J.L. Parker, P.M. Claesson, and P. Attard, 'Bubbles, Cavities, and the Long-Ranged Attraction between Hydrophobic Surfaces', J. Phys. Chem. 98, 8468-8480 (1994).
See A. Carambassis, L. C. Jonker, P. Attard, and M. W. Rutland, ``Forces Measured Between Hydrophobic Surfaces due to a Sub-microscopic Bridging Bubble'', Phys. Rev. Lett. 80, 5357-5360 (1998).
Also Physical Review Letters Focus, `Small Attractive Bubbles'
See also James W. G. Tyrrell and P. Attard, ``Images of Nanobubbles on Hydrophobic Surfaces and their Interactions'', Phys. Rev. Lett. 87, 176104 (2001).
and
James W. G. Tyrrell and P. Attard,
``Atomic Force Microscope Images of Nanobubbles on a Hydrophobic Surface
and Corresponding Force-Separation Data'',
Langmuir (in press)
What is colloid science, and why are surface forces important?
Of particular interest are the so-called surface forces that act between colloid particles. Colloid science is a branch of physical chemistry that deals with the behavior of small particles in fluids. The colloids are typically between 1 micrometer and 10 nm in size, commonly solids (but in emulsions they are liquid drops, and in foams and froths they are bubbles) and the continuous phase is usually liquid (but it can be gas, as in smokes, aerosols, and fogs). Common examples of colloidal systems are salad dressings and mayonnaise (oil in water emulsion), whipped cream (air and fat globules in water), waste water, (dirt particles, oil droplets), creams and gels used in pharmacy and by beauticians, medicines, paints, coal and mineral slurries, oil emulsification and solubilisation, mineral grinds used for ore recovery.
The important thing about a colloid dispersion is whether it is stable, which is to say that the particles remain dispersed and move about in it in a liquid-like fashion. In an unstable dispersion the particles adhere to each other and form clumps, which either precipitate and sediment, or float on top of the liquid as a scum. Depending upon the application, either a stable or an unstable dispersion is required, and great effort is made to understand and to control their stability.
It is the forces between the colloid particles that determines what happens. Colloids have been treated since the turn of the century on the basis of two types of forces: van der Waals attractions and electric double layer repulsions. The former arises from correlated electronic and dipolar fluctuations in the particles and in the intervening liquid. It is present between all materials, although its strength does vary. The double layer repulsion occurs when the particles have a net charge, which is generally on their surface. This repulsion is exponentially screened by the intervening ions in the water, with a decay length typically from 100nm in 'pure' water, to 1nm in concentrated (.1M) electrolyte solutions. In general terms the double layer repulsion dominates at long range, and the van der Waals attraction at short. So the interaction potential between colloid particles has a deep primary minimum at contact, and a potential barrier due to the double layer at 1--10nm from contact. If this barrier is high enough, the thermal collisions between the particles does not provide enough energy to overcome it, and so the particles can never stick together: the dispersion will be stable. Conversely, if the barrier is small compared to the thermal energy kT, the particles fall into the deep van der Waals well at contact and can never get out again: the dispersion is unstable.
Since the van der Waals attraction is largely unaffected by electrolyte, the easiest way to make a stable dispersion unstable is to add salt, since this screens the double layer repulsion. This is exactly what happens when the mouth of a river silts up as it enters the ocean. The fine colloidal clay particles were dispersed in the fresh water of the river due to their double layer repulsion (if they are small enough gravitational energy is negligible compared to the thermal energy, so they don't sink as individuals). When the river hits the salty ocean, the repulsion is screened and the clay particles clump together due to the van der Waals attraction. When the clumps get large enough mass, gravity becomes significant and causes them to sink to the bottom as silt.
In the case of foams, such as used in fighting fires, the bubbles of gas (eg. carbon dioxide) are separated by a thin film of water. What keeps the bubbles separate is an electric double layer repulsion due to the surface charge that develops at the gas-water interface of the bubbles. The foam is said to be strong if it does not collapse (ie if the water film does not drain away and if the bubbles do not coalesce). A strong or stable foam is obviously desirable to blanket the fire, starving it of oxygen.
Another good example is paint. Water soluble paints consist of pigmented latex particles in water. In the can it is a liquid-like (ie a stable dispersion). If the dispersion becomes unstable in the can the paint is useless. When it is put on the wall it dries, which is to say the water evaporates concentrating the dispersion and forcing the particles into intimate contact with each other and with the wall. They stick irreversibly, since adding water again does not wash them off. From the above, the reason that they adhere irreversibly is because they have been forced into the primary van der Waals minimum, and adding water again is not going to get them out. (I like to use paint as an example because it lets me joke that my work is as interesting as watching paint dry.)
