Вы находитесь на странице: 1из 44

Advances in Colloid and Interface Science

86 2000. 3982

Dynamics of protein and mixed


proteinrsurfactant adsorption layers at the
waterrfluid interface
R. Miller a,U , V.B. Fainerman b, A.V. Makievski a,b,
J. Kragel
a , D.O. Grigoriev a,c , V.N. Kazakov d,
O.V. Sinyachenko d
a

MPI fur
Max Planck Campus, D-14476 Golm, Germany
Kolloid- und Grenzflachenforschung,

b
International Medical Physicochemical Centre and Institute of Technical Ecology,
16 Ilych A enue, Donetsk 340003, Ukraine
c
Institute of Chemistry, St. Petersburg State Uni ersity, Uni ersitetskiy Pr. 2, 1989084
St. Petersburg, Russia
d
Donetsk Medical Uni ersity, 16 Ilych A enue, Donetsk 340003, Ukraine

Abstract
The adsorption behaviour of proteins and systems mixed with surfactants of different
nature is described. In the absence of surfactants the proteins mainly adsorb in a diffusion
controlled manner. Due to lack of quantitative models the experimental results are discussed partly qualitatively. There are different types of interaction between proteins and
surfactant molecules. These interactions lead to proteinrsurfactant complexes the surface
activity and conformation of which are different from those of the pure protein. Complexes
formed with ionic surfactants via electrostatic interaction have usually a higher surface
activity, which becomes evident from the more than additive surface pressure increase. The
presence of only small amounts of ionic surfactants can significantly modify the structure of
adsorbed proteins. With increasing amounts of ionic surfactants, however, an opposite effect
is reached as due to hydrophobic interaction and the complexes become less surface active
and can be displaced from the interface due to competitive adsorption. In the presence of
non-ionic surfactants the adsorption layer is mainly formed by competitive adsorption
between the compounds and the only interaction is of hydrophobic nature. Such complexes
are typically less surface active than the pure protein. From a certain surfactant concentraU

Corresponding author. Tel.: q49-331-5679252; fax: q49-331-5679202.


E-mail address: miller@mpikg golm.mpg.de R. Miller..

0001-8686r00r$ - see front matter 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 1 - 8 6 8 6 0 0 . 0 0 0 3 2 - 4

40

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

tion of the interface is covered almost exclusively by the non-ionic surfactant. Mixed layers
of proteins and lipids formed by penetration at the waterrair or by competitive adsorption
at the waterrchloroform interface are formed such that at a certain pressure the components start to separate. Using Brewster angle microscopy in penetration experiments of
proteins into lipid monolayers this interfacial separation can be visualised. A brief comparison of the protein adsorption at the waterrair and waterrn-tetradecane shows that the
adsorbed amount at the waterroil interface is much stronger and the change in interfacial
tension much larger than at the waterrair interface. Also some experimental data on the
dilational elasticity of proteins at both interfaces measured by a transient relaxation
technique are discussed on the basis of the derived thermodynamic model. As a fast
developing field of application the use of surface tensiometry and rheometry of mixed
proteinrsurfactant mixed layers is demonstrated as a new tool in the diagnostics of various
diseases and for monitoring the progress of therapies. 2000 Elsevier Science B.V. All
rights reserved.
Keywords: Mixed adsorption layers; Liquidrfluid interfaces; Adsorption kinetics; Penetration kinetics;
Proteinrsurfactant mixtures; Proteinrlipid mixtures; Dynamic interfacial tensions; Surface viscoelasticity; Brewster angle microscopy; Maximum bubble pressure tensiometry; ADSA

Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2. State-of-the-art theoretical models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.1. Thermodynamic models for protein layers . . . . . . . . . . . . . . . . . . . . . . . . 42
2.2. Thermodynamic model for mixed proteinrsurfactant layers . . . . . . . . . . . . . 45
2.3. Fundamentals of adsorption kinetics at a liquid interface . . . . . . . . . . . . . . 49
2.4. Adsorption kinetics modelling for protein systems . . . . . . . . . . . . . . . . . . . 50
3. Adsorption kinetics of proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4. Simultaneous adsorption from mixed aqueous solutions . . . . . . . . . . . . . . . . . . . 57
5. Proteinrlipid adsorption from two separate solutions at the joint interface . . . . . . 61
6. Penetration of proteins into lipid layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
7. Protein adsorption at waterroil interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
8. Peculiarities in the protein adsorption at drop and bubble surfaces . . . . . . . . . . . 66
9. Dilational elasticity of a protein adsorption layer . . . . . . . . . . . . . . . . . . . . . . . 69
10. Surface tension and elasticity of blood serum as example of a natural proteinrsurfactant mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
11. Summary, conclusions, outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

1. Introduction
Proteinrsurfactant mixtures are widely used in many technologies, for example
for the stabilisation of emulsions and foams in the food industry w1x or for coating

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

41

processes in the photographic industry w2x. Mixtures of proteins and surfactants are
also common to all biological systems. For example, blood serum is a mixture of
human serum albumin with a number of compounds, including low-molecular
surface-active molecules. The dynamic surface tension of such biological fluids is
used as a diagnostic and therapeutic tool, for example for renal diseases and other
pathologies w3x. This implies the scientific and practical importance of studies on
the adsorption behaviour and its dynamics of mixed proteinrsurfactant systems
which are intensively studied by many groups using various experimental techniques w412x.
However, only very few systematic studies dealing with this problem have been
published so far. The effect of low molecular non-ionic surfactants, such as
alcohols, acids and oxyethylated ethers, on the adsorption dynamics of a protein
depends on their concentration and surface activity and on the solution conditions,
such as temperature, ionic strength and pH.
Non-ionic surfactants, for example the highly surface-active oxyethylated
TWEEN 20 w13x, produce almost no effect on the surface tension of -lactoglobulin solution at short adsorption times, but leads to a significant decrease of the
equilibrium surface tension. One can assume that these types of surfactants adsorb
competitively with the protein. However, no definite predictions can be made
concerning the effect of ionic surfactants on the adsorption dynamics because in
this case the inter-ion interaction between surfactants and proteins can result in
the formation of proteinrsurfactant complexes w14x. For very small added amounts
of ionic surfactants 100 times lower than the protein concentration . even a small
increase in the surface tension can be observed. For larger amounts of the same
surfactant the surface or interfacial tension generally decreases. In certain concentration regions for some proteins an anomalous adsorption behaviour was observed
w15x.
In the present review mainly experimental studies of the adsorption behaviour of
mixtures of various proteins with different ionic and non-ionic surfactant are
shown. As proteins the most frequently studied human serum albumin HSA.,
bovine serum albumin BSA., -casein -CS. and -lactoglobulin -LG. are
chosen. The surfactants on which the analysis is focussed here are selected such
that their surface activity is comparable. On the basis of a qualitative analysis,
schematic pictures for the different types of adsorption layers are drawn. First the
state-of-the-art of the theoretical description is reviewed, where mainly models for
pure proteins are given. Then the kinetics of adsorption of proteins in the absence
and presence of surfactants are described in Sections 3 and 4. Sections 5 and 6
focus on proteinrlipid systems and competitive as well as penetration systems are
reviewed. Results for the adsorption of proteins at the waterroil interface and
peculiarities for studies at drops and bubbles are given in Sections 7 and 8, while
Section 9 deals with the dilational rheology of protein layers. Section 10 is
dedicated to an interesting new field of research, the cross-linking of surface
science and medicine, where fundamental knowledge on mixed proteinrsurfactant
layers at a liquid interface is required to use measured interfacial phenomena as a

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

42

tool of medical diagnostics. Ideas on the structure of proteinrsurfactant complexes


as inspiration for future experimental and theoretical studies to describe such
mixed systems are given in Section 11.

2. State-of-the-art theoretical models


The practical importance of the adsorption of proteins at fluid interfaces
stimulated the development of different models under equilibrium and dynamic
conditions as demonstrated in recent reviews w1618x. Besides the statistical
theories and scaling models of de Gennes w19x, for example, there are many
thermodynamic models which are based on the Butler equation w20x for the
chemical potentials of the components in the bulk phase and at the interface as the
starting point. Examples for such models are those of Joos w21x, Ter-MinassianSaraga w22x and Lucassen-Reynders w23x which describe various details of adsorption layers at interfaces.
2.1. Thermodynamic models for protein layers
The Butler equation is suitable for describing the specific demand, , of
different adsorbed molecules at an interface. For surfactant solutions this can be
expressed by the adsorption relation of Joos which is the quantitative equivalent
for an interface of the principle of BraunLe Chatelier w24x. This adsorption
relation says that when an adsorbed molecule may occupy different parts of the
interface, then at small surface pressures, , a maximum molar surface area, , is
occupied, whereas minimum is achieved at large . This means that the
adsorption layer thickness increases with increasing protein concentration. An
illustration of the idea is given schematically in Fig. 1.
With increasing surface coverage, i.e. stronger competition between the adsorbed protein molecules, their molar area becomes smaller until finally a minimum
area is reached. A thermodynamic model has been recently derived by Fainerman
et al. w25x using the Butler equation as starting point, the results of which can be
summarised as follows.
The equation of state of the adsorption layer as dependence of surface pressure
s o y on surface concentration reads o and are the surface tensions of
the solvent and solution, respectively.
sy

RT

ln 1 y .

1.
n

where the total adsorption is defined by s


given by

i , and the average molar area is


is1

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

43

Fig. 1. Idea of the protein adsorption model: a. schematic of the conformational changes of adsorbed
protein molecules; b. molar area occupied by one adsorbed molecule; the states 14 correspond to
increasing surface pressure or surface coverage , i.e. decreasing average molar surface area .
n

i q1.exp wy i 1 . r RT .x

i i
s

iG1

s 1

is1
n

2.

iG1

exp w y i 1 . r RT .x

is1

The total adsorption can be expressed by the adsorption in the state of


minimum partial molar surface area 1 , the adsorption in the ith state i by the
total adsorption
n

s 1

i exp
is1

i s

i y 1 . 1
RT

i exp  y w i y 1 . 1 x r w RT x4
n

3.

4.

i exp  y w i y 1. 1 x r w RT x4
is1

The adsorption isotherm as dependence of surface pressure on protein bulk


concentration, c, reads
b1 c s

1 y exp w y . r RT .x
n

5.

i exp wy i 1 . r RT .x
is1

Here is a constant which determines the variation in surface activity of the


protein molecule in the ith state with respect to the state 1 characterised by a

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

44

minimum partial molar area 1 s min , bi s b1 i . The value i can be either


integer or fractional and the increment is defined by i s r 1. For s 0 one
obtains bi s b1 s const., while for ) 0 the bi increases with increasing i . The
system of Eqs. 1. 5. completely describes an ideal adsorption layer of proteins.
For a non-ideal adsorption layer instead of Eq. 1. we obtain the equation of
state containing a term representing the interaction with the parameter a while it
is assumed that molecules in all conformations have the same interaction behaviour.. Taking into account a non-ideality adsorption layer and the contribution
of the DEL one can transform the equation of state for the surface layer 1. into
the following relationship w26x
sy

RT

ln 1 y . q a .

