You are on page 1of 24

Advances in Colloid and Interface Science

80 1999. 27]50

Bubble nucleation from gas cavities } a review


S.F. Jones, G.M. Evans, K.P. GalvinU
Department of Chemical Engineering, The Uni ersity of Newcastle, Callaghan,
New South Wales 2308, Australia

Abstract

This review is concerned with the nucleation of bubbles in solutions supersaturated with a
gas, in particular the bubble nucleation that occurs at specific sites, as a cycle. A classifica-
tion system for the kinds of nucleation that occur is defined and discussed in order to place
this specific form of nucleation into a better defined context. It is noted that in the absence
of pre-existing gas cavities, bubble nucleation requires exceedingly high levels of supersatu-
ration. It is argued that the nucleation observed in most instances, which is often at low
levels of supersaturation of 5 or less, is invariably associated with the existence of metastable
gas cavities in the walls of the container or the solution bulk, prior to the system being made
supersaturated. Here, the nucleation energy barrier for each gas cavity is very much lower
than for the classical case, given that less interfacial free energy is needed for the cavity to
grow to the critical size when the system is made supersaturated. Once a system contains gas
cavities with radii of curvature greater than the critical nucleation radius, bubbles are
produced in a steady fashion without the need to scale a nucleation energy barrier. This
non-classical form of nucleation is the main focus of the paper. Issues concerning the
formation of these gas filled cavities, and their stability are examined. Q 1999 Elsevier
Science B.V. All rights reserved.

Keywords: Nucleation; Bubbles; Supersaturated solution; Bubble growth; Bubble detach-


ment

Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.1. Definition of supersaturation and thermodynamic parameters . . . . . . . . . . . . . 29

U
Corresponding author. Tel.: q61 49 216194; fax: q61 49 216920.

0001-8686r99r$ - see front matter Q 1999 Elsever Science B.V. All rights reserved.
P I I: S 0 0 0 1 - 8 6 8 6 9 8 . 0 0 0 7 4 - 8
28 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

2. Review of nucleation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.1. Type I classical homogeneous nucleation . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.2. Type II classical heterogeneous nucleation . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.3. Type III pseudo-classical nucleation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.4. Type IV non-classical nucleation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3. Classical nucleation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.1. Evidence for type I and II nucleation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2. Evidence for type III and IV nucleation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4. Gas entrapment mechanism and gas cavity stability . . . . . . . . . . . . . . . . . . . . . . . 38
5. The cycle of bubble production } nucleation time . . . . . . . . . . . . . . . . . . . . . . . 41
6. Bubble growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
7. Bubble detachment from a solid substrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
8. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

1. Introduction

There are many industrial practices and natural processes involving either
supersaturated or superheated liquids for which the study of bubble formation and
growth are important. In nature, where examples are numerous, these can have
dramatic consequences. The eruption of Lake Nyos in Cameroon in 1986, which
resulted in the loss of more than a thousand lives, was caused by the sudden
release of carbon dioxide gas dissolved in the deeper regions of the lake w1x. This
sudden gas release is very similar to the explosive degassing of magma encountered
in volcanoes. At the more personal level, divers may succumb to the bends if
bubbles of nitrogen form in their blood while surfacing w2x.
Bubble production is important in commercial electrolytic processes w3x, in
cavitation w4x, in the effervescence of carbonated beverages such as sparkling wine,
beer and soft drinks w5x, and in electric power generation, during the production of
high pressure steam w6x. In liquid waste treatment, dissolved air is sometimes used
to produce bubbles that selectively carry particulate species to the surface w7x. In
the petro-chemical industry the foaming of liquid plastics is achieved through the
nucleation and growth of bubbles via a diffusion process w8x. Conversely, vacuum
degassing processes are used to remove bubbles formed in molten steel or glass w9x.
Bubbles are formed when a pure homogeneous liquid undergoes a phase change.
The tendency to produce the new phase is quantified by the degree of superheat,
and the process is governed principally by the diffusion of heat. In such a pure
system there are no concentration gradients. A chemical process, such as electroly-
sis, may also result in bubble production. In this second case, both temperature and
concentration gradients are important. A third process for autogenous bubble
generation is gas desorption, the subject of this review. Here, the tendency to
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 29

generate bubbles is quantified in terms of the supersaturation, and the rate of the
bubble growth governed principally by the concentration gradient, assuming the
heat of desorption is negligible, and the accompanying transfer of the solvent is
also negligible.
This review is concerned with the production of bubbles during gas desorption.
The review also draws upon the important work on boiling that is relevant to the
issues raised. It is argued that at low levels of supersaturation pre-existing gas
cavities are responsible for the formation of the bubbles, a view shared by many,
including Dean w10x, Harvey et al. w11,12x, and Bankoff w13x. These gas cavities exist
at various levels of stability and respond to changes in the thermodynamic condi-
tion of the solution. The review also covers work on the consecutive processes of
bubble nucleation, growth and detachment that occur at specific cavities on a
substrate in contact with a supersaturated liquid. These physical processes, which
are interrelated as a cycle, are the subject of a comprehensive research paper that
follows this review w14x.
The purpose of this review is to examine a specific form of nucleation charac-
terised by a steady production of bubbles from an existing gas cavity, to be referred
to as type IV nucleation. Here, the bubble nucleation continues provided the
minimum radius of curvature of the meniscus in the cavity is greater than the
critical nucleation radius, as determined by the classical theory of nucleation. The
review begins with a discussion of bubble nucleation, and suggests a new classifica-
tion for the broad range of nucleation often encountered. By introducing a system
for classifying the nature of the nucleation, more consistency in the reporting of
results should be possible. The mechanisms involved in establishing a gas filled
cavity, and the stability of the liquid meniscus are examined. The issue of the
nucleation time, referred to by a number of workers as the waiting time or delay
time is also addressed.
In the final section of the review the physics of the bubble growth is examined
briefly. A complete description requires a consideration of many factors including
diffusion of heat and gas molecules, convection, and inertial, viscous and surface
tension forces. A useful analysis is that produced by Scriven w15x, which accounts
for a majority of the growth normally observed. His analysis focuses on the growth
caused by molecular diffusion analogous to heat diffusion. of gas molecules
through the liquid and the liquid]gas interface. The outward convective movement
of the interface results in a higher molecular diffusion, relative to the interface,
and hence a higher growth rate than expected for diffusion alone. The review then
deals with the physics of the bubble detachment, the final stage of the bubble
growth cycle.

1.1. Definition of supersaturation and thermodynamic parameters

Before proceeding, it is important to define the term, supersaturation, used for


quantifying the tendency of a system to produce bubbles. An example of saturation
data is shown in Fig. 1 as a function of the temperature for the system consisting of
carbon dioxide gas dissolved in water at a pressure of 1.013 = 10 5 Pa w16x.
30 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

Fig. 1. Solubility of carbon dioxide as a function of the temperature, at a pressure of 1.013 = 10 5 Pa


w16x.

