Advances in Colloid and Interface Science 179-182 (2012) 85–98

Contents lists available at SciVerse ScienceDirect

Advances in Colloid and Interface Science

journal homepage: www.elsevier.com/locate/cis

Investigating forces between charged particles in the presence of oppositely charged

polyelectrolytes with the multi-particle colloidal probe technique
Michal Borkovec a,⁎, Istvan Szilagyi a, Ionel Popa a, 1, Marco Finessi a, Prashant Sinha a,
Plinio Maroni a, Georg Papastavrou b
a
Department of Inorganic, Analytical, and Applied Chemistry, University of Geneva, Sciences II, 30, Quai Ernest-Ansermet, 1205 Geneva, Switzerland
b
Physical Chemistry II, University of Bayreuth, 95440 Bayreuth, Germany

a r t i c l e i n f o a b s t r a c t

Available online 1 July 2012 Direct force measurements are used to obtain a comprehensive picture of interaction forces acting between
charged colloidal particles in the presence of oppositely charged polyelectrolytes. These measurements are
Keywords: achieved by the multi-particle colloidal probe technique based on the atomic force microscope (AFM). This
Surface forces novel extension of the classical colloidal probe technique offers three main advantages. First, the technique
Polyelectrolytes works in a colloidal suspension with a huge internal surface area of several square meters, which simplifies
Dendrimers
the precise dosing of the small amounts of the polyelectrolytes needed and makes this approach less sensitive
Water–solid interface
Latex particles
to impurities. Second, the particles are attached in-situ within the fluid cell, which avoids the formation of
Atomic force microscopy nanobubbles on the latex particles used. Third, forces between two similar particles from the same batch
Direct force measurements are being measured, which allows an unambiguous determination of the surface potential due to the symme-
try of the system. Based on such direct force measurements involving positively and negatively charged latex
particles and different polyelectrolytes, we find the following forces to be relevant. Repulsive electrostatic
double-layer forces and attractive van der Waals forces as described by the theory of Derjaguin, Landau,
Verwey, and Overbeek (DLVO) are both important in these systems, whereby the electrostatic forces domi-
nate away from the isoelectric point (IEP), while at this point they vanish. Additional non-DLVO attractive
forces are operational, and they have been identified to originate from the electrostatic interactions between
the patch-charge heterogeneities of the adsorbed polyelectrolyte films. Highly charged polyelectrolytes
induce strong patch-charge attractions, which become especially important at low ionic strengths and high
molecular mass. More weakly charged polyelectrolytes seem to form more homogeneous films, whereby
patch-charge attractions may become negligible. Individual bridging events could be only rarely identified
from the retraction part of the force profiles, and therefore we conclude that bridging forces are unimportant
in these systems.
© 2012 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
2. Multi-particle colloidal probe technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3. Amidine latex particles in the presence of poly(styrene sulfonate) (PSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4. Sulfate latex particles in the presence of poly(amido amine) (PAMAM) dendrimers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5. Sulfate latex particles in the presence of linear poly(ethylene imine) (LPEI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
6. Conclusions and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

1. Introduction

⁎ Corresponding author. Tel.: +41 22 379 6405; fax: +41 22 379 6069.
E-mail address: michal.borkovec@unige.ch (M. Borkovec).
Polyelectrolytes are charged polymers and they are extensively used
1
Present address: Department of Biological Sciences, Columbia University, New as additives in various industrial applications, such as waste-water
York, 10027, USA. treatment, papermaking, or material processing [1–3]. They are mostly