4 RT
F

2 RTc . 1r2 w ch y 1 x

6.

where F is the Faraday constant, z is the number of non-bound unit charges in the
protein molecule, s zF 0r2 RT, 0 is the electric potential of the surface, is
the dielectric permittivity of the medium and c is the total concentration of ions
within the solution. For protein solutions at high ion concentrations, the Debye
length s RTrF 2 c .1r2 is small. This means that for protein solutions the DEL
thickness can be smaller than the adsorption layer thickness. Therefore, the
concentration of ions in Eq. 6. is just their concentration within the adsorption
layer. It follows from Eq. 6. that for large c the approximation < 1 can be
used. Thus, one obtains an equation of state for non-ideally charged surface layers
of a protein w26x
sy

RT

ln 1 y . q a y a el . .

7.

where a el s z 2 Fr 8 RTc .1r2 , and for the adsorption isotherm of Eq. 5. we


now get
i exp  ya2 2 w i 1r . y 1 x y i 1r . y 2 a
q2 a elrz . q w a elrz . x

b1 c s

i 1 y .

i ir

8.

The simplifications discussed by Fainerman and Miller w26x, a el 4 a, a elrz f ya,


allow to simplify the equation of state for protein surface layers wEq. 7.x and the
adsorption isotherm equation wEq. 8.x to
sy

RT

ln 1 y . y a el 2 2

9.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

b1 c s

45

10.

1 y . 1r

The total adsorbed amount in these equations can be expressed via the adsorption
in state 1
n

s 1

i exp
is1

1 i y 1.

exp y

i y 1 . 1
RT

11 .

where the first exponential factor arises due to the non-ideality of entropy of
mixing.
Knowing the model parameters 1 s min , , n s 1 q max y min .r, ,
a el and b1 the two dependencies c . and . result, and hence the important
surface pressurerbulk concentration isotherm can be obtained and compared to
experimental data, as is shown in Fig. 2 for the proteins under discussion here. The
experimental data are described by the thermodynamic model using the parameters given in Table 1.
As one can see, the theory describes the experimental data properly and only
above a certain critical concentration is there a deviation which shows the limits of
the model. For G 2025 mNrm the theoretical model given by Eqs. 9. 11.
predicts an unrealistically sharp increase of surface pressure with a small increase
of protein concentration, and simultaneously a slight increase in adsorption. This
contradicts the experimental data which show that, starting from some protein
concentration, the value remains almost constant, while the adsorption continues to increase. This leads to an increased coverage up to an almost complete
saturation of the adsorption layer at high protein concentration. The difference
between the model and the experimental data could be attributed to the formation
of a second adsorption layer. The fact that the surface pressure of concentrated
protein solutions is independent of the bulk concentration can also be satisfactorily
explained in the framework of a monolayer model, considering a weakening of the
inter-ion interactions in concentrated monolayers w26x or the possibility of twodimensional aggregation of adsorbed protein molecules w27x.
2.2. Thermodynamic model for mixed protein r surfactant layers
There are no thermodynamic models for mixtures of a protein with a surfactant
in the literature so far. Let us derive here, however, a simple but quantitative
model for such systems, using the assumption of an ideal layer with respect to the
adsorption enthalpy and the same starting point as above for a protein solution
without any surfactant. For i different surfactants or proteins. which exist in j
different states at the interface the following very general equations of state
sy

RT

ln 1 y i j
i, j

12 .

46

Protein

ma x
nm2 rmolecule.

min
nm2rmolecule.

nm2 rmolecule.

BSA, HSA

80 " 20

40 " 10

40 " 20

-LG

20 " 5

6"2

1.0 " 0.5

-CS

100 " 20

5"2

2"1

Values in parentheses are for the waterrn-tetradecane interface.

b
lrmol.

ae l
r.

Molecular
weight

2.8 = 10y8
6.5 = 10y8 .a

180 " 40

69 000

2.8 = 10y8

40 " 10

18 400

6.9 = 10y6
9.4 = 10y7 .a

70 " 20

24 000

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

Table 1
Isotherm parameters of some selected proteins wwith respect to Eqs. 2. 4., 6. and 7.x

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

47

Fig. 2. Surface pressure isotherms of proteins: BSA B., -LG . and -CS I..

and adsorption isotherm


i1

bi1 c i s

1 y i j

13.

i1 r

i, j

can be obtained w28,29x, where the average molar area of all interfacial states and
components is defined by

i j
ij

i, j

14 .

i j
i, j

The ratio of the adsorption in any arbitrary state j to the adsorption in a certain
state k e.g. the state with minimum area value. is expressed by the generalised
relationship w30x
i j
i k

s exp

ik y i j

i j

/ /
ik

exp

ik y i j .
RT

15.

which can be referred to as the generalised Joos formula. For the non-ideal
behaviour of the system which contains a single component capable of existing in a
different state, the rigorous thermodynamic expressions are far more complicated
than those listed above see, e.g. w30x.. The relevant mathematical formalism
becomes yet more involved if the ionisation contribution into the surface pressure
of the adsorption layer and into the chemical potentials of surface active ions
located in the diffuse region of the double electric layer is taken into account w30x.
Therefore, to describe the adsorption behaviour of the proteinrsurfactant mixture,
one should introduce the assumptions which simplify the problem significantly.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

48

The mixture of protein component 1. existing in the state with minimal surface
area i.e. when the surface pressure values are not too low. and the surfactant
existing in the single state component 2. characterised by the ideal behaviour in
the monolayer, was considered by Miller et al. w31x. The equation of state for a
mixed monolayer was shown to be
sy

RT

ln 1 y . y a el 12 21

16 .

while the expressions for the adsorption isotherm of protein and surfactants are
b 1 c1 s

b2 c2 s

1 1

17 .

1 y . 1r
2 2

18.

1 y . 2r

where s 1 q 2 . The average molar area of adsorbed components 1 and 2 can


be expressed by w31x
s

1 1 q 2 2

19 .

1 q 2

The relation between the adsorptions of protein and surfactant can be derived
from the adsorption isotherms of Eqs. 17. and 18.:
1 1
2 2

b 1 c1
b2 c2

1 y .

1 y 2 .r

20.

For 1 4 2 at a given ratio of the concentrations in the solution bulk, the


portion of protein in the surface layer decreases sharply with the increase of the
total adsorption . For a el s 0, Eq. 16. reduces to the known relationship
proposed by Joos w24x for the mixture of two surfactants in an ideal monolayer
1 1
2 2

b 1 c1
b2 c2

exp y

1 y 2 .

21.

RT

At low surfactants concentrations and adsorptions, as 1r ( 1 and 2r < 1,


an approximation follows from Eqs. 17. and 18.
1 1 s b1 c1 1 y b 2 c 2 . r 1 q b1 c1 .

and

2 2 s b 2 c 2

22 .

Using Eq. 16. and the corresponding equation of state for the protein solution,
one obtains the expression for the surface pressure jump for the mixture

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

12 s

RT
1

ln

1
1 y b2 c2

RTa el

49

b 1 c1
1 q b 1 c1

2 b2 c2 y b2 c2 .

23 .

where 12 is the extra decrease of surface tension for the solution of component
1, caused by the addition of component 2. It can be seen that, for certain
relationships between the parameters entering Eq. 23., the first term in this
equation can be neglected as compared with the second term. This results in the
negative value of the surface pressure jump of the mixture, that is, the surface
tension of the proteinrsurfactant mixture can exceed that characteristic to the
system where no surfactant is added. Moreover, calculating the extreme value of
the second term of Eq. 23. one sees that the surface tension maximum of mixture
is located at b 2 c 2 s 1. In the region where the surfactant concentration is high, we
calculate from Eq. 20. that almost all protein molecules are expelled from the
surface layer. This constitutes a crucial difference between surfactantrprotein
mixtures and mixtures of different surfactant or proteins. For such mixtures, the
exponential factor can be excluded from Eq. 21.. A simple expression follows for
21 that is, for the variation of surfactant surface tension caused by the addition
of protein. in the system studied, the inequalities 2 < 1 and 2 2 < 1 1 hold.
Transforming Eqs. 16. 18. one obtains
21 s

RTa el
2

1 1 . 2 (

RTa el
2

b 1 c1 . 2

1
1 q b2 c2

2 1 r

24 .

From a generalised SzyszkowskiLangmuir equation for two component mixtures


the following relationship for 12 results
12 s

RT

ln 1 q

b2 c2
1 q b 1 c1

25 .

As the power index in Eq. 24. is extremely high, the value of 21 does not
exceed 1 mNrm. This approximate theoretical model which attempts to explain the
anomalous behaviour of the proteinrsurfactant mixture was recently confirmed
using the HSArC 10 DMPO mixture as an example w31x.
2.3. Fundamentals of adsorption kinetics at a liquid interface
All quantitative descriptions of adsorption kinetics processes are so far based on
the model derived by Ward and Tordai w32x. Differences in the models developed
on this basis consist mainly of the boundary and initial conditions as reviewed
recently w33x. The diffusion-controlled adsorption model of Ward and Tordai
assumes that the step of transfer from the subsurface to the interface is fast
compared to the transport from the bulk to the subsurface. It is based on the
following equation

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

50

c
t

sD

2 c

at

x2

x ) 0, t ) 0.

26.

As very suitable boundary condition Ficks first law is used at the surface located at
xs0

sjsD

x s 0, t ) 0.

at

27 .

For a mixture of different surface-active molecules a diffusion equation for each


component is needed. To complete the transport problem one additional boundary
condition and an initial condition are necessary. The typical boundary conditions is
an infinite bulk phase
lim c x,t . s c o

t)0

at

28 .

and the usual initial condition is a homogenous concentration distribution and a


freshly formed interface and zero adsorption
c x,t . s c o

at

ts0

29.

0 . s 0.

30 .

For surfactant mixtures the conditions are chosen equivalent to those for a single
surfactant system. The solution of this initial and boundary condition problem is
c 0,t . s c o y

't
'D H0

d t y .
dt

d'

31 .

or the equivalent relationship


D

t. s 2

c o't y

H0't c 0,t y .d'

32 .

The use of Eqs. 31. and 32. to describe the adsorption kinetics of a surfactant
from solution requires an additional relation c . w33x.
2.4. Adsorption kinetics modelling for protein systems
This is true also for proteins. However, proteins can undergo conformational
changes in the adsorption layer, as is demonstrated in Fig. 3.
Due to increasing adsorption, the degree of unfolding becomes smaller with
advancing time. At low bulk concentrations, however, the area at the interface
covered by adsorbed molecules is small and the molecules can occupy a maximum
interfacial area. A kinetics model should take all this into consideration. Therefore
a comprehensive kinetic adsorption theory for proteins can certainly not be worked

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

51

Fig. 3. Changes of protein configuration at the interface during adsorption at different protein bulk
concentrations, according to Wustneck
et al. w34x.

out in the very near future. Here we can give only some general conclusions which
result from the evolution of protein molecules in different states in an adsorption
layer.
The very first contact of a protein molecule with the interface is characterised by
a surface area 1. The adsorbed molecules rearrange then at the interface to
establish the equilibrium value i0 , which is governed by 0 , i.e. ., and the
number of different molar areas. For an ideal equilibrium adsorption layer the
amount of molecules adsorbed in the ith state is given by w25x
i0 s

cb1 i

1 y . i 1r

33 .

