pubs.acs.org/Langmuir © 2009 American Chemical Society
Effect of Adsorbed Polymers on Bubble-Particle Attachment Audrey Beaussart, Luke Parkinson, Agnieszka Mierczynska-Vasilev, John Ralston, and David A. Beattie* Ian Wark Research Institute, ARC Special Research Centre for Particles and Material Interfaces, University of South Australia, Mawson Lakes, SA 5095, Australia Received August 24, 2009. Revised Manuscript Received October 13, 2009 The influence of adsorbed dextrin-based polymers on the attachment of a rising air bubble to a talc surface has been investigated. Liquid film rupture and dynamic contact angle studies have highlighted the major role that adsorbed polymers can play in bubble-particle attachment. No direct link was established between the equilibrium contact angle of polymer-treated talc surfaces and talc flotation recovery. However, clear correlations were observed between the flotation recovery of polymer-treated talc and the measured wetting film rupture time and rate of dewetting for a bubble attaching to a talc basal plane surface treated with the polymers. The retardation of the three-phase contact line expansion caused by the adsorbed polymers was found to have the largest influence on the bubble-particle attachment. The effect of the morphology (coverage, distribution, and shape) of the adsorbed layer on the wetting film rupture and the motion of the receding water front is discussed.
Introduction The flotation process for the separation of minerals is optimized by altering the hydrophobicity of the different solid phases present in the system, usually by addition of reagents into the mineral suspension.1,2 Chemicals, such as surfactants and polymers, are added to obtain the differences in wettability necessary for selective separation of the minerals. Until now, the influence of polymers (termed depressants as they depress/reduce the recovery of minerals) has been probed by measuring the enhancement in wettability and the decrease of flotation recovery of the unwanted minerals when increasing the amount of adsorbed polymer.3-11 However, reduction of the final contact angle is only one aspect of the role that polymers play in the prevention of bubble-particle attachment. The successful attachment of a hydrophobic mineral particle to an air bubble relies on a number of subprocesses.12,13 First, the liquid film between the particle and the bubble will drain to a critical thickness. The thin film will then rupture, creating a threephase contact line (TPCL), which will expand until a stable wetting perimeter is formed. The time for the subprocesses to take place (thinning, rupture, and TPCL movement during the dewetting of the surface) defines the so-called induction time. *Corresponding author. E-mail:
[email protected]. (1) Pugh, R. J. Int. J. Miner. Process. 1989, 25, 101–130. (2) Bulatovic, S. M. Handbook of flotation reagents: chemistry, theory and practice, 1st ed.; Elsevier: Boston, 2007. (3) Beattie, D. A.; Huynh, L.; Kaggwa, G. B.; Ralston, J. Int. J. Miner. Process. 2006, 78, 238–249. (4) Mierczynska-Vasilev, A.; Ralston, J.; Beattie, D. A. Langmuir 2008, 24, 6121–6127. (5) Bacchin, P.; Bonino, J. P.; Martin, F.; Combacau, M.; Barthes, P.; Petit, S.; Ferret, J. Colloids Surf., A 2006, 272, 211–219. (6) Shortridge, P. G.; Harris, P. J.; Bradshaw, D. J.; Koopal, L. K. Int. J. Miner. Process. 2000, 59, 215–224. (7) Ansari, A.; Pawlik, M. Miner. Eng. 2007, 20, 609–616. (8) Ansari, A.; Pawlik, M. Miner. Eng. 2007, 20, 600–608. (9) Bulatovic, S. M. Miner. Eng. 1999, 12, 341–354. (10) Rath, R. K.; Subramanian, S.; Laskowski, J. S. Langmuir 1997, 13, 6260– 6266. (11) Subramanian, S.; Laskowski, J. S. Langmuir 1993, 9, 1330–1333. (12) Nguyen, A. V.; Schulze, H. J.; Ralston, J. Int. J. Miner. Process. 1997, 51, 183–195. (13) Yoon, R. H. Int. J. Miner. Process. 2000, 58, 129–143.