0001-8686/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.cis.2012.06.005
86 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98
used to control the stability, rheology, or adhesion properties of colloi-
dal suspensions [4–8]. Frequently, optimal results are achieved with
polyelectrolytes bearing the opposite charge than the suspended parti-
cles. In papermaking, for example, polyelectrolytes are added to achieve
rapid aggregation of cellulose fibers and filler particles (e.g., silica).
Cationic polyelectrolytes are used in this application since the respec-
tive solid components are negatively charged [9,10]. Concentrated
suspensions of calcium carbonate can be made to flow easily by adding
polyelectrolytes [11,12]. Anionic polyelectrolytes are used in this case,
since calcium carbonate particles are positively charged at the neutral
pH conditions used [13,14].
Due to the importance of these processes in industry, numerous
researchers have addressed how polyelectrolytes affect the properties
of oppositely charged particle suspensions [7,15–21]. Strong electro-
static attraction between the particles and the polyelectrolytes leads
to strong adsorption affinity, and therefore it is essential to under-
stand how adsorbed polyelectrolytes modify the surface properties.
Substantial effort was devoted to investigate properties of such
polyelectrolyte films experimentally [22–24] as well as theoretically
[25–33]. Less attention was paid on the interaction forces acting
between water–solid interfaces coated with such polyelectrolyte films,
even though these interaction forces determine the stability or rheology
of colloidal suspensions. Researchers have rather studied stability and
rheology of colloidal suspensions in the presence of oppositely charged
polyelectrolytes directly [7,15–21]. While such studies are of substantial
practical relevance, they reveal information on interaction forces only
indirectly.
The reasons why studies of interaction forces between surfaces
with adsorbed polyelectrolytes have been lagging behind can be easily
understood. Techniques to measure forces between interfaces across
liquids were developed only relatively recently [34,35]. The first reli-
able and widely available instrument for direct force measurement
was the surface forces apparatus (SFA) [35–39]. This apparatus mea-
sures forces and distances between two curved macroscopic mica
plates. While the apparatus has a moderate force resolution of typically
100 nN, the relevant force regime remains accessible due to the rela-
tively large curvature radius of few cm. On the other hand, the SFA
offers an unsurpassed distance resolution by measuring the distances
between the mica plates with multiple beam interferometry with
sub-Å precision. The SFA was the first technique to provide a direct
experimental confirmation of the forces inherent to the theory devel-
oped by Derjaguin, Landau, Verwey, and Overbeek (DLVO) [40,41].
This theory stipulates that surface forces across water are dominated
by electrostatic double-layer repulsion and van der Waals attraction.
Early reliable measurements involving surfaces with adsorbed poly-
electrolytes were carried out with the SFA as well [38,39,42]. A related
technique referred to as MASIF (i.e., measurement and analysis of
surface interaction forces) was developed to measure forces between
a macroscopic glass sphere and a planar substrate [43,44]. This tech-
nique demonstrated that forces between oppositely charged substrates
were attractive right after the addition of a polyelectrolyte, but they be-
came dominated by double-layer repulsion at later times.
The second important development in direct force measurements
was the colloidal probe technique based on the atomic force micro-
scope (AFM) [45,46]. This technique replaces the sharp AFM-tip
on the cantilever with a colloidal particle. Such colloidal probes
are normally prepared by gluing a particle of few μm in diameter to
a tip-less cantilever with a micromanipulator. This colloidal probe
is approached to the sample by means of the AFM scanner, and
the force can be measured through the deflection of the cantilever.
With a modern AFM one can typically achieve a force resolution of
10–50 pN, and by subsequent averaging of the force profiles, the
noise can be reduced by a factor of ten or more. The advantage of the
technique is that a wide range of particles and surfaces can be investi-
gated. The disadvantage is, however, that the separation distances are
normally not determined independently but through the onset of the
constant compliance region, where the cantilever is in contact with
the substrate. While an independent measurement of the separation
distance with total internal reflection microscopy (TIRM) was described
recently [47], most researchers continue to determine the distances
from the contact point between solid surfaces through the onset of the
constant compliance region. The latter technique leads to sub-nm preci-
sion for solid substrates. The colloidal probe technique has been used by
various authors to study interaction forces in the presence of polyelec-
trolytes [48–51]. These researchers have reported that strong repulsive
forces are acting between adsorbed polyelectrolyte films, which appear
to originate from the overlap of the diffuse part of the double-layer.
The classical colloidal probe technique measures forces between a
spherical colloidal particle and a planar substrate, and therefore this
system is inherently asymmetric (i.e., sphere-plane geometry). While
the sphere-plane geometry is ideally suited to study asymmetric sys-
tems, the study of symmetric systems is impossible in this geometry.
However, asymmetry effects can be minimized by choosing identical
materials for the probe particle and the substrates or by suitable
coatings. The best way to exclude asymmetry artifacts is by measuring
forces between two colloidal particles (i.e., sphere–sphere geometry)
[52–55]. Measurements in the sphere–sphere geometry also require
that the two particles are centered laterally, but such an alignment
can be achieved with an AFM-scan or an optical microscope.
More recently, other techniques for direct force measurements
involving colloidal particles have been developed. The most relevant
ones include total internal reflection microscopy (TIRM) [43,56–58]
and optical tweezers techniques [59–62]. TIRM measures the fluctua-
tions in the separation distance of a particle near the water–solid
interface through the evanescent field (i.e., sphere–plane geometry).
The optical tweezers technique positions one particle near a second
one, and from the particle trajectories one obtains information about
the interaction potential. While these techniques have an excellent
force resolution in the sub-pN regime, their positional resolution is on
the order of few nm so far. Moreover, controversies concerning appro-
priate analysis of the video images have been resolved only recently
[60,62]. These techniques are operational only in a handful of laborato-
ries worldwide, but they have already provided interesting results
involving interactions between surfaces in concentrated polyelectrolyte
solutions [58] or with grafted polyelectrolyte layers [61,63].
The available surface area is another relevant aspect to consider in
force measurements between surfaces with adsorbed polyelectro-
lytes. Since polyelectrolytes adsorb strongly, already small adsorbed
quantities (down to μg/m 2) may influence the surface charge and
thus the interaction forces substantially. Therefore, one must be able
to precisely control small amounts of added polyelectrolytes. The
problem is less severe with SFA and MASIF, since the instruments
operate with macroscopic surfaces with a surface area of few cm 2.
On the other hand, the colloidal probe techniques involve one or two
colloidal particles with a surface area of few μm2 only. Based on this
comparison, one can understand why attractive forces, which occur for
small amounts of adsorbed polyelectrolytes, have been first observed
with SFA and MASIF. With the colloidal probe technique only repulsive
forces, which are characteristic for larger adsorbed amounts, were ini-
tially found [48,49]. For the small surface areas involved in the colloidal
probe technique, sufficiently accurate control of the polyelectrolyte dose
may seem impossible. However, the recently developed multi-particle
colloidal probe technique circumvents the mentioned difficulties. With
this technique one can measure the forces between two similar colloidal
particles precisely and at the same time control the dose accurately.
2. Multi-particle colloidal probe technique
This novel variant of the colloidal probe technique is most useful
for the measurement of forces between colloidal particles of few μm
in diameter [64,65]. Fig. 1 illustrates the four-step procedure, which
can be easily implemented with any AFM mounted on an inverse
 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98
optical microscope. (a) The particle suspension is first prepared with
the proper additives and thoroughly degassed. (b) Simultaneously, a
glass slide and a tip-less AFM cantilever are appropriately silanized
to allow an attachment of the particles. (c) The silanized glass slide
is inserted into the AFM fluid cell and the particle suspension is
injected. Thereby, particles will stick to the glass slide. (d) The cell
is flushed with degassed background electrolyte solution and the
silanized cantilever is inserted. The deposited particles are localized
with the optical microscope and the cantilever is brought into the
field of view. With the AFM scanner, one now positions the cantilever
over one of the deposited particles and brings it in contact with the
cantilever. By inspecting the retracted cantilever with the optical mi-
croscope, one can verify that the particle attachment was successful.
With this newly prepared probe, one approaches another deposited
particle, and both particles are centered through the optical micro-
scope with the AFM scanner. The centering can be done with an
excellent precision of 30–50 nm by observing the interference fringes
appearing in the optical microscope. Finally, the force profile between
the particles is recorded by repeated approach–retract cycles as with
the classical colloidal probe technique. The reproducibility of the
measurements can be verified by repeating the process for different
particle pairs.
The advantages of the technique are threefold. First, nanobubbles
do not interfere with the measurements, as they are normally
not present in particle suspensions, and if necessary, they can be
removed by degassing. Nanobubbles were shown to strongly in-
fluence forces measured with the classical colloidal probe tech-
nique between latex particles or other hydrophobic substrates. Such
bubbles form easily during the drying-rewetting process, but they
are difficult to remove, and they may remain stable for several days
[66,67].
Second, due to the internal surface area in a colloidal suspension
of at least few m 2, this method is relatively insensitive to impurities.
By the same token, the dose of any additives adsorbing to the particle
surfaces can be controlled with high precision. The classical colloidal
probe involves normally one or two colloidal particles with an area
of few μm 2, which can be easily contaminated and the dose is difficult
to control.
Third, forces between two similar particles from the same batch are
being measured. In this sphere–sphere geometry, one may obtain the
surface potential unambiguously due to the symmetry of the system.
This technique may also be used to calibrate the surface potential
(a)
(c)
Fig. 1. Principle of the multi-particle colloidal probe technique. (a) Preparation of the particle suspension. (b) Silanization of the tip-less cantilever and glass slide (red). (c) Injection of the
suspension into the AFM fluid cell, which leads to deposition of the particle on the slide. (d) Pick-up of a particle by the cantilever, rinsing, and force measurement.
87
of the probe particle, which may be subsequently used in a regular
colloidal probe experiment in the sphere–plane geometry.
The critical aspect of the multi-particle colloidal probe technique is
the right silanization procedure to make the particles stick. We have
been successful in attaching latex particles to well cleaned silica or
glass substrates that have been silanized with hexamethyldisilazane
or 3-(ethoxydimethylsilyl)propylamine in the gas phase. Centering
of the particles of few μm in diameter can be achieved with the AFM
scanner and the optical microscope to sufficient precision.
The multi-particle colloidal probe technique was used to inves-
tigate interaction forces between charged latex particles in the
presence of oppositely charged polyelectrolytes (see Fig. 2). Posi-
tively charged amidine latex particles in the presence of negatively
charged poly(styrene sulfonate) (PSS) will be discussed first [64,68],
while negatively charged sulfate latex particles in the presence of
poly(amido amine) dendrimers (PAMAM) or linear poly(ethylene
imine) (LPEI) follow subsequently [65,69,70]. As the interactions
depend sensitively on the amount of adsorbed polyelectrolyte, the
use of multi-particle colloidal probe technique is essential, since
the large surface area permits to control the adsorbed amount precisely.
Based on these results, a comprehensive picture of the interaction forces
between charged particles in the presence of adsorbed polyelectrolytes
will be put forward.
3. Amidine latex particles in the presence of poly(styrene
sulfonate) (PSS)
Let us first discuss forces between positively charged amidine latex
particles in the presence of negatively charged linear poly(styrene
sulfonate) (PSS, see Fig. 2). The amidine-functionalized poly(styrene)
particles have a diameter of about 3 μm. Bare particles as well as parti-
cles with a saturated layer of adsorbed PSS are highly charged, and are
strongly repulsive as shown in Fig. 3. These forces originate from the
repulsion of the diffuse parts of the electric double-layer, as indicated
by their exponential nature and substantial ionic strength dependence
(see Fig. 4). From such measurements the surface potentials can be
extracted, and they are summarized in Fig. 5. We mostly report forces
between particles normalized to the effective radius
R1 R2
Reff ¼ ð1Þ
R1 þ R2
(b)
(d)
88 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98

(a) (c)

(b)

Fig. 2. Adsorption of polyelectrolytes on oppositely charged particles. (a) Anionic polyelectrolyte on a positively charged particle and (b) cationic polyelectrolyte on a negatively
charged one. (c) Polyelectrolytes discussed here include anionic poly(styrene sulfonate) (PSS), and cationic linear poly(ethylene imine) (LPEI) or poly(amido amine) dendrimers
(PAMAM). For simplicity, only the first dendrimer generation (G1) is shown.

