Langmuir 2005, 21, 11283-11288 11283

Physicochemical Characterization of an Anatase TiO2

Surface and the Adsorption of a Nonionic Surfactant: An
Atomic Force Microscopy Study
Cathy E. McNamee,*,† Yoshinobu Tsujii,‡ and Mutsuo Matsumoto‡
Department of Chemical Engineering, Kyoto UniversitysKatsura, Nishikyo-ku, Kyoto
615-8510, and Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan

Received July 4, 2005. In Final Form: September 11, 2005

Atomic force microscopy was used to characterize an anatase TiO2 surface, prepared by the helical vapor
preparation method. The forces between two bare TiO2 surfaces were measured in the presence of water
at various pH values. This TiO2 isoelectric point (iep) was characterized by the presence of only a van der
Waals attraction and was measured at pH 5.8; this value is similar to that for a rutile TiO2 surface. The
adsorption mechanism of a nonionic surfactant molecule to this anatase TiO2 surface was investigated by
measuring the forces beween two such TiO2 surfaces at their iep pH in the presence of linear dodecanol
tetraethoxylate (C12E4), a poly(ethoxylene oxide) n-alkyl ether. C12E4 was seen by the presence of steric
forces to adsorb to the uncharged TiO2 surface. For low surfactant concentrations, C12E4 adsorbed with
its hydrophobic tail facing the TiO2 substrate, to reduce its entropically unfavorable contacts with water.
Additional surfactant adsorption occurred at higher surfactant concentrations by the hydrophobic and
hydrophilic interactions between the surfactant tails and heads, respectively, and gave sub-bilayers. A
two-step adsorption isotherm was subsequently proposed with four regions: (1) submonolayer, (2) complete
monolayer, (3) sub-bilayer, and (4) bilayer. The absence of a long-range repulsive force between the two
TiO2 surfaces in the presence of the C12E4 surface aggregates indicated that a C12E4 nonionic surfactant
aggregate did not possess charge.

Introduction surfaces but give a monolayer structure on hydrophobic

surfaces.14 To control and improve the adsorption of a
Self-assembled monolayers (SAMS) and other organic-
surfactant to a metal oxide surface, such as silica or
polymeric templates are used in many applications,
titanium dioxide, it is necessary to understand the forces
ranging from environmental purification1 to ceramics and
involved in such a system. The forces between two
medical2 and energy storage and transfer3-5 applications.
hydrophilic or hydrophobic surfaces have been well
They can be made by the specific adsorption of surfactants
characterized for ionic surfactants. However, only the
or DNA onto metals and oxides.6-9
forces between two hydrophobic surfaces in the presence
The structure of surfactants adsorbed at a surface
of a nonionic surfactant in water have been characterized.12
depends on the physical properties of the surface and the
The forces between a hydrophilic surface and a nonionic
surrounding solution.10-15 For example, ionic surfactants
surfactant are still not well understood, as the smaller
adsorb in a micellar or bilayer fashion on hydrophilic
forces due to the nonionic surfactants are generally hidden
by the stronger electrostatic double-layer forces resulting
* To whom correspondence should be addressed. Phone:
+81 75 383 2692. Fax.: +81 75 383 2652. E-mail: cathymcn@
from a charged hydrophilic surface. Such information is
cheme.kyoto-u.ac.jp. vital for the formulation of new materials and the
† Kyoto UniversitysKatsura. expansion of their applications.
‡ Kyoto University. Nonionic surfactants are noncharged in an aqueous
(1) Gardner, K. H.; Arias, M. S. Environ. Sci. Technol. 2000, 34, 160. medium. However, such materials have been seen to give
(2) Kohler, N.; Fryxell, G. E.; Zhang, M. J. Am. Chem. Soc. 2004,
126, 7206.
charged particles after the adsorption of a high concen-
(3) Hoffmann, M. R.; Martin, S. T.; Choi, W. Y.; Bahnemann, D. W. tration of nonionic surfactants on TiO2 particles16-18 and
Chem. Rev. 1995, 95, 69. may form charged aggregates, such as those from nonionic
(4) Oregan, B.; Gratzel, M. Nature 1991, 353, 737. phospholipid liposomes.19 To date, there have been two
(5) Wang, R.; Hashimoto, K.; Fujishima, A.; Chikuni, M.; Kojima, E.;
Kitamura, A.; Shimohigoshi, M.; Watanabe, T. Nature 1997, 388, 431. main reasons proposed for the observation of these charged
(6) Alivisatos, A. P.; Johnsson, K. P.; Peng, X. G.; Wilson, T. E.; Loweth, aggregates of nonionic surfactant. If potential-determining
C. J.; Bruchez, M. P.; Schultz, P. G. Nature 1996, 382, 609. ions, e.g., K+ and Ca2+, are present in the system due to
(7) Aizenberg, J.; Black, A. J.; Whitesides, G. M. Nature 1999, 398,
495.
impurities or for other reasons, then a charged surface
(8) Donners, J. J. J. M.; Nolte, R. J. M.,; Sommerdijk, N. A. J. M. J. may result.20 Alternatively, nonionic surfactants may form
Am. Chem. Soc. 2002, 124, 9700. charge-possessing materials, due to the aggregation of
(9) Spatz, J. P. Angew. Chem., Int. Ed. 2002, 41, 3359. the negative electric dipoles of the ethylene oxides in the
(10) Király, Z.; Börner, R. H. K.; Findenegg, G. H. Langmuir 1997,
13, 3308.
(11) Tiberg, F.; Jönsson, B.; Tang, J.; Lindman, B. Langmuir 1994, (16) Smith, G. A.; Zulli, A. L.; Grieser, M. D.; Counts, M. C. Colloids
10, 2294. Surf., A 1994, 88, 67.
(12) Kjellin, U. R. M.; Claesson, P. M.; Linse, P. M. Langmuir 2002, (17) Schepers, F. J.; Toet, W. K.; van de Pas, J. C. Langmuir 1993,
18, 6745. 9, 956.
(13) Fragneto, G.; Lu, J. R.; McDermott, D. C.; Thomas, R. K.; Rennie, (18) Chen, X.; Cheng, H.; Ma, J. Powder Technol. 1998, 99, 171.
A. R.; Gallagher, P. D.; Satija, S. K. Langmuir 1996, 12, 477. (19) Király, Z.; Börner, R. H. K.; Findenegg, G. H. Langmuir 1997,
(14) Grant, L. M.; Ederth, T.; Tiberg, F. Langmuir 2000, 16, 2285. 13, 3308.
(15) Drach, M.; Narkiewicz-Micha_ek, J.; Rudziñski, W.; Findenegg, (20) Nelson, B. P.; Candal, R.; Corn, R. M.; Anderson, M. A. Langmuir
G. H.; Király, Z. Phys. Chem. Chem. Phys. 2002, 4, 2307. 2000, 16, 6094.