Under dynamic conditions i - i0 and protein molecules increase their number of


adsorbed segments each occupying an area 1 , while when i ) i0 earlier adsorbed segments rearrange to leave the interface. The transfer between the different
states may be described by a first-order reaction
kq
i

kq
iq1

ki

k iq1

iy1 l
i l
iq1
y
y

34 .

A rate constant with q indicates an increase, a constant marked by y leads to a


decrease of the partial molar surface of the protein by a value s 1. From Eq.
30. follows the equation of adsorption kinetics for a protein in i states
di
dt

q .
q
y
s yi ky
i q k iq1 q iy1 k i q iq1 k iq1 q I i

35.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

52

with Ii being the diffusion flux of the ith state from the bulk of the solution to the
surface. In the present model Ii s 0 for all i G 2. For i s 1 we get from Eq. 31.
d1
dt

q
s 2 ky
2 y 1 k 2 q D

/
x

36 .
xs0

The diffusion flux Ii can be obtained from Eq. 31.. As a good approximation we
can use w35x
I1 s D

/
x

s w c o y c 1 .x
xs0

/
t

1r2

37.

The protein concentration in the sublayer c 1 . can be determined via the


adsorption isotherm Eqs. 5. 8.. Eq. 35. is quite complicated for further analysis
so that some simplifications have to be introduced. From experimental experience
w34,3640x we know that the proteins follow a diffusion-controlled adsorption
mechanism at the airrwater interfaces, at least in the range of small surface
pressure F 2 mNrm. Particularly from refs w36,39,40x, we can deduce that the
time tU at which the surface pressure starts to increase, i.e. the time to reach a
certain adsorption m , fulfils the relation c 2 tU ( const. in the protein concentration interval c s 0.050.001 grl. This confirms a diffusion model in the range of
small for which we get as approximation from Eq. 32. neglecting the integral.
w x 0 s 2 c o

Dt

38 .

Usually at surface pressure ) 2 mNrm the kinetics of adsorption is slower than


predicted by Eq. 38.. Therefore it can be concluded that, at least for low
concentrations, the rearrangement or conformational changes and desorption of
adsorbed segments of the protein molecules are fast enough and do not control the
overall adsorption process, while for higher surface concentrations or surface
pressures, these processes start to be of importance.
Processes of unfolding and rearrangement with increasing saturation of the
adsorption layer may be explained by the influence of the protein concentration on
the adjusting time of adsorption equilibrium. These processes may cause some
strong effects on the surface rheology too. Also the effect of compression and
dilation of the surface layer on the process of adsorption have to be discussed in
the frame of the model of different molecular states of the protein at the interface.
For mixed proteinrsurfactant systems the adsorption equilibrium is understood
only qualitatively. About the adsorption kinetics there are no comprehensive
attempts at a theoretical description. For particular cases with respect to composition and time interval, a semi-quantitative interpretation is possible by separating
the effect of the two components w41x. This holds true also for penetration
experiments where a protein adsorbs at an interface covered partly by an insoluble
surfactant layer. In particular, this field of theoretical modelling deserves more

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

53

attention in the near future in order to provide a basis for understanding these
biologically and medically important mixed systems.

3. Adsorption kinetics of proteins


The adsorption kinetics, mainly studied by dynamic surface tension measurements, shows many features very much different from that of typical surfactants.
The interfacial tension isotherms and dynamic surface pressure curves for the
standard proteins like BSA, HSA, -casein and -lactoglobulin were measured by
many authors using various techniques, however, mainly for the solutionrair
interface only w4149x. First the beginning of the adsorption process shows
pronounced periods of time where adsorption proceeds, but the interfacial tension
does not change. This period is called induction time.
Once the interfacial tension starts to decrease due to adsorption it often exhibits
extremely steep dependencies, caused by the shape of the adsorption isotherm,
which is also very steep in a relatively narrow concentration interval. The model
equations wEqs. 9. 11.x give a qualitative explanation of the nature of the
induction time. In agreement with the dependencies on as given by
Fainerman and Miller w26x, the surface tension begins to decrease only from a
certain monolayer coverage . min . This minimum coverage is determined by
the adsorption model parameters. For standard proteins like BSA, HSA, -casein
and -lactoglobulin . min is in the range 0.10.2. The time necessary to
reach this minimum coverage is called induction time. The reason for its existence
is the rather small contribution of the surface layer entropy to the surface pressure.
One can see from Eq. 9. that the contribution of entropy of mixing to the surface
pressure, determined by the factor RTr is two orders of magnitude lower than
the factor RTr for surfactants. This essentially distinguishes the adsorption
behaviour of proteins and surfactants: for the latter no induction time exists.
In this section first a discussion of the problems related to low concentrations
used in experiments with proteins is given. Then an example of extremely long
induction times is discussed observed for very low protein concentration. For the
standard proteins the typical dynamic surface tensions are shown then, however, a
quantitative interpretation fails so far due to lack of theoretical models.
In several experimental techniques the adsorption at the interface, or sometimes
even at the surface of the container, tubes and connectors of dosing systems, can
lead to a depletion of protein in the bulk. Estimations have shown that the protein
mass in the bulk of a drop and in the adsorption layer, are comparable for drops of
a radius 1.5 mm and a bulk concentration of c - 20 mgrl w50x. The use of the
pendant drop method, however, may be considerably extended to small surface
pressures, usually - 2 mNrm w36,3840x and hence to low bulk concentrations
when taking into account the protein mass balance.
On the other hand, it becomes possible to determine the protein adsorption for
small directly from drop experiments, as was shown by Miller et al. w50x, which
makes this technique complementary to experiments such as radiotracer tech-

54

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

Fig. 4. Dynamic surface tensions log t . for -CS at different low bulk concentrations exhibiting the
phenomenon of an induction time, c -CS s 10y1 0 ., 2 = 10y1 1 I., 10y1 1 ^. molrcm3 , according
to Miller et al. w50x.

niques or ellipsometry w51,52x. On the basis of the above-mentioned thermodynamic model, the results from drop measurements have been compared to data
from other experiments for -CS and -LG and very good agreement was found
w53x.
The typical feature of protein adsorption kinetics, the existence of an induction
period, was analysed semi-quantitatively by Miller et al. w50x using -CS as model
protein. The general physical picture of the induction period is that there is
minimum adsorption of a protein necessary at the interface until the surface
pressure starts increasing. The flexible -CS molecule is known to unfold and
rearrange at low adsorption layer coverage w51,52x.
To ensure that no artefacts simulate the induction time, different complementary methods have been used, such as maximum bubble pressure method, drop
volume technique, and pendant drop technique, which differ in the time scales
from 0.001 s up to 10 5 s. Dynamic surface tensions of -CS solutions at concentrations from 10y1 1 to 10y7 molrcm3 were measured. For the lowest concentrations
the results are shown in Fig. 4.
A decrease of surface tension of approximately 2 mNrm is found at c s 10y1 1
molrcm3 after approximately 5000 s, and for c s 2 = 10y1 1 molrcm3 this surface
pressure is reached after approximately 700 s. According to the theoretical model
derived by Miller et al. w50x, a diffusion coefficient for -CS of approximately
D s 1.5 = 10y6 cm2rs and an adsorption of s 0.2 mgrm2 at s 2 mNrm is
obtained. Similar results have been obtained from radiotracer and ellipsometry
measurements w38,51,52,54x. The minimum adsorption layer thickness min of -CS
is calculated to be approximately 0.5 nm. Such a thickness corresponds to a flexible
unfolded protein chain at the interface.
Using a diffusion coefficient of 10y6 cm2rs for HSA, the time induction time.
and the minimum adsorption min can be estimated at which the surface tension

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

55

Fig. 5. Dynamic surface tension t . of -LG at the waterrair interface, phosphate buffer at pH 5,
22C: c -LG s 10y1 0 ., 10y9 B., 5 = 10y9 ., 2 = 10y8 '. molrcm3 , according to Wustneck
et

al. w34x.

Fig. 6. Dynamic surface tension t . of -CS at the waterrair interface, phosphate buffer at pH 7,
22C: c -CS s 10y1 1 ., 5 = 10y1 1 B., 10y1 0 '., 10y9 ., 10y8 I. and 2 = 10y8 ^. molrcm3 ,
according to Wustneck
et al. w34x.

56

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

starts to decrease. Assuming that min does not depend on the protein bulk
concentration for HSA of 7.25 = 10y1 0 and 1.45 = 10y9 molrcm3 , the corresponding induction times are approximately 20 s and 5 s, respectively. These results agree
very well with those reported by Miller et al. w55x, i.e. the process of unfolding
seems to be very quick. A different situation was found for -CS where a
relaxation time of unfolding of 500 s was estimated w50x.
The adsorption kinetics of proteins have been first systematically investigated by
Graham and Phillips w38x and later by other authors, as reviewed in w34x. The
typical course of the dynamic surface tension for some -LG and -CS solutions
are shown in Figs. 5 and 6.
Different models based on effective diffusion coefficients Deff have been applied
for the interpretation of these data, however, the results show Deff far from
physically reasonable values. In particular with decreasing concentration, values of
Deff are found several orders of magnitude higher than expected from the size and
shape of the molecules. The reason for these discrepancies is surely that the
models had been based on a Langmuir isotherm as no other quantitative models
exist. Moreover, no assumption of any changes of the conformation of adsorbed
molecules at the interface has been made. Future theoretical work has to be
focussed on a generalisation of the adsorption kinetics models using the new
thermodynamic models and the kinetics of protein unfoldingrrefolding in the
interfacial layers, as discussed by Eqs. 34. 37..
For HSA solutions the dynamics of surface tension decrease is rather different
from that of surfactants. One can see from Fig. 7 that for HSA concentrations
c F 10y1 0 molrcm3, almost no surface tension decrease was observed during the
first 200 s and it took more than 10 h for equilibrium to be attained w49x. For higher
concentrations the induction time as discussed above decreases quickly and the
dynamic surface tension decreases in a way observed for usual surfactants. The

Fig. 7. Dynamic surface tension of HSA at the waterrair interface, phosphate buffer solution at pH 5,
temperature 22C: c HS A s 2 = 10y1 1 ., 3 = 10y1 1 `., 5 = 10y1 1 ., 7 = 10y1 1 ., 10y1 0 '.,
5 = 10y1 0 ^., 10y9 B., 10y8 I. molrcm3 , according to Miller et al. w31x.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

57

comparison of the data with existing theoretical models, however, does not yield
reasonable results so that also for HSA better kinetic models have to be derived
taking into consideration the change in the molar area with increasing surface
coverage w31x.