Supersaturation may be achieved by raising the temperature of the system. Point


A represents a saturated solution at temperature TA with a saturation mole
fraction equal to X b . Suppose the temperature of the solution is raised to a value
TB at point B. The system is then in a supersaturated state, with the mole fraction
of the dissolved component equal to X b . The desorption of carbon dioxide from
the water causes the state of the system to move gradually from point B to B9,
where the new saturation mole fraction is X i . Lubetkin and Blackwell w17x defined
the saturation ratio as,

Xb
as 1.
Xi

and the supersaturation as,

ssay1 2.

Further, it is useful to incorporate the Henrys law constant, H, into the above
expressions in order to view the subject of supersaturation from the perspective of
pressure. According to Henrys law, the value of H varies as a function of
temperature in terms of the changes in the equilibrium concentration, X i , shown in
Fig. 1, as,
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 31

Pi
Hs 3.
Xi

where Pi is the pressure, 1.013 = 10 5 Pa. At temperature TB , the difference in the


carbon dioxide equilibrium pressure for the two states, B and B9, is,

Xb
D P s Pb y Pi s H X b y X i . s Pi y 1 s Pi s
Xi / 4.

where Pb is the equilibrium carbon dioxide pressure at B. In a gas cavity, the


pressure elevation, D P, due to the radius of curvature, R9, of the meniscus, is
described by the Laplace equation,

2g
DP s 5.
R9

where g is the interfacial tension between the solution and the gas. Assuming the
vapour pressure of the solvent is negligible, equating Eq. 4. and Eq. 5. provides a
method for calculating the critical radius of curvature, R9U s R9, for any level of
supersaturation, s . A bubble, in a solution supersaturated by a gas, is then in
thermodynamic equilibrium with the solution if its radius is equal to this critical
radius value. The critical radius is identical to that used in the classical theory of
nucleation.

2. Review of nucleation

In this review, the term nucleation is used generically to denote any process that
leads autogenously to the formation of a bubble. There seems to be some
confusion in the literature, because of the generic nature of the term. For this
reason, some effort will be made to define mechanistically four types of nucleation.
It is noted that homogeneous nucleation within the liquid bulk and even heteroge-
neous nucleation on molecularly smooth surfaces requires very high levels of
supersaturation. Here it is necessary for the newly formed gas phase to tear apart
the liquid, and hence overcome its enormous cohesivertensile strength. At low
levels of supersaturation, however, gas bubbles will only nucleate within the liquid
at pre-existing metastable gas cavities, in the vessel surface in contact with the
liquid, on suspended particles, or from metastable micro-bubbles in the liquid bulk.
It is shown later that many authors have reported nucleation rates much higher
than calculated using the classical theory. The most likely reason for the discrep-
ancy is the pre-existence of gas filled cavities in their system. This means that their
initial system state, prior to it being made supersaturated, consists of a large
number of bubble nuclei, with a broad range of menisci curvatures. A gas cavity
which already contains a meniscus with a curvature equivalent to the critical value,
32 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

Fig. 2. Type I and II classical homogeneous and heterogeneous nucleation.

has its nucleation energy barrier lowered to zero. Other menisci of smaller radii
will have a finite nucleation energy barrier. However, these energy levels will be
much lower than the nucleation energy barrier calculated for a classical nucleation
event. For a classical event, a bubble surface must be formed at a position where
previously no gas existed. The energy barrier involved is therefore much higher and
hence the calculated classical nucleation rate very much lower.
Four possible types of nucleation are described, and are also illustrated in Figs. 2
and 3.

2.1. Type I classical homogeneous nucleation

This involves nucleation in the liquid bulk of a homogeneous solution. There are
no gas cavities present prior to the system being made supersaturated. Hence, the
required level of supersaturation is very high, in excess of 100 or more. Once a
bubble is produced, it rises to the surface of the liquid. Additional bubbles are
unlikely to form at the same location.

2.2. Type II classical heterogeneous nucleation

This form of nucleation is essentially the same as type I, and should normally
require comparable levels of supersaturation. Initially, the system contains no gas
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 33

Fig. 3. Type III pseudo classical and type IV non-classical nucleation.

cavities, in the bulk or in the surface of the container. The system is suddenly made
supersaturated, for example by a sudden pressure reduction, resulting in a classical
nucleation event. A bubble may then form in a pit in the surface of the container,
on a molecularly smooth surface, or on a particle in the bulk. The bubble then
grows, and detaches, leaving behind a portion of its gas. The production of the first
bubble is referred to as type II nucleation. At this point, the rate of bubble
production decreases significantly because the level of supersaturation is lower,
and because the fluctuations brought about by the sudden pressure reduction are
less significant. Arguably, the only reason why bubble production continues at this
point is because of the existence, now, of gas filled cavities. This latter type of
nucleation that follows the initial event is referred to as either type III or IV.
34 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

2.3. Type III pseudo-classical nucleation

This nucleation includes homogeneous and heterogeneous nucleation at pre-ex-


isting gas cavities at the surface of the container, at the surface of suspended
particles, and as metastable micro-bubbles in the solution bulk. At the moment the
system is made supersaturated, the radius of curvature of each meniscus is less
than the critical radius, as determined by the classical theory. Hence, for each
cavity there exists a finite nucleation energy barrier which must be overcome.
When supersaturated, local fluctuations in supersaturation are responsible for
bringing to life the nucleation sites. Type III nucleation is achievable at low levels
of supersaturation. After it occurs at a fixed cavity in the surface of the container
wall, type IV nucleation may follow. Type III nucleation at cavities in the bulk,
however, may not lead onto type IV nucleation because the growing bubble causes
the site to rise to the surface of the liquid.

2.4. Type IV non-classical nucleation

This nucleation is considered non-classical because there is no nucleation energy


barrier to overcome. The nucleation usually occurs at pre-existing gas cavities in
the surface of the container or elsewhere in the liquid bulk, and may follow type II
or type III nucleation events. Pre-existing gas cavities housing menisci with radii of
curvature greater than the critical nucleation value provide a stable source for
bubble nucleation. Over time, as the supersaturation decreases, the critical nucle-
ation radius, R9U , increases to a value equal to the radius of a given cavity
meniscus, and bubble production from that cavity ceases. This kind of nucleation is
responsible for sustaining the cycle of bubble production in carbonated beverages,
for example, long after the bottle is opened, or the liquid poured into the glass.