10.1021/la0517890 CCC: $30.25 © 2005 American Chemical Society

Published on Web 10/14/2005
11284 Langmuir, Vol. 21, No. 24, 2005 McNamee et al.

nonionic surfactant molecule. The question of whether a

nonionic surfactant gives a charged aggregate can be
answered if we measure the forces between two initially
uncharged hydrophilic surfaces, i.e., surfaces that are at
their isoelectric point (iep), in the presence of nonionic
surfactants and without any other possible potential-
determining ions.
Generally, silica has been used as a model surface of a
metal oxide. However, its iep occurs at low pH. Thus, the
forces between a nonionic surfactant and a noncharged
surface cannot be measured in the absence of strong
electrostatic forces without the addition of an acid to
appropriately adjust the solution pH to give the iep. Such
added ions may, however, specifically adsorb at the
substrate surface, thus hiding the adsorption mechanism
of the nonionic surfactants. The rutile form of TiO2 has
been reported to have an iep when the solution pH is 5.6.21
Thus, using TiO2 allows us to measure the effect of the
adsorption of a nonionic surfactant on forces between two,
noncharged hydrophilic surfaces at a nonadjusted pH of Figure 1. Tapping mode AFM image of the TiO2 substrate in
the aqueous surfactant solution. water.
A flat, smooth surface is required in these measure-
ments, as roughness may cause artifacts in the force data. to give a conductance of 18.2 MΩ cm and a total organic carbon
A silica plate that has been sputtered with an anatase content of 5 ppm was used to prepare the solutions.
TiO2 film gives the desired ideal surfaces. The iep of the The TiO2 (anatase) substrates for the AFM force measurements
anatase form of titanium dioxide in an aqueous solution were obtained from Prof. Toshinobu Yoko of the Institute for
has still not been well characterized. However, the iep of Chemical Research, Kyoto University, Japan. The TiO2 plates
the rutile form of titanium dioxide has been studied and were prepared by a helical plasma vapor deposition method.27
is reported to occur when the solution pH is 5.6.21 Above The TiO2 film was of a single color, indicating a uniform thickness
and therefore a flat surface. The thickness of the sputtered TiO2
this pH, the surface is negatively charged, due to the film was approximately 100 nm, and the film had a negligible
surface OH groups undergoing basic reactions. Below the defect. AFM images showed an RMS roughness of 2.1 nm over
iep pH, the surface is positively charged due to acidic 2 µm; see Figure 1 for an example of one such image.
reactions of the surface OH groups. Although both rutile The TiO2 particles for the probe in the AFM force measure-
and anatase are crystalline and have an [TiO6] octahedron ments were acquired from Ube Nittoh Kasei, Japan. They were
structure, they possess different modes of chemical made by using a sol-gel method to coat silica particles with
arrangement and linkage.22,23 For example, the distance TiO2. The final TiO2 particles had a radius of 3.25 µm, as
between the O-O atoms in anatase is up to 0.1 Å smaller determined by scanning electron microscopy.
than in rutile,24 and anatase is more photoactive than 2.2. Surface Force Measurement Technique. 2.2.1. Prepa-
rutile, meaning its hydroxyl groups are oxidized by UV ration. The TiO2 particles for the probe were dispersed in ethanol