4. Simultaneous adsorption from mixed aqueous solutions


While studies on mixed proteinrsurfactant systems have been qualitative over a
long period of time, in recent years more systematic investigations on specific
features have been performed: food proteins mixed with surfactants w14,31,34,56x or
with lipids w5759x.
For the investigation of mixed systems, various surfactants have been chosen in
the literature. The adsorption isotherms of the following surfactants are displayed
in Fig. 8: the cationic hexadecyl trimethyl ammonium bromide CTAB.; the anionic
sodium dodecyl and tetradecyl sulphate SDS, STS.; and the non-ionic surfactants
decyl dimethyl phosphine oxide C 10 DMPO. and polyoxyethylene 20 sorbitan
monolaurate Tween 20.. SDS and Tween 20 are the most frequently used
compounds in this area of research. As one can see, their surface activity is quite
different and thus a reasonable comparison of their effect on mixed
proteinrsurfactant adsorption layers is difficult w60x. On the other hand, the
surface activities of CTAB, STS and C 10 DMPO are comparable, which was the
reason for the extensive studies started recently w61x.
The dynamic surface tension of -LGrCTAB mixtures at the waterrair inter-

Fig. 8. Surface pressure isotherms at the waterrair interface of the surfactants discussed in this study:
C 10 DMPO `., CTAB I., STS ^., Tween 20 B. and SDS ..

58

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

Fig. 9. Dynamic surface tension of -LGrCTAB mixtures at the waterrair interface at 25C,
phosphate buffer solution: c -LG s 10y9 molrcm3 ; cCTAB s 10y10 I., 5 = 10y9 `., 10y8 ^., 10y6
molrcm3 ., according to Miller et al. w61x.

Fig. 10. Schematic of the proteinrsurfactant interaction changing with the ionic surfactant concentration: a. free protein molecule; b. free surfactant molecules. Steps of interaction: i. first interaction via
ionic interaction; ii. increasing ionic interaction until all charges are saturated; iii. starting hydrophobic interaction; and iv. increased hydrophobic interaction.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

59

face is shown in Fig. 9 for a constant -LG concentration and increasing amounts
of surfactant. While the dynamic curves t . for the two lowest CTAB concentrations are more or less identical to the data for pure -LG, the highest studied
concentration yields dynamic surface tensions completely controlled by the surfactant. In the intermediate range, the contribution of the proteinrsurfactant complex
formed in such mixed systems dominates w14,60,61x. The adsorption kinetics for this
mixing ratio is thus controlled by the adsorbing complex having a surface activity
different from that of the protein molecules.
The principle of interaction between protein molecules and ionic surfactants is
shown schematically in Fig. 10. At low surfactant concentration a first interaction
starts due to electrostatic interaction state I., which proceeds until the available
charges are saturated by the surfactant ions state II.. In this state the complexes
are much less soluble than the protein and also have the highest surface activity. It
can happen that in this state a part of the complexes precipitate w2x.
With further increase of the surfactant bulk concentration, increasing amounts
of surfactant molecules interact with the complexes via hydrophobic interactions,
making the complex step-by-step more hydrophilic and hence more soluble in
water and less surface active. This hydrophobic interaction proceeds until the
surface activity of the complexes reaches a comparatively low value. Due to
competition in the adsorption layer, more and more complexes are replaced by
surfactant molecules so that finally, typically at the CMC of the surfactant in term
of free surfactant ., the adsorption layer is mainly formed by the surfactant. This
picture is confirmed by studies of various other parameters of the mixed adsorption
layers e.g. w60x., where data on surface shear viscosity, adsorption layer thickness

Fig. 11. Dynamic surface pressure of HSA at a concentration of c HS A s 10y1 0 molrcm3 mixed with
the non-ionic surfactant C 10 DMPO at various concentrations; phosphate buffer solution at pH 7,
temperature 22C: cC 10 DMPO s 1 = 10y9 ., 1 = 10y8 `., 4 = 10y8 '., 7 = 10y8 ^., 10y7 B.,
2 = 10y7 I., 4 = 10y7 ., 7 = 10y7 ., 10y6 molrcm3 ., according to Miller et al. w31x.

60

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

and surface mobility have been determined independently. These data agree
exactly with the findings for the equilibrium and dynamic surface tensions.
The mixed proteinrnon-ionic surfactant systems behave significantly differently.
For example, the dynamic surface tensions for mixed HSArC 10 DMDO solutions at
22C are shown in Fig. 11. From the first addition of surfactant it becomes clear
that there is a competitive adsorption and with increasing surfactant concentration
the kinetics become more and more surfactant-like.
When one looks into the equilibrium data as given by Miller et al. w31x one can
see that for cC 10 DMPO ) 10y7 molrcm3 , the isotherms of the pure surfactant and
the mixed system constant protein concentration of 10y1 0 molrcm3 . are almost
identical and the adsorption of HSA can be assumed to be negligible.
In the concentration range cC 10 DMPO s 10y9 10y7 . molrcm3 , the surface tension of the mixtures exceed the values for the pure HSA solution, although from
Eq. 25. we get a substantial decrease of the surface tension of the solution due to
the addition of the second component, except when b1 c1 < b 2 c 2 . For 4 = 10y8
molrcm3 - cC 10 DMPO - 10y7 molrcm3 this surface tension excess amounts only to
approximately 1 mNrm, however, for c F 4 = 10y8 molrcm3, the exceeds that
of the pure HSA solution by 34 mNrm. The theoretical model of Eq. 25. cannot
explain such increase of surface tension, i.e. negative value of 12 . However, this
phenomenon is explained quite well by Eq. 23.. The location of the maximum at
the curve corresponds approximately to the value estimated from Eq. 23., that is,
b 2 c 2 f 1. For low values of a el , the mixing of components 1 and 2 cannot lead to

Fig. 12. Schematic of the proteinrsurfactant interaction changing with the non-ionic surfactant
concentration. Steps of interaction: i. first interaction via hydrophobic interaction; ii. increasing
hydrophobic interaction; and iii. increased hydrophobic interaction.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

61

an increase in surface tension. Therefore, an anomalous behaviour of surface


tension for HSArC 10 DMPO mixtures results from the large free charge of protein
molecules w25,62x.
This phenomenon can also be explained rather simply from a physical point of
view. The interaction with the non-ionic surfactant leads to less surface-active
complexes, which then cause the increase in surface tension. The effect of increasing amounts of non-ionic surfactants in a protein solution is shown schematically in
Fig. 12. The simultaneous adsorption is almost a pure competitive adsorption,
enhanced even by the stepwise hydrophilisation of the protein due to hydrophobic
interaction with the surfactant.
Surface shear viscosity studies support these findings w31x. For concentrations
cC 10 DMPO F 2 = 10y8 molrcm3, the surface layer possesses a rather high viscosity,
characteristic for pure HSA solutions while at a concentration of 7 = 10y8
molrcm3, the shear viscosity decreases sharply to almost zero, characteristic for
pure surfactant solutions. Therefore, both tensiometric and rheologic studies
indicate that the compatibility of HSA and C 10 DMDO in the mixed monolayer is
very poor, in contrast to mixtures of surface active homologues w6365x, where any
addition of a second component results always in an extra surface tension decrease
for the mixture, as given by Eq. 25..
To summarise, it can be seen from the above qualitative discussion of the
adsorption behaviour of proteinrsurfactant mixture that these systems can exhibit
rather unusual features. In general, for the arbitrary concentrations of the components, surfactant ionisation in the monolayer and the intermolecular interaction of
the components in the bulk and in the monolayer, one can expect that equilibrium
and dynamic mixed proteinrsurfactant monolayers could display various and even
unpredictable features.
5. Protein r lipid adsorption from two separate solutions at the joint interface
Investigations on the formation of mixed lipidrprotein layers at interfaces are
very helpful in understanding the nature of interactions in such interfacial layers
w6670x. Such interactions can change the hydrophobicity or hydrophilicity and
conformation of proteins significantly w66,69,70x. The interaction between protein
and phospholipids can be studied by interfacial tension measurements at the
waterrchloroform interface via the adsorption of the single components from
separate bulk phases. This means that a lipid, such as DPPC, adsorbs from the
chloroform phase while the protein, such as -LG, adsorbs from the aqueous phase
w71x. Compared to conventional studies where surfactants were mixed with the
protein in an aqueous solution, the present type of investigation provides two
advantages. It avoids the formation of complexes in the bulk phase due to
electrostatic binding or hydrophobic interaction and subsequent adsorption of
these complexes, and it excludes the influence of lipidrprotein interaction in
solution on the adsorption mechanisms.
The interfacial tensions of such systems, for example -dipalmitoyl phosphatidyl

62

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

Fig. 13. Dynamic surface pressure t . for a mixed adsorption of -LG from an aqueous 0.1 mgrl
solution outside the chloroform drop containing different DPPC concentrations: c -LG s 5.4 = 10y1 2
molrcm3 ; c DP PC s 10y10 B., 5 = 10y1 0 I., 10y9 ., 5 = 10y9 ^., 10y8 '., 5 = 10y8 `.
molrcm3 , according to Li et al. w71x.

choline DPPC. mixed with -LG, was measured by the pendent drop and drop
volume techniques w71x. Figs. 13 and 14 show selected results obtained at two -LG
concentrations in a broad DPPC concentration interval.
In a wide concentration interval the adsorption kinetics of the mixed system is
influenced by both components. However, at low protein concentration, c -LG s
5.4 = 10y1 2 molrcm3 , the adsorption rate of the mixed interfacial layer is mainly
controlled by the DPPC. When c -LG is increased up to c -LG s 4.3 = 10y1 1

Fig. 14. Dynamic surface pressure t . for a mixed adsorption of -LG from an aqueous solution
outside the chloroform drop containing DPPC at different concentrations: c -LG s 4.3 = 10y1 1
molrcm3 ; c DP PC s 10y10 `., 10y9 I., 5 = 10y9 ^., 10y8 '., 5 = 10y8 . molrcm3 , according
to Li et al. w71x.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

63

Fig. 15. Surface pressure as a function of 't for a mixed adsorption of HSA from a 5.0 = 10y1 1
molrcm3 aqueous solution outside a chloroform drop containing DPPC different concentrations:
1 = 10y8 B., 2 = 10y8 I., 3 = 10y8 ., 4 = 10y8 `., 5 = 10y8 '., 6 = 10y8 ^., 8 = 10y8
. molrcm3 ; according to Wu et al. w73x.

molrcm3 the interfacial activity of the protein abruptly increases, and at low lipid
concentrations, 10y1 1 molrcm3 - c DPPC - 10y1 0 molrcm3 , the DPPC has very
little effect on the whole adsorption process, i.e. the adsorption rate is dominated
by the protein adsorption. This behaviour is also observed at higher DPPC
concentrations when the protein concentration is further increased to 4.3 = 10y1 1
molrcm3 Fig. 14. or even at c -LG s 1.9 = 10y1 0 molrcm3 w71x. When the lipid
concentration c DPPC exceeds 3 = 10y8 molrcm3, the adsorption behaviour is very
similar to that of the pure DPPC independent of the protein concentration w72x.
A complementary and very extensive study of four phospholipids DPPE, DPPC,
DMPE, DMPC. mixed with one of three proteins -LG, -CS, and HSA, respectively. was performed by Wu et al. w73x in order to generalise the conclusions about
the role of the phospholipid and protein structure on the equilibrium as well as
dynamic interfacial tension behaviour. For this aim the lipid concentration was
increased such that an almost saturated adsorption layer was reached. It was shown
that the head group has a much stronger effect on the equilibrium adsorption state
than the chain length.
For the dynamic interfacial tensions the lipid structure plays a minor role. An
example for simultaneous adsorption of DPPC from the chloroform and the
protein HSA from the aqueous phase is shown in Fig. 15.
A quantitative analysis of the dynamic interfacial tensions is possible only at the
higher lipid concentrations where the protein adsorption plays a minor role.
Independent of the lipid and protein structure, the adsorption kinetics of the whole
mixed system appears to follow a diffusion-controlled mechanism. This might be
different for the waterrair interface where, in contrast to the waterrchloroform
interface, interfacial phase separation was observed in penetration experiments.