3. Classical nucleation

The classical view of nucleation dates back to Volmer and Weber w18x, Farkas
w19x, Becker and Doring
w20x, Doring
w21x and Zeldovich w22x. Extensive reviews
have also been made of this early work by Volmer w23x, Frenkel w24x, and Cole w25x.
Other papers have examined the vaporisation of pure substances w26,27x and the
nucleation of bubbles from solutions containing dissolved gas w28x. More recent
reviews have been undertaken by Sides w29x and Lubetkin w30,31x.
The type I and II nucleation are adequately described by the so-called classical
model. The model is based around the following: when the system is supersatu-
rated, the bulk free energy per unit of liquid volume, D g , associated with
transferring molecules to the new phase is negative, and hence thermodynamically
favourable. So, creation of the new phase lowers the bulk free energy. But, the
process also results in the production of a significant interfacial free energy and, in
the case of heterogeneous nucleation, a third energy associated with the tension of
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 35

the contact line. Initially, because the surface to volume ratio of the embryo is very
high, the combined result is actually a net increase in the free energy of the system.
Beyond a critical embryo radius of curvature, R9U , however, the combined free
energy decreases, and hence the embryo is free to grow. This critical radius of
curvature is,

y2g
R9U s 6.
D g

The total free energy required at the critical position is known as the nucleation
energy barrier, W. The number of critical sized nuclei, n, is described using the
following Boltzman distribution,

n s Nexp yWrkT . 7.

where N is the number of molecules. The addition of just one molecule to a


critical nucleus is sufficient to trigger spontaneous growth. So, the rate at which a
gas molecule is transferred to each critical nucleus provides a measure of the
nucleation rate. A further pre-exponential factor is therefore applied to Eq. 7. to
obtain the rate.
Embryos smaller than the critical size readily disappear altogether, unless the
local energy fluctuations remain large enough, for long enough w27x. At exceedingly
high levels of supersaturation, the critical embryo size is correspondingly very
small, and hence the probability of a sufficient number of gas molecules coming
together to achieve a successful nucleation event very much greater. The presence
of a solid substrate simply lowers the interfacial free energy of the nucleus,
assuming the interfacial energy between the solid and the liquid is lower than the
interfacial energy between the liquid and the gas. In effect, the solid acts as a
catalyst, lowering the size of the nucleation energy barrier. Although the radius of
curvature of the critical size nucleus is unchanged, the number of molecules
required to form the nucleus is reduced. Consequently, the nucleation becomes
more probable. Heterogeneous nucleation, involving water in contact with smooth
hydrophobic surfaces, still requires supersaturations of 100 or more w32x. This is
hardly surprising given that the net free energy reduction is invariably quite
modest.
Although the classical view of the type I and II nucleation continues to provide a
useful conceptual understanding, there remains a number of difficulties. Firstly,
the model is based on an extrapolation of thermodynamic parameters such as
interfacial tension to the molecular level. According to Lubetkin w31x, however,
these parameters do show remarkable resilience at length scales of the order of 1
nm. Secondly, the application of thermodynamics to describe systems not in
equilibrium is not an ideal approach. Thirdly, in the case of boiling and gas
desorption, the question of mechanical stability needs to be considered. In these
cases it is necessary to pry open a cavity of vapour and gas respectively, and hence
overcome the cohesive or tensile strength of the liquid.
36 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

Enormous pressures, as quantified by the Laplace equation, are required for


mechanical stability of bubbles with a radius of a few nanometres. Presumably, the
conditions are extreme enough to place the gas properties of the embryo in a
critical fluid state. What does this mean, especially in terms of the interface
between the bulk liquid which is at a low pressure, and the embryo which is
presumably at a high pressure? Is there a gradation between the properties of the
liquid and those of the embryo? If so, what principles govern the nature of the
interface and the mechanical stability. Bowers et al. w33x, for example, proposed a
model based on the formation of nuclei, with unstable structures and poorly
defined interfaces, called blobs.
There seems to be one essential truth concerning the type I and II classical
nucleation, both homogeneous and heterogeneous, despite the short-comings in
the underlying theory. In experiments in which significant care has been taken to
eliminate gas cavities from the liquid environment, bubble production requires
exceedingly high levels of supersaturation.

3.1. E idence for type I and II nucleation

In this section evidence is presented in support of the view that types I and II
nucleation require high levels of supersaturation.
Fisher w34x measured the tensile strength of water at smooth rigid interfaces and
at interfaces containing conical sites with sharp and rounded bottoms. In order to
obtain reproducible results, he found it necessary to have complete wetting
between the liquid and the solid phase. Hemmingsen w35x, taking extreme care to
eliminate any gas cavities in his system, showed that bubble nucleation in glass
capillaries required supersaturation levels of 100 and more.
Wilt w36x used the classical nucleation theory to show that, at levels of supersatu-
ration of 5 and less, no bubble should form in the bulk or on a smooth planar
surface unless the contact angle is exceedingly high at more than 1758. This
supported the view that almost all heterogeneous nucleation at low levels of
supersaturation occurs at pre-existing gas cavities. Finkelstein and Tamir w37x
developed a reliable method for determining the threshold decompression at which
a bubble would form homogeneously from a saturated solution under pressure.
They investigated different gases in water and found a value of the pressure
difference for each of them ranging between 13 and 42 MPa. The corresponding
value obtained using classical nucleation theory was 142 MPa, at least three times
larger than those experimentally obtained. This result has also been obtained by
Yang and Maa w38x in their studies of micro bubbles during boiling.
Ryan and Hemmingsen w32x showed that the level of supersaturation required
for heterogeneous nucleation to occur on hydrophobic and hydrophilic smooth
particles is about the same. Very smooth micro particles, without any detectable
pores, crevices or other surface features present, were used to reduce the tendency
for gas cavities to form. Low rates of bubble formation were obtained for smooth
hydrophobic surfaces at high supersaturations type I and II., whereas very high
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 37

rates of formation occurred at rough hydrophobic surfaces at lower supersatura-


tions type III and IV.. This tends to indicate that the presence of cavities is
essential for bubble initiation at low levels of supersaturation.

3.2. E idence for type III and IV nucleation

In this section evidence is presented to demonstrate that type III and IV


nucleation requires relatively low levels of supersaturation, certainly much lower
than predicted by the classical theory.
Chen et al. w39x determined the onset of heterogeneous nucleation on particles
and measured the nucleation rate to be much higher than predicted, concluding
that the particle surface served as a catalyst. They pointed out that Lubetkin w40x
observed similar enhanced nucleation rates in his experiments, because of a
lowering of the free energy barrier.
It is evident that the nucleation threshold obtained in many studies is much
lower than that expected from classical nucleation theory. Part of the problem in
accounting for the array of results is the lack of recognition by previous workers of
the kinds of nucleation possible. It is inappropriate to use equations for describing
a classical homogeneous nucleation event when the nucleation involved is of types
III and IV. It is now widely accepted that most observed nucleation originates from
pre-existing gas cavities w41x.
Harvey et al. w11,12,42,43x were perhaps the first to demonstrate the effect of
pre-existing gas cavities on bubble nucleation. When a blunt rod was moved
through a liquid, bubble generation was greater if the rear end of the rod was
either rough or dirty. They postulated the existence of preformed nuclei, referred
to as Harvey nuclei. These nuclei have a radius of curvature greater than the radius
of the critical nucleus. In order to deactivate these gas cavities, they placed the test
system under high static pressures to condense any existing vapour and dissolve any
existing gas. In all the pressurised samples, nucleation occurred at superheats of
the order of 100 K type II., as opposed to unpressurised samples in which
nucleation only required low levels of superheating.
Dean w10x described an interesting experiment in which glass, broken in air and
located in a supersaturated liquid, provides a source for bubble nucleation.
However, glass broken in the liquid does not result in a source for nucleation.
Clark et al. w44x showed that bubble nucleation only occurred at pits and scratches.
They studied active pits ranging in size from ; 8 to 75 m m. Westerheide and
Westwater w45x reported that nucleation on their micro-electrode occurred at
similar specific sites, with consecutive bubbles forming at a given site.
Chirkov and co-workers w46]49x investigated experimentally and theoretically the
conditions for spontaneous formation of hydrogen bubbles during electrolysis.
They obtained nucleation at supersaturations of the order of 10. However, the
classical theory predicted a required level of ; 10 5. Similar results were obtained
by Dapkus and Sides w50x using aqueous sulfuric acid. Chirkov and co-workers
suggested that a nucleus smaller than that of the critical size could easily be
38 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