A final example is froth flotation. Many ore minerals are hydrophobic (eg Galena, which is lead sulfide), and the mining industry commonly recovers the valuable ore from the worthless rock surrounding it by grinding it up to colloidal dimensions, making a dispersion, and bubbling air through it. By suitable additives, they can control things so that the minerals attach to the air bubbles and float to the top, forming a froth, and the rock is left behind. Again it is the surface forces that determine this selective attraction and adhesion between the ore and the air bubble.
So then understanding and controlling the van der Waals and the double layer force is very important in colloid science. In fact surface forces occur quite widely, (such as in crack propagation, glue adhesion, protein tertiary structure, self-assembly of biological membranes, strengths of alloys and ceramics), which provides additional motivation for the work. Up until comparatively recently it was believed that the van der Waals and the double layer forces were the only forces operative. However, in the last 30 years many new forces have been described. (For example, in the case of paint mentioned above the repulsion is not provided by an electric double layer but by steric hindrance due to polymer brushes on the surface of the particles.) The main experimental evidence for these has come from direct measurement of the forces between surfaces in liquids originally with the surface forces apparatus, and increasingly with the atomic force microscope. (One surface is placed on a spring, and the distance to the other particle is controlled by expansion of a piezo-crystal, and the deflection of the spring is measured either interferometrically, electronically, or optically, ideally with molecular resolution.) One of these new forces ---and the most puzzling--- is the long-ranged attraction measured between hydrophobic surfaces.
The reason that most people have been trying to measure and to pin down this force is that it is so dramatically different to the other known forces. It is orders of magnitude stronger than the van der Waals attraction, and similarly its range can extend beyond that of the double layer repulsion. Obviously the existence of such a force has dramatic implications for the stability of hydrophobic colloids, for example, the latex paint particles mentioned above, oil droplets and air bubbles, sulfide mineral particles (iron, lead), and even the structure of proteins, (since these have hydrophobic patches.) People want to know when this attraction is occurring, and they want to be able to control it.
Beyond the practical applications of the hydrophobic attraction are the conceptual challenges that such a long range force poses. It is necessary to understand what is going on, and to fit this force into the existing framework of liquid state physics. Such a long-ranged force is really difficult to accept in the context of statistical mechanics, and either one has to find an artifact in the experiments, (or a different interpretation of them), or one has to modify the existing theories.
TOPWhat is the long-ranged hydrophobic attraction, and what could possibly cause it?
Summary: The work seeks to elucidate the physical mechanism for the long-range attraction measured between hydrophobic surfaces. The closely related work is the theoretical suggestions that it could be due to electrostatic correlations [P. Attard, J. Phys. Chem. 93, 6441 (1989)], a separation-induced spinodal [D.R. Berard, P. Attard, and G.N. Patey, J. Chem. Phys. 98, 7236 (1993)], or bridging sub-microscopic bubbles [J.L. Parker, P.M. Claesson, and P. Attard, J. Phys. Chem. 98, 8468 (1994)]. The last paper is mainly experimental and gives references to other measurements.
Detail: Our direct motivation was to measure the forces between hydrophobic surfaces in water in an effort to pin down the mechanism that gives rise to certain long range attractions that had been measured over the years. (This hydrophobic attraction is measured between microscopic to macroscopic surfaces over ranges of 10-100nm, and is unrelated to the hydrophobic effect that operates between non-polar molecules driving the formation of clathrate cages and aggregation over sub-nanometer distances.) Hydrophobic surfaces are those that repel water, so that a droplet of water placed on it has a high contact angle (forms beads) and does not wet it. Examples include teflon surfaces such as those used in non-stick frypans, paraffin or waxy surfaces, or even possibly the surface of an oil droplet or of an air bubble. That an attraction orders of magnitude stronger and longer ranged than the classic van der Waals attraction exists between such surfaces has important practical implications, e.g. froth flotation, which depends upon hydrophobic mineral particles adhering to air bubbles, the stability of foams, (via the coalescence of air bubbles), and the stability of dispersions of hydrophobic colloid particles such as latexes. Despite its practical importance, and despite the measurements of the effect dating back over 20 years, no consensus has emerged about the physical origin of the force. The force is so controversial that comparisons have been made with polywater: the force has been measured between surfaces separated by 300 nm, which corresponds to about 1000 water molecules in width, and the idea that surfaces can induce order in liquids extending so far from the surfaces contradicts very fundamental theories of the liquid state.