13290 DOI: 10.1021/la903145h
Formation of a stable particle-bubble aggregate requires a collision time that is longer than the induction time.14 Therefore, a number of studies have been devoted to measurements of the factors influencing the kinetics of the subprocesses of the particle-bubble attachment. A detailed investigation on the role of surface roughness and hydrophobicity on the stability of the wetting film has been recently conducted.15,16 In addition, the kinetics of the motion of the TPCL has been studied experimentally and theoretically.17,18 Although an adsorbed polymer layer can considerably alter the characteristics of a mineral surface, the influence of polymers on induction time for hydrophobic mineral surfaces has not been experimentally studied. Adsorbed surfactant layers have been studied in their ability to slow down thin film rupture and TPCL dewetting movement by Tjus et al.19 for bubbles pressed against model hydrophobic surfaces (consisting of LB films and methylated quartz surfaces) using the methodology of Scheludko et al.20 In addition, wetting and dewetting of polymer-coated surfaces have been studied theoretically21 and experimentally.22,23 The work presented in this Letter is the first direct measurement of all stages of bubble-particle attachment for a real hydrophobic mineral surface in the absence and presence of polymer depressants. The work leads on directly from our earlier investigation of the equilibrium adsorption properties of three dextrin-based polymers on talc (studied using in situ atomic force microscopy (14) Schulze, H. J.; Radoev, B.; Geidel, T.; Stechemesser, H.; T€opfer, E. Int. J. Miner. Process. 1989, 27, 263–278. (15) Krasowska, M.; Zawala, J.; Malysa, K. Adv. Colloid Interface Sci. 2009, 147-148, 155–169. (16) Krasowska, M.; Malysa, K. Int. J. Miner. Process. 2007, 81, 205–216. (17) Nguyen, A. V.; Alexandrova, L.; Grigorov, L.; Jameson, G. J. Miner. Eng. 2006, 19, 651–658. (18) Phan, C. M.; Nguyen, A. V.; Evans, G. M. Langmuir 2003, 19, 6796–6801. (19) Tjus, K.; Pugh, R. J.; Herder, P.; Eriksson, J. C.; Stenius, P. Colloids Surf. 1988, 34, 95–99. (20) Sheludko, A.; Chal’ovska, S.; Fabrikant, A. Spec. Discuss. Faraday Soc. 1970, 1, 112–17. (21) Halperin, A.; De Gennes, P. G. J. Phys. (Paris) 1986, 47, 1243–7. (22) Muller, P.; Sudre, G.; Theodoly, O. Langmuir 2008, 24, 9541–9550. (23) Cohen Stuart, M. A.; De Vos, W. M.; Leermakers, F. A. M. Langmuir 2006, 22, 1722–1728.
Published on Web 10/27/2009
Langmuir 2009, 25(23), 13290–13294
Beaussart et al.
Letter
imaging) and the effect of the polymer on the static contact angle of polymer-treated talc particles.4 The data presented here increase our understanding of the depression mechanism of polymers by identifying their effect on each elementary step of the bubble-particle attachment process and highlight the crucial importance of considering the dynamic properties of the interface. The selection of freshly cleaved talc as the solid enables us to correlate our observations directly with mineral flotation data. Links between the morphology and physical properties of the adsorbed polymer layer and the rupture/TPCL expansion time are also established.
Experimental Section Materials. Three polysaccharide-based polymers supplied by Penford Australia were used in this study: Dextrin TY (regular wheat dextrin), carboxymethyl (CM) Dextrin, and hydroxypropyl (HP) Dextrin. The general structure of dextrin and the substituted groups (primarily at C6 on the glucopyranose rings) is depicted in Figure 1. CM and HP Dextrin have a degree of substitution less than 10% (i.e., D.S. < 0.3). The molecular weight averages, determined by size exclusion chromatography (SEC),3 are 5000, 34 000, and 64 000 g mol-1 for Dextrin TY, CM Dextrin, and HP Dextrin, respectively. Solid polymer samples, as received from Penford, were used to prepare stock solutions of 2000 mg L-1. The appropriate mass of solid polymer was dissolved in background electrolyte solution and stirred overnight to ensure complete hydration of the dextrin molecules. CM Dextrin was prepared using a gelling procedure developed by Penford. A solid paste was formed by adding a small amount of Milli-Q water to the appropriate mass of polymer powder. A 2% KOH solution was then slowly added to the stirring mixture until the formation of a gel. Once gelled, the background solution was slowly added until the desired concentration was obtained. All polymer solutions were optically clear. Talc particles were purchased from Merck, Germany (>99% pure). The Brunauer-Emmett-Tellet (BET) surface area was measured at 2.9 m2 g-1. The apparent particle size distribution, determined using a Malvern Instrument Mastersizer apparatus, was 0.5-100 μm, with D10 of 3.5 μm, D50 of 15 μm, and D90 of 52 μm. Talc rock mineral from Delaware (U.S.A.) was obtained from the Mineralogy Department of the South Australian Museum. Freshly cleaved talc surfaces were obtained by cleaving the top layer of a flat section of the mineral sample by gently adhering and peeling a piece of sticky tape. Particles and flat mineral surface exhibit negligible quantities of impurities as determined by X-ray photoelectron spectroscopy (XPS) analysis.24 Induction Time and Contact Angle Measurements. Induction time and receding contact angle measurements of bare talc and polymer-treated talc surfaces were determined using the apparatus depicted in Figure 2. Freshly cleaved talc samples were adhered using a cyanoacrylate glue to a glass slide and immersed in polymer solutions of the desired concentration for 30 min prior to measurements. The talc samples were then gently rinsed with background electrolyte prior to attachment to the holder and subsequent immersion into a silica and poly(tetrafluoroethylene) (PTFE) cuvette containing 200 cm3 of 10-2 M KCl electrolyte, maintained at pH 9. These sample preparation conditions mirror those used in our earlier atomic force microscopy investigation of these polymers adsorbed on talc (adsorption and then solution replacement with electrolyte).4 In addition, the studied polymers do not adsorb to the liquid-vapor interface (confirmed with dynamic surface tension measurements from pendant drop shape analysis using a Dataphysics OCAH200 instrument) so their removal from solution will not alter their effect on bubbleparticle attachment.