generated by fluctuations in the solution properties. However, they found it


necessary to modify the Laplace equation.
Dean w10x was one of the first to postulate the existence of metastable micro-
bubbles. These menisci could exist in gas filled cavities or in the bulk of the
solution. During gas desorption, existing gas cavities and bubbles in contact with
the liquid should decrease in volume. In turn, their Laplace pressure should rise
which, in turn, should increase the tendency for the cavities to disappear al-
together. So how do micro-bubbles achieve a form of metastability? Presumably,
surface active molecules adsorb at the gas-liquid interface of these micro-
cavitiesrbubbles forming a permeable skin, making it more difficult for interfacial
mass transfer. The varying-permeability model, based on these ideas, was intro-
duced by Ward et al. w51x and Yount et al. w52x. Young w4x reported the work of
some investigators who have observed such nuclei microscopically. Recently Bunkin
et al. w53x obtained evidence of clusters formed by charged micro-bubbles.
The crevice model has received considerable attention by Strasberg w54x, Apfel
w55x, Winterton w56x, Crum w57x and several others. The basis of the model is that
hydrophobic impurities are believed to congregate at the bottom of crevices,
preventing the water from making contact with the solid. Atchley and Prosperetti
w41x re-examined the crevice model as applied to cavitation in order to account for
the loss of mechanical stability of the nucleus during cavitation. More recently,
Vinogradova et al. w58x concluded that nucleation was more likely on hydrophobic
surfaces simply because of the high concentration of gas filled sub-micron cavities
close to the surface.

4. Gas entrapment mechanism and gas cavity stability

Type III and IV nucleation depends upon the existence of gas filled cavities.
Some attention has been given to the mechanism of gas entrapment in cavities and
to the behaviour of the meniscus inside the cavity. This has led to a focus on the
stability of these pre-existing gas cavities. The stability of the gas cavities depends
on the radius of curvature of the meniscus, which in turn depends on the contact
angle of the gasrsolutionrsolid interface, and the geometry of the nucleation site.
As noted above, stability can also be enhanced by the presence of surfactants
adsorbed at the meniscus.
A sudden and large pressure reduction can initiate a classical nucleation event,
producing a gas filled cavity, as occurs in type II nucleation. These gas cavities are
referred to as Harvey nuclei, and are formed as a result of previous nucleation
events. However, because a very high level of supersaturation is needed, this
mechanism is considered uncommon. Gas or air. filled sites are readily generated
simply by causing the liquid to flow over the vessel surface during the filling of the
container.
The mechanism of gas entrapment by a liquid front advancing over a solid
surface was examined by Bankoff w59x. He considered the simple case of a
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 39

Fig. 4. Conditions for entrapment of gas in the advance of a semi-infinite liquid sheet across a groove
a. no gas is trapped, b. a gas cavity is produced w25,59,60x.

semi-infinite two-dimensional sheet of liquid advancing unidirectionally across a


substrate, on which it has an advancing contact angle u , over a groove of half angle
b w25,60x. The first case is shown schematically in Fig. 4a for u - 2 b . Here, as the
liquid passes over the groove it advances down the front wall, thus expelling the gas
in the cavity and fully wetting the conical groove. The second case is shown in Fig.
4b. Here, if u ) 2 b , the liquid should advance across the cavity before a portion of
the liquid can reach the base. In this case a gas filled cavity is formed.
Carr w60x concluded that the radius of curvature of the meniscus formed as a
result of an advancing sheet of liquid, depends on the height of the liquid wall, the
diameter of the cavity mouth, the static and dynamic advancing contact angle, and
the half angle of the groove. Eddington and Kenning w61x quantified the sizes of
the nucleation sites by determining their effective size. The actual size, of course,
was related to the effective size by the contact angle, and hence the wettability of
the substrate.
40 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

Fig. 5. Meniscus emerging from a conical cavity w62x.

Griffith and Wallis w62x, who considered the case of a nucleus emerging from a
conical cavity as shown in Fig. 5 for a contact angle of 908, investigated the stability
of the meniscus. They showed that the cavity volume varied directly with the radius
of curvature of the cavity meniscus, depending only on b and u . In fact, as the
bubble within the cavity grows, the radius of curvature increases. However, as the
bubble emerges at the cavity mouth, the radius of curvature actually decreases to a
minimum before increasing again. If the level of supersaturation is too low, the
nucleus will fail to emerge from the cavity, unless there is assistance from some
local fluctuation in the thermodynamic state w25x.
Tong et al. w63x examined the effects of dynamic contact angles and hysteresis on
the vapourrgas entrapment process. They defined criteria for determining whether
the critical radius was within the cavity or at the mouth of the cavity for liquids
with intermediate contact angles. It was shown that for highly wetting liquids the
critical bubble radius is usually less than the cavity radius. They suggested that the
changes in the contact angle, brought about by the changes in the contact line
velocity and hence the dynamic advancing and receding contact angles, might
explain the variations observed in experiments on bubble nucleation.
Lorenz, cited by Tong et al. w63x, and Wang and Dhir w64x investigated the vapour
trapping process following departure of a bubble from a conical cavity. Lorenz
developed geometrical relationships to describe the radius of a meniscus in a gas
cavity of a given volume, concluding that the ratio of the radius to that of the cavity
mouth was dependent on the static contact angle, and the cavity half angle. Kant
and Weber w65x investigated nucleate boiling from cylindrical pits and demon-
strated that only cavities of a certain size range can entrap vapour and become
active and nucleate bubbles.
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 41

