Dynamic contact angle studies of self-assembled thin films from fluorinated alkyltrichlorosilanes

Article abstract of DOI:10.1021/jp013820m

Dynamic contact angle studies on self-assembled thin films from C₇F₁₅CH₂OCH₂CH₂CH₂ SiCl₃ reveal a range of interesting behavior. Solution-based processing conditions have been identified that allow preparation of essentially monolayer films on quartz exhibiting water adv/rec contact angles of 119/104 ?± 2?° and extremely low contact angle hysteresis (hexadecane adv/rec = 74/70?°, heptane adv/rec = 56/55?°) with hydrocarbon liquids. This compound provides an example of a fluorinated trichlorosilane that is able to deliver low-hysteresis films by deposition at room temperature. Adsorption of silane oligomers, formed by hydrolysis and condensation reactions in solution, was also found to occur, slower than but competitive with monolayer formation. This process became more significant as dip coating times increased. Ellipsometric data on silicon wafers confirmed that film thicknesses increased with dip time, while AFM imaging showed that the oligomeric material was deposited in the form of particulates. The effects of this process on water dynamic contact angles are discussed. We also compare contact angles on these films with those on films prepared from C_nF_2n+1CH₂CH₂SiCl₃ (n = 6, 8, 10) and draw some conclusions regarding structure-property effects in these systems. Finally, we propose a mechanism that can account qualitatively for the bulk of the results observed here. Its central feature is reaction of the fluorinated alkyltrichlorosilane with surface-adsorbed water to yield a self-assembled monolayer consisting of silanetriol molecules hydrogen-bonded to the substrate. Effects of high humidity aging on dynamic contact angles of these films suggest that they are at most only lightly cross-linked when prepared under conditions utilized here.

Full text of DOI:10.1021/jp013820m

4746
J. Phys. Chem. B 2002, 106, 4746-4754
Dynamic Contact Angle Studies of Self-Assembled Thin Films from Fluorinated
Alkyltrichlorosilanes1
,†
†
‡
Mark J. Pellerite,* Erika J. Wood, and Vivian W. Jones
M Company, 3M Center 201-1W-28, St. Paul, Minnesota 55144
ReceiVed: October 16, 2001; In Final Form: February 10, 2002
3
7 2 2 2 2 3
Dynamic contact angle studies on self-assembled thin films from C F15CH OCH CH CH SiCl reveal a range
of interesting behavior. Solution-based processing conditions have been identified that allow preparation of
essentially monolayer films on quartz exhibiting water adv/rec contact angles of 119/104 ( 2° and extremely
low contact angle hysteresis (hexadecane adv/rec ) 74/70°, heptane adv/rec ) 56/55°) with hydrocarbon
liquids. This compound provides an example of a fluorinated trichlorosilane that is able to deliver low-
hysteresis films by deposition at room temperature. Adsorption of silane oligomers, formed by hydrolysis
and condensation reactions in solution, was also found to occur, slower than but competitive with monolayer
formation. This process became more significant as dip coating times increased. Ellipsometric data on silicon
wafers confirmed that film thicknesses increased with dip time, while AFM imaging showed that the oligomeric
material was deposited in the form of particulates. The effects of this process on water dynamic contact
angles are discussed. We also compare contact angles on these films with those on films prepared from
n 2 2 3
C F2n+1CH CH SiCl (n ) 6, 8, 10) and draw some conclusions regarding structure-property effects in these
systems. Finally, we propose a mechanism that can account qualitatively for the bulk of the results observed
here. Its central feature is reaction of the fluorinated alkyltrichlorosilane with surface-adsorbed water to yield
a self-assembled monolayer consisting of silanetriol molecules hydrogen-bonded to the substrate. Effects of
high humidity aging on dynamic contact angles of these films suggest that they are at most only lightly
cross-linked when prepared under conditions utilized here.