whereby R1 and R2 are the radii of the two particles. This normaliza-
tion is made since the Derjaguin approximation stipulates that such
normalized forces are independent of the particle size. This approxi-
mation is expected to be excellent, since the radius of the particles
is much larger than the range of the interaction potential. The particle
radii are determined by averaging particle diameters measured under
different angles from the digital images taken with the optical micro-
scope. The resulting accuracy in the determination of the particle
radius is about 0.2 μm.
Let us first discuss the forces between bare latex particles. These
force profiles are shown in Fig. 3a at an ionic strength of 1.1 mM
in KCl electrolyte at pH 4. DLVO theory asserts that forces between
highly charged particles are repulsive at larger separations due to
double-layer forces originating from the overlap of the diffuse layers
[40,41,71]. These forces can be described in terms of the exact nu-
merical solution of the Poisson–Boltzmann (PB) equation involving
two charged plates [72]. They primarily depend on the ionic strength
and the surface potential. At large distances the forces decay exponen-
tially with the Debye length κ −1, which is given by
2 2e2 N A I
κ ¼
εε0 kT
where e is the elementary charge, NA the Avogadro's number, I the
ionic strength, ε the dielectric constant of water, ε0 the permittivity
of vacuum, k the Boltzmann constant, and T the absolute temperature.
The forces shown in Fig. 3a have a decay length of about 7.9 nm. This
value is in good agreement with the Debye length κ −1 ≃ 9.2 nm
calculated from the solution composition. This type of agreement is
typical for such measurements, and the expected decay lengths agree
with the experimental ones within 15%. The diffuse layer potential
turns out to be ψd ≃ 62 mV. Note that the sign of the potential cannot
be inferred from force measurements in symmetric systems. Since
the particles are functionalized with positively charged amidine groups,
the potential must be positive at this pH.
The calculated forces are compared for three boundary conditions,
namely constant charge (CC), constant potential (CP), and constant
regulation (CR) [72–75]. The latter boundary condition invokes an
additional regulation parameter p, whereby the conditions of CC
correspond to p = 1 and CP to p = 0. Note that CP is not a lower
bound of the force, and therefore the regulation parameter may
become negative [76]. However, such conditions are not expected
to occur in the present system. In the case presented here, we find a
relatively low value p ≃ 0.27, which indicates that the surfaces regu-
late their charge upon approach substantially, probably by surface
ionization reactions or adsorption of chloride ions.
Fig. 4a shows the corresponding forces in a semi-logarithmic
representation, whereby the exponential nature of the repulsive force
becomes obvious. With increasing ionic strength, the decay length
decreases. At the same time, the intercept diminishes and therefore
ð2Þ
the surface potential decreases too. Fig. 5 shows this dependence
on the ionic strength in detail. The surface potential decreases with
increasing ionic strength. This dependence can be well described by
PB theory, which predicts the charge–potential relationship
 
2kT eσ
ψd ¼ asinh ð3Þ
e 2kTεε0 κ

a b

Fig. 3. Forces between highly charged particles compared with Poisson–Boltzmann (PB) theory with different boundary conditions, including constant charge (CC), constant
potential (CP), and constant regulation (CR). (a) Bare amidine latex particles and (b) same particles with a saturated layer of poly(styrene sulfonate) (PSS).
 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98 89

a b

Fig. 4. Forces between highly charged particles compared with PB theory with constant regulation boundary conditions. (a) Bare amidine latex particles and (b) same particles with
a saturated layer of poly(styrene sulfonate) (PSS) at several ionic strengths.

whereby σ denotes the surface charge density. The solid line shown in
Fig. 5a corresponds to σ≃ 5.9 mC/m2. The fact that this charge density
is substantially smaller than the corresponding conductometric charge
density of 350 mC/m 2 indicates that a large fraction of the charged
amidine groups are neutralized by specifically adsorbed chloride ions.
The diffuse layer potentials measured by AFM are in good agreement
with surface potentials obtained from electrophoresis. The electropho-
retic mobilities were converted to surface ζ-potentials by means of
the standard electrokinetic model developed by O'Brien and White
[77]. The use of this model is necessary, as simpler electrophoretic
theories may lead to spurious results, especially for the larger particles
and low ionic strengths used here.
Forces between particles with a saturated adsorbed PSS layer
follow very similar patterns. The polymer with the molecular mass
of 2260 kg/mol is adsorbed at a dose of 4.9 mg/g latex at an ionic
strength 1.1 mM. This dose is sufficiently high to ensure a saturated
layer, meaning that an increase of this dose will not lead to higher
adsorbed amount. Independent studies have established that very
thin layers with a thickness of a few nm form under these condi-
tions [78,79]. Forces between the particles coated with PSS are
shown in Fig. 3b. One observes strongly repulsive forces as for the
bare latex particles, which indicates that the coated particles are
also highly charged. The surface potential and regulation parameter
are ψd ≃ − 38 mV and p ≃ 0.11. The adsorption of the negatively
charged PSS to the positively charged latex particles leads to charge
reversal or overcharging. This phenomenon will be discussed below
in more detail. The small value of the regulation parameter suggests
that the charge varies substantially upon approach, which is probably
due to the substantial amount of cations adsorbed within the poly-
electrolyte layer. In contrast to the force profiles between the bare
latex particles, one observes the presence of a short ranged attractive
force. This attractive force will be addressed in more detail later.
Fig. 4b clearly demonstrates that the repulsive forces are due to dif-
fuse layer overlap. The forces decay exponentially and the corresponding
decay lengths agree well with the ones calculated from the ionic
strength of the solution. The magnitude of the fitted surface potentials
also decreases with increasing ionic strength, as indicated in Fig. 5b.
The fit to the PB model given by Eq. (3) yields the charge density of
σ≃−2.6 mC/m2. This surface charge density of a relatively low mag-
nitude indicates that counterions play an important role in the charge
balance.
Forces between the amidine particles with increasing dose of PSS
are shown in Fig. 6 with the best fits obtained by the PB model and
CR boundary conditions. The forces between the bare particles are
strongly repulsive due to diffuse layer overlap as discussed above. As
the PSS dose is being increased, the particles still interact by double-
layer forces, but their amplitude decreases. This decrease indicates a
decrease in the magnitude of the diffuse layer potential. At the same
time, short ranged attractive forces set in. At a dose of 0.9 mg/g, the dif-
fuse layer repulsion disappears completely, and only the attractive force
remains. As the dose increases further, the diffuse layer force sets in
again. Its magnitude increases until it saturates at a dose of 1.3 mg/g.
This increase of the force can be rationalized by the fact that the
adsorbed PSS reverses the overall surface charge of the particles.
When the PSS dose is increased even further, the forces remain con-
stant. This behavior indicates that the adsorption of PSS saturates.

a b

Fig. 5. Comparison of surface potentials obtained from direct force measurements by AFM and electrophoretic mobility (EM) as a function of the ionic strength. (a) Bare amidine
latex particles, and (b) the same particles with a saturated layer of poly(styrene sulfonate) (PSS).
90 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98
a b
Fig. 6. Forces between positively charged amidine latex particles with increasing dose of PSS at an ionic strength of 1.1 mM. (a) Below and at the isoelectric point (IEP) and (b) at IEP
and above.
Beyond that point, the adsorbed PSS amount remains constant, and any
excess PSS remains dissolved in solution. At doses lower than 1.3 mg/g,
however, the polyelectrolyte adsorbs on the particle surface quantita-
tively [80,81].
The best way to illustrate the characteristic charge reversal is
by plotting the surface potentials versus the PSS dose as shown in
Fig. 7. The respective charge densities were obtained from the surface
potential with the relation given in Eq. (3). Surface potentials obtained
by AFM and electrophoresis are in relatively good agreement. The
potential vanishes at the isoelectric point (IEP) and at this point the
forces are fully attractive. Since the signs of the potentials obtained
by AFM cannot be determined, they were adjusted according to the
fact whether the data were measured below or above IEP. This figure
further illustrates that the charge reversal is independent of the PSS
molecular mass.
The fact that the polyelectrolyte adsorbs beyond the IEP and
induces a charge reversal indicates the presence of strong attractive
polyelectrolyte–surface interactions, which cannot be explained by
simple smeared-out electrostatic model. The classical explanation
is that additional non-electrostatic forces (e.g., van der Waals,
Fig. 7. Surface potential (left axis) and surface charge density (right non-linear axis) of
positively charged amidine latex particles with increasing dose of PSS with different
molecular masses at an ionic strength of 1.1 mM at pH 4. Comparison between direct
force measurements by AFM (open symbols) and electrophoretic mobility (EM, gray
symbols). The top panel shows pictorially the structure of the adsorbed PSS layer
with increasing PSS dose.
hydrophobic) are present, which lead to strong adsorption and charge
reversal. Another explanation is that the adsorbed layer is heteroge-
neous due to repulsion between the adsorbed polymer chains [82].
While such charge correlation effects are certainly present, they seem
to lead to less effective overcharging, and therefore both effects are like-
ly to be important. When one compares the number of charges originat-
ing from the polyelectrolyte at IEP with number of charges on the
particles, one finds that they are roughly the same. Therefore, the ad-
sorption is approximately stoichiometric, which means that one PSS
charge neutralizes one charge on the particle surface.
Let us now discuss the attractive forces between the particles near
the IEP. The results are summarized in Fig. 8. Two characteristic
trends can be observed. The attraction weakens with increasing
ionic strength of the solution and with decreasing molecular mass
of PSS. Due to this dependence, we suspect that the attractive force
has two different contributions. First, the classical van der Waals
attraction, which is inherent to DLVO theory, and an additional
non-DLVO force due to electrostatic patch-charge attraction.
At higher ionic strength and low molecular mass, the force can be
interpreted solely by van der Waals interactions [40,41,71]
F H
¼− ð4Þ
Reff 6ðh þ δÞ2
where h is the separation distance and H the Hamaker constant. The
parameter δ quantifies a shift of the plane of origin, and this transfor-
mation can be used to include effects of roughness in an approximate
fashion. With the theoretical value of the Hamaker constant for
polystyrene of H = 9.0 × 10 −21 J [83], one can rationalize the measured
attraction with a shift of δ ≃ 1.1 nm. The latter value corresponds quite
well to twice the mean square roughness, which was estimated to be
about 0.4 nm by AFM imaging. The value of δ is also comparable to
the thickness of the adsorbed PSS layer [78,79].
The additional non-DLVO attraction can be modeled with an addi-
tive exponential force
F −qh
¼ −Ae ð5Þ
Reff
where q −1 is the decay length and A the amplitude. The observed
forces are fully consistent with this simple approach. For high mo-
lecular mass PSS, one observes a decay length that is independent
of the ionic strength and has a value of q − 1 ≃ 2.4 nm. The amplitude
decreases strongly with increasing ionic strength. However, with
increasing molecular mass both the decay length and the amplitude
increase.
 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98 91