5. The cycle of bubble production nucleation time

Much of the experimental evidence on bubble nucleation is inferred from


observations made of the bubble production itself. Consider now type IV nucle-
ation at a specific site on the base of the vessel. A bubble will suddenly appear and
grow according to some physical law. At a certain size it will detach, thereby
providing an opportunity for a new bubble to form, grow and detach at the same
site. There is evidence to suggest that the time taken for the new bubble to form at
the site is very important. Because, in many cases, this period is very brief, most
workers have neglected this part of the cycle, and simply focused on the observ-
able, the growing bubble. This time period, referred to by some as the waiting
time and the delay time, is referred to by the present authors as the nucleation
time.
Burman w66x observed the production of bubbles of carbon dioxide from a glass
capillary tip in a supersaturated solution, and noticed that the growth was steady,
continuous and consistent. His graphs of the bubble diameter vs. the absolute time
showed successive bubbles grew in a similar fashion, taking the same time to grow,
detaching at the same size, and taking the same time to re-appear. Burman w66x,
despite examining a series of about 30 bubbles, failed to investigate the waiting
time between the bubbles. Although the liquid]gas interface inside the capillary
was visible, he also failed to study its motion. More recently, Volanschi et al. w3x
investigated the nucleation of bubbles from an artificial site, also observing bubble
trains with regular spacings in between.
Buehl and Westwater w67x examined the growth of successive bubbles growing at
single nucleation sites. They observed irregular bubble growth from a given site
under constant conditions. The randomness was at odds with the views of Dean
w10x and clearly different to the results of Volanschi et al. w3x and Burman w66x.
Arguably, the randomness was a consequence of bubble coalescence associated
with using a high supersaturation, and hence having nucleation sites close together.
Hsu and Graham w68x produced an explanation for the waiting time of bubbles
formed on a heated surface by boiling. The bubble departure ruptured the thermal
boundary layer, and a new bubble was formed only after the effects of the
disturbance had disappeared. Hsu w69x developed a model based around the
so-called waiting time required for the liquid to attain the necessary superheat.
The nucleation proceeds when the thermal boundary layer thickness reaches a
critical fraction of the nucleus radius.
Similarly, Westwater w70x noted that the nucleation time should be dependent on
the time required for the boundary layer to be restored following bubble detach-
ment. In the case of a gas evolving solution, the time taken for the liquid around
the site of the preceding bubble to return to near its original concentration of
dissolved gas is crucial. Jones et al. w14x based their understanding of the nucleation
time on the above ideas of Westwater w70x and Hsu w69x.
Williams and Mesler w71x studied the effect of surface orientation on the delay
time of bubbles from artificial sites during nucleate boiling. They also reported on
the related work of Jakob w72x who found that the bubble growth times were
42 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

comparable to the delay times. They noted that no theoretical explanation for the
delay time existed. They referred to a statement made by Zuber w73x that the
delay time is a function of the local heat flux and of the interfacial and
hydrodynamic conditions which exist at the cavity. Williams and Messler noted,
however, that a prediction based on these effects was difficult. Their experimental
results based on vertical and horizontal sites indicate that the delay time is an
important variable in nucleate boiling, with the delay time much longer and the
bubble size much larger on the vertical sites.
Lubetkin w74,75x used an acoustic device as a method for counting the number of
bubbles produced during nucleation, following the supersaturation of carbonated
water. He also measured the nucleation time required for a bubble to appear. He
showed that the inverse of this time correlated well with the nucleation rate.
Accurate estimates of the time required for nucleation may depend upon
knowing the initial rate of growth. A change in the growth law of the bubble at the
sub-micron scale may result in significant inaccuracy when the bubble growth data
is extrapolated back to a bubble diameter of zero in order to obtain t s 0 w31x.
However, if the time period associated with the extrapolation back to a diameter of
zero is small, relative to the nucleation time, the error should also be small.

6. Bubble growth

Once the nucleation process has been completed, the bubble is free to grow, and
eventually detach from the substrate. The growth rate is influenced by a number of
factors such as the rate of molecular diffusion to the interface of the bubble, liquid
inertia, viscosity and surface tension. Although the significant factors governing the
initial growth are unclear, it is apparent that molecular diffusion eventually governs
the final growth. A generally accepted theoretical description of the growth of a
bubble requires the coupling of the equations for continuity, motion, conservation
of the diffusing species, and heat transfer.
Much of the experimental and theoretical work on bubble formation and growth
has led to the prediction of scaling relationships of the form, R ; t a, where R is
the bubble radius, and t denotes the time period of the growth. A broad range of
exponent values, a, have been reported, reflecting the difficulty of conducting
well-controlled bubble growth experiments and also the possibility of different
factors such as the Laplace pressure, inertial and viscous forces controlling the
growth. For example, in studies of nucleate boiling of water, Scriven w15x obtained
0.50, and Saddy and Jameson w76x obtained 0.75. For bubble growth from a solution
supersaturated with carbon dioxide, Buehl and Westwater w67x and Burman w66x
obtained 0.5, and Strenge et al. w77x reported exponent values of 0.312 - a - 0.512
for the boiling of pentane.
Bisperink and Prins w5x, using a model cylindrical cavity of 80.3 m m, examined
the growth of single bubbles in carbonated liquids. The solution was made
supersaturated by applying a vacuum to produce a pressure of 0.36 atmospheres,
which in turn resulted in a more significant Laplace pressure. These researchers
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 43

developed a computational model of the bubble growth, taking into account the
mass transfer by diffusion, and the effect of dynamic surface tension. Interestingly,
they obtained bubble growth predominantly of the form R ; t 0.5, but they did not
note this important result.
Scriven w15x presented a very useful solution to the problem of describing bubble
growth, covering a broad range of vapourrgas densities. A transition between
phases of different densities produces in the surrounding fluid a radial convective
motion which modifies the concentration and temperature fields and hence the
motion of the front. Scriven w15x established the influence of radial convection on
spherically symmetric phase growth controlled by diffusion. As the growth rate
increases, the front moves into the concentration field, producing a higher rate
relative to the front, resulting in a higher gradient and hence a thinner boundary
layer thickness.
Scrivens solution is usually presented as,

R s 2 b'kt 8.

where R is the bubble radius, t is the time, b a dimensionless growth parameter


and k a diffusion term. For growth by heat conduction k is the thermal diffusivity
and for growth governed by molecular diffusion k is the diffusivity. Scriven
provided a table of b values vs. a quantity f  , b 4 , a superheat or supersaturation
parameter. That is,

r L Co y Csat .
f s f ,b 4 s 9.
rG r L y Csat .
rG
s1y 10.
rL

where rG and r L are the density of the gas and the liquid, respectively, and Co
and Csat are mass based concentrations of the bulk and equilibrium concentrations
of the gas in the liquid. This important result, valid for a negligible Laplace
pressure, shows that the radius is proportional to the square root of the growth
time.
For an overall summary of the factors governing the growth of a bubble, and the
assumptions generally made, we refer the reader to the work of Bankoff w78x.

7. Bubble detachment from a solid substrate

This section is concerned with the forces that act on a bubble in contact with a
substrate during bubble growth. There are a number of forces that need to be
considered, some which assist with the bubble adhesion, and others that assist with
44 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

the detachment w30x. That is,

Fd q Fs s Fi q Fp q Fb . 11 .

