence as bubbles are generated at a capillary is presented. Results of coalescence prevention in presence of surfactants and salt are also included.

  1. Download : Download full-size image

Introduction

Flotation is widely used to separate particles according to their hydrophobicity. Separation is achieved by dispersing air into bubbles that collide with and attach to hydrophobic particles. Water is not an easy media in which to generate small bubbles and it is necessary to add reagents called frothers (surfactants) in most flotation systems. Thus, the fine bubble size (typically 0.1–4 mm) results from the interaction of the air delivery system (e.g., through the impeller in a mechanical flotation machine) and frother.

Even though bubbles drive flotation performance, little is known about how frothers act in their production. Analysis of froth formation and stability emphasizes mechanisms that retard coalescence [1], [2]. This coalescence prevention explanation has been extended to bubble generation [3].

At the moment of bubble formation the surface concentration (adsorption density) of surfactant is likely well below the equilibrium value but it is not zero. Forces arising from the presence of surfactant can therefore be anticipated. While frothers are the main reagent used in fine bubble production, some salts at high concentration (>0.4 ionic strength) also produce fine bubbles in flotation systems [4]. A study of both frothers and salts should help identify the mechanism(s).

To study coalescence at the moment of bubble production demands measurements over the time frame of milliseconds. In this paper we introduce a novel approach to the problem using the sound bubbles emit when formed [5] and employ a stress model to qualitatively interpret the findings.

Section snippets

Total stress

The final act of coalescence is thinning to rupture of the liquid (water) film between contacting bubbles. In the presence of surfactants forces opposing film thinning are generated. These forces, or ‘stress,’ may be seen as the resistance of the film to deformation. One component of stress derives from the surface tension gradients that accompany disturbances in the distribution of surfactant, Gibbs–Marangoni elasticity (E) [6], [7]; a second component derives from the reduction in surface

Apparatus

The experimental set-up (Fig. 1) comprises a 30 L acrylic tank where air bubbles are injected through a glass capillary tube. Gas flow rate is measured and regulated with a mass flow meter controller (Sierra, model 840DL1V1). The acoustic emissions are measured with a hydrophone (Lab-core System), which has a wide frequency range, 5 to 85,000 Hz. The signal passes from the hydrophone to an amplifier before being transferred to a computer. The acoustic emissions were recorded with the freeware

Validation

The method was validated by testing the Minnaert equation [Eq. (4)]. Bubbles of diameter 2.4 and 6.0 mm were generated in water. The signals were processed by Fourier analysis to determine the peak frequency corresponding to each bubble size (Minnaert frequency). Fig. 2 shows the result.

The peak frequencies were 2583 Hz for the 2.4 mm bubbles and 1033 Hz for the 6.0 mm bubbles. From Eq. (4) the product ‘fd’ should be constant: this is the case, 6.199 mm/s vs. 6.198 mm/s, respectively. Taking

Reliability

The procedure calls for establishing the gas rate at the transition from non-coalescence to coalescence, i.e., identifying the transition from (a) to (b) in Fig. 9. Three full repeat tests were conducted for pentanol to establish the precision. Fig. 10 shows the results, including error bars representing the standard deviation. The boundary divides coalescence events that occur above the line from the non-coalescence region below the line: the larger the region below the line the more effective

The method

The challenge of recording bubble formation and associated coalescence events over 1–2 ms was met by an acoustic technique. It provided an effective alternative to imaging, reducing storage requirements and avoiding tedious inspection of videos at reduced equipment cost (by about 95% in the present case). The method also lends itself to trying on opaque systems, for example, solid suspensions relevant to flotation studies.

Inducing coalescence by increasing gas rate at a capillary was judged

Conclusions

The characteristic sound trace found when bubbles coalesce at a capillary tip allowed development of a novel technique to study bubble coalescence. The technique uses a hydrophone to record sounds, which are shown to be related to coalescence events. Signal recognition is straightforward and the technique has a resolution high enough to discriminate events that occur within 1 to 2 ms.

The gas flow rate marking the boundary between coalescence and non-coalescence was determined as a function of

Acknowledgments

This work was conducted under the Chair in Mineral Processing funded through an NSERC (Natural Sciences and Engineering Research Council of Canada) CRD (Collaborative Research and Development) grant now sponsored by Vale Inco, Xstrata Process Support, Teck Cominco, Agnico-Eagle, COREM, SGS Lakefield Research, Shell Canada and Flottec.

W. Kracht would also like to thank the Chilean Government for the Chilean National Scholarship (Beca Presidente de la República) and the Universidad de Chile for

References (40)

  • J.J. Quinn et al.

    Miner. Eng.

    (2007)
  • U. Hofmeier et al.

    J. Colloid Interface Sci.

    (1995)
  • J. Lucassen et al.

    Chem. Eng. Sci.

    (1972)
  • E. Dickinson

    Colloids Surf. B Biointerfaces

    (1999)
  • H. Schott

    J. Pharm. Sci.

    (1995)
  • J.A. Finch et al.

    Miner. Eng.

    (2006)
  • L. Wang et al.

    Colloids Surf. A Physicochem. Eng. Aspects

    (2006)
  • J.A. Finch et al.

    Miner. Eng.

    (2008)
  • P.K. Weissenborn et al.

    J. Colloid Interface Sci.

    (1996)
  • P. Dawson

    Eur. J. Radiol.

    (2002)
  • G. Drogaris et al.

    Chem. Eng. Sci.

    (1983)
  • G. Keitel et al.

    Chem. Eng. Sci.

    (1982)
  • B.A. Comley et al.

    Int. J. Miner. Proc.

    (2002)
  • S. Gélinas et al.

    Miner. Eng.

    (2007)
  • Y.S. Cho et al.

    Int. J. Miner. Proc.

    (2002)
  • R. Tuckermann

    Atmos. Environ.

    (2007)
  • V. Machon et al.

    Trans. IChemE

    (1997)
  • R.A. Grau et al.

    Int. J. Miner. Proc.

    (2005)
  • P.J. Harris
  • R.J. Pugh

    Adv. Colloid Interface Sci.

    (1996)
  • Cited by (39)

    • Review on research progress in boiling acoustics

      2022, International Communications in Heat and Mass Transfer
    • Total petroleum hydrocarbon removal from oil-sand by injecting charged nanobubbles

      2022, Environmental Nanotechnology, Monitoring and Management
    • Passive acoustic emission monitoring to detect bubble coalescence in the presence of solid particles

      2017, Minerals Engineering
      Citation Excerpt :

      Unlike the work of Kracht and Finch who found no partial coalescence in the presence of frother the present study reveals a small region of partial coalescence in the MIBC system. At MIBC concentrations below 0.1 mM (10 ppm), the Kracht and Finch (2009) data show a lower transition flow rate while at higher concentrations the literature transition data generally fall in the partial coalescence region. Fig. 5 shows the results for increasing sodium chloride concentration.

    View all citing articles on Scopus
    View full text