a b

Fig. 8. Attractive force profiles near the IEP for positively charged amidine latex particles in the presence of PSS. (a) Ionic strength dependence for PSS of a molecular mass of
2260 kg/mol. (b) Dependence on the molecular mass at an ionic strength of 0.1 mM. The thick line indicates the van der Waals (vdW) force.

These findings can be rationalized in an approximate fashion with-
in the patch-charge model [84–86]. This model considers interactions
between unevenly charged surfaces with a periodic structure across an
electrolyte solution within Debye–Hückel theory. At larger distances,
the model predicts an exponential attraction as given in Eq. (5). For a
square lattice, the decay length is given by [86]
π2
2 2
q ¼κ þ
a mass dependence can be rationalized with the relation a ∝M 1/2,
where 2a is the lattice constant. This relation suggests that the decay
length remains constant at low ionic strength and is solely determined
by the lattice spacing. At higher ionic strengths, the decay length de-
creases and approaches the Debye length. Eq. (6) is consistent with
the experimental data shown in Fig. 9a with 2a ≃ 15 nm. This length
scale corresponds to a characteristic size of the surface charge heteroge-
neities. The above relation equally suggests that at higher salt levels the
decay length should decrease with the ionic strength. The molecular
ð6Þ

Fig. 9. Parameters entering the attractive exponential force observed for positively charged amidine latex particles neutralized by PSS. Results of calculations based on the
patch-charge model are shown as solid lines. The ionic strength dependence is represented in the left column, while the molecular mass dependence is given in the right one.
(a) Decay length and (b) amplitude.
92 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98
(a)
(b)
(c)
(d)
Fig. 10. Approach and retract parts of the individual force curves near IEP for positively
charged amidine latex particles neutralized by PSS. The retract parts mostly feature
(a) short-ranged adhesion events. Some force curves also indicate the existence single
molecule bridging events, such as (b) chain stretching events, (c) chain peeling events,
and (d) multiple chain events.
which follows from the fact that the dose needed to reach the IEP is
independent of the molecular mass.
For a single small patch within the unit cell, the amplitude can be
approximately described by [86]
4πΔσ 2 θ2
A≃ f ðq=κ Þ ð7Þ
ε0 εq ⋅
where σ is the difference between the charge density of the patch and
the bare surface, θ is the fractional surface coverage. The function is
f(x) = 1 for CC boundary conditions and f(x) = x 2 for CP conditions.
From independent experiments, one can estimate Δσ ≃ − 22 mC/m 2
and θ ≃ 0.35. The molecular mass dependence of the amplitude can
be estimated by realizing that the gyration radius Rgof PSS increases
with the molecular mass M as Rg ∝ M υ where υ ≃ 0.6 as typical for neu-
tral polymers in good solvents [87]. Assuming that adsorbed PSS
a b
Fig. 11. Forces between negatively charged sulfate latex particles with increasing dose of PAMAM G10 dendrimers at an ionic strength of 1.1 mM. (a) Below and at the isoelectric
point (IEP) and (b) at and above IEP.
maintains the same extension in the adsorbed state as in solution,
the surface coverage obeys θ ∝ M 2υ − 1. From these dependencies it is
possible to calculate the amplitude for both boundary conditions as
shown in Fig. 9. While the agreement is far from being quantitative,
the model reproduces the observed trends and the order of magni-
tude of the amplitude correctly.
The simple model of patch-charge interactions used here invokes
numerous simplifying assumptions. Most importantly, the model as-
sumes a periodic square lattice, while the arrangement of PSS on the
surface is random, and eventually resembles a liquid-like structure.
Charge regulation appears to have substantial effect on the strength
of the patch-charge attraction, but is not considered except for the
cases of CC and CP conditions. Experimental results are closer to the
CC conditions, which correspond to the weakest force predicted by
the patch-charge model. The model for the forces presently used
assumes that the lattices are arranged laterally such that the maximum
attraction occurs, while fluctuations and the disordered surface ar-
rangement may reduce the strength and range of these forces further.
Finally, the size of the interacting areas on the particles may also be
important. While all these effects may play a role, the approximate
agreement between model and experiment suggests that the additional
attractions observed are mainly due to the patch-charge mechanism.
While several authors have argued that patch-charge attraction in-
deed represents an important contribution to the interactions between
surfaces with adsorbed polyelectrolytes [6,15,64,70,88,89], others have
suggested that additional attractive forces originate from polymer
strands bridging the two surfaces [1,44,90–92]. Distinguishing these
two mechanisms experimentally has proved difficult. However, clear
evidence concerning this point can be obtained from the retract part
of the force curves in the present colloidal probe experiments. Fig. 10
summarizes few typical force curves. Most retract curves only feature
a short range adhesion event (Fig. 10a). Such events are characteristic
for van der Waals and patch-charge interactions discussed above,
but could also reflect additional short-range forces. Less than 3% of the
retract force curves show clear signatures of polymer bridging. Some
events originate from the pulling of a single polymer chain bridging
the two particles and lead a characteristic spikes (Fig. 10b). A second
type of event can be rationalized by individual chains peeling from
the surface and they feature characteristic plateaus (Fig. 10c). Other
more complex events may have contributions from several chains
(Fig. 10d). Even in such complex events the contributions of individual
chains can be sometimes disentangled [93]. While the existence of
bridging events can be unequivocally demonstrated, one finds that
they are extremely rare, and their number is always below 3%. From
 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98
Fig. 12. Surface potential (left axis) and surface charge density (right non-linear axis)
of negatively charged sulfate latex particles with increasing dose of PAMAM G6 and
G10 dendrimers at an ionic strength of 1.1 mM. Comparison between direct force mea-
surements by AFM (open symbols) and electrophoretic mobility (EM, gray symbols).
this observation we conclude that bridging forces are unimportant in
the PSS–amidine latex system, and that additional attractions originate
from patch-charge forces.
4. Sulfate latex particles in the presence of poly(amido amine)
(PAMAM) dendrimers
The second situation to be discussed concerns forces between
negatively charged sulfate latex particles in the presence of positively
charged poly(amido amine) (PAMAM) dendrimers (see Fig. 2). The
sulfate-functionalized poly(styrene) particles have again a diameter
of about 3 μm. PAMAM dendrimers are cationic, highly branched
polyelectrolytes, which strongly adsorb to negatively charged sur-
faces. The molecular mass of the dendrimers can be widely varied
through their generations, which are denoted as G1, G2, etc. The
highest available generation of PAMAM dendrimers is G10, which
has a molecular mass near 930 kg/mol. The adsorption of dendrimers
is normally irreversible, and desorption does not occur, except for the
lowest generations and for highly acidic conditions [94].
The bare sulfate latex particles are negatively charged. The forces
are strongly repulsive and dominated by diffuse layer overlap. An
example of such an exponential force profile is shown in Fig. 11.
The strength of these repulsive forces can be rationalized with a
diffuse layer charge density of σ ≃ − 3.0 mC/m 2. The fact that the
particles are negatively charged has been confirmed with electropho-
retic mobility measurements and is consistent with their surface func-
tionalization. When PAMAM dendrimers are added to the system, the
a b
Fig. 13. Attractive force profiles near the IEP for negatively charged sulfate latex particles neutralized with cationic PAMAM dendrimers. (a) Ionic strength dependence of G10
dendrimers. (b) Dependence on the dendrimers generation at an ionic strength of 0.1 mM. The thick line indicates the van der Waals (vdW) force.
93
surface charge of the particle is being progressively neutralized and
short ranged attractive forces set in. When the IEP is reached, only the
attractive forces are present. When the dose is being increased
beyond the IEP, the particles accumulate positive charge until the
saturation point is being reached. Beyond this point, the forces remain
insensitive to further addition of dendrimers.
The situation is analogous to the case of PSS discussed above. The
cationic dendrimers adsorb on the negatively charged surface. Initially,
the surface charge is progressively reduced until the IEP is reached. Be-
yond the IEP, the amount of adsorbed dendrimers increases further,
whereby the surface becomes positively charged. At an even higher
dose, the adsorption process saturates, and the adsorbed amount does
not increase any further. In this case, however, one finds that that
depending on the generation 2–8 charges on the dendrimers are neces-
sary to neutralize one charge on the surface. This super-stoichiometric
adsorption indicates that a large number of anions co-adsorb with the
dendrimers.
The diffuse layer potentials and charge densities extracted from
the force curves are shown in Fig. 12. These measurements are com-
pared with surface potentials obtained from electrophoresis. In spite
of some scatter, one observes a relatively good agreement between
both sets of measurements. Initially, the particles are negatively
charged. With increasing amount of positively charged dendrimers
their surface charge is neutralized until the IEP is reached. This pro-
cess continues through the IEP, and dendrimers adsorb beyond this
point, which makes the particles positive. In spite of their globular
structure, PAMAM dendrimers in the presence of sulfate particles
strongly resemble linear PSS in the presence of amidine particles.
The main difference is that the resulting adsorbed layers remain
rather dilute even at saturation, especially at low ionic strengths.
The dendrimers are highly charged, and therefore they repel strongly.
This repulsion is also operational during the adsorption process,
which causes that dendrimers avoid adsorbing in the neighborhood
of already adsorbed dendrimers. These electrostatic repulsions lead
to liquid-like ordering of the adsorbed layers, as can be demonstrated
by AFM imaging [95–97]. These surfaces represent good examples
of surfaces with lateral patch-charge heterogeneities, since the
dendrimers are positively charged and the particle surface in between
is negative.
Fig. 13 shows the attractive forces near the IEP. One observes
that these forces depend strongly on the ionic strength as well as on
the dendrimer generation. At lowest generations and higher ionic
strengths the attractive forces are the weakest. These forces are con-
sistent with van der Waals interactions between polystyrene particles
across water as given by Eq. (4) provided one introduces a shift of the
plane of origin of δ ≃ 0.85 nm. This distance is well comparable to the
mean square roughness of the particles of 1.3 nm as determined by
94 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98