A tunable microbubble device was used to release single bubbles of ultrapure N2 gas (BOC Gases) of diameter 120 ( 5 μm. Bubbles rose through the center of the cuvette at terminal velocity for 50 mm25 prior to interaction with the talc surface. The film drainage, rupture, and dewetting processes were observed by stereomicroscope (Olympus SZ4511TR) and recorded via highspeed CCD camera (Photron 1024 PCI) recording at 1000-27 000 Hz, depending on the surface. The entire apparatus was isolated on an antivibration table. Due to the very gradual onset of hydrodynamic resistance due to the solid surface, the determination of a true zero for the start of film drainage (where the bubble first “senses” the solid) is difficult. Therefore, the start of film drainage was taken as the time at which each bubble is half its diameter away from the surface. This allowed for a consistent determination of the onset of film thinning to a precision of HP Dextrin (approximately 84°) > Dextrin TY (approximately 88°). The small reduction in contact angle for these polymers on talc is similar to measurements of a maize-based dextrin on talc and molybdenite using the captive bubble methodology.24 Particle film contact angle measurements for the TY, CM, and HP Dextrin on talc4 produce larger changes in the measured receding contact angle, but these data are complicated by the influence of the talc edges present when dealing with particulate talc samples (the edge surface of talc accounts for approximately 10% of the surface area and is hydrophilic), and the effects of sample drying prior to analysis which may not be ideal for analysis of minerals with polymer layers that retain a high percentage of hydration water when adsorbed (e.g., polysaccharide polymers4,27). Desorption of the polymer upon dewetting does not occur for these polymers on talc, as ex situ AFM imaging demonstrates,28,29 and cannot be used to account for the small reduction in talc hydrophobicity. (27) Hedin, J.; Loefroth, J.-E.; Nyden, M. Langmuir 2007, 23, 6148–6155. (28) Beattie, D. A.; Huynh, L.; Mierczynska-Vasilev, A.; Myllynen, M.; Flatt, J. Can. Metall. Q. 2007, 46, 349–358. (29) Kaggwa, G. B.; Huynh, L.; Ralston, J.; Bremmell, K. Langmuir 2006, 22, 3221–3227.
DOI: 10.1021/la903145h
13293
Letter
Beaussart et al.