Introduction
CnF2n+1CH CH SiCl
3
2
2
1
; n ) 6, 8, 10
Alkylsilanes RSiX3 (R ) alkyl, X ) Cl, alkoxy) on siliceous
surfaces and other substrates represent one of the most
extensively investigated classes of self-assembling monolayer
property relationships and the effect on film properties of the
linking group structure between the perfluoroalkyl segment and
2
22,34,38,45-47
systems. Since the pioneering work of Sagiv and co-workers,
the silyl headgroup.
virtually every tool available to the surface chemist has been
This study presents results from dynamic contact angle
measurements on self-assembled films derived from fluorinated
trichlorosilane 2 on quartz. Contact angle data before and after
3-35
brought to bear in the study of these systems.
The picture
2
,10,25
that has emerged
regarding the structure of these mono-
layers revolves around a two-dimensional cross-linked alkyl-
siloxane network, formed by hydrolysis of the trichlorosilane
and condensation of the resulting silanol groups, partially bound
to the substrate through additional condensation with surface
silanols (Figure 1), although the extent and distribution of
siloxane linkages is currently subject to debate. This view has
recently been challenged, however,36 so despite the voluminous
literature in this area, research is extremely active and basic
mechanistic questions remain to be answered.
C F CH OCH CH CH SiCl
3
7
15
2
2
2
2
2
aging in dry and humid environments, along with ellipsometry
and AFM data on coated silicon wafers, are discussed within
the framework of a mechanistic hypothesis for the formation
and structure of these self-assembled thin films. Also presented
are some limited dynamic contact angle data on films from 1
obtained on quartz under the same conditions, to probe the
significance of structure-property interactions in these films.
Where possible, we attempt to place the data from this study,
Fluorinated alkylsilanes have been somewhat less well-
48
studied than their hydrocarbon counterparts, even though such
materials are finding application as low-surface-energy films
and the picture that emerges therefrom, in the context of results
49
22,37
for optical substrates and lubricating treatments for micro-
from previous work
on mechanism and structure in fluori-
37
electromechanical systems (MEMS) devices. For instance,
dynamic contact angles in these films are not well characterized,
and most of the work reported to date has involved silanes of
the general structure 1; little systematic data exist on structure-
nated alkyltrichlorosilane self-assembly. We prefer the general
term “thin films” because, as we show, any description of self-
assembled films from 2 strictly in terms of monolayers is not
only overly confining, but at times incorrect.
Experimental Section
*
To whom correspondence should be directed. E-mail: mjpellerite1@
mmm.com. Fax: 651-737-5335.
†
‡
Materials. All solvents and reagents were standard com-
mercial grade and used as received. Methyl perfluorobutyl ether
Advanced Materials Technology Center.
Corporate Analytical Technology Center.
1
0.1021/jp013820m CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/09/2002

Contact Angle Studies of Fluorinated Alkyltrichlorosilanes
J. Phys. Chem. B, Vol. 106, No. 18, 2002 4747
(0.43 mol) potassium hydroxide in 35 g of water. The mixture
was allowed to stir for 4 h, during which time it split into two
immiscible liquid phases. The lower layer was isolated in a
separatory funnel and washed three times with water to remove
any remaining solvent. The crude product was dried over
anhydrous magnesium sulfate, filtered, and vacuum distilled to
yield 103 g (76% yield) of 1H,1H-pentadecafluorooctyl allyl
ether, C7F15CH2OCH2CHdCH2, bp 88-89 °C/30 mmHg. The
1
19
product exhibited H and F NMR spectra consistent with this
structure. This procedure was found to afford higher, more
consistent yields than the method reported in the literature.51
The above allyl ether was converted to silane 2 by heating
in a screw-top vial at 60 °C overnight a mixture of allyl ether,
2
0% molar excess of trichlorosilane, and 100 ppm Pt catalyst
(added as a 1 wt % solution of bis(divinyltetramethyldisiloxane)-
Pt(0) in dichloromethane). The crude reaction product was
purified by vacuum distillation to afford trichlorosilane 2, bp
5
1
1
10-117 °C/2 mmHg (lit. 125 °C/7.5 mmHg), in yields of
1
typically 70-85%. NMR (in CDCl3): H, δ 1.5 (m, 2H); δ 1.9
13
(
m, 2H); δ 3.6 (t, J ) 6 Hz, 2H); δ 3.9 (t, J ) 14 Hz, 2H); C,
2
9
δ 20.8, δ 22.9, δ 68.1, δ 73.4; Si, δ 13.1. HRMS: obsd, m/z
+
5
38.9491 (M - Cl) ; calc for C11H8Cl2F15OSi, 538.9481. The
product was contaminated by <5% â-isomer, C7F15CH2OCH2-
CH(CH3)SiCl3. The major low-boiling byproduct was C7F15-
CH2OCHdCHCH3 (mixture of cis and trans isomers), formed
by rearrangement of the starting allyl ether. Yields of this
byproduct varied from run to run in the range of 15-30%.