Fig. 14. Parameters entering the attractive exponential force observed for negatively charged sulfate latex particles neutralized by PAMAM dendrimers. Results of calculations based
on the patch-charge model are shown as solid lines. The ionic strength dependence for PAMAM G10 is represented in the left column, while the generation dependence at an ionic
strength of 0.1 mM is given in the right column. (a) Decay length and (b) amplitude.

AFM imaging. The strengthening of the attractions with decreasing
ionic strength and increasing dendrimers generation can be interpreted
with patch-charge interactions. This interaction is expected to be expo-
nential as described by Eq. (4). The solid lines shown in Fig. 13 indicate
that the best fit to the sum of the van der Waals force and the exponen-
tial force law is fully consistent with the experimental data.
The square-lattice patch-charge model can be again used to ratio-
nalize the data semi-quantitatively. Fig. 14 compares the observed
decay length q −1 and the magnitude of the amplitude for different
ionic strengths and generations. In this situation, one clearly sees
that the decay length decreases with increasing ionic strength as
predicted by the patch-charge model. The nearest-neighbor spacing
can be accurately estimated from the adsorbed amount at IEP to be
2a ≃ 49 nm. When this value is inserted into Eq. (6) one obtains
the right trend, but the decay length is overestimated at low ionic
strength. The disagreement may be again related to the fact that
the arrangement of the dendrimers on the surface is not crystalline,
but liquid-like. The charge density difference can be estimated as
σ≃30 mC=m2 from the diffuse layer charge of the bare particles and
the effective charge of the dendrimers. The surface coverage θ ≃ 0.14
can be obtained from the adsorbed amount and the radius of the
adsorbed dendrimers determined by AFM. With these values Eq. (7)