According to Eq. 11., the drag force Fd and the surface tension force Fs are
responsible for holding the bubble on the substrate, and the inertial, Fi , pressure,
Fp , and buoyancy, Fb , forces are responsible for pulling the bubble off the
substrate. The relative velocity between the surrounding fluid and the bubble,
caused by the bubble growth, produces a drag force, which tends to retard the
effect of the buoyancy force, Fb . When the liquid]gas interface decreases in
velocity, the liquid inertia has a tendency to lift the bubble from the substrate. At
low growth rates, the static equilibrium result,

Fs s Fp q Fb 12.

is applicable. For a bubble of contact angle u s a , shown in Fig. 6, it follows that


2g
2p R d g sin u s y Dr gh p R 2d q Dr gV
/ 13.
R

where R d is the contact radius of the bubble at detachment, h is the bubble height
as measured downwards from the apex, R the radius of curvature at the apex, and
V the bubble volume. The surface adhesion force is balanced by the pressure force
the first bracketed term on the right hand side. and the buoyancy force the
second term on the right hand side.. To approximate the bubble geometry as a
spherical cap, 2grR 4 Dr gh. In other words, the bubble geometry will be visually
distorted by gravity if the hydrostatic pressure is comparable to the Laplace
pressure in the first term on the right hand side.. Assuming 2 R ; h, the two
pressures are comparable when R 2 ; gr Dr g . which, for water, occurs when
R ; 3000 m m. Therefore, the assumption of spherical-cap geometry should be
reasonable if R - 500 m m. It should be noted, however, that the Laplace pressure,
relative to a pressure of 1 atmosphere, is generally small when the bubble radius
exceeds 10 m m. Therefore, the Laplace pressure is significant enough to maintain
the spherical-cap geometry, but sufficiently small to neglect when describing the
growth of bubbles with a radius greater than 10 m m. So, neglecting the gravitatio-
nal terms in Eq. 13.,
2g
2p R d g sin u s p R 2d 14.
R
giving the trivial result,
Rd
Rs 15.
sin u
for a bubble of spherical cap geometry of constant internal pressure elevation,
2grR.
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 45

Fig. 6. Representation of bubble at the critical detachment condition.

Clearly, the gravitational force must be considered in order to describe the


detachment condition. A detailed analysis of the Laplace equation, taking into
account the force of gravity, leads to the principal curvatures at the contact
position, 1rR1 and 1rR 2 . A possible detachment condition occurs when these
cancel out w79x. There is then no pressure difference across the interface at the
contact position. That is,

2g 1 1
% Dr gh s g
/ q s0
R R1 R2 / 16.

Although this equation does not govern fundamentally the condition for instabil-
ity, it provides a useful approximation. For a bubble in contact with a horizontal
substrate, the contact angle should remain unaffected by any change in the
meniscus geometry. However, for a bubble anchored at the mouth of a cavity, the
contact angle is much less certain. Combining Eqs. 13. and 16. gives,

2p R d g sin u s Dr gV 17.

The bubble size is then governed by the critical condition when the surface
tension force matches the net weight force. Assuming the bubble is a sphere, with
its foot attached to a small cavity,

4
2p R d g sin a s Dr g p R 3 18.
3
46 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

where R d is the contact radius, a the effective contact angle, and R the bubble
radius.
Rammig and Weiss w80x used artificial nucleation sites with mouth diameters of
27]70 m m. Considering the bubble detachment process, they assumed that at low
superheat a static force balance applies, in which buoyancy and surface tension are
the dominant forces. Hence they found that the bubble departure radius scaled to
the cubic root of the cavity mouth radius. For a spherical cap bubble, R d s Rsin u ,
and a s u . Hence,
p Dg sin 2u s Dr gf 1 u . D 3 19 .

where f 1 u . depends on the contact angle, D is the bubble diameter, twice the
radius of curvature, R. Eq. 19., which is based on Tates Law for bubble
detachment, was employed by Lubetkin w75x.

8. Conclusion

It is argued in this paper that bubble nucleation, under most circumstances


involving no special surface preparation and only low supersaturation of order 10.,
is a consequence of pre-existing gas cavities in the surface of the container wall, in
the surface of suspended particles, or in the form of metastable micro-bubbles. The
pressurisation of a system for long periods should lead to the dissolution of these
gas cavities. However, there is evidence to suggest that the meniscus within the
surface cavity or the solution bulk achieves a meta-stability, most probably as a
result of surfactant species adsorbed at the gas-liquid interface. Research is needed
in order to further examine the stability of such micro-bubbles.
Further research is also needed in order to determine the likely distribution of
pre-existing gas cavities in a liquid solution which is subjected to equilibrium
conditions. A system containing gas cavities is not in a genuine state of equilibrium
so long as the gas cavities remain. When the system is suddenly made supersatu-
rated there exists a critical nucleation radius, R9U . A portion of the pre-existing gas
cavities will have a meniscus with a radius of curvature greater than R9U , and a
remaining portion less than R9U . The larger radii will nucleate and produce a
bubble according to the type IV classification, with a probability of 1.0. The smaller
radii will generate bubbles according to the type III classification, with a lower
probability that approaches zero as the size decreases.
A nucleation model built around these ideas is needed. The model needs to
encompass the geometry of the sites, and the variations in the radius of curvature
that occur as the embryo grows inside the cavity. This embryonic growth needs to
be examined in connection with models for describing bubble growth generally. In
the context of describing the cycle of bubble production at a given site, the
influence of the departing bubble on the concentration field surrounding the gas
cavity must also be examined. In turn, this will provide an understanding of the
physics governing the nucleation time. The classical view of nucleation is simply
inappropriate when the initial state of a system is one containing gas cavities.
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 47

Nomenclature
a exponent in bubble growth law
f1 u . geometrical coefficient dependent on contact angle
g acceleration due to gravity m sy2 .
h bubble height m.
k diffusion parameter in Scrivens equation m2 sy1 .
K Boltzmanns constant J Ky1 .
n number of critical size nuclei in a given volume
Co bulk concentration of gas in liquid kg my3 .
Csat equilibrium concentration of gas in liquid kg my3 .
D bubble diameter m.
Fd vertical drag force on growing bubble N.
Fi vertical inertial force due to liquid around growing bubble N.
Fp vertical force due to Laplace pressure at base of bubble N.
Fs vertical force due to surface tension at contact line N.
H Henrys law constant Pa mol fraction.y1 .
N number of molecules available in a given volume for nucleation
Pi system pressure Pa.
Pb pressure required for equilibrium with bulk liquid Pa.
DP pressure difference Pa.
R bubble radius m.
R9 radius of curvature m.
R9U critical radius of embryo m.
R1 principle radius of curvature m.
R2 principle radius of curvature m.
Rd bubble contact radius at detachment m.
T system temperature K.
V bubble volume m3 .
W nucleation energy barrier J.
Xi saturation mole fraction
Xb bulk mole fraction
a saturation ratio
b half angle of gas cavity groove
b dimensionless growth parameter in Scrivens equation
g interfacial tension N my1 .
dimensionless density in Scrivens equation
f supersaturation parameter in Scrivens equation
rG gas density kg my3 .
rL liquid density kg my3 .
s supersaturation
u contact angle
Dr density difference kg my3 .
D gv bulk free energy per unit of liquid volume Pa.
48 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

Acknowledgements

The authors wish to acknowledge the scholarship support of the Australia


Research Council and the Research Management Committee of the University of
Newcastle.