and then ultrafiltered with pure ethanol several times. They
light to form OH- ions and OH radicals easier due to its were then stored in ethanol in a glass vessel. The preparation
slightly higher Fermi level.25 Thus, there may be differ- of the cantilever required the particles to be first spread on a
ences in the extent and nature of the surface hydroxyl cleaned glass slide plate. The tip end of the micropyramid of a
groups of the two TiO2 types and therefore their iep’s. V-shaped cantilever (Olympus Optical Co., Ltd.) was then glued
In this study, atomic force microscopy (AFM) was used using 5 min curing epoxy resin (Araldite, Ciba-Geigy Japan, Ltd.),
to characterize a sputtered anatase TiO2 surface, prepared and a single TiO2 particle from the glass plate was transferred
by the helical vapor preparation method. The adsorption onto the tip end.
mechanism of a nonionic surfactant molecule, C12E4, to The Cleveland method28 was used to determine the spring
constant (k) of the cantilever (k ) 0.16 N/m). We determined the
an anatase TiO2 surface was then investigated at the iep spring constant to be within (10% of the value supplied by the
of the TiO2. The question as to whether a nonionic manufacturer.
surfactant aggregate is charged was then discussed by The TiO2 plates were cleaned prior to a force experiment by
looking at the forces measured between two, initially rinsing them with ethanol and then with copious amounts of
uncharged TiO2 surfaces that have been covered by C12E4 distilled water.
aggregates. The TiO2 probe, which was attached to the cantilever, was
cleaned before the setup and commencement of an AFM force
experiment in 1 part 31% H2O2, 1 part 28% NH3, and 2 parts
Experimental Section
distilled water solution for 5 min. It was then rinsed with ethanol
2.1. Materials. Sodium nitrate (high purity, Nacalai Tesque, and then with distilled water.
Japan) was used as the supporting electrolyte and linear 2.2.2. Measurements. The surface forces between a TiO2
dodecanol tetraethoxylate (>99% purity, Nikko Chemicals Co. substrate and a TiO2 particle in an aqueous solution were
Ltd., Japan; C12H25(OC2H4)4OH, C12E4; cmc ) 6.4 × 10-5 M26) measured at room temperature (25 ( 1 °C) as a function of their
as the nonionic surfactant. Water that was distilled and deionized distance using an atomic force microscope (Seiko Instruments,
Inc., SPA400). The piezo actuator (Seiko Instruments, Inc., PZT
FS20A, z scan size 20 µm) was calibrated and its nonlinearity
(21) Larson, I.; Drummond, C. J.; Chan, D. Y. C.; Grieser, F. J. Am.
Chem. Soc. 1993, 115, 11885.
compensated as described elsewhere.29 The forces between the
(22) Cheng, H.; Ma, J.; Zhao, Z.; Qi, L. Chem. Mater. 1995, 7, 663.
(23) Cromer, D. T.; Herrington, K. J. Am. Chem. Soc. 1955, 77, 4708. (27) Takahashi, M.; Tsukigi, K.; Dorjpalam, E.; Tokuda, Y.; T., Yoko,
(24) Burdett, J. K.; Hughbanks, T.; Miller, G. J.; Richardson, J. W.; T. J. Phys. Chem. B 2003, 107, 13455.
Smith, J. V. J. Am. Chem. Soc. 1987, 109, 3639. (28) Cleaveland, J. P.; Manne, S.; Bocek, D.; Hansma, P. K. Rev. Sci.
(25) Wold, A. Chem. Mater. 1993, 5, 280. Instrum. 1993, 64, 403.
(26) Rosen, M. J.; Cohen, A. W.; Dahanayake, M.; Hua, X. J. Phys. (29) Yamamoto, S.; Ejaz, M.; Tsujii, Y.; Matsumoto, M.; Fukuda, T.
Chem. 1982, 86, 541. Macromolecules 2000, 33, 5602.
Physicochemical Characterization of a TiO2 Surface Langmuir, Vol. 21, No. 24, 2005 11285