a b

Fig. 15. Forces between negatively charged sulfate latex particles with an increasing dose of LPEI with a molecular mass of 250 kg/mol at an ionic strength of 1.1 mM and pH 4.
(a) Below and at the isoelectric point (IEP) and (b) at and above IEP.
 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98
Fig. 16. Surface potential (left axis) and surface charge density (right non-linear axis)
of negatively charged sulfate latex particles with increasing dose of LPEI at an ionic
strength of 1.1 mM and pH 4. Comparison between direct force measurements by
AFM (open symbols) and electrophoretic mobility (EM, gray symbols).
yields the results for CC and CP boundary conditions shown in Fig. 14.
The observed amplitudes agree well with the CC boundary condition.
The effect of the molecular mass on the patch-charge interaction
can be addressed as follows. The lattice constant a can be again related
to the nearest-neighbor spacing between the dendrimers, which can
be obtained from the adsorbed amount. The amplitude can be calcu-
lated from the adsorbed mass at IEP determined from electrophoresis
and by approximating the dendrimers radii by the corresponding
values is solution. The results are shown in Fig. 14. The decay length
is again overestimated by the model, especially at higher generations.
However, the observed amplitude is in good agreement with the
values obtained from the CC boundary conditions.
A comparison between the present interpretation based on the
patch-charge model with a recent computer simulations is most in-
structive [65]. That study modeled the dendrimers as charged hard
spheres interacting with a screened Coulomb potential. These simula-
tions equally predict a charge reversal and the occurrence of additional
short-ranged attractive forces near the IEP. Again, the predicted range of
the attraction is significantly larger than the one observed experimen-
tally, quite in analogy to the estimate based on the patch-charge
model. A possible reason for this discrepancy is that the patch-charge
model assumes that the dendrimers are arranged in a square lattice,
while they have a liquid-like structure on the surface. The simulations
assume that dendrimers are mobile at the interface, which makes
them crystallize upon approach. In reality, however, the dendrimers
are likely to be frozen in a liquid-like structure and to be immobile on
the surface. In this situation, the surface is laterally more disordered,
a b
Fig. 17. Attractive force profiles near the IEP for negatively charged sulfate latex particles neutralized with cationic LPEI. (a) Ionic strength dependence for LPEI of molecular mass of
(a) 250 kg/mol and (b) 2.5 kg/mol. The thick line indicates the van der Waals (vdW) force.
95
which may explain the shorter range of the attractions. Thus, the source
of the discrepancy between experiment and simulations could be simi-
lar as for the patch-charge model.
5. Sulfate latex particles in the presence of linear poly(ethylene
imine) (LPEI)
The last example to be discussed is the case of negatively charged
sulfate latex particles in the presence of linear poly(ethylene imine)
(LPEI). The particles used were the same as described in Section 4.
LPEI is a linear weak polyelectrolyte, which is neutral in basic conditions
and positively charged in acidic conditions [98–102]. The experiments
were again carried out at pH 4.0 and at these conditions the degree of
ionization of LPEI is about 65% [98].
Fig. 15 shows the forces between the sulfate latex particles for
different doses of LPEI added. The forces between the bare sulfate
latex particles are repulsive and long-ranged due to their negative
surface charge and resulting overlap of the diffuse layers. As the dose
of LPEI is increased, the repulsive forces weaken and a short-ranged
attraction sets in. At the IEP, which is located at 0.28 mg/g, the
long-ranged repulsive forces disappear and only the attractive part
can be seen. As the dose is increased, the long-ranged repulsive forces
set in again. Their strength increases further until they remain con-
stant beyond the dose of about 1.1 mg/g.
This dependence of the forces on the polyelectrolyte dose is analo-
gous to the previous examples, and can be again explained by a charge
reversal process. Fig. 16 shows the respective surface potentials and
charges obtained by AFM and they are compared with results obtained
by electrophoresis. Reasonably good agreement is obtained between
these two methods. The surface charge density is negative at first,
and increases with increasing LPEI dose. The surface is fully neutral-
ized at IEP, and further LPEI addition leads to overcharging and accu-
mulation of a positive charge. The adsorption process saturates at
a dose of 1.1 mg/g, and upon further addition the adsorbed amount
remains constant. The adsorption is again super-stoichiometric and
about 8 charges on the LPEI are necessary to neutralize one charge
on the surface.
The attractive forces near the IEP are shown in Fig. 17. The exper-
imental data for different ionic strengths and molecular masses are
compared with the van der Waals forces. They were calculated from
Eq. (4) by introducing a shift of δ ≃ 0.85 nm as above. The forces
depend neither on the ionic strength nor on the molecular mass,
and one concludes that no additional attractive interactions beyond
van der Waals forces are present. This behavior is in sharp contrast
with the two previous cases, where the attractive forces significantly
depend on these parameters. These additional forces were interpreted
in terms of patch-charge attraction, and thus we suspect that no such
96 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98
(a) (b)
Fig. 18. Approach and retract parts of the individual force curves near IEP for sulfate latex
particles in the presence of LPEI. The retract parts mostly feature (a) short-ranged adhe-
sion events, but (b) few force curves indicate the existence single molecule bridging
events.
additional forces are present in the LPEI system. Since patch-charge in-
teractions are induced by surface charge heterogeneities, the adsorbed
LPEI layers must be therefore rather homogeneous. A possible reason
why this layer could be more homogeneous than the previously dis-
cussed ones will be given in the next section.
Since no additional interactions seem to be present, one should not
detect any bridging forces upon the retraction of the particles. Fig. 18
shows that retraction force curves normally show short-range adhe-
sion. Only in some cases, individual bridging events can be observed.
However, such events are very infrequent, and their probability is
typically b2%. This observation confirms again that bridging forces
are unimportant in this system.
6. Conclusions and outlook
The present review discusses the advantages of the novel
multi-particle colloidal probe technique to directly measure interac-
tion forces between charged colloidal particles in the presence of
oppositely charged polyelectrolytes. Three systems have been stud-
ied in detail so far, and they involved amidine and sulfate function-
alized polystyrene latex particles and poly(styrene sulfonate) (PSS)
[64,68], poly(amido amine) (PAMAM) dendrimers [65,70], and linear
poly(ethylene imine) (LPEI) [69]. The technique permits to study
reproducibly the small polyelectrolyte doses needed to reach near the
isoelectric point (IEP), where the forces become attractive. To our
knowledge, this regime cannot yet be explored with alternative tech-
niques available for direct force measurements.
The key advantage of the multi-particle colloidal probe technique
is that one works with a suspension with several m 2 of inner surface
area, which permits to control the dose of added polyelectrolytes pre-
cisely, and further makes the method less sensitive to impurities.
Classical colloidal probe technique uses just individual colloidal parti-
cles with a surface area of few μm 2. In multi-particle colloidal probe
technique the particles are attached to the cantilever in-situ in the
AFM fluid cell, and one easily eliminates the presence of nanobubbles.
Such bubbles frequently form during the drying–rewetting step
through the mounting of hydrophobic latex particles on the AFM canti-
lever. Finally, as the technique measures forces between two colloidal
particles from the same suspension in the sphere–sphere geometry,
one can study forces in symmetric systems, and equally address its
variations by exploring different pairs of particles.
Based on the discussed results, we present a generic picture of how
interactions between charged colloidal particles are influenced by op-
positely charged polyelectrolytes. Thereby, one should address two
aspects separately, namely polyelectrolyte adsorption and resulting
interaction forces between surfaces with adsorbed polyelectrolyte
films.
Polyelectrolytes interact strongly with oppositely charged sur-
faces, principally due to electrostatic attractions, which lead to the
formation of very thin adsorbed films [78,79]. The adsorption process
saturates as soon as the repulsive electrostatic forces between the
individual adsorbed polyelectrolytes become sufficiently strong. This
saturation normally happens at relatively low surface concentrations,
which are typically around 0.1–1 mg/m 2. At these surface concentra-
tions, however, the charge density of the adsorbed polyelectrolyte
film exceeds by far the charge of the particle surface, and a charge
reversal (or overcharging) has taken place. The reason that the
adsorption proceeds beyond the charge reversal point is not only
related to charge correlations within the adsorbed film, but also to
the presence of additional attractive van der Waals or hydrophobic
forces, which promote adsorption of the polyelectrolytes irrespective
of their charge. One important observation illustrating the importance
of such additional forces is that polyelectrolytes do not desorb even at
salt levels exceeding 1 M [94], while it is established theoretically that
polyelectrolytes should desorb at these conditions if only electrostatic
forces would be operational [29,32,33]. Another consequence of the
low adsorbed amount is that the adsorbed films are strongly hetero-
geneous, consisting of individual adsorbed polyelectrolytes that are
unevenly distributed on the oppositely charged particle surface.
Such patch-charge distributions can be visualized by AFM for the
adsorbed PAMAM dendrimers [65,70,95], but also for linear polyelec-
trolytes [64,103]. This regime is probably similar to the dilute 2-d
Wigner regime proposed on theoretical grounds earlier [33]. When