References
w1x W.C. Evans, Lake Nyos, knowledge of the fount and the cause of disaster. Nature 379 1996. 21.
w2x H.D. Van Liew, M.E. Burkard, Bubbles in circulating blood: stabilization and simulations of cyclic
changes of size and content. J. Appl. Physiol. 75 1993. 1379]1385.
w3x A. Volanschi, D. Oudejans, W. Olthuis, P. Bergveld, Gas phase nucleation core electrodes for the
electrolytical method of measuring the dynamic surface tension in aqueous solutions. Sensors
Actuators B 35r36 1996. 73]79.
w4x F.R. Young, Cavitation, McGraw-Hill, London, 1989.
w5x C.G.J. Bisperink, A. Prins, Bubble growth in carbonated liquids. Colloids Surf. A: Physicochem.
Eng. Asp. 85 1994. 237]253.
w6x W. Lixin, O. Fan, Hydrodynamic characteristics of the process of depressurization of saturated
water. Chin. J. Chem. Eng. 2 1994. 211]218.
w7x J.P. Malley, J.K. Edzwald, Concepts for dissolved-air flotation treatment of drinking waters. J.
Water SRT-Aqua 40 1991. 7]17.
w8x J.H. Han, C.D. Han, Bubble nucleation in polymeric liquids I. Bubble nucleation in concentrated
polymer solutions, J. Polym. Sci. B Polym. Phys. 28 1990. 711]741; Bubble nucleation in polymeric
liquids II. Theoretical considerations, J. Polym. Sci. B Polym. Phys. 28 1990. 743]761.
w9x J. Szekely, G.P. Martins, On spherical phase growth of multicomponent systems. Trans. Metall.
Soc. AIME 245 1969. 1741]1747.
w10x R.B. Dean, The formation of bubbles. J. Appl. Phys. 15 1944. 446]451.
w11x E.N. Harvey, D.K. Barnes, W.D. McElroy, A.H. Whiteley, D.C. Pease, K.W. Cooper, Bubble
formation in animals } I. Physical factors. J. Cell. Comp. Physiol. 24 1944. 1]22.
w12x E.N. Harvey, A.H. Whiteley, W.D. McElroy, D.C. Pease, D.K. Barnes, Bubble formation in animals
] I. Gas nuclei and their distribution in blood and tissues. J. Cell. Comp. Physiol. 24 1944. 23]34.
w13x S.G. Bankoff, Ebullition from solid surfaces in the absence of a pre-existing gaseous phase, Trans.
ASME 1957. 735]740.
w14x S.F. Jones, G. Evans, K.P. Galvin, The cycle of bubble production from a gas cavity in a
supersaturated solution, Adv. Colloid Interface Sci. 80 1999. 51]84.
w15x L.E. Scriven, On the dynamics of phase growth. Chin. Eng. Sci. 10 1959. 1]13.
w16x P.G.T. Fogg, W. Gerrard, Solubility of Gases in Liquids, John Wiley, Chichester, 1991, pp.
242]243.
w17x S.D. Lubetkin, M. Blackwell, The nucleation of bubbles in supersaturated solutions. J. Colloid
Interf. Sci. 26 1988. 610.
w18x M. Volmer, A. Weber, Nucleus formation in supersaturated systems. Z. Phys. Chem. 119 1926.
277.
w19x L. Farkas, The velocity of nucleus formation in supersaturated vapors. Z. Physik. Chem. 125 1927.
236.
w20x R. Becker, W. Doring,
Ann. Phys. 24 1935. 719]752.
w21x W. Doring,
Die Uberhitzungsgrenze und Zerreissfestigkeit, Z. Phys. Chem. 36 1937. 371; 38 1938.
292.
w22x Ya.B. Zeldovich, On the theory of new phase formation: cavities. Acta Physico. Chim. URSS 18
1943. 1.
w23x M. Volmer, Kinetics of Phase Formation, ATI N881935 F-TS-7068-RE., Clearing house for federal
and technical information, Springfield, Virginia, translated from Kinetic der Phasenbildung, Theo-
dor Steinkopff, Dresden, 1939.
S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50 49

w24x J. Frenkel, Kinetic Theory of Liquids, Clarendon Press, Oxford, 1946.


w25x R. Cole, Boiling nucleation. Adv. Heat Transfer 10 1974. 85]166.
w26x J.L. Katz, M. Blander, Condensation and boiling: corrections to homogeneous nucleation theory
for non ideal gases. J. Colloid Interf. Sci. 42 1973. 496]502.
w27x M. Blander, J.L. Katz, Bubble nucleation in liquids. AIChE J. 21 1975. 833]848.
w28x T.W. Forest, C.A. Ward, Effect of a dissolved gas on the homogeneous nucleation pressure of a
liquid. J. Chem. Phys. 66 1977. 2322]2330.
w29x P.J. Sides, in: White, Bockris, Conway Eds.., Phenomena and Effects of Electrolytic Gas Evolution
in Modern Aspects of Electrochemistry, vol. 18, Plenum Press, New York, 1986, pp. 303]354.
w30x S.D. Lubetkin, in: D.J. Wedlock Ed.., Bubble Nucleation and Growth, Controlled Particle, Droplet
and Bubble Formation, Butterworth Heinemann, Oxford, 1994, pp. 159]190.
w31x S.D. Lubetkin, The fundamentals of bubble evolution. Chem. Soc. Rev. 24 1995. 243]250.
w32x W.L. Ryan, E.A. Hemmingsen, Bubble formation in water at smooth hydrophobic surfaces. J.
Colloid Interf. Sci. 157 1993. 312]317.
w33x P.G. Bowers, B.-L. Kedma, R.M. Noyes, Unstable supersaturated solutions of gases in liquids and
nucleation theory. J. Chem. Soc. Faraday Trans. 92 1996. 2843]2849.
w34x J.C. Fisher, The fracture of liquids. J. Appl. Phys. 19 1948. 1062]1067.
w35x E.E. Hemmingsen, Cavitation in gas-supersaturated solutions. J. Appl. Phys. 46 1975. 213]218.
w36x P.M. Wilt, Nucleation rates and bubble stability in water-carbon dioxide solutions. J. Colloid Interf.
Sci. 112 1986. 530]538.
w37x Y. Finkelstein, A. Tamir, Formation of gas bubbles in supersaturated solutions of gases in water.
AIChE J. 31 1985. 1409]1419.
w38x Y.C. Yang, J.R. Maa, Microbubbles in water and boiling superheat. J. Colloid Interf. Sci. 120 1987.
87]93.
w39x C.-C. Chen, L.-C. Hung, H.-K. Hsu, Heterogeneous nucleation of water vapor on particles of SiO 2 ,
Al 2 O 3 , TiO 2 , and carbon black. J. Colloid Interf. Sci. 157 1993. 465]477.
w40x S.D. Lubetkin, The kinetics of nucleation of adamantane crystals from the vapor. J. Chem. Phys. 89
1988. 7502.
w41x A.A. Atchley, A. Prosperetti, The crevice model of bubble nucleation. J. Acoust. Soc. Am. 86 1989.
1065]1084.
w42x E.N. Harvey, D.K. Barnes, W.D. McElroy, A.H. Whiteley, D.C. Pease, Removal of gas nuclei from
liquids and surfaces. J. Am. Chem. Soc. 67 1945. 156]157.
w43x E.N. Harvey, W.D. McElroy, A.H. Whiteley, On cavity formation in water. J. Appl. Phys. 18 1947.
162]172.
w44x H.B. Clark, P.S. Strenge, J.W. Westwater, Active sites for nucleate boiling. Chem. Eng. Prog. Symp.
Ser. Heat Transfer, Chicago 55 1959. 103]110.
w45x D.E. Westerheide, J.W. Westwater, Isothermal growth of hydrogen bubbles during electrolysis.
AIChE J. 7 1961. 357]362.
w46x Y.G. Chirkov, A.G. Pshenichnikov, An estimate of the probability of spontaneous gas-bubble
nucleation in supersaturated solutions, translated from Elektrokhimiya, 211. 1985. 119]122.
w47x Y.G. Chirkov, A.G. Pshenichnikov, Influence of the electrode surface on gas-bubble nucleation in
supersaturated solutions during electrolysis, translated from Elektrokhimiya, 225. 1986a. 618]622.
w48x Y.G. Chirkov, V.I. Rostokin, A.G. Pshenichnikov, Gas-bubble nucleation in supersaturated solu-
tions: gathering the gas into a cavity by diffusion, translated from Elektrokhimiya, 224. 1986b.
529]531.
w49x Y.G. Chirkov, V.I. Rostokin, A.G. Pshenichnikov, Formation of subcritical equilibrium gas-bubble
nuclei in supersaturated solutions, translated from Elektrokhimiya, 232. 1987. 237]241.
w50x K.V. Dapkus, P.J. Sides, Nucleation of electrolytically evolved hydrogen at an ideally smooth
electrode. J. Colloid Interf. Sci. 111 1986. 133]151.
w51x C.A. Ward, P. Tikuisis, R.D. Venter, Stability of bubbles in a closed volume of liquid-gas solution.
J. Appl. Phys. 53 1982. 6076]6084.
w52x D.E. Yount, E.W. Gillary, D.C. Hoffman, A microscopic investigation of bubble formation nuclei.
J. Acoust. Soc. Am. 76 1984. 1511]1521.
50 S.F. Jones et al. r Ad . Colloid Interface Sci. 80 (1999) 27]50