The theories that have been proposed fall into three categories. (I do not count the poly-structural theory alluded to above, as I don't think that this was ever taken seriously by any one other than its proponents; there have been a number of other suggestions that fall into the same category that I similarly won't mention.) It was originally proposed by me [P. Attard, J. Phys. Chem. 93, 6441 (1989)] that the surfaces felt each other by electrostatic correlations in the intervening water and electrolyte. The strength was said to arise from the metastable nature of the water or electrolyte next to the surfaces, and the decay length of the force, which determines its range, was half the Debye length, which goes like the square root of the electrolyte concentration. This idea was developed in various ways by other, mainly dealing with what gave rise to the strength of the force [cf R. Podgornik, J. Chem. Phys. 91, 5840 (1989). Y. H. Tsao, D. F. Evans, and H. Wennerstrom, Langmuir 9, 779 (1993). S. J. Miklavic, D. Y. C. Chan, L. R. White, and T. W. Healy, J. Phys. Chem. 98, 9022 (1994). O. Spalla and L. Belloni, Phys. Rev. Lett. 74, 2515 (1995).] What all these theories have in common is that the decay length should depend strongly on the electrolyte concentration. The experimental evidence is mixed: there are about three papers which appear to quantitatively confirm the prediction, and there are about seven papers which strongly contradict the idea.
The second theory suggests that the water between the hydrophobic surfaces is in a meta-stable state, and that the attraction arises from the approach to a separation induced spinodal [D.R. Berard, P. Attard, and G.N. Patey, J. Chem. Phys. 98, 7236 (1993); P. Attard, C.P. Ursenbach, and G.N. Patey, Phys. Rev. A 45, 7621 (1992)]. A variation of the proposal deals with the spinodal decomposition of a binary mixture rather than liquid-vapor transition. The long-range nature of the attraction in this theory arises from the well-known divergence of the correlation length near criticality, since the critical point is just the extremum of the spinodal line. The experimental support for this notion comes from the observation of spontaneous cavitation of the water between the hydrophobic surfaces when they are separated from contact; it is easy to show that for contact angles of 100 degrees or so the vapor cavity is more stable than the liquid for separations out to micrometers. Hence when the surfaces are brought toward contact in the liquid, they are in fact in a metastable situation. Unfortunately it is difficult to make a quantitative prediction with the theory, and no definitive experimental test of it has been proposed or performed.
The third theory postulates that there are sub-microscopic bubbles attached to the surfaces, of the order of 100nm in diameter, and one is actually measuring an attraction as the bubbles adhere, bridge, and spread along the two surfaces. [J.L. Parker, P.M. Claesson, and P. Attard, J. Phys. Chem. 98, 8468 (1994).] The theory was based on the observation of steps or discontinuities in the force curves at large separations, which were interpreted as the instant of attachment. The difficulty with the theory is that the evidence is rather indirect, (the putative bubbles are too small to see directly). Further, on thermodynamic grounds such sub-microscopic bubbles have a high internal gas pressure and should not exist (or at best be meta-stable), [P. Attard, Langmuir 12, 1693 (1996)]. However, the idea is supported by certain experiments that show the range of the force to decrease when the water is de-aerated [J. Wood and R. Sharma, Langmuir 11, 4797 (1995). L. Meagher and V. S. J. Craig, Langmuir 10, 2736 (1994)]. The good feature of the theory is that it avoids having to postulate a long-ranged attraction, since it instead attributes the range of the force to the physical size of the bubbles.
TOPHow were the AFM measurements done?
The device used was the atomic force microscope (AFM), which can measure separations with sub-nanometer resolution, and forces with sub-piconewton resolution. A hydrophobic colloid particle of 10 micrometer diameter was attached to a cantilever spring and the force of interaction with a hydrophobic substrate was measured in water and in various electrolytes. We simultaneously visually observed the system with a microscope and noted the existence of long-lived, sub-micron bubbles attached to the surfaces in non-contact regions. Our force measurements showed long-ranged attractions with several distinct features that unambiguously show the presence of a bridging sub-microscopic bubble. The forces were essentially unchanged upon increasing the electrolyte concentration, which rules out electrostatic correlations as a viable mechanism for this system.
We concluded that a class of the long-ranged attractions measured between hydrophobic surfaces are attributable to the presence of sub-microscopic bubbles attached to the surfaces.
TOPWhat is the significance of the research for future work?