substrate and probe were measured using a circular Teflon liquid

cell with a radius of 1 cm, in which the substrate was placed.
Approximately 2 mL of the solution was added to the liquid cell,
which was then placed on the piezo actuator. The liquid cell was
in contact with the cantilever holder, leaving almost no space for
the liquid to evaporate.
The method of Ducker and others30 was used for the force
measurements. Briefly, the cantilever with the TiO2 probe fixed
on the cantilever holder was positioned in the solution to face the
substrate. The change in the deflection of the cantilever (∆x) was
measured as a function of the piezo displacement, using the
differential intensity of the reflection of the laser beam off the
cantilever onto a split photodiode. The force (F) was calculated
from Hooke’s law, F ) k∆x, where k is the spring constant of the
cantilever. The force was normalized to give the interaction free Figure 2. Compression force curves between two TiO2 surfaces
energy (Gf) by the Derjaguin approximation31 (eq 1). This was in 1 mM NaNO3 at pH 7.5, 5.8, and 4.5: O, pH 5.8; 4, pH 4.5;
possible since the radius of the TiO2 probe (R) was much greater 0, pH 7.5; s, 20 mV, 1 mM; ---, 40 mV, 0.9 mM; ‚‚‚, VDW only
than the separation distance used. with A ) 2.6 × 10-20 J.

F/R ) 2πGf (1) Figure 2 shows the compression force curves between
two TiO2 surfaces in 1 mM NaNO3 at pH 7.5, 5.8, and 4.5.
Zero separation was characterized from the position of the
The measurements were performed in the presence of the
onset of the linear compliance region in the force profile, which supporting electrolyte NaNO3, to ensure a constant ionic
resulted from the further compression of both surfaces after concentration. A repulsive force was detected for pH 7.5
contact. The surface separation, D, was determined from the and 4.5, whereas only an attractive force, commencing at
displacement of the substrate relative to this constant compliance approximately 15 nm, was observed for pH 5.8. These
region. The zero force position between the surfaces was defined force profiles were independent of the speed of the
from the baseline of no deflection of the TiO2 probe cantilever at compression force mode, which was varied from 3618 to
large probe-substrate separations. 6 nm s-1 (speeds corresponding to compression times of
2.3. AFM Tapping Mode Imaging. The images of the 0.5-30 s). The decay lengths, 1/κ, of the forces in Figure
adsorbed surfactants at the TiO2 surfaces in water were obtained
using a Digital Instruments NanoScope III multimode microscope
1 gave the actual concentration of ions in solution. These
in tapping mode and a tapping mode fluid cell. The Si3N4 were determined by plotting log(F/R) against D. Decay
cantilevers (OMCL-AC240TS-C2, Olympus) used were coated lengths of 10.1 nm were obtained, which were within 5%
with aluminum reflex and had a spring constant of 2 N m-1, a of the theoretical Debye length34 of 9.6 nm at 298 K for
resonance frequency of 70 kHz, and a tetrahedrally shaped tip a 1 mM 1:1 electrolyte aqueous solution. This good
with a height of 14 µm and a tip radius of <10 nm. The TiO2 correspondence showed the reliability of the measure-
substrates were attached to the piezo actuator, and the cantilever ments. The surface potentials (Ψ0) of the force curves were
was held by the fluid cell, so that the laser light could be reflected determined by fitting the experimental force curves with
onto the back surface of the cantilever to the photodetector. The the summation of an electrostatic force, calculated using