w53x N.F. Bunkin, A.V. Kochergin, A.V. Lobeyev, B.W. Ninham, O.I. Vinogradova, Existence of charged
submicrobubble clusters in polar liquids as revealed by correlation between optical cavitation and
electrical conductivity. Colloids Surf. A: Physicochem. Eng. Asp. 110 1996. 207]212.
w54x M. Strasberg, Onset of ultrasonic cavitation in tap water. J. Acoust. Soc. Am. 31 1959. 163]176.
w55x R.E. Apfel, The role of impurities in cavitation-threshold determination. J. Acoust. Soc. Am. 48
1970. 1179]1186.
w56x R.H.S. Winterton, Nucleation of boiling and cavitation. J. Phys. D: Appl. Phys. 10 1977. 2041]2056.
w57x L.A. Crum, Nucleation and stabilisation of microbubbles in liquids. Appl. Sci. Res. 38 1982.
101]115.
w58x O.I. Vinogradova, N.F. Bunkin, N.V. Churaev, O.A. Kiseleva, A.V. Lobeyev, B.W. Ninham,
Submicrocavity structure of water between hydrophobic and hydrophilic walls as revealed by optical
cavitation. J. Colloid Interf. Sci. 173 1995. 443]447.
w59x S.G. Bankoff, Entrapment of gas in the spreading of liquid over a rough surface. AIChE J. 4 1958.
24]26.
w60x M.W. Carr, A Study of the Kinetics of Nucleation of Growth and Detachment of Carbon Dioxide
and Chlorine Bubbles Using Pressure Release Bubble Nucleation and the Quartz Crystal Micro
Balance, Ph.D. Thesis, University of Bristol, 1993.
w61x R.I. Eddington, D.B.R. Kenning, The effect of contact angle on bubble nucleation. Int. J. Heat
Mass Trans. 22 1979. 1231]1236.
w62x P. Griffith, J.D. Wallis, The role of surface conditions in nucleate boiling. Chem. Eng. Prog. Symp.
Ser. Heat Transfer, Storrs 56 1960. 49]63.
w63x W. Tong, A. Bar-Cohen, T.W. Simon, S.M. You, Contact angle effects on boiling incipience of
highly-wetting liquids. Int. J. Heat Mass Trans. 33 1990. 91]103.
w64x C.H. Wang, V.K. Dhir, On the gas entrapment and nucleation site density during pool boiling of
saturated water. J. Heat Transfer Trans. ASME 115 1993. 670]679.
w65x K. Kant, M.E. Weber, Stability of nucleation sites in pool boiling. Exp. Thermal Fluid Sci. 9 1994.
456]465.
w66x J.E. Burman, Bubble Growth in Supersaturated Solution, Ph.D. Thesis, Imperial College, Univer-
sity of London, 1974.
w67x W.M. Buehl, J.W. Wetswater, Bubble growth by dissolution: influence of contact angle. AIChE J.
12 1966. 571]576.
w68x Y.Y. Hsu, R.W. Graham, An Analytical and Experimental Study of the Thermal Boundary Layer
and the Ebullition Cycle in Nucleate Boiling, NASA Tech. note TN-D-594, 1961.
w69x Y.Y. Hsu, On the size range of active nucleation cavities on a heating surface, J. Heat Transfer,
Trans. ASME 1962. 207]216.
w70x J.W. Westwater, in: R. Davies Ed.., Cavitation in Real Liquids, Elsevier, Amsterdam, 1964, p. 34.
w71x D.D. Williams, R.B. Mesler, The effect of surface orientation on delay time of bubbles from
artificial sites during nucleate boiling. AIChE J. 13 1967. 1020]1024.
w72x M. Jakob, Heat Transfer, Wiley, New York, 1949, pp. 631]632.
w73x N. Zuber, Appl. Mechanics Rev. 17 1964. 663.
w74x S.D. Lubetkin, Measurement of bubble nucleation rates by an acoustic method. J. Appl. Elec-
trochem. 19 1989. 668]676.
w75x S.D. Lubetkin, The nucleation and detachment of bubbles. J. Chem. Soc. Faraday Trans. 1 85
1989. 1753]1764.
w76x M. Saddy, G.J. Jameson, Prediction of departure diameter and bubble frequency in nucleate
boiling in uniformly superheated liquid. Int. J. Heat Mass Trans. 4 1971. 1771]1783.
w77x P.H. Strenge, A. Orell, J.W. Westwater, Microscopic study of bubble growth during nucleate
boiling. AIChE J. 7 1961. 578]583.
w78x S.G. Bankoff, Diffusion-Controlled Bubble Growth, in: T.B. Drew, J.W. Hoopes, Jr., T. Vermeulen
Eds.., Advances in Chemical Engineering, vol. 6, Academic Press, New York, 1966, pp. 1]60.
w79x E.A. Boucher, M.J.B. Evans, Pendent drop profiles and related capillary phenomena. Proc. R. Soc.
Lond. A. 346 1975. 349]374.
w80x R. Rammig, R. Weiss, Growth of vapour bubbles from artificial nucleation sites. Cryogenics 31
1991. 64]69.

You might also like