Our conclusion that bubbles were responsible for the measured attractions fundamentally changes the way in which hydrophibc surfaces are viewed. In future experiments the possibility of bubbles must be taken into account, and their effect on other measured phenomena should be considered. That the range of the force is due to the physical size of the bubbles rescues the very fundamental notion that surfaces perturb liquids over very short ranges. The work will likely stimulate further research on the effects of de-aeration and gassing, and on bubble coalescence.
The existence of such sub-microscopic bubbles is of interest in its own right. In the first place they will have a dramatic effect the interaction of real hydrophobic colloidal particles. The second, and to my mind deeper, consequence is that the existence of such bubbles challenges some well-established notions of macroscopic thermodynamics. The resolution of the paradox may lie in generalizing thermodynamics to small length scales, or else calculating the expected life-times of such sub-microscopic bubbles, if indeed they are metastable. In either case there will be general implications for nucleation and heterocoagulation theory.
TOPWhat is a separation-induced spinodal, and what is spontaneous cavitation?
Hydrophobic surfaces hate water, and vice versa. When two hydrophobic surfaces are brought close to contact, the water wants to leave. One way to do this is to cavitate, which is to say it vaporises. One of course gets water vapor in this cavity rather than vacuum, and you could call it steam, except that this would be confusing since the experiments are at room temperature and atmospheric pressure. Obviously liquid water does not normally vaporise, and it is only because of the extreme antipathy between it and the hydrophobic surfaces that it becomes favorable to do so. This cavitation is induced by the close proximity of the two surfaces, and it was first observed experimentally by Christenson and Claesson [H. K. Christenson and P. M. Claesson, Science 239, 390 (1988)]. Simple thermodynamic arguments using the contact angle of a water drop on the hydrophobic surface show that the free energy of a bridging cavity is lower than that of liquid water when the surfaces are separated as far as micrometers. [D.R. Berard, P. Attard, and G.N. Patey, J. Chem. Phys. 98, 7236 (1993)]. The fact that such cavities are not observed as the two surfaces approach contact from far apart indicates that the liquid between them must be metastable and that there is some barrier preventing cavitation.
The theory for the long-ranged hydrophobic attraction relies upon this notion of induced cavitation, and also the behavior of bulk liquids as they approach the spinodal line. As many people know, there is a curve in the phase diagram called the coexistence line where liquid and vapour coexist. (If you plot density on the abscissa and pressure on the axis the curve is parabolic with the tip up, and the critical point is at its apex). Inside this line is the spinodal parabola, which meets the coexistence curve at the critical point, but otherwise is separated from it by a finite distance. The region between the spinodal and the coexistence curve is the metastable region; it is possible to super-heat a liquid without it boiling, or to super-cool a gas without it condensing, and in both cases one has a metastable fluid. In the absence of nucleation sites, the metastable fluid can be macroscopically long-lived. It is not possible to proceed past the spinodal line and remain as one phase. At the spinodal one can regard the molecules as self-nucleating, and one can do nothing to prevent the phase transition occurring.
The spinodal line is the true continuation of the critical point. The compressibility is infinite at the spinodal (as at the critical point), which means that density fluctuations become macroscopic, which is the same as saying that the correlation length diverges. One can show that the force between two colloids in a near-critical or a near spinodal fluid is attractive and long-ranged (it decays with the bulk correlation length) [P. Attard, C.P. Ursenbach, and G.N. Patey, Phys. Rev. A 45, 7621 (1992)].
The connection between the spinodal attractions in the bulk that I just mentioned and the measured long-range attractions between hydrophobic surfaces is the observed cavitation. The question we asked ourselves is this: if the water cavitates at some small separation, and is metastable at larger separations, is there a spinodal separation? That is, is there a separation beneath which one cannot prevent cavitation occurring? One expects just as in the bulk case, that if such a separation-induced spinodal exists, then the forces would be attractive, and the correlation length, and the consequent range of the force, would get very large. We did computer simulations on a Lennard-Jones liquid confined between hard-walls {D.R. Berard, P. Attard, and G.N. Patey, J. Chem. Phys. 98, 7236 (1993)]. The liquid did indeed cavitate at small separations, and there was indeed a spinodal separation. Approaching this separation we found that the attractions were much stronger than the van der Waals attraction, and longer ranged.
Qualitatively, then the idea of a separation-induced spinodal can indeed account for the measured hydrophobic attractions. The problem is that we haven't done the simulations for water, and so we cannot prove quantitatively that the mechanism is viable. This remains the current status of this theory.
TOPWhat is polywater?