the polyelectrolytes are weakly charged, the adsorbed amounts are
larger and the adsorbed films are expected to be more homogeneous,
and one should approach the semidilute 2-d Wigner regime.
The mentioned theoretical arguments assume thermodynamic
equilibrium [29,32,33]. However, in the real situation adsorption of
polyelectrolytes is basically irreversible. Once a polyelectrolyte chain
has adsorbed, it will not desorb except under extreme conditions, and
any rearrangement or migration on the surface is unlikely [51,94]. In
particle suspensions, the irreversibility of the adsorption process has
another important consequence. When the dose is sufficiently high,
the adsorbed layer will always saturate, and any excess polyelectrolyte
will remain dissolved in solution. At lower dose, however, the ad-
sorption will be always quantitative and leads to the formation of
under-saturated but otherwise stable adsorbed films. Formation
of such films is only limited by adsorption kinetics, which is rapid
due to attractive polyelectrolyte–particle interactions [80].
Based on this picture of the adsorbed polyelectrolyte layers, we
can proceed to discuss the resulting interaction forces. Since adsorbed
polyelectrolyte films are thin and highly charged, the interactions at
larger distances are dominated by electrostatic double-layer forces
inherent to DLVO theory. Such forces originate from the overlap of
diffuse layers, and they can be quantified by PB theory. One finds
that the magnitudes of the diffuse layer charge densities of adsorbed
polyelectrolyte films are in good agreement with results from electro-
phoresis and that they are in the range of 2–5 mC/m 2. These relatively
low charge densities indicate that the majority of the charges of
adsorbed polyelectrolyte chains are neutralized by coadsorbed coun-
terions. As the polyelectrolyte dose is increased, repulsive diffuse
layer forces decrease in strength until they vanish at the isoelectric
point (IEP) and then they increase again in strength until the adsorbed
layer is saturated. In particular, forces between saturated films of
adsorbed polyelectrolytes are dominated by such repulsive diffuse
layer forces.
With increasing polyelectrolyte dose, additional short-range at-
tractive forces set in. These forces can be most easily studied close to
the IEP, where the repulsive diffuse layer forces are absent. However,
additional attractive forces are also present away from the IEP. At
IEP, these forces can be sometimes rationalized by simple van der
Waals forces expected from DLVO theory. In other cases, additional
attractive non-DLVO forces are present. The strength of the additional
component increases with decreasing ionic strength and increasing
molecular mass of the polyelectrolyte. This additional component
can be interpreted to occur from the patch-charge distribution on
 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98
(a)
Fig. 19. Schematic representation of the different systems presented here. (a) Very heterogeneous layer formed by PAMAM dendrimers. (b) Moderately heterogeneous layer with
PSS. (c) Relatively homogeneous LPEI film.
the surface. When two surfaces with an inhomogeneous charge distri-
bution approach, they will always arrange laterally such that a posi-
tively charged patch will face a negatively charged one, leading to a
net electrostatic attraction. When the patches are laterally correlated,
these forces will be stronger. In the case of periodic structures,
such forces have been quantified within the Debye–Hückel theory
[86]. The results of this theory are in semi-quantitative agreement
with experimental values for the strength and range of the attractive
component, suggesting that the patch-charge mechanism is indeed
dominant in the present systems.
These additional patch-charge attractions are most prominent for
adsorbed layers of PAMAM dendrimers, they are weaker for the PSS
films, and finally they are experimentally non-identifiable for the
LPEI films (see Fig. 19). The surface patch-charge heterogeneities
are most developed in the PAMAM system, due to the globular
and highly branched nature of the polyelectrolyte. The adsorbed
dendrimer films are strongly correlated laterally with a pronounced
liquid-like structure. Therefore, patch-charge forces should be strong
in this system. AFM imaging suggests that adsorbed PSS films are
also laterally heterogeneous, but due to the linear structure of the
polyelectrolyte they are more disordered and more homogeneous.
The fact that patch-charge forces are weaker in this system than in
the dendrimer system is thus not surprising. In the LPEI system, the
patch-charge forces seem to be negligible with respect to the van
der Waals force, which means that the adsorbed LPEI film has to be
even more homogeneous than the PSS film. The PAMAM and PSS
systems probably resemble the dilute 2-d Wigner regime, while the
LPEI system is closer to the semidilute 2-d Wigner regime. These
two regimes were proposed theoretically some time ago [33].
The presence of additional attractive non-DLVO forces that strength-
en with increasing molecular mass and decreasing ionic strength can
be also inferred from colloidal stability studies [81,104,105]. In these
studies, smaller but otherwise similar latex particles have been investi-
gated in the presence of the same polyelectrolytes. The early-stage
aggregation rates are measured with time-resolved dynamic light
scattering. The overall pattern of the interaction forces discussed
here is fully confirmed by these studies. Near the IEP, the aggregation
is rapid because of the presence of attractive forces, while away from
the IEP the aggregation slows down due to buildup of charge and
resulting double-layer repulsions. However, two effects are present
that signal the presence of attractive non-DLVO forces. First, the aggre-
gation at low ionic strength may become even faster than in the pres-
ence of excess salt near IEP. Second, the aggregation rate is enhanced
when the molecular mass of the polyelectrolyte is increased. In
the PAMAM dendrimer system, one indeed observes a substantial
enhancement of the aggregation rate at IEP at low salt as well as
with increasing molecular mass of the dendrimers experimentally
[104] and in computer simulations [106]. Similar effects have been
found in the PSS system, albeit the enhancement is somewhat weaker
[81]. However, these effects have not been observed in the LPEI system
[105]. Therefore, colloidal stability studies confirm that patch-charge
attractions are strongest in the PAMAM system, followed by the PSS
system, and that these forces are negligible in the LPEI system. A recent
study involving weakly charged polyelectrolytes also confirms that
patch-charge effects are less important in such systems [107]. Based
on colloidal stability studies, Gregory [15] first identified this mechanism
97
(b) (c)
and suggested the importance of patch-charge attraction in oppositely
charged particle–polyelectrolyte systems.
Some authors have suggested that bridging forces could be also
important between surfaces coated with adsorbed polyelectrolytes
[1,44,90–92]. The retract part of the measured force profiles presented
here allows to quantify the frequency of such bridging events un-
equivocally. No bridging events can be identified in the dendrimer
system, which is consistent with the compact architecture of the den-
drimers. For the linear polyelectrolyte systems, namely PSS and LPEI,
single molecule bridging events can be identified, but their frequency
is very low (typically below few %). From these observations we con-
clude that bridging forces are unimportant in the systems presented
here. On the other hand, it is well established that bridging forces are
important for neutral polymers, especially at incomplete coverage
[71,108]. For this reason, bridging forces ought to become relevant
with decreasing charge density of the polyelectrolyte and at higher
ionic strengths. At this point, however, the nature of this crossover re-
mains unclear to us.
Acknowledgment
This article is dedicated to the 65th birthday of John Ralston. MB
and IS would like to thank for his hospitality during the stay at the
Ian Wark Institute and to the University of South Australia for sup-
port. This research was financed by the Swiss National Science Foun-
dation, University of Geneva, Swiss Federal Office for Education and
Science, and COST Action D43.
References
[1] Bolto B, Gregory J. Water Res 2007;41:2301–24.
[2] Horn D, Linhart F. Retention aids. London: Blackie Academic and Professional;
1996.
[3] Dautzenberg H, Jaeger W, Kotz J, Philipp B, Seidel C, Stscherbina D. Polyelectro-
lytes: formation, characterization and application. New York: Hanser Publishers;
1994.
[4] Howe AM, Wesley RD, Bertrand M, Cote M, Leroy J. Langmuir 2006;22:4518–25.
[5] Claesson PM, Dedinaite A, Rojas OJ. Adv Colloid Interface Sci 2003;104:53–74.
[6] Borkovec M, Papastavrou G. Curr Opin Colloid Interface Sci 2008;13:429–37.
[7] Leong YK, Scales PJ, Healy TW, Boger DV. Colloids Surf A 1995;95:43–52.
[8] Elimelech M, Gregory J, Jia X, Williams RA. Particle deposition and aggregation:
measurement, modeling, and simulation. Oxford: Butterworth-Heinemann
Ltd.; 1995.
[9] Freudenberg U, Zimmermann R, Schmidt K, Behrens SH, Werner C. J Colloid
Interface Sci 2007;309:360–5.
[10] Kobayashi M, Skarba M, Galletto P, Cakara D, Borkovec M. J Colloid Interface Sci
2005;292:139–47.
[11] Geffroy C, Persello J, Foissy A, Cabane B, Tournilhac F. Rev Inst Fr Petrole 1997;52:
183–90.
[12] Tobori N, Amari T. Colloids Surf A 2003;215:163–71.
[13] Nystrom R, Linden M, Rosenholm JB. J Colloid Interface Sci 2001;242:259–63.
[14] Eriksson R, Merta J, Rosenholm JB. J Colloid Interface Sci 2007;313:184–93.
[15] Gregory J. J Colloid Interface Sci 1973;42:448–56.
[16] Ashmore M, Hearn J. Langmuir 2000;16:4906–11.
[17] Walker HW, Grant SB. Colloids Surf A 1996;119:229–39.
[18] Killmann E, Adolph H. Colloid Polym Sci 1995;273:1017–79.
[19] Kjoniksen AL, Joabsson F, Thuresson K, Nystrom B. J Phys Chem B 1999;103:
9818–25.
[20] Pefferkorn E. Adv Colloid Interface Sci 1995;56:33–104.
[21] Yu WL, Bouyer F, Borkovec M. J Colloid Interface Sci 2001;241:392–9.
[22] Furusawa K, Kanesaka M, Yamashita S. J Colloid Interface Sci 1984;99:341–8.
[23] Cosgrove T, Obey T, Vincent B. J Colloid Interface Sci 1986;111:409–18.
[24] Cohen Stuart MA, Fleer GJ, Lyklema J, Norde W, Scheutjens JMHM. Adv Colloid
Interface Sci 1991;34:477–535.
98 M. Borkovec et al. / Advances in Colloid and Interface Science 179-182 (2012) 85–98