Discussion The data presented in this work highlight the dominant role of wetting film rupture and three phase contact line movement on the flotation recovery of hydrophobic minerals. However, the data do not fully explain why the three polymers have such dramatic effects on bubble-surface and bubble-particle attachment, and why there is such contrast in effect between the three polymers. The effect of the polymer on thinning/rupture time can be linked to the adsorbed polymer layer morphology reported in our earlier work4 and the contact angle data obtained in this work. Dextrin TY was found to adsorb as randomly dispersed hemispheres with significant gaps between the polymer domains. This morphology appears to result in as fast a rupture of the liquid film as for bare talc, with the rapid rupture most likely due to the large domain-domain separation distances that expose large areas of the hydrophobic talc surface (also reflected in the small hydrophobicity reduction for Dextrin-TY-treated talc). In contrast, CM and HP Dextrin appear to be equally efficient in slowing thin film rupture. The high surface area coverage for CM Dextrin and small domain spacing between polymer domains for HP Dextrin (exposing only small gaps of bare talc to the approaching air bubble) appear to stabilize the wetting film, presumably through alteration of the bubble-surface forces.30 In terms of TPCL movement, the modest reductions in the receding contact angles as a result of the presence of the adsorbed layers argue against hydrophobicity reduction being behind the slower dewetting of the treated talc surface.31 It is therefore likely that the adsorbed layer morphology is also behind the difference in TPCL expansion time. HP Dextrin and Dextrin TY both adsorb as hemispherical domains randomly dispersed on bare talc. The number of domains was found to be higher in the case of HP Dextrin, with smaller interdomain regions of bare talc. Both polymers retard the movement of the TPCL, presumably through pinning of the contact line as it encounters each polymer domain.32 However, it is likely that the TPCL would move more smoothly on the Dextrin-TY-coated surface due to the large areas of exposed talc, whereas the numerous protrusions created by adsorption of HP Dextrin would probably generate a greater degree of pinning of the TPCL and slow down the movement of the receding water front. This interpretation is supported by experiments on Langmuir-Blodgett monolayers and multilayers of carboxylic acids by Semal et al.,33 where surfaces with increased microroughness gave rise to greater contact angle hysteresis and enhanced slowing of TPCL movement. Patchy coating and pinning of contact lines cannot explain the effect of CM Dextrin on the speed of movement of the dewetting front, as CM Dextrin does not have the same morphology on talc as the other two polymers; it forms a near-complete layer, with small pits. These pits will not greatly affect the TPCL movement (30) Pugh, R. J. Int. J. Miner. Process. 1989, 25, 131–146. (31) Fetzer, R.; Ralston, J. J. Phys. Chem. C 2009, 113, 8888–8894. (32) Priest, C.; Sedev, R.; Ralston, J. Phys. Rev. Lett. 2007, 99. (33) Semal, S.; Blake, T. D.; Geskin, V.; De Ruijter, M. J.; Castelein, G.; De Coninck, J. Langmuir 1999, 15, 8765–8770.
13294 DOI: 10.1021/la903145h
and are unlikely to be responsible for the slowing of the TPCL. There are two potential explanations behind the observation of a complete layer slowing down the dewetting. First, if the adsorbed polymer has some affinity for the liquid-vapor interface, bridging between the polymer on the surface and the liquid-vapor interface may cause some slowing of the dewetting front.23 This explanation cannot apply in this case, as none of the polymers are surface active for the liquid-vapor interface (see Experimental Section). Second, it is possible that the complete film presents some friction for the TPCL as it moves over the surface, due to its soft, deformable nature. When a liquid film recedes on a soft substrate, the wetting front is slowed down due to the imbalance of the forces at the surface and the energy dissipated by the strain/ relaxation cycle to which the surface is exposed34-37 (i.e., for a dynamic contact angle smaller or larger than 90°, there will be a component to the surface tension normal to the surface which can cause deformation on soft substrates during the TPCL movement). In the case of CM Dextrin, even though the layer is only a few tens of nanometres in thickness, the same mechanism may well operate, as the adsorbed layer will be softer than the talc substrate, in analogy to the soft deformable polysaccharide layers adsorbed from aqueous solution on hydrophobic and hydrophilic surfaces as observed using quartz crystal microbalance with dissipation monitoring (QCM-D) measurements.27
Conclusions A bubble-surface collision apparatus has been used to probe the effect of adsorbed dextrin-based polymers on the subprocesses of bubble-surface attachment. The data were compared with froth flotation experiments. The time of thin film rupture and dewetting for polymer-treated talc were found to be in excellent agreement with the effectiveness of the polymers in reducing talc flotation recovery, highlighting the major influence of adsorbed polymer on the dynamics of the bubble-particle attachment process. This study has highlighted the importance of studying polymer effects on dynamic dewetting of mineral surfaces. Acknowledgment. Financial support for this study was received from the Australian Research Council and AMIRA International sponsors of Project P498B, which include Penford Australia, Cytec, Anglo Platinum, CP Kelco, Rio Tinto, and Xstrata. Russell Schumann of Levay and Co. Environmental Services is acknowledged for his for assistance with SEC measurements. The authors would like to thank Professor Allan Pring (South Australian Museum, Adelaide) for the provision of the talc sample and Iliana Sedeva for pendant drop analysis of polymer solutions. A.B. thanks the Ian Wark Research Institute and AMIRA for her scholarship support. (34) Shanahan, M. E. R.; Carre, A. Langmuir 1995, 11, 1396–1402. (35) Long, D.; Ajdari, A.; Leibler, L. Langmuir 1996, 12, 5221–5230. (36) Tomasetti, E.; Rouxhet, P. G.; Legras, R. Langmuir 1998, 14, 3435–3439. (37) Voue, M.; Rioboo, R.; Bauthier, C.; Conti, J.; Charlot, M.; De Coninck, J. J. Eur. Ceram. Soc. 2003, 23, 2769–2775.
Langmuir 2009, 25(23), 13290–13294