Trichloro-1H,1H,2H,2H-perfluorododecylsilane, C10F21CH2-
CH2SiCl3. A screw-top vial equipped for magnetic stirring
was charged with 5.46 g of 1H,1H,2H-perfluoro-1-dodecene
(C10F21CHdCH2, 0.01 mol), 0.38 g of 1 wt % solution of bis-
(divinyltetramethyldisiloxane)Pt(0) in dichloromethane, and 1.59
g (0.012 mol) of trichlorosilane. The vial was capped and the
mixture stirred until the exotherm had subsided. The reaction
mixture was then heated in an oil bath at 60 °C for 24 h, and
the product, a solid at room temperature, was isolated by bulb-
to-bulb vacuum distillation. The yield of C10F21CH2CH2SiCl3
52
was 6.13 g (90%), bp 100-105 °C/0.05 mmHg (lit. 85-86
1
°
C/0.6 mmHg). Analysis of the product by H NMR showed
<
0.5% residual olefin and no detectable â-isomer.
Substrates. Quartz rods (2 mm diameter) were obtained from
Figure 1. Simplified schematic illustration of self-assembled monolayer
formation from alkyltrichlorosilane or fluorinated alkyltrichlorosilane
on siliceous surface.
the 3M Co. Central Research Glass Shop and cut to 55 mm
length. Silicon wafers were 〈100〉 orientation, P-type, boron-
doped, 75 mm diameter, obtained from Silicon Sense, Inc.
(HFE-7100), 1H,1H-pentadecafluorooctanol (C7F15CH2OH), and
(Nashua, NH.)
bis(divinyltetramethyldisiloxane)platinum(0) (23 wt % Pt in
siloxanes) were obtained from 3M Co. The fluorinated alcohol
used here was a ∼70:30 mixture of linear and branched C7F15
Substrate and Film Preparation. Substrate Cleaning. Quartz
rods were degreased by immersion for several minutes in an
ultrasound bath containing a 1:1 (v/v) mixture of ethanol and
chloroform, then allowed to dry in air. Final cleaning was
accomplished by either 10 min exposure in an air plasma
1
9
50
chains by F NMR analysis. Silanes 1 (n ) 6, 8) were
obtained from Gelest, Inc., Tullytown, PA, while 1H,1H,2H-
perfluoro-1-dodecene (C10F21CHdCH2) was obtained from
Lancaster Synthesis, Windham, NH. Water-saturated heptane
was prepared by adding 0.1 g of water to a 1 L bottle of
commercial reagent-grade heptane, shaking the mixture, and
allowing it to stand overnight before use. When decanting the
heptane for use in preparation of coating solutions, care was
taken not to sample the water phase. Water (18.2 MΩ) for
contact angle measurements was drawn from a Millipore
deionized water filtration system. Hexadecane was of com-
mercial anhydrous grade and used as received.
(Harrick (Ossining, NY) PDC-3xG plasma cleaner/sterilizer, set
at high power) or 5 min exposure in a home-built UV/ozone
chamber. Silicon wafers were cut into quarters and cleaned by
5
were found to yield hydrophilic substrates giving water contact
angles of essentially zero.
min exposure in the UV/ozone chamber. These treatments
Preparation and Characterization of Self-Assembled Films.