surfactant aqueous solution was injected into the fluid cell using
the nonlinear solution of Ohshima and others35 with the
a standard syringe and sealed using a silicone O-ring.
The images were determined using a tapping mode method constant surface charge boundary condition, and a van
with an imaging scan rate of 1.39 Hz, a resolution of 512 data der Waals force, using the Hamaker constant of 2.6 ×
points per linen and 512 lines. All the images are shown here as 10-20 J for a TiO2-water-TiO2 case.36 The pH variance
unmodified (i.e., no filtering), except for flattening along the scan of 7.5, 5.8, and 4.5 gave Ψ0 values of 40, 0, and 20 mV,
lines. respectively. The charge on an anatase TiO2 surface
changes with pH, due to acid/base reactions of the surface
3. Results and Discussion OH group. The nonexisting charge at pH 5.8 shows that
3.1. Physicochemical Characterization of an Ana- this pH defines the iep of this TiO2. The iep of anatase is
tase TiO2 Surface. Figure 1 shows the force profile of similar to that for a single rutile TiO2 crystal, which occurs
the surface of the TiO2 sputtered film on silica, as imaged at pH 5.6 ( 0.2.37 The fact that the iep’s of rutile and this
by tapping mode AFM. Here, the globular structure anatase were similar shows that their structural differ-
inherent of an anatase surface32,33 could be observed, ences are too small to measurably affect the chemical
verifying the correct TiO2 type. The iep of the anatase reactivity of the surface OH groups by acid/base reactions.
TiO2 surfaces was subsequently determined by measuring 3.2. Physicochemical Effect of the Adsorption of
the force curves in the presence of 1 mM NaNO3 at several C12E4 to the TiO2 Surface. In this part of the experiment
pH values. The NaNO3 acted as a supporting electrolyte, we investigated the forces that evolve when a nonionic
ensuring a constant ionic concentration. As the rutile form surfactant is adsorbed onto an initially uncharged TiO2
of TiO2 has its iep at pH 5.6, it was imagined that the surface. In this way, we could gain information about the
surface oxide groups on the anatase may react to pH interaction mechanism of nonionic surfactants and metal
similarly to those on a rutile surface. Thus, the pH values oxide surfaces. We also addressed the question of whether
above, near, and below this value were chosen; a value of adsorbed aggregates of nonionic surfactants possess a
5.8 was used instead of 5.6, as this was the pH of the charge.
water we used. To address these points, we minimized/eliminated all
the charging sources in our system other than our
(30) Ducker, W. A.; Senden, T. J.; Pashley, R. M. Nature 1991, 353,
239. (34) Ohshima, H. In Electrical Phenomena at Interfaces; Ohshima,
(31) Israelachvili, J. N. Intermolecular and Surface Forces; Academic H., Furusawa, K., Eds.; Marcel Dekker: New York, 1988; p 3.
Press: London, 1994; p 161. (35) Ohshima, H.; Healy, T. W.; White, L. R. J. Colloid Interface Sci.
(32) Jellison, G. E.; Boatner, L. A.; Budai, J. D.; Jeong, B.-S.; Norton, 1982, 89, 484.
D. P. J. Appl. Phys. 2003, 93, 9537. (36) Israelachvili, J. N. Intermolecular and Surface Forces; Academic
(33) Karunagaran, B.; Rajendra Kumar, R. T.; Viswanathan, C.; Press: London, 1994; p 190.
Mangalaraj, D.; Narayandass, S. K.; Mohan Rao, G. Cryst. Res. Technol. (37) Larson, I.; Drummond, C. J.; Chan, D. Y. C.; Grieser, F. J. Am.
2003, 38, 773. Chem. Soc. 1993, 115, 11885.
11286 Langmuir, Vol. 21, No. 24, 2005 McNamee et al.