[25] Hesselink FT. J Colloid Interface Sci 1977;60:448–66. [66] Attard P. Adv Colloid Interface Sci 2003;104:75–91.
[26] van de Steeg HGM, Cohen Stuart MA, de Keizer A, Bijsterbosch BH. Langmuir [67] Ducker WA. Langmuir 2009;25:8907–10.
1992;8:2538–46. [68] Popa I, Gillies G, Papastavrou G, Borkovec M. J Phys Chem B 2010;114:3170–7.
[27] Borisov OV, Zhulina EB, Birshtein TM. Macromolecules 1994;27:4795–803. [69] Finessi M, Sinha P, Szilagyi I, Popa I, Maroni P, Borkovec M. J Phys Chem B
[28] Linse P. Macromolecules 1996;29:326–36. 2011;115:9098–105.
[29] Muthukumar M. J Chem Phys 1987;86:7230–5. [70] Popa I, Papastavrou G, Borkovec M. Phys Chem Chem Phys 2010;12:4863–71.
[30] Hariharan R, Biver C, Mays J, Russel WB. Macromolecules 1998;31:7506–13. [71] Russel WB, Saville DA, Schowalter WR. Colloidal dispersions. Cambridge:
[31] Joanny JF. Eur Phys J B 1999;9:117–22. Cambridge University Press; 1989.
[32] Netz RR, Andelman D. Phys Rep 2003;380:1–95. [72] Pericet-Camara R, Papastavrou G, Behrens SH, Borkovec M. J Phys Chem B
[33] Dobrynin AV, Deshkovski A, Rubinstein M. Macromolecules 2001;34:3421–36. 2004;108:19467–75.
[34] Derjaguin BV, Rabinovich YI, Churaev NV. Nature 1978;272:313–8. [73] Carnie SL, Chan DYC. J Colloid Interface Sci 1993;161:260–4.
[35] Israelachvili JN, Adams GE. J Chem Soc Faraday Trans 1 1978;74:975–1001. [74] Behrens SH, Borkovec M. Phys Rev E 1999;60:7040–8.
[36] Pashley RM. J Colloid Interface Sci 1981;83:531–46. [75] Behrens SH, Borkovec M. J Chem Phys 1999;111:382–5.
[37] Richetti P, Kekicheff P. Phys Rev Lett 1992;68:1951–4. [76] Borkovec M, Behrens SH. J Phys Chem B 2008;112:10795–9.
[38] Dahlgren MAG, Waltermo A, Blomberg E, Claesson PM, Sjostrom L, Akesson T, [77] O'Brien RW, White LR. J Chem Soc Faraday Trans 2 1978;74:1607–26.
et al. J Phys Chem 1993;97:11769–75. [78] Seyrek E, Hierrezuelo J, Sadeghpour A, Szilagyi I, Borkovec M. Phys Chem Chem
[39] Tadmor R, Hernandez-Zapata E, Chen N, Pincus P, Israelachvili JN. Macromolecules Phys 2011;13:12716–9.
2002;35:2380–8. [79] Hierrezuelo J, Szilagyi I, Vaccaro A, Borkovec M. Macromolecules 2010;43:
[40] Derjaguin B, Landau LD. Acta Phys Chim 1941;14:633–62. 9108–16.
[41] Verwey EJW, Overbeek JTG. Theory of stability of lyophobic colloids. Amsterdam: [80] Kleimann J, Gehin-Delval C, Auweter H, Borkovec M. Langmuir 2005;21:3688–98.
Elsevier; 1948. [81] Gillies G, Lin W, Borkovec M. J Phys Chem B 2007;111:8626–33.
[42] Claesson PM, Ninham BW. Langmuir 1992;8:1406–12. [82] Grosberg AY, Nguyen TT, Shklovskii BI. Rev Mod Phys 2002;74:329–45.
[43] Claesson PM, Ederth T, Bergeron V, Rutland MW. Adv Colloid Interface Sci [83] Bevan MA, Prieve DC. Langmuir 1999;15:7925–36.
1996;67:119–83. [84] Richmond P. J Chem Soc Faraday Trans 2 1974;70:1066–73.
[44] Poptoshev E, Rutland MW, Claesson PM. Langmuir 1999;15:7789–94. [85] Richmond P. J Chem Soc Faraday Trans 2 1975;71:1154–63.
[45] Ducker WA, Senden TJ, Pashley RM. Nature 1991;353:239–41. [86] Miklavic SJ, Chan DYC, White LR, Healy TW. J Phys Chem 1994;98:9022–32.
[46] Butt HJ. Biophys J 1991;60:1438–44. [87] Tanahatoe JJ, Kuil ME. J Phys Chem A 1997;101:8389–94.
[47] Clark SC, Walz JY, Ducker WA. Langmuir 2004;20:7616–22. [88] Leong YK. Colloid Polym Sci 2001;279:82–7.
[48] Biggs S. Langmuir 1995;11:156–62. [89] Bouyer F, Robben A, Yu WL, Borkovec M. Langmuir 2001;17:5225–31.
[49] Pericet-Camara R, Papastavrou G, Behrens SH, Helm CA, Borkovec M. J Colloid [90] Salmi J, Osterberg M, Stenius P, Laine J. Nord Pulp Pap Res J 2007;22:249–57.
Interface Sci 2006;296:496–506. [91] Akesson T, Woodward C, Jonsson B. J Chem Phys 1989;91:2461–9.
[50] Block S, Helm CA. Macromolecules 2009;42:6733–40. [92] Chen KL, Mylon SE, Elimelech M. Environ Sci Technol 2006;40:1516–23.
[51] Kirwan LJ, Maroni P, Behrens SH, Papastavrou G, Borkovec M. J Phys Chem B [93] Papastavrou G, Kirwan LJ, Borkovec M. Langmuir 2006;22:10880–4.
2008;112:14609–19. [94] Longtin R, Maroni P, Borkovec M. Langmuir 2009;25:2928–34.
[52] Toikka G, Hayes RA, Ralston J. Langmuir 1996;12:3783–8. [95] Pericet-Camara R, Papastavrou G, Borkovec M. Langmuir 2004;20:3264–70.
[53] Larson I, Drummond CJ, Chan DYC, Grieser F. J Phys Chem 1995;99:2114–8. [96] Cahill BP, Papastavrou G, Koper GJM, Borkovec M. Langmuir 2008;24:465–73.
[54] Rentsch S, Pericet-Camara R, Papastavrou G, Borkovec M. Phys Chem Chem Phys [97] Pericet-Camara R, Cahill BP, Papastavrou G, Borkovec M. Chem Commun 2007:
2006;8:2531–8. 266–8.
[55] Considine RF, Hayes RA, Horn RG. Langmuir 1999;15:1657–9. [98] Smits RG, Koper GJM, Mandel M. J Phys Chem 1993;97:5745–51.
[56] Prieve DC. Adv Colloid Interface Sci 1999;82:93–125. [99] Borkovec M, Koper GJM. J Phys Chem 1994;98:6038–45.
[57] Behrens SH, Plewa J, Grier DG. Eur Phys J E 2003;10:115–21. [100] Borkovec M, Koper GJM. Macromolecules 1997;30:2151–8.
[58] Biggs S, Dagastine RR, Prieve DC. J Phys Chem B 2002;106:11557–64. [101] Borkovec M, Koper GJM, Piguet C. Curr Opin Colloid Interface Sci 2006;11:280–9.
[59] Crocker JC, Grier DG. Phys Rev Lett 1994;73:352–5. [102] Koper GJM, Borkovec M. Polymer 2010;51:5649–62.
[60] Baumgartl J, Arauz-Lara JL, Bechinger C. Soft Matter 2006;2:631–5. [103] Roiter Y, Jaeger W, Minko S. Polymer 2006;47:2493–8.
[61] Drechsler A, Synytska A, Uhlmann P, Elmahdy MM, Stamm M, Kremer F. [104] Lin W, Galletto P, Borkovec M. Langmuir 2004;20:7465–73.
Langmuir 2010;26:6400–10. [105] Szilagyi I, Rosicka D, Hierrezuelo J, Borkovec M. J Colloid Interface Sci 2011;360:
[62] Gutsche C, Keyser UF, Kegler K, Kremer F. Phys Rev E 2007;76:031403. 580–5.
[63] Elmahdy MM, Synytska A, Drechsler A, Gutsche C, Uhlmann P, Stamm M, et al. [106] Trulsson M, Forsman J, Akesson T, Jonsson B. Langmuir 2009;25:6106–12.
Macromolecules 2009;42:9096–102. [107] Sadeghpour A, Seyrek E, Szilagyi I, Hierrezuelo J, Borkovec M. Langmuir 2011;27:
[64] Popa I, Gillies G, Papastavrou G, Borkovec M. J Phys Chem B 2009;113:8458–61. 9270–6.
[65] Popa I, Papastavrou G, Borkovec M, Trulsson M, Jonsson B. Langmuir 2009;25: [108] Sun G, Butt HJ. Macromolecules 2004;37:6086–9.
12435–8.