64

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

We will only briefly touch on this type of experiment here in Section 6, as it will be
discussed in greater detail in another paper of this issue w74x.

6. Penetration of proteins into lipid layers


Since the phenomenon of monolayer penetration was first reported by Schulman
and Hughes w75,76x, kinetics and equilibrium properties of many penetration
systems have been intensively investigated w7786x. Penetration systems consisting
of proteins and lipids, because of their particular importance in nature, are of
particular interest most of all in biophysics, biochemistry and medicine. Because of
its ideal amphiphilic properties, phospholipids form stable monolayers at the
airwater interface and can therefore be used as the simplest model for studying
protein penetration. In recent years, new methods have extended the knowledge of
various two-dimensional monolayer phases. With the recently developed Brewster
angle microscopy BAM., which allows direct visualisation of the long-range
orientation order, owing to the optical anisotropy induced by the tilted aliphatic
chains w8789x, it was possible to observe that -LG penetration can induce a
first-order phase transition in a fluid-like Langmuir monolayer of dipalmitoyl
phosphatidyl choline DPPC.. A corresponding theoretical model was derived to
describe this interesting phenomenon w90x. Interesting results on -LG penetration
into a DPPC monolayer have been obtained recently w91x. A mechanism was
developed based on the strong interaction between the lipid molecules that leads
to a separation of the compounds at the interface. Using a typical surfactant, a
C 10 DMPO, instead of the protein, the same features where observed for the DPPC
monolayers w92x. Thus, the phase transition of the lipid monolayer is not the result
of a specific effect of the penetrating protein molecules, but a general property of
penetration layers. The conditions, however, under which a penetration system
undergoes phase transition are not yet clear. In a further contribution to this issue,
the penetration process of proteins into spread lipid monolayers is described
thermodynamically in detail and experimental results are given supporting the
theoretical models w74x.

7. Protein adsorption at water r oil interfaces


The adsorption behaviour of proteins at waterroil interfaces should follow the
same general physical picture, however, specific differences have to be expected
caused by the protein structure and chemistry. At the waterrair interface the
protein molecules try to expose the hydrophobic parts upon the air phase which
leads to an unfolding of the molecules. This process can proceed as long and as far
as there is time and space at the interface available. Thus, the process of unfolding
is restricted by a competitive adsorption of other molecules, proteins or surfactants. Once a molecule is spread over a free interfacial area unfolded. it is unclear

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

65

Fig. 16. Dynamic surface pressure of HSA solutions at pH 7 for two concentrations: 2 = 10y1 1
molrcm3 B,I., 2 = 10y1 0 molrcm3 ',^.. Closed symbols, waterrair interface; open symbols,
waterrtetradecane interface.

to what extent these molecules can refold, and what will be the conformation of
the molecules after refolding.
For waterroil interfaces the situation is different, as the adsorbing protein
molecules can penetrate into the hydrophobic oil phase with the hydrophobic parts
of the molecule. This means that this type of unfolding can proceed even at a
comparatively strong competition at the interface due to adsorption of other
molecules, as the unfolding does not happen at the interface but within the oil bulk
phase. Again, the question of refolding due to increased competition or compression of the interfacial layer is difficult to answer. This question has been addressed
by many authors and is discussed in terms of interfacial protein denaturation. This
question was summarised recently by MacRitchie w93x.
Comparative studies have been performed by Graham and Phillips w38,51x and
more recently by Makievski et al. w53x for HSA, -casein and -lactoglobulin at the
solutionrair and solutionroil interfaces. The dynamic interfacial pressure as a
function of time for HSA, -casein and -lactoglobulin at the solutionrair and
solutionroil interfaces for various protein concentrations are discussed by
Makievski et al. w53x. The data indicate that the nature of the interface significantly
affects the dynamics of the adsorption process and the equilibrium adsorption
characteristics. In particular, for HSA and -casein, both the rate of interfacial
tension increase as well as the equilibrium values of at the solutionroil
interface are higher than the corresponding values at the solutionrair interface.
For -lactoglobulin, however, comparable values are observed at high bulk concentrations, while at low bulk concentrations the observed increase in was more
pronounced for the solutionrair interface.
In Fig. 16 an example of the dynamic surface pressure as a function of time t .
is given for two HSA concentrations, measured by the pendent drop technique at

66

Fig. 17.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

c isotherm for HSA at the waterrair '. and waterrtetradecane interface ^. at pH 7.

the waterrair and waterrtetradecane interfaces. The differences in the kinetics


and in the absolute values are easily recognised.
To estimate the equilibrium surface pressures in order to construct the isotherms
the curves s t . were extrapolated to t . The values obtained from two
different extrapolation procedures, lim 1r 't . and lim 1rt ., show differt
t
ences smaller than "0.5 mNrm w94x. In Fig. 17 the extrapolated equilibrium
surface pressure isotherms for HSA at pH 7 at the solutionrair and solutionroil
interfaces are plotted as a function of the initial protein concentration in the
solution.
One can see that the surface pressure isotherms measured for the waterroil
interface have significantly higher values as discussed above. This may be explained
by a larger adsorption layer thickness at the waterroil interface as compared to the
waterrair interface due to penetration of protein loops or tails into the oil phase.
The data of Makievski et al. w53x agree with those in the literature with respect to
the location of the isotherm, however, they are incompatible with the data
published by Graham and Phillips w51x in respect to the maximum surface pressure
at the solutionroil interface approx. 20 mNrm.. In particular, at high concentration the surface pressure for HSA reaches values of approximately 28 mNrm. For
-casein, values of even approximately 32 mNrm have been reported by Makievski
et al. w53x while the data reported by Graham and Phillips were only 24 mNrm.

8. Peculiarities in the protein adsorption at drop and bubble surfaces


It was already discussed above in Section 3 that the use of small solution
volumes can lead to a significant depletion of the bulk phase due to protein
adsorption, as the absolute concentrations common for such systems are extremely
low. It was also mentioned that this disadvantage inherent to single drop studies
could be turned into an advantage as the loss of protein mass can be correlated to
the adsorbed amount at the interface. In order to evaluate the correctness of this

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

67

idea and to estimate the accuracy of such a procedure, experiments with drops and
bubbles have been recently performed w95x. The main difference between these
differences is the limited and more or less unlimited bulk volume, respectively.
The adsorption of a surfactant or protein from inside a drop or from a large
external solution bulk to a bubblerdrop surface, is characterised by some additional features, which can result in a different adsorption behaviour in particular for
proteins. Firstly, the conditions for the transport by diffusion to a spherical surface
are not only different due to the available reservoir but also due to a depletion of
molecules from the bulk leading to a decrease in the bulk concentration. This
effect can be significant particularly for proteins where the concentrations of
interest are in the range of 10y1 2 to 10y8 molrcm3. In such cases the mass of the
adsorbed protein or surfactant can be of the same order as the mass of protein
inside the drop. Secondly, there are differences due to the curvature of the
droprbubble surface: in a drop, the radial field diverges decelerating the diffusion
while transport from outside a drop or bubble is enhanced due to a converging
radial field. The third difference to be mentioned is that the waterrair interface
represents an environment for adsorbing proteins on a drop surface different from
that of a bubble surface, which can have an impact on the protein conformation.
The analysis of the differences in the diffusion process have been performed by
Makievski et al. w95x. For the diffusion from an infinite bulk to a spherical drop or
bubble. surface the following approximate solution for the dynamic adsorption is
obtained w9698x
s

cDt
r

q 2c

Dt

39 .

where r is the drop radius.


The adsorption from inside a spherical drop at its surface of area A can be
expressed by an average protein concentration c as a function of time. in the drop
volume V w50x
s

V
A

c y c. s

r
3

c y c.

40 .

An approximate solution for the adsorption from inside a drop was given in w50x
sy

cDt
r

q 2c

Dt

41 .

As compared to a flat surface, where the second term in Eqs. 39. and 41. is a first
approximation for the adsorption process, the diffusion to a bubble surface from an
infinite solution leads to an increased adsorption rate wpositive sign of the curvature term in Eq. 39.x and hence to a larger adsorption, in contrast to the diffusion
from inside a drop where a deceleration of the adsorption rate is obtained
wnegative sign of the curvature term in Eq. 41.x. For a diffusion coefficient of

68

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

Fig. 18. Dynamic surface tension of -casein solutions as a function of time: 5 = 10y1 2 molrcm3
',^., 5 = 10y1 1 molrcm3 ,`., 10y9 molrcm3 B,I.. Closed symbols, drop experiments; open
symbols, bubble experiment, according to Miller et al. w50x.

D s 10y6 cm2rs, a typical value for -casein, the first terms in Eqs. 39. and 41.
exceed 10% when t ) 1000 s. For larger diffusion coefficients the difference
between the two cases becomes significant at shorter times. In Fig. 18 an impressive example is given for -casein solutions at three different concentrations.
The differences are significant and even at comparatively high bulk concentrations the reasons mentioned above lead to a distinct change in the adsorption
kinetics. It can be suspected that the minimum surface tension at a drop surface
reached at sufficiently large bulk concentration will be remarkably higher than that
at a bubble surface. This will be mainly caused by the difference in the adsorption
rate which restricts unfolding at the bubble surface and enhances it at a drop
surface.
As mentioned above, the equilibrium surface tensions of the -casein solutions
obtained by extrapolation of the dynamic tensions to t from bubble experiments are in good agreement with the data of Graham and Phillips w51x cf. Fig. 2..
At the same time, the data obtained from the drop method and plotted as a

Fig. 19. Surface pressure isotherm of -casein: data of Graham and Phillips w51x B., obtained from
pendent bubble `. and pendent drop .. Theoretical line corresponds to present bubble experiments
and data of Graham and Phillips.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

69

function of the initial protein concentration, give much lower values for the surface
pressure . For -CS this is shown in Fig. 19.
For c - 10y7 molrl this effect can be ascribed to the redistribution of the
protein between the bulk and the drop surface. The values of -casein adsorption
calculated from the mass balance condition wEq. 36.x are shown to be in a perfect
agreement with the results directly measured by Miller et al. w50x.
When the initial concentration of the protein within the drop exceeds 5 = 10y7
molrl, the loss of mass caused by adsorption is rather small and it is impossible to
explain the higher -values by this argument. Here it is certainly the difference in
adsorption rate that is responsible for diverging results: when the -casein adsorption in the drop is very slow the molecules have sufficient time to become
completely unfolded and spread over the solutionrair interface.
From Miller et al. Fig. 18, one can see that at any concentration the adsorption
process at the drop surface proceeds much more slowly than for the bubble
method. Thus, at low values the probability for a complete unfolding of the
-casein molecule is higher as compared to the bubble surface. However, for
medium and high values, one can expect that a refolding of molecules, which
were previously unfolded, could happen. This segment-by-segment refolding process
is certainly quite slow, and can be described in future by a theory given in principle
by Eqs. 34. 36..