The polywater episode occurred in the sixties . It was at its time what cold fusion has become more recently. There is a beautiful book by Felix Franks, titled 'Polywater', which is a very readable account of its history. He makes the point that a sort of mass hysteria periodically grips science, and I think in fact he estimates that it recurs in different guises every twenty or thirty years. (Before polywater it was N-rays.)
Polywater itself arose in colloid science. Basically measurements of water in glass capillaries showed properties very different to bulk water. It was attributed to a new state of water, called polywater. It was said that the walls of the capillary induced structure in the water that could extend out to large distances (micrometers). The measurements were driven by the Russian school under Derjaguin. The cold war contributed to fears that a seed of polywater would nucleate a change to this new phase of anomalous water. If such seeds were placed in the rivers and lakes of the US it would be brought to its knees; hence the military jumped on the band-wagon with funding, and the stampede was on. It took a few years, but eventually the anomalies were traced to impurities in the capillaries, and a sort of soft polymeric layer that silica forms in water.
The connection with the hydrophobic attraction is this. Polywater appeared to show that surfaces could induce long-range order in otherwise disordered liquids. This contradicted long-standing and well-established theories of the liquid state, which show that surface-induced order only propagates one or two molecular layers into the liquid away from the surfaces. Likewise, the measured hydrophobic attraction at 100nm separation or so superficially appear to show that water is disrupted or changed from its bulk properties by the surfaces over 1000 or so molecular layers. Indeed one of the earliest theories for the attraction proposed such a layer-to-layer propagation of the ordered water [J. C. Eriksson, S. Ljunggren, and P. M. Claesson, J. Chem.\Soc. Faraday Trans. II 85, 163 (1989)].
In the early days many people dismissed the experiments on the hydrophobic attraction because the extreme long-ranged nature seemed to require just such an unacceptable ordering. Comparisons with polywater were made, and suggestions such as contamination, polymer bridging and even glue dissolution were put forward. The situation was not helped by the fact that the data between different laboratories did not always agree. Many of the protagonists were not on speaking terms. (Actually this was the best case scenario; in some cases the confrontations were painful to witness.) Controversy and polemic raged, and it would be fair to say that emotions still run high.
I myself have no doubts in the existence of the force (I know many of the different experimentalists and their track records). However I also recognize that some of the data is irreproducible, and one has to exercise a deal of caution and selectivity. (One thing that is nice about the bubble explanation is that it has a random character to it that could account for the irreproducibility of the force.) My own motivation has been not to dismiss the experiments, but rather to find a viable explanation for them that avoids the unacceptable structural mechanism discussed above. The three theories that I have developed account for one or other feature that have been observed in the experiments. Whether a single mechanism will eventually be shown to account for all the data, or whether there is more than one 'hydrophobic attraction' remains to be seen.
TOPWhat makes Olympic swimmers in suits faster?
The surprising answer to this question appears to be given by quite unrelated experiments reported in a recent Physical Review Letter. James Tyrrell and Phil Attard of the University of South Australia set out to obtain molecular scale images of water repellent surfaces using a scanning probe microscope. The motivation for the work was to find `nanobubbles', which are thought to be the cause of the long range attractions between such hydrophobic surfaces in water (See Physical Review Letters Focus, `Small Attractive Bubbles'). Such nanobubbles, which are smaller than the wavelength of light, had not previously been seen, and there existence and longevity was in doubt. The striking images produced by Tyrrell and Attard reveal bubbles about 30nm in height, which is comparable to the range of the attractions that they measured. The bubbles had lifetimes of the order of hours, and rapidly reformed when disturbed.
The most surprising feature of the images was that the surfaces were literally covered in nanobubbles, with few bare patches, and very little of the surface was in direct contact with water. This observation has dramatic implications for traditional hydrodynamics, where a drag force arises from so-called `stick' boundary conditions, (ie. the flow is assumed stationary at a solid surface). In the case of a fluid interface free boundary conditions are more appropriate, and consequently the amount of shear in the flow and the drag is reduced. The results of Tyrrell and Attard suggest that hydrophobic surfaces are more like a fluid than a solid interface due to their covering of nanobubbles, and consequently they exhibit much less hydrodynamic drag. It has long been known that the flow in narrow hydrophobic capillaries is greater than predicted by stick boundary conditions, which can now be explained as due to a covering of the nanobubbles found by Tyrrell and Attard. And it appears that elite swimmers are also taking advantage of the phenomenon by using swimming suits manufactured from hydrophobed fibres.
TOPBack to publications, research, or to Phil Attard's home page.