Immediately after cleaning, substrates were allowed to stand
for 20 min in a 100% relative humidity chamber, then removed
and immediately dipped for the desired exposure time (0.5-60
min) in a freshly prepared solution of 0.1 or 0.3 wt % fluorinated
trichlorosilane in heptane (1.2 or 3.6 mM 2). Self-assembled
films made under optimized conditions utilized 0.1 wt %
solutions of silane in water-saturated heptane; as-received
Trichloro-3-(1H,1H-pentadecafluorooctyloxy)propyl-
silane (C7F15CH2OCH2CH2CH2SiCl3, 2). To a stirred solution
of 124 g (0.31 mol) of 1H,1H-pentadecafluorooctanol and 62 g
(0.51 mol) of allyl bromide in 1240 g of dimethyl sulfoxide at
room temperature was added dropwise a solution of 28.1 g of

4748 J. Phys. Chem. B, Vol. 106, No. 18, 2002
Pellerite et al.
heptane was also used in some experiments as noted in the text.
All treatments were performed at room temperature. Samples
were found to be autophobic and emerged completely dry from
the coating solutions, except for those processed at the longest
dip coating times for which only partial autophobicity was
observed. Contact angle measurements were performed on fresh
coating samples with no additional processing, except for coated
silicon wafers that were measured before and after solvent
rinsing, as noted below. Wilhelmy measurements were also
repeated on coated quartz rod samples after storage in 0 or 100%
relative humidity chambers for various periods.
Film Characterization. Contact Angle Measurements. Dy-
namic contact angles on coated quartz rod substrates were
measured using a Cahn Instruments (Madison, WI) DCA-322
Wilhelmy balance and a stage speed of 90 µ/s. In this
technique,53 a microbalance measures the force exerted on a
vertically mounted sample by the wetting meniscus as the liquid
reservoir is raised and lowered over about the first 1 cm of the
sample length. Using instrument software, a least-squares fit
of the linear regions of a plot of advancing and receding
meniscus force versus stage displacement was extrapolated to
zero depth of immersion (to eliminate buoyancy effects), and
the advancing and receding forces at this point, Fadvzdi and F zdi,
were determined. These values were then used to compute the
rec
5
3a,b
dynamic contact angles θadv and θrec using eq 1
with
Figure 2. Wilhelmy dynamic contact angle data for first two sets of
three adv/rec cycles for self-assembled film from silane 2 on quartz
rod, 0.3 wt % silane in heptane, 3 min dip.
Fzdi ) Pγ cos θ
(1)
experimentally determined values of the sample perimeter P
and the surface tension γ of the wetting liquid. P was calculated
from P ) πd, where d is the diameter of the quartz rod measured
with a micrometer. Surface tensions were measured using a
freshly cleaned, hydrophilic quartz rod and eq 1 assuming cos
θ ) 1. Three complete advancing/receding contact angle cycles
were always run to assess film stability upon immersion in the
wetting medium, and generally two sets of measurements were
run to generate stabilized samples for oil contact angle
determinations (see text). A fresh sample of wetting liquid was
used for each three-cycle measurement. Surface tensions of
water, hexadecane, and heptane wetting liquids were always
measured at the start of each session as an internal check of the
system. Measured values were typically within 0.5 dyn/cm of
CA). Probe tips (Olympus, Japan) with a nominal force constant
of 40 N/m were used. Typical amplitude dampening settings
for imaging were 0.75 (set-point amplitude/free amplitude) with
a scan rate of 1 Hz.
Results and Discussion
Ultimate Contact Angles for Self-Assembled Films from
Silane 2 on Quartz. Water contact angles, particularly receding
values, on quartz rods treated with silane 2 were found to be
strongly dependent on coating conditions. Some of this depend-
ence was systematic, such as that on dip time in the coating
solution (vide infra), while much of it could be capricious and
difficult to reproduce. The procedure outlined in the Experi-
mental Section represents the combination of conditions that
in our hands yielded the most consistent results. Fresh solutions
54
the accepted literature values. Static and dynamic contact
angles on coated silicon wafers were measured using an AST
Products (Billerica, MA) VCA-2500XE video contact angle
2
8
(VCA) apparatus. Estimated uncertainties in the contact angles
of 0.1 wt % silane in water-saturated heptane, UV/ozone
cleaning, and preconditioning of cleaned substrates in 100%
were (2° for the Wilhelmy measurements, (1° for static and
advancing VCA measurements, and (2° for receding VCA
measurements.