be exponential, short-ranged, and fast decaying in nature.

Additionally, their extension increases with an increased
surfactant concentration. To test the possibility of these
forces resulting from an electrostatic double layer of
adsorbed, charged surfactant ions, we calculated the Debye
length of the hypothetical electrostatic force curves when
we used ionic concentrations between 1 × 10-8 and 1 ×
10-5 M ionssvalues corresponding to the range of sur-
factant concentrations we used. We found that the Debye
length decreases from 251 to 91 nm when we increase the
concentration from 1 × 10-8 to 1 × 10-5 M and take into
account the inherent ionic concentration by virtue of this
system pH. The facts that the ranges of our measured
Figure 3. Compression forces between two TiO2 surfaces in repulsive forces were much smaller than these values and
the presence of C12E4 at a solution pH of 5.8 when the that their extension increased with an increased surfactant
concentration of C12E4 was varied between 1 × 10-8 and 1 × concentration indicated that these are steric forces39 and
10-5 M: 0, 1 × 10-8 mM C12E4; O, 1 × 10-7 mM C12E4; 3, 5 not electrostatic forces.
× 10-7 mM C12E4; pentagon, 7 × 10-7 mM C12E4; 4, 1 × 10-6 We can obtain information about the structure of the
mM C12E4; ], 1 × 10-5 mM C12E4.
adsorbed surfactant molecules and their adsorption
mechanism by looking at the change in the steric force as
surfactants. We therefore chose to measure the forces at the surfactant concentration is increased. For the low
the iep of TiO2, i.e., pH 5.8. This removed the electrostatic surfactant concentrations of 1 × 10-8 and 1 × 10-7 M, the
component attributable to the electrostatic charging of range of the steric, repulsive force was only 1 nm,
the TiO2 surfaces in water. This pH also had the advantage indicating that the thickness of a possible adsorbed layer
of being the pH of the water we used. This removed the at a TiO2 surface would only be 0.5 nm. As the error in
necessity of using salts to modify the solution pH, which the AFM force measurements is approximately 1 nm, the
may have otherwise specifically adsorbed onto the TiO2 origin of this steric, short-ranged repulsion may be either
surface. Finally, we performed this part of the experiment a water layer that can hydrogen-bond to the OH groups
without a supporting electrolyte (ionic salt), to avoid on the TiO2 surface40 (the thickness of a water molecule
possible specific adsorption of an ion to the TiO2 surface,20,38 is approximately 0.31 nm12) or a small number of adsorbed
which could also have caused charging of the surface. surfactant molecules. Considering the latter adsorption
Figure 3 shows the compression forces between two TiO2 mechanism, the C12E4 molecules may adsorb at the
surfaces in the presence of C12E4 at a solution pH of 5.8 TiO2-water interface with their hydrophobic tail facing
when the concentration of C12E4 was varied from 1 × the TiO2 surface. In this way, the number of unfavorable
10-8 to 1 × 10-5 M. A higher surfactant concentration was contacts of the hydrophobic groups with water would be
not used, to avoid the influence of micelles and oscillatory reduced. The degree of tilt of the C12E4 molecules with
forces. These observed force profiles were independent of respect to the TiO2 surface (R) can be inferred from
the speed of the compression force mode, which was varied comparing the thickness of the adsorbed C12E4 layer at
from 3600 to 6 nm s-1 (speeds corresponding to compres- the TiO2 surface (Y, calculated from the steric force portion
sion times of 0.5-30 s). In the case of 1 × 10-8 and 1 × of the force curve) with the length of a C12E4 molecule
10-7 M C12E4, two force types were observed. An attractive (L), which is 3.11 nm:13
force began when the surface separation was between 10
and 15 nm, and a repulsive force was observed at the
small separations of 1 nm. Similarly, for 5 × 10-7 M C12E4, R ) tan-1(Y/L) (2)
an attractive force was seen to commence when the surface
separation was approximately 11 nm and a repulsive force The C12E4 molecules were calculated to adsorb tilted
was observed at the small separations of 2.4 nm. The effect with an angle of 12° with respect to the horizontal. At
of using the higher C12E4 concentrations of 7 × 10-7, 1 such low tilt angles, only a submonolayer would be formed.
× 10-6, and 1 × 10-5 M was to increase the range of the As the concentration of C12E4 was increased to 5 × 10-6
repulsive force; the commencing separation distances were and 7 × 10-5 M, the steric forces shifted to larger
3.4, 7, and 9 nm, respectively. No long-range attractive separations, indicating adsorption layers with thicknesses
force could be seen for these higher surfactant concentra- of 1.2 and 1.7 nm, respectively. These values were also
tions. lower than the length of a C12E4 molecule, indicating
The attractive forces observed at the low surfactant that a submonolayer must be forming and that the
concentrations of 1 × 10-8, 1 × 10-7, and 5 × 10-7 M are molecules are tilted with angles of 21° and 28° with respect
presumably caused by the van der Waals forces that result to the horizontal.
between two TiO2 surfaces in water. This is because the At the higher C12E4 concentrations of 1 × 10-6 and 1
attraction commences at a value similar to that observed × 10-5 M, the steric forces shifted to larger separations,
in the absence of C12E4 at the iep of TiO2 (see the previous indicating adsorption layers with thicknesses of 3.5 and
section). A reason no attraction was seen at higher 4.5 nm, respectively. These values are significantly larger
concentrations is because the range and strength of that than the size of a water molecule, meaning the steric force
repulsive force are large compared to those of the van der must therefore be the result of adsorbed C12E4 molecules.
Waals force of this system. If we consider the size of a C12E4 molecule, the estimated
The repulsive forces may be steric or electrostatic double- approximate thickness of a bilayer (B) would be between
layer forces. Their origin may be better understood by the thickness of a C12E4 micelle in water and 2 times the
plotting the forces on a log(F/R) versus separation distance length of a C12E4 molecule (2L), i.e., between 4.5 and 6.2
scale; see the inset of Figure 2. These forces are seen to
(39) Israelachvili, J. N. Intermolecular and Surface Forces; Academic
Press: London, 1994; p 294.
(38) Ronson, T. K.; McQuillan, A. J. Langmuir 2002, 18, 5019. (40) Suda, Y. Langmuir 1988, 4, 147.
Physicochemical Characterization of a TiO2 Surface Langmuir, Vol. 21, No. 24, 2005 11287