9. Dilational elasticity of a protein adsorption layer


Rheological studies are an additional source of information about the structure
of adsorption layers, in particular of proteins at liquid interfaces. While shear
rheology yields only qualitative structure information, dilational rheology is based
on the thermodynamics and kinetics of the respective adsorption layers and is
hence a complementary method to describe a protein adsorption layer quantitatively.
The surface dilational modulus is defined as the increase in surface tension for
a small increase in surface area A by
s

d
dln A

42.

The surface dilatational modulus or viscoelasticity modulus. is a complex number


and incorporates a real and imaginary part, which correspond to the elasticity and
viscosity, respectively w99x. In the simplest case the modulus is purely elastic the
Gibbs elasticity modulus. with a limiting value
sy

d
dln

43.

This limiting value is reached if no exchange of protein with the bulk solution
s const.. exists and the surface tension is in equilibrium. Deviation from this

70

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

Fig. 20. Dynamic surface pressure and drop area A as a function of time for a 5 = 10y1 0 molrcm3
HSA solution at the waterrtetradecane interface at three subsequent trapezoidal deformations of the
drop surface.

simple limit occurs when relaxation processes at the surface or near the surface set
in. The viscoelasticity is much more sensitive to small changes in the adsorption
layer coverage than any dynamic or equilibrium interfacial tension, as it is proportional to the slope of the adsorption isotherm drd.
Dilational rheological experiments of liquid interfaces can be performed by
various methods w99x. Transient relaxation methods are possibly most suitable for
protein layers due to the extremely low bulk concentrations w100,101x. Such
relaxation experiments can be performed with pendent drops or bubbles as
described by Makievski et al. w53x. After 100200 min pre-adsorption time rapid
during 35 s. compression or expansion of the droprbubble surface by 510%
have been produced by respective drop volume changes. After a 1030 min
relaxation the surface of the drop was restored to its initial size. The whole
operation performed for the solutionrair and solutionroil interfaces was repeated
several times. An example of such experiments is illustrated in Fig. 20 for a HSA
solution at the waterrtetradecane interface. Besides the change in interfacial
tension also the change in the drop surface area is shown.
As one can see slow relaxation processes set in after a compression or expansion
of the interfacial layer, certainly due to conformational changes, as the time for
diffusional exchange would be much slower at the present bulk concentration of
5 = 10y1 0 molrcm3 HSA.
To estimate easily the dilational elasticity it is convenient to represent in a

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

71

dimensionless form
U s

sy

dln
dln A

44.

Neglecting any diffusional exchange from Eq. 39., it follows that


U s

dln
dln

45.

For some particular cases, Eq. 45. yields rather simple expressions. For example,
for an ideal gaseous monolayer s RT , Eq. 45. yields U s 1. Another much
more realistic approximation is obtained when the first term on the right-hand side
of Eq. 6. is small as compared to the second term and can be neglected. This leads
to the approximate expression
s

RT

a el 2 2

46 .

Introducing Eq. 46. into Eq. 45., one obtains U s 2. If the contribution of the
first term in Eq. 2. remains significant, which is the case for low and medium
surface pressures, then we get U ) 2. On the contrary, transitional relaxations
between the adsorption states would lead to a decrease in and hence in a
decrease of U .
To calculate the dilational elasticity from experiments the differential quotient
in Eq. 44. is replaced by a finite difference which contains experimentally
available values only
U s y

A
A

47.

The dependence of U on obtained for HSA at c 0 - 10y9 molrcm3 at both


interfaces, waterrair and waterrtetradecane are presented in Fig. 21. One can see
that for ) 5 mNrm, the experimental U values only slightly depend on the
surface pressure and lay between 1 and 3, mainly between 1.5 and 2.5. One can
thus conclude that Eq. 2. agrees satisfactorily with the data obtained in these
transient relaxation experiments. The values of U for HSA at the solutionrair
interface exceed those at the solutionroil interface. The value of U averaged over
all measured data at the solutionrair interface was 2.9, and at the solutionroil
interface U s 1.8. The increase in HSA concentration at c 0 ) 10y9 molrcm3
resulted in a slight decrease of the dilational elasticity at both interfaces.
Benjamins et al. w102x calculated the elasticity for BSA interfacial layers at the
solutionrair and solutionrvegetable oil interfaces from results of Langmuir trough
experiments with a movable barrier and from interfacial tension changes of slowly

72

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

Fig. 21. Dependence of the dimensionless visco-elastic modulus U on the surface pressure for HSA
solutions. Open symbols, waterrair interface; closed symbols, waterrtetradecane interface.

oscillating drops oscillation frequency was 0.1 Hz in both experiments, amplitude


of surface area variations F 20%.. For the waterroil interface and - 10 mNrm
the obtained values for U are in the range 2.52.8 and independent of surface
pressure, and then decrease down to 1 when increases up to 17 mNrm. These
results are in good agreement with the approximated value of U s 2, and the
experimental data for HSA. The elasticity values by Benjamins et al. w102x for the
solutionrair interface were also somewhat higher, quite similar to the presented
HSA data.

Fig. 22. Dependence of dynamic surface tension for two blood serum samples taken from patients
suffering from different pathologies B,I and ^,', respectively. plotted on the logarithm of the
surface lifetime. Closed symbols, MBPM; open symbols, ADSA data.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

73

Fig. 23. Dependence of surface tension for blood serum sample on ty1 r2 . Open symbols, ADSA data;
closed symbols, MBPM data; solid line, extrapolation of the MBPM results.

10. Surface tension and elasticity of blood serum as example of a natural protein r
surfactant mixtures
All biological liquids of the human organism contain surface-active compounds,
such as proteins, lipids, and molecules of other natures. The main surface-active
compound of serum is albumin HSA. with a concentration in blood of 3550 grl.
The results of the measurements performed with two samples of blood serum,
taken from two patients suffering from different pathologies, are presented in Fig.
22 w103x.
The experimental results from the two methods are combined here in order to
cover a broader time interval: values from the maximum bubble pressure method
MBPM. were measured in the effective time range from 0.01 to 50 s, while the
values from the drop shape method ADSA . cover the range from 10 to 1000 s.
The mutual consistency of the methods is demonstrated by the fact that in the
overlapping interval from 10 to 50 s, the results are in an agreement to within
experimental error. It is also seen that more than half of the total surface tension
decrease falls into the MBPM interval. On the other hand, the ADSA data
essentially complement the MBPM results in the long lifetime range. It can be also
shown that at times longer than 1000 s, almost no change of the serum surface
tension can be observed.
To calculate the equilibrium surface tension from the MBPM results ., the
dependence of the surface tension vs. ty1 r2 was extrapolated to the infinite time,
according to the procedure described in Refs. w3,32x. For example, the lower
tensiogram of Fig. 22 is plotted in these co-ordinates in Fig. 23 w103x.
For this case, the value of estimated from the MBPM data the point of
intersection of the extrapolation straight line with the ordinate. is 57.7 mNrm,

74

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

while the slope of this line is s drdty1 r2 s 19.8 mNrm sy1 r2 . It is seen that
the slope of the extrapolation line coincides with the slope of the linear part of the
ADSA curve in the range above 0.2 sy1 r2 . However, at lower values of ty1 r2 , the
ADSA data are lower than those obtained by the MBPM. A similar extrapolation
to infinite time, performed for the ADSA data in the range ty1 r2 - 0.06 sy1r2 .
leads to the equilibrium value U s 47.8 mNrm and the slope of the linear part to
U s 140 mNrm sy1 r2 .
The shape of the experimental dependence as presented in Fig. 23, with a
narrow interval with an inflection point and two linear parts of different slopes, is
typical for mixtures of surfactants characterised by different activities and concentrations w104,105x. Curves quite similar to that presented in Fig. 23 were obtained
both experimentally and theoretically . for mixtures of two surfactants, when the
concentration of one of the two was 10100 times higher than the concentration of
the second surfactant, while the adsorption activity of the first surfactant was to the
same degree lower that that of the second component of the mixture w104,105x.
Therefore it can be concluded, that for ty1 r2 ) 0.13 sy1r2 i.e. for time values
below 60 s. the surface-active compound with the main mass portion in the blood
serum the blood serum albumin. is adsorbed. However, at lifetimes above 100 s
blood serum components of higher surface activity adsorb, having concentrations
10 or 100 times lower than the albumin concentration. It follows from the theory
w104x that the slope is proportional to the square of the adsorption, and inverse
proportional to the surfactant concentration. Comparing the two values of
obtained above 19.8 mNrm sy1 r2 and 140 mNrm sy1r2 . one can conclude that
the concentration of highly surface active impurities in the sample of blood serum
studied is extremely low.
The viscoelasticity modulus obtained in stress experiments of the surface layer
can be calculated from Eq. 42.. In these experiments the expansion of the drop is
followed by the relaxation of the surface tension to its initial value. The change of
surface tension which follows the stress deformation can be described by an
exponential dependence
t s exp tr .

48.

where is the initial jump of the surface tension, t is the time elapsed after the
deformation, is the relaxation time. The relaxation time characterises the ability
of the monolayer to restore its initial state, that is, this value reflects the kinetics of
adsorption from the solution and the processes involved in the rearrangement of
the state of adsorbed molecules.
The results obtained from stress experiments with the samples of blood serum
referred to in the discussion of Fig. 22 are presented in Fig. 24 w103x.
The measurement time in the dynamic regime was 1200 s adsorption after the
formation of a fresh drop surface .. At this time moment, an expansion of the drop
surface by 78% was made. The smaller the deformation is, the more rigorous is
the correspondence of the data with Eq. 42., while finite differences are applied
instead of differentials. On the other hand, if the deformation is too small, then the

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

75

Fig. 24. Dynamic surface tension of blood serum measured using ADSA for the same samples as in
Fig. 22; the stress deformation caused by an expansion of the drop was imposed 1200 s after the start of
the experiment with a fresh drop.

errors in the surface tension measurements become comparatively large. Thus,


deformations of 510% are optimal. It is seen from Fig. 24 that the surface
pressure jumps and the viscoelasticity modulus. are different for the blood serum
samples studied. For this fluid, the range of the viscoelasticity modulus is quite
broad: 1080 mNrm, while the usual values of the relaxation time are between 50
and 300 s. Thus, the application of the axisymmetric drop shape analysis complementary to the maximum bubble pressure method allows not only to understand
more clearly the results obtained by the two methods, but provides also a deeper
insight into the tensiometric and rheological characteristics of the interfacial layers
formed by human biologic fluids.
Finally, the characteristics of blood serum tensiograms and rheological parameters for the samples taken from three females are compared in Table 2: 1. a
healthy person; 2. a patient suffering from kidney disease; and 3. a patient
suffering from neoplastic disease of reproductive organs.
It is seen that the combined use of the two methods enables one to differentiate

Table 2
Tensiometric and rheometry parameters of some selected samples of blood serum
U

Patient
no.

mNrm.

mNrm sy1r2 .

a
mNrm.

a
mNrm sy1r2 .

mNrm.

s.