5
6
relative humidity before coating all contributed to improved
reproducibility. Additional treatments after coating such as
6,15
Ellipsometry Measurements on Silicon Wafers. Using a
Gaertner (Skokie, IL) L116A single-wavelength ellipsometer
operating at 6328 Å and 70° incident angle, ellipsometric
parameters Ψ and ∆ were measured on the same four spots
before and after coating. Film thicknesses were determined from
rinsing with water or solvents, or heating, provided no benefit.
Under these optimized conditions, 1-3 min dip times gave the
highest, most consistent water contact angles. Also, in general,
contact angles would vary during the first set of Wilhelmy cycles
and stabilize during the second set, as illustrated in Figure 2. A
proposed origin of this effect is discussed below. Table 1 shows
the water dynamic contact angles obtained for stabilized coatings
prepared under these conditions on quartz, along with hexa-
decane and heptane contact angles obtained from further
measurements on previously stabilized samples.
5
5
these values using software supplied with the instrument, a
three-layer model (air/film/substrate), and film refractive index
nf ) 1.36 (the value measured for the neat trichlorosilane 2).
Ellipsometric measurements were made before and after vigor-
ous agitation of the coated samples submerged for 1 min in a
beaker of methyl perfluorobutyl ether, then in a beaker of
While extremely low hysteresis (θadv - θrec)57 on the order
of a few degrees or less could be obtained with the oils, about
15° was the lowest that could be obtained with water. Neverthe-
less, the water receding angle of 104° (averaged over many runs)
observed here is among the highest reported in the literature
for self-assembled fluorinated alkyltrichlorosilane films on
2
-propanol.
AFM Measurements. Coated silicon wafer samples that had
been rinsed as described above were imaged using tapping-mode
atomic force microscopy on a Nanoscope IIIa multimode
scanning probe microscope (Digital Instruments, Santa Barbara,

Contact Angle Studies of Fluorinated Alkyltrichlorosilanes
J. Phys. Chem. B, Vol. 106, No. 18, 2002 4749
TABLE 1: Dynamic Contact Angles for Stabilized
Self-Assembled Films from Fluorinated Trichlorosilanes on
Quartz Rods
adv/rec CA (deg)a
compound
water
hexadecane
74/70
heptane
56/55
C
C
C
C
7
6
8
F
F
F
15CH
13CH
17CH
2
2
2
OCH
2
CH
3
2
2
CH SiCl
3
119/104
113/94
119/100
118/95
CH
CH
2
SiCl
SiCl
2
3
62/59
64/58
10
2 2
F21CH CH SiCl
3
a
Scatter in contact angle measurements was typically (2°.
smooth58 substrates. This value is somewhat lower than the 110°
reported for solvent-cast and annealed films of a fluoroalkyl
acrylate polymer,59 and significantly lower than the 116°
60
reported for a vacuum-deposited C20F42 thin film and the 117°
exhibited by a C10F21(CH2)11SH self-assembled monolayer on
gold.61
Figure 3. Water advancing and receding contact angles for self-
assembled films from silane 2 on quartz rods, 0.3 wt % in heptane, as
a function of dip time.
For water and hydrocarbon wetting liquids, we observed a
correlation between advancing contact angle and hysteresis:
Water, which gave the highest advancing angles, also yielded
the highest hysteresis. Analogous behavior has been noted on
TABLE 2: Ellipsometric Thickness and Water Contact
a
Angles for Self-Assembled Films from Silane 2 on Silicon
Wafers before and after Solvent Rinsing
4
0
thin films prepared using monofunctional fluoroalkylsilanes.
thickness (Å)b
static/adv/rec water CA (deg)
Fadeev and McCarthy,62a in an extensive study of wettabilities
on grafted films of hydrocarbon trialkylsilanes on silicon, report
similar results for measurements with water, methylene iodide,
and hexadecane. However, hysteresis was similar with the latter
two liquids, despite methylene iodide giving larger contact
angles. We observed consistently larger hysteresis with hexa-
dip time
before
rinse
after
rinse
before
rinse
after
rinse
(min)
1
3
0
0
0
19
26
32
44
∼80
15
17
19
27
45
116/121/108
115/121/110
111/122/104
108/122/81
111/123/87
115/121/110
115/123/107
113/123/103
113/127/95
122/131/89
1
3
6
54
decane (γ ) 27.5 dyn/cm ) than with heptane (γ ) 20.1 dyn/
54
cm ). The correlation between advancing or static contact angles
and wetting liquid surface tension is of course well-known and
a
Wafers cleaned and prepared as described in text; coatings prepared
by dipping for indicated times in 0.1 wt % 2 in water-saturated heptane.