nm, respectively. Our measured adsorbed thicknesses here

are between the value of a monolayer with the surfactant
in an extended conformation and that of a bilayer. This
suggests that this is a sub-bilayer region, where the TiO2
surface is covered with a C12E4 monolayer and patches
of a second layer of adsorbed C12E4 molecules. We were
not able to directly obtain images of the adsorbed C12E4
molecules on the TiO2 at the different concentrations by
tapping or contact mode AFM to verify this hypothesis,
due to the low resolution in such a system and the small,
but nonzero roughness of the TiO2 surface (RMS ) 2.1 nm
over 2 µm). (It should be noted that, although this
roughness of 2.1 nm is too large to image the small
Figure 4. Decompression forces between two TiO2 surfaces in
adsorption layers of the surfactants (thicknesses between the presence of C12E4 at a solution pH of 5.8 when the
1 and 4 nm) on TiO2, the TiO2 surface was flat enough concentration of C12E4 was varied between 1 × 10-8 and 1 ×
over the small areas required for the force curve mea- 10-5 M: 0, 1 × 10-8 mM C12E4; O, 1 × 10-7 mM C12E4; 3, 5
surements to obtain the force curves.) If we look at the × 10-7 mM C12E4; pentagon, 7 × 10-7 mM C12E4; 4, 1 × 10-6
AFM studies that investigated the adsorption of ionic mM C12E4; ], 1 × 10-5 mM C12E4.
surfactants to an atomically smooth mica substrate,
however, we see images that show a change in the
adsorption conformations of the molecules from a sub-
monolayer to a perforated bilayer (sub-bilayer) and then
to adsorbed cylinders as the surfactant concentration was
increased from values much below the cmc to values near
the cmc.41 In the case of our nonionic surfactant, C12E4,
bilayers on silica were observed by ellipsometry when the
surfactant concentration was (0.6-0.9)cmc.42 Additionally,
studies of similar poly(ethoxylene oxide) n-alkyl ethers
adsorbing at hydrophilic silica surfaces have reported that
islands of bilayers could be detected by light scattering at
volume fractions much lower than those required for full
bilayer formation.43 Thus, we can presume that the
surfactant conformation for 1 × 10-6 and 1 × 10-5 M is Figure 5. Schematic diagram depicting the adsorption of
indeed a sub-bilayer. Interestingly, the C12E4 concentra- C12E4 to the TiO2 surface.
tion required to form these islands of bilayers (sub-
bilayers) is only 0.02(cmc) and 0.16(cmc). As the measured substrate than be surrounded by water molecules. The
thicknesses of the adsorbed surfactant layers are an presence of the attractions in the decompression force
average of all the thicknesses over the interaction areas, curves for concentrations between 1 × 10-8 and 5 × 10-8
the percentage of C12E4 bilayers (P) existing at the TiO2 M C12E4 can be attributed to the adsorption conformation
surface for each concentration can be approximated via of the surfactants. At low concentrations, the surfactant
adsorbs nearly horizontal to the substrate, leaving many
(Y - L) points of contact with a water molecule, thereby giving a
P) × 100 (3) hydrophobic decompression force. As the surfactant
B-L
concentration increases, the molecule would adsorb in an
In this way, the percentage of bilayers for 1 × 10-6 and increasingly more vertical manner, leaving less or almost
1 × 10-5 M C12E4 is calculated to be between 13% and no contact points possible with a water molecule. It is for
29% and between 45% and 100% when B thicknesses of this reason that we see almost no decompression attractive
6.2 and 4.5 nm are used. force for 7 × 10-8 M C12E4. As the concentration of C12E4
More information as to the mechanism and conformation was increased to form bilayers, we also did not see
of the adsorbed surfactant molecules can be obtained from attractive forces in the decompression force curves. This
the decompression force curves measured between the is attributable to the fact that the surface is now
two TiO2 surfaces in the presence of the surfactants; see hydrophilic, as a consequence of the ethoxylene molecules
Figure 4. In the case of the hydrophobic chains facing the facing the water phase. The second layer of C12E4
water edge, we would see an increase in the size of the molecules must therefore be adsorbing as a consequence
jumps in the decompression force curves as the adsorbed of hydrogen bonding of the ethoxylene groups of the already
C12E4 concentration increased. However, the fact that adsorbed C12E4 molecules and the hydrophobic attraction
we see a decrease in the decompression attractions as the between their C12E4 chains. As the thickness of the
concentration is increased from 1 × 10-8 to 7 × 10-8 M adsorbed layers for concentrations g1 × 10-6 M C12E4
C12E4 (concentrations giving submonoayer formations) indicated a sub-bilayer formation, the second layer of
indicates that the hydrophilic portion of the surfactant adsorbing C12E4 molecules must adsorb such that their
(ethoxylate chain) must be facing the water and the C12 groups do not contact the TiO2 substrate but contact
hydrophobic tails facing the TiO2 surface. This adsorption part of the already adsorbed C12E4 monolayer molecules.
would therefore be entropically driven, as it would be more A sketch of this proposed adsorption is depicted in Figure
favorable for the hydrophobic tails to face the TiO2 5.
Using the information above, an isotherm depicting the
(41) Blom, A. B.; Duval, F. P.; Kovács, L.; Warr, G. G. Langmuir adsorption of the nonionic surfactant C12E4 onto TiO2
2004, 20, 1291. can be proposed to follow a two-step adsorption isotherm;
(42) Triberg, F.; Landgren, M. Langmuir 1993, 9, 927.
(43) Cummins, P. G.; Staples, E.; Penfold, J. J. Phys. Chem. 1990, see Figure 6. The adsorption regions would be submono-
94, 3740. layer formation (region 1) f complete monolayer coverage
11288 Langmuir, Vol. 21, No. 24, 2005 McNamee et al.