1
2
3

61.4
54.9
57.3

12.3
24.2
22.1

53.7
49.5
42.2

80.3
89.1
278.4

27.9
28.6
50.0

64.9
95.7
73.3

From ADSA experiments at longer adsorption times.

76

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

more precisely between the pathologies. For both pathologies, the value is
lower, and the value is two times higher than those characteristic for the healthy
person. At the same time, for patient 3, a significant decrease of U is observed
the difference between U and is 15.1 mNrm, as compared with the value of
7.7 mNrm for the healthy person., and the dilational viscosity modulus is almost
twice the increase as compared with the data obtained for patients 1 and 2. On the
contrary, for the patient 2 the difference between and U is small 5.4 mNrm.,
while the U value is essentially higher three times higher as compared with
patients 1 and 3, respectively., while the value is high. One of the possible
explanations for the change in the parameters U and U for patient 3 may be an
increased concentration of sialic acid in the blood common to the neoplastic
disease of reproductive organs w106108x.

11. Summary, conclusions, outlook


The present overview discusses experimental results obtained for selected proteins, namely HSA, BSA, -casein and -lactoglobulin, and mixed systems with
surfactants and lipids. The surfactants on which we report here have been selected
such that their surface activity is comparable, i.e. the cationic CTAB, the anionic
STS, and the non-ionic surfactant C 10 DMPO have been mainly referred to in this
comparative study.
The equilibrium and dynamic adsorption behaviour of pure protein solutions can
be described very well by a recently developed thermodynamic model and the
respective diffusion controlled adsorption theory based on the corresponding
equation of state.
For the mixed systems, quantitative theoretical models do not exist, and thus the
adsorption kinetics data cannot discussed quantitatively, however, the equations
derived in Section 2.2 allow a semi-quantitative understanding of mixed adsorption
layers. The assumption that different types of interaction between protein and
surfactant molecules, and a specific surface activity and conformation of the
proteinrsurfactant complexes exist, is in good agreement with the adsorption data
obtained. The complexes formed by proteins and the ionic surfactants can be
higher surface active than the protein alone due to a modification of the structure
of adsorbed proteins. Thus even -casein becomes rather compact at the interface
at a certain STS or CTAB concentration. In the presence of non-ionics the
adsorption layer is mainly formed by competitive adsorption between the compounds. Hydrophobic interaction can also support the formation of complexes.
However, such complexes are typically less surface-active than the pure protein so
that they do not participate in the competitive adsorption. This type of complex
supports the depletion of proteins from the interface at increasing amounts of
surfactants, which is observed for non-ionic as well as ionic surfactants.
In addition to adsorption dynamics studies also penetration kinetics experiments
are of interest for the better understanding of protein layers mixed with surfactants

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

77

Fig. 25. Relaxation fluxes in a mixed proteinrsurfactant layer under harmonic area changes, relaxation between the two states A and B due to diffusion flux a. and due to bindingrrelease of surfactants
from the formed complexes b., respectively.

at liquid interfaces. In particular the penetration of proteins into lipid layers are of
much practical relevance for biological and medical studies.
The strength of interaction between protein and the different surfactants can be
studied by relaxation experiments. Using harmonic perturbations in a broad
frequency interval it should be possible to separate relaxation due to transport
Fig. 25a. from relaxation due to bound surfactants Fig. 25b.. Fig. 25 shows
schematically the two different relaxation fluxes for a mixed proteinrsurfactant
adsorption layer.
There are a number of open questions the answer of which requires more
systematic studies. For example, in the studies of the liquidrliquid interface the
solvent for the lipids should be varied as the chosen chloroform can certainly
destruct proteins. Using a more gentle solvent it can be better cleared up whether
lipid and protein also separate at the waterroil interface. For adsorbed mixed
proteinrsurfactant layers studied so far it has not yet been observed that the

78

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

components separate at the interface. Techniques for the investigation of this


question exist, however, experiments are not known so far.
It will also be interesting to learn under which conditions the modification of a
protein by specific interaction with ionic surfactants can make it possible that the
resulting complex transfers from the aqueous into the adjacent oil phase. This
phenomenon seems possible as the ionic interaction leads to a strong hydrophobisation of the protein molecule so that the changed conformation would make it
soluble in oil.
These are only a few of the questions resulting immediately from the presented
studies. All the given results are important to improve the scientific basis on mixed
proteinrsurfactants systems for many practical purposes. One of them, as briefly
discussed in Section 10, is the use of tensiometry and rheology as tools in medical
diagnostics and therapy control. This overlap of two scientific areas seems to be
extremely interesting and fruitful and there is great hope that it will develop
further as successfully as it did in the short period of the last 5 years. The book by
Kazakov et al. w3x is an impressive token of this fortunate scientific cross-linking.
Another contribution in the present special issue of the journal is dedicated to this
new type of interfacial studies of high medical relevance w109x.

Acknowledgements
The work was financially supported by the Max-Planck-Gesellschaft MIL1., by
projects of the Deutsche Forschungsgemeinschaft Mi 418r9-1, Mi 418r7-1. and
the European Union INCO Copernicus..

Notation
A
ADSA
ael
BAM
BSA
b1
c
c
CTAB
C10 DMPO
D
Deff
DMPC
DMPE
DPPC
DPPE

drop surface area


axisymmetric drop shape analysis
interfacial interaction parameter
Brewster angle microscopy
bovine serum albumin
adsorption constant
bulk concentration of surfactant or protein
average protein concentration
hexadecyl trimethyl ammonium bromide
dodecyl dimethyl phosphine oxide
diffusion coefficient
effective diffusion coefficients
dimyristoyl phosphatidyl choline
dimyristoyl phosphatidyl ethanolamine
dipalmitoyl phosphatidyl choline
dipalmitoyl phosphatidyl ethanolamine

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

HSA
j
ki
MBPM
n
R
r
SDS
STS
T
t
Tween 20
V
x

-LG
-CS

i
min

U
s drdty1 r2

79

human serum albumin


mass flux
rate constant
maximum bubble pressure method
number of adsorption states
gas law constant
drop radius
sodium dodecyl sulphate
sodium tetradecyl sulphate
absolute temperature
time
polyoxyethylene 20 sorbitan monolaurate
drop volume
spatial coordinate normal to the interface
model parameter
-lactoglobulin
-casein
molar area difference between two states
average surface concentration
surface concentration in state i
minimum adsorption
surface tension
equilibrium surface tension
dilational elasticity
dimensionless dilational elasticity
slope of the dependence ty1 r2 .
surface pressure
molar surface area
average molar surface area
minimum molar surface area
relaxation time

References
w1x E. Dickinson, J.M. Rodriguez Patino Eds.., Food emulsions and foams: interfaces, interactions
and stability, Special Publication No. 227, Royal Society of Chemistry, 1999.
w2x R. Wustneck,
J. Kragel,
Proteins at liquid interfaces wmonographx, in: D. Mobius,
R. Miller

Eds.., Studies of Interface Science, vol. 7, Elsevier, Amsterdam, 1998, pp. 433490.
w3x V.N. Kazakov, O.V. Sinyachenko, V.B. Fainerman, U. Pison, R. Miller, Dynamic surface tension
of biological liquids in medicine, in: D. Mobius,
R. Miller Eds.., Studies in Interface Science,

vol. 8, Elsevier, Amsterdam, 1999.


w4x I.I. Harris, K.G.A. Pankhurst, R.C.M. Smith, Trans. Faraday Soc. 43 1947. 506.
w5x J.P.S. Arora, S.P. Singh, Y.K. Singhal, Tenside Surfactants Detergents 21 1984. 197.
w6x J.P.S. Arora, S.P. Malik, S. Jain Smt, R.P. Singh, Tenside Surfactants Detergents 24 1987. 217.
w7x E. Dickinson, V.B. Galazka, Food Hydrocolloids 5 1991. 281.
w8x J.L. Courthaudon, E. Dickinson, Y. Matsumura, A. Williams, Food Structure 10 1991. 109.

80

w9x
w10x
w11x
w12x
w13x
w14x
w15x
w16x
w17x
w18x
w19x
w20x
w21x
w22x
w23x
w24x
w25x
w26x
w27x
w28x
w29x
w30x
w31x
w32x
w33x
w34x
w35x
w36x
w37x
w38x
w39x
w40x
w41x
w42x
w43x
w44x
w45x
w46x
w47x
w48x
w49x
w50x
w51x
w52x
w53x

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982


J.L. Courthaudon, E. Dickinson, D.G. Dalgleish, J. Colloid Interface Sci. 145 1991. 390.
D.C. Clark, P.J. Wilde, D.R. Wilson, R. Wustneck,
Food Hydrocolloids 6 1992. 173.

D.C. Clark, A.R. Mackie, P.J. Wilde, D.R. Wilson, Faraday Discuss. 98 1994. 253.
B.S. Murray, A. Ventura, C. Lallemant, Colloids Surfaces A 143 1998. 211.
J. Kragel,
R. Wustneck,
D.C. Clark, P.J. Wilde, R. Miller, Colloids Surfaces A 98 1995. 127.

N.J. Turro, X.-G. Lei, K.P. Ananthapadmanabhan, M. Aronson, Langmuir 11 1995. 2525.
R. Wustneck,
J. Kragel,
R. Miller, P.J. Wilde, D.C. Clark, Colloids Surfaces A 114 1996. 255.

G.J. Fleer, J.M.H.M. Scheutjens, Adv. Colloid Interface Sci. 16 1982. 341.
F. MacRitchie, Adv. Colloid Interface Sci. 25 1986. 341.
V.B. Fainerman, E.H. Lucassen-Reynders, R. Miller, Colloids Surfaces A 143 1998. 141166.
P.-G. de Gennes, Adv. Colloid Interface Sci. 27 1987. 189.
J.A.V. Butler, Proc. R. Soc. Ser. A 138 1932. 348.
P. Joos, Biochim. Biophis. Acta 375 1975. 1.
L. Ter-Minassian-Saraga, J. Colloid Interface Sci. 80 1981. 393.
E.H. Lucassen-Reynders, Colloids Surfaces A 91 1994. 79.
P. Joos, G. Serrien, J. Colloid Interface Sci. 145 1991. 291.
V.B. Fainerman, R. Miller, R. Wustneck,
J. Colloid Interface Sci. 183 1996. 25.

V.B. Fainerman, R. Miller, Proteins at liquid interfaces wmonographx, in: D. Mobius,


R. Miller

Eds.., Studies of Interface Science, vol. 7, Elsevier, Amsterdam, 1998, pp. 51102.
V.B. Fainerman, R. Miller, Langmuir 15 1999. 1812.
V.B. Fainerman, R. Miller, R. Wustneck,
A.V. Makievski, J. Phys. Chem. B 100 1996. 7669.