63
manifests itself in the Zisman plot. However, it is not obvious
that such a general relationship should also exist between contact
angle hysteresis and surface tension. Perhaps this is indicative
of less-than-perfect monolayers, since contact angles on high-
quality self-assembled thiol monolayers on gold typically show
low hysteresis with both water and oils. One example of this is
b
Rinses done in methyl perfluorobutyl ether then 2-propanol. Scatter
in ellipsometric thicknesses was (1 Å for dip times of 1-3 min, (2-4
Å for longer times. Thickness for 60 min dip unrinsed film was difficult
to reproduce.
An analogous dependence was observed by Menawat and co-
6
1
64
provided by recent data of Fukushima and co-workers for
monolayers prepared from C10F21(CH2)11SH on gold. These
films gave adv/rec contact angles of 83/75° with hexadecane
and 122/117° with water. While this hexadecane hysteresis of
workers in a study of several hydrocarbon silanes, in which
silane concentration was varied at fixed dip time. Their observed
increase in hysteresis was attributed to multilayer adsorption at
higher concentrations, with the maximum in receding angle
thought to coincide with the highest-quality, best-packed mono-
layer. For our fluorinated system, although monolayer for-
mation from monomeric silane appears to be the faster
process, self-assembly does not simply stop at the mono-
layer; instead, hydrolysis and oligomerization of the silane in
solution and adsorption of oligomeric siloxane species37 also
occurs in competition with monolayer formation. As discussed
below, we believe that this phenomenon is responsible for the
contact angle dependence on dip times seen in Figure 3 and
Table 2.
Oligomer adsorption can account for our observation that for
short dip times, the water contact angles changed during the
first set of three DCA cycles on a fresh sample, finally
stabilizing during the second set (Figure 2). We believe this is
due to removal of small amounts of weakly bound oligomers
by the water meniscus during the first set of cycles. At short
dip times, the levels of adsorbed oligomers are low, and the
excess material can still be removed by rinsing, unlike on films
prepared at longer dip times. This effect is also seen in
ellipsometry data on coated silicon wafers (vide infra). However,
the amounts of material involved here appear to be too small
to detect in surface tension measurements on water samples used
in the contact angle determinations; no reduction from the value
for pure water was ever observed.
8° is only slightly larger than that (4°) observed here for 2 films
on quartz, the water hysteresis is substantially lower (5° vs 15°
in our study.) This could suggest that water contact angle
hysteresis is the more sensitive probe of monolayer quality, but
we also note with caution that this comparison involves
completely different amphiphile/substrate systems.
Finally, we examined the effects of adding small amounts of
alkylamines to coating solutions of 2, as well as preconditioning
cleaned quartz rod substrates in an atmosphere of ammonia
before dip coating. The former caused severe problems with
coating solution stability, while the latter had no discernible
effect on contact angles of the resulting self-assembled films.
4
4
Tripp and Hair reported that amines can enhance formation
and properties of films prepared from (3,3,3-trifluoropropyl)-
trichlorosilane on silicon wafers. However, we found no
evidence for such effects operating in our system.
Dependence of Water Dynamic Contact Angles on Dip
Time. Coating time exerted dramatic effects on water dynamic
contact angles of self-assembled thin films derived from silane
2
. Illustrative data are shown in Figure 3. Both advancing and
receding angles rose rapidly with dip time, the receding angle
going through a maximum around 1-3 min and dropping off
at longer times, while the advancing angle continued to increase
and leveled off after ∼10 min.

4750 J. Phys. Chem. B, Vol. 106, No. 18, 2002
Pellerite et al.
Figure 4. AFM images and profile analyses (along the indicated horizontal lines) of self-assembled films from silane 2 on silicon wafers prepared
at various dip times (0.1 wt % silane in water-saturated heptane) and solvent rinsed as described in the text.