surfaces in the presence of C12E4 when the C12E4

concentrations are high enough to cause surface ag-
gregates. If we look at Figure 3, we see that no long-range
repulsive forces were observed in our systems when they
were measured in the presence of C12E4 and at the pH
that caused the iep of TiO2. This was the case, even when
high enough C12E4 concentrations were used that ensured
the presence of surface surfactant aggregates, i.e., a
surfactant concentration of 1 × 10-5 M C12E4. This
absence indicates that the TiO2 surfaces remain un-
charged.12 Thus, we can conclude that an aggregation of
these nonionic surfactants does not comprise a charge
and the adsorption of such nonionic surfactants to a surface
Figure 6. Schematic diagram depicting the adsorption of does not appear to affect the surface charge of a substrate.
C12E4 to a TiO2 surface. Regions 1, 2, 3, and 4 represent a 4. Conclusions
submonolayer, a complete monolayer, a sub-bilayer, and a
bilayer. In region 1, the C12E4 molecules adsorb in horizontal, The charge on an anatase TiO2 surface can be controlled
tilted, and vertical orientations to the TiO2 substrate at low, by pH, due to acid/base reactions of the surface OH group.
middle, and high concentrations. The iep occurs at pH 5.8, a value similar to that for a
single rutile TiO2 crystal. The difference in the TiO2
(region 2) f sub-bilayer formation (region 3) f bilayer crystallinities of anatase and rutile therefore does not
formation (region 4). Region 1 would occur for 0 M e C12E4 appear to affect the surface properties of TiO2.
concentration < 1 × 10-6 M, with the orientation of the A nonionic surfactant molecule was seen to adsorb to
molecule changing from horizontal to near-vertical to the an uncharged TiO2 surface, driven by hydrophobic inter-
substrate as the concentration increases. Region 3 would actions. Additional adsorption occurs at higher concen-
take place when bilayer patches occur and form when the trations, helped by hydrophobic interactions between the
surfactant concentrations are much less than the cmc, surfactant tails and the hydrogen bonding between the
i.e., 1 × 10-6 M e C12E4 concentration e 1 × 10-5 M. A ethoxylene headgroups, to give sub-bilayers. A two-step
full bilayer would result in region 5 for concentrations adsorption isotherm was subsequently proposed, with
just below, at, or above the cmc. This adsorption mech- submonolayer formation (region 1) f complete monolayer
anism is similar to the one reported for the adsorption of coverage (region 2) f sub-bilayer formation (region 3) f
ionic surfactants to a charged substrate. There, the two- bilayer formation (region 4).
step adsorption isotherm showed a variation in adsorption An aggregate of the C12E4 nonionic surfactant was also
regions with surfactant concentration of monolayer f shown not to possess charge.
complete monolayer coverage f bilayer formation f
Acknowledgment. This study has been financially
“micelle-like formation” (surfactant coverage near the
supported by Science Promotion Grant No. 10640558 of
cmc).44
the Ministry of Education, Japan, and the Humboldt
The question as to whether a nonionic surfactant Foundation, Germany. We also thank Prof. Toshinobu
aggregate is charged can be answered by looking at the Yoko of Kyoto University for supplying the TiO2 plates
forces measured between two, initially uncharged TiO2 and Donose Constantin of Kyoto University for help in
measuring the AFM images.
(44) Sharma, B. G., Basu, S., Sharma, M. M. Langmuir 1996, 12,
6506. LA0517890