V.B. Fainerman, R. Miller, R. Wustneck,


J. Phys. Chem. B 101 1997. 6479.

V.B. Fainerman, E.H. Lucassen-Reynders, R. Miller, Colloids Surfaces A 143 1998. 141.
R. Miller, V.B. Fainerman, A.V. Makievski, J. Kragel,
R. Wustneck,
Colloids Surfaces A 161

2000. 151.
A.F.H. Ward, L. Tordai, J. Phys. Chem. 14 1946. 453.
S.S. Dukhin, G. Kretzschmar, R. Miller, Dynamics of adsorption at liquid interfaces. Theory,
experiment and application, in: D. Mobius,
R. Miller Eds.., Studies of Interface Science, vol. 1,

Elsevier, Amsterdam, 1995.


R. Wustneck,
J. Kragel,
R. Miller et al., Food Hydrocolloids 10 1996. 395.

R. Miller, P. Joos, V.B. Fainerman, Adv. Colloid Interface Sci. 49 1994. 249.
S. Ghosh, H.B. Bull, Biochemistry 2 1963. 411.
E. Tornberg, in: Akila Pour-El Ed.., ACS Symposium Series, N92, Functionality and Protein
Structure, 1972, p. 105.
D.E. Graham, M.C. Phillips, J. Colloid Interface Sci. 70 1979. 403.
K. Kalischewski, K. Schugerl,
Colloid Polymer Sci. 257 1979. 1099.

J.A. de Feijter, J. Benjamins, in: E. Dickinson Ed.., Food emulsions and foams, Special
Publication No. 58, R. Soc. Chem. 1987. 72.
R. Miller, Trends Polymer Sci. 2 1991. 47.
A.J.I. Ward, L.H. Regan, J. Colloid Interface Sci. 78 1980. 389.
E. Tornberg, G. Lundh, J. Colloid Interface Sci. 79 1981. 76.
J.A. Feijter, J. Benjamins, F.A. Veer, Biopolymers 17 1978. 1760.
F.K. Hansen, R. Myrfold, J. Colloid Interface Sci. 176 1995. 408.
B.C. Tripp, J.J. Magda, J.D. Andrade, J. Colloid Interface Sci. 173 1995. 16.
B.S. Murray, Ph.V. Nelson, Langmuir 12 1996. 5973.
B.S. Murray, Colloids Surfaces A 125 1997. 73.
G. Gonzalez, F. MacRitchie, J. Colloid Interface Sci. 32 1970. 55.
R. Miller, V.B. Fainerman, R. Wustneck,
J. Kragel,
D.V. Trukhin, Colloids Surfaces A 131 1998.

225.
D.E. Graham, M.C. Phillips, J. Colloid Interface Sci. 70 1979. 415.
J. Benjamins, J.A. de Feijter, M.T.A. Evans, D.E. Graham, M.C. Phillips, Discuss. Faraday Soc.
59 1978. 218.
A.V. Makievski, R. Miller, V.B. Fainerman, J. Kragel,
R. Wustneck,
in: E. Dickinson, J.M.

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982

w54x
w55x
w56x
w57x
w58x
w59x
w60x
w61x
w62x
w63x
w64x
w65x
w66x
w67x
w68x
w69x
w70x
w71x
w72x
w73x
w74x
w75x
w76x
w77x
w78x
w79x
w80x
w81x
w82x
w83x
w84x
w85x
w86x
w87x
w88x
w89x
w90x
w91x
w92x
w93x
w94x
w95x

81

Rodriguez Patino Eds.., Food Emulsions and Foams: Interfaces, Interactions and Stability,
Special Publication No. 227, Royal Society of Chemistry 1999, p. 269.
R. Douillard, M. Daoud, J. Lefebvre, Ch. Miner, Y. Lecanny, J. Coutret, J. Colloid Interface Sci.
163 1994. 277.
R. Miller, Z. Policova, R. Sedev, A.W. Neumann, Colloid Surfaces A 76 1993. 179.
J. Kragel,
D.C. Clark, P.J. Wilde, R. Miller, Prog. Colloid Polymer Sci. 98 1995. 239.

J.B. Li, J. Zhao, R. Miller, Nahrung-Food 42 1998. 233.


J.B. Li, J. Zhao, J. Wu, R. Miller, Nahrung-Food 42 1998. 231.
J.B. Li, J. Kragel,
A.V. Makievski, V.B. Fainerman, R. Miller, H. Mohwald,
Colloids Surfaces A

142 1998. 355.


J. Kragel,
R. Wustneck,
F. Husband et al., Colloids Surfaces B 12 1999. 399.

R. Miller, V.B. Fainerman, A.V. Makievski, D.O. Grigoriev, P.J. Wilde, J. Kragel,
in: E.

Dickinson, J.M. Rodriguez Patino Eds.., Food Emulsions and Foams: Interfaces, Interactions
and Stability, Special Publication No. 227, Royal Society of Chemistry, 1999, p. 207.
A.V. Makievski, V.B. Fainerman, M. Bree, R. Wustneck,
J. Kragel,
R. Miller, J. Phys. Chem. B

102 1998. 417.


V.B. Fainerman, S.V. Lylyk, Kolloid. Zh. 45 1983. 500.
E. Hutchinson, J. Colloid Sci. 3 1948. 413.
M.J. Rosen, X.Y. Hua, J. Colloid Interface Sci. 86 1982. 164.
M.A. Bos, T. Nylander, T. Arnebrant, D.C. Clark, in: Food Emulsifiers and their Applications,
1997, p. 95.
M.A. Bos, T. Nylander, Langmuir 12 1996. 2791.
K. Kurihara, Y. Katsuragi, Nature 365 1993. 213.
D.G. Cornell, J. Colloid Interface Sci. 88 1982. 536.
D.C. Clark, D.L. Patterson, J. Agric. Food Chem. 37 1989. 1455.
J.B. Li, H. Chen, J. Wu, J. Zhao, R. Miller, Colloids Surfaces A 1999. in press.
J.B. Li, R. Miller, H. Mohwald,
Colloids Surfaces A 114 1996. 113.

J. Wu, J.B. Li, J. Zhao, R. Miller, Colloids Surfaces A 2000. in press.


D. Vollhardt, S. Siegel, V.B. Fainerman, Adv. Colloid Interface Sci. 86 2000. 103.
J.B. Schulman, A.H. Hughes, Biochem. J. 29 1935. 1243.
J.B. Schulman, Trans. Faraday Soc. 33 1937. 1116.
B.A. Pethica, Trans. Faraday Soc. 51 1955. 1402.
M.A. McGregor, G.T. Barnes, J. Colloid Interface Sci. 54 1976. 439.
M.A. McGregor, G.T. Barnes, J. Colloid Interface Sci. 60 1977. 408.
K. Motomura, I. Hayami, M. Aratono, R. Matuura, J. Colloid Interface Sci. 87 1982. 333.
I. Panaiotov, L. Ter-Minassian-Saraga, G. Albrecht, Langmuir 1 1985. 395.
D. Vollhardt, M. Wittig, Colloids Surfaces 47 1990. 233.
S. Siegel, D. Vollhardt, Colloids Surfaces A 76 1993. 197.
Q. Jiang, C.J. OLenick, J.E. Valentini, Y.C. Chiew, Langmuir 11 1995. 1138.
J.M. Rodriguez Patino, M.R. Rodriguez Nino,
Colloids Surfaces A 103 1995. 91.
S. Sundaram, K.J. Stebe, Langmuir 13 1997. 1729.
D. Honig,
D. Mobius,
J. Phys. Chem. 95 1991. 4590.

S. Henon, J. Meunier, Rev. Sci. Instrum. 62 1991. 936.


D. Honig,
G.A. Overbeck, D. Mobius,
Adv. Mater. 4 1992. 419.

V.B. Fainerman, J. Zhao, D. Vollhardt, A.V. Makievski, J.B. Li, J. Phys. Chem. B 103 1999.
8998.
J. Zhao, D. Vollhardt, G. Brezesinski, S. Siegel, J. Wu, J.B. Li, R. Miller, @ Colloids Surfaces A
2000. in press.
J. Zhao, D. Vollhardt, J. Wu, R. Miller, S. Siegel, J.B. Li, Colloids Surfaces A 2000. in press.
F. MacRitchie, Proteins at liquid interfaces wmonographx, in: D. Mobius,
R. Miller Eds.., Studies

of Interface Science, vol. 7, Elsevier, Amsterdam, 1998, pp. 149177.


V.B. Fainerman, A.V. Makievski, R. Miller, Colloids Surf. A 87 1994. 61.
A.V. Makievski, G. Loglio, J. Kragel,
R. Miller, V.B. Fainerman, A.W. Neumann, J. Phys. Chem.

B 103 1999. 9557.

82

w96x
w97x
w98x
w99x
w100x
w101x
w102x
w103x
w104x
w105x
w106x
w107x
w108x
w109x

R. Miller et al. r Ad ances in Colloid and Interface Science 86 (2000) 3982


V.B. Fainerman, Kolloid. Zh. 43 1981. 926.
S.-Y. Lin, K. McKeigue, C. Maldarelli, AIChE J. 36 1990. 12.
L. Liggieri, F. Ravera, M. Ferrari, A. Passerone, R. Miller, J. Colloid Interface Sci. 186 1997. 46.
R. Miller, R. Wustneck,
J. Kragel,
G. Kretzschmar, Colloids Surfaces A 111 1996. 75.

R. Miller, G. Loglio, U. Tesei, K.-H. Schano, Adv. Colloid Interface Sci. 37 1991. 73.
R. Miller, V.B. Fainerman, J. Kragel,
G. Loglio, Curr. Opin. Colloid Interface Sci. 2 1997. 578.

J. Benjamins, A. Cagna, E.H. Lucassen-Reynders, Colloids Surfaces A 114 1996. 245.


A.F. Vozianov, V.N. Kazakov, O.V. Sinyachenko, V.B. Fainerman, R. Miller, Interfacial tensiometry and rheometry in nephrology, Izd. Donetsk. Med. Univ., Donetsk, Ukraine, 1999.
V.B. Fainerman, R. Miller, Colloids Surfaces A 97 1995. 65.
A.V. Makievski, V.B. Fainerman, R. Miller, M. Bree, L. Liggieri, F. Ravera, Colloids Surfaces A
122 1997. 269.
V. Bhuvarahamurthy, N. Balasubramanian, S. Vijayakumar, S. Govindasamy, Int. J. Gynecol.
Obstet. 48 1995. 49.
V. Bhuvarahamurthy, S. Govindasamy, Mol. Cell Biochem. 144 1995. 35.
P.S. Patel, G.N. Rawal, D.B. Balar, Gynecol. Oncol. 50 1993. 294.
V.N. Kazakov, A.F. Vozianov, O.V. Sinyachenko, D.V. Trukhin, V.I. Kovalchuk, U. Pison, Adv.
Coll. Interface Sci. 86 2000. 1.

Вам также может понравиться