Additional evidence for the connection between contact angle
hysteresis and oligomer adsorption can be found in the contact
angle and ellipsometry data for self-assembled films from silane
AFM images on the above silicon wafer samples shed some
light on the form taken by the adsorbed oligomeric material.
Figure 4 shows images of solvent-rinsed wafers dip coated for
1, 10, and 60 min (ellipsometric thicknesses 16, 21, and 43 Å,
respectively.) It is clear that the film thickness in excess of one
monolayer is caused by adsorption of small particulate islands
that are even found at short dip times. Profile analysis of the
images shows a distribution of 15-60 Å feature heights at 1
min dip time, 15-90 Å at 10 min, and up to 150 Å at 60 min.
In addition to becoming more numerous, the particulates appear
to be growing as dip time increases.
2
on silicon wafers, as shown in Table 2. Compiled there are
static, advancing, and receding water contact angles and
ellipsometric film thicknesses on wafer samples treated at dip
coating times ranging from 1 to 60 min, both before and after
rinsing with methyl perfluorobutyl ether (a good solvent for
the fluorinated silane and oligomers) and 2-propanol. Here again,
as was seen in the contact angle data on quartz rods, receding
contact angles decreased with increasing dip times; the ellip-
sometric data show clearly that film thicknesses also increased.
Even 1 min dip time gave unrinsed films in excess of one
monolayer thick, suggesting that oligomer adsorption occurs at
a rate competitive with that for monolayer formation. Note also
that the maximum receding contact angles on rinsed films
occurred for those prepared at short dip times and giving
thicknesses in the monolayer range65 of 15-17 Å. Furthermore,
since rinsing of short dip time samples yielded films of
essentially monolayer thickness, this implies that the bulk of
the excess material was weakly adsorbed and easily removed
by rinsing. On the other hand, while rinsing of longer dip time
samples removed a portion of the film thickness in excess of
one monolayer, it did not remove all of it, nor did it restore the
high receding contact angles seen on samples prepared at short
dip times. This underscores the role of the oligomers in the
contact angle dependence on dip time. It also suggests that
oligomers adsorbed at long dip times are more strongly bound
Adsorption of particulates was also observed recently by
3
7
Bunker and co-workers in AFM imaging of self-assembled
films from silane 1 (n ) 8) delivered from isooctane, but there
are significant differences in behavior between the two cases.
For instance, in Figure 4 the particulates appear to be adsorbing
in quasi-cluster type structures at long dip times. This was not
37
observed for 1, where the particulate distribution appeared to
be fairly random. Origins of this difference are unknown but
are under study. Also, the features observed by Bunker and co-
workers were significantly larger than those seen here, being
3
7
up to 300 nm in width and 25 nm in height. This could arise
from differences in solubility of the two silanes in the solvents
employed for film formation. Bunker and co-workers present
convincing evidence for formation of aggregated, inverse
micellar structures in solutions of silane 1 in isooctane. Although
silane 2 would exhibit better solubility than 1 in hydrocarbon
media due to the lower fluorine content, it is entirely likely that
aggregated structures form in the silane solutions in wet heptane
used in this work. Indeed, the solutions do show a Tyndall
(perhaps covalently, via condensation of oligomer silanol groups
with silanols exposed at defects in the monolayer) than those
adsorbed at short dip times. Or, this could be a manifestation
of oligomer molecular weights increasing (and hence their
solubility in the rinse solvent dropping) as reaction time
increases. In any case, taken together, our data provide good
support for the main features of the above mechanism. Also,
they emphasize the importance of obtaining receding contact
angle measurements when comparing a series of film samples.
Table 2 shows that static and advancing angles do not always
reveal the full story, and receding angles are much more
sensitive to subtle changes in film structure.5
6
6
effect, suggesting the presence of colloidal species. We have
not undertaken the light scattering experiments that would be
necessary to confirm this and explore the system further. Also,
our coatings are prepared on a time scale significantly shorter
than that required for trichlorosilane hydrolysis and particulate
nucleation in the experiments of Bunker and co-workers.37
A
detailed analysis of the AFM images obtained for self-assembled
films from silane 2 is beyond the scope of this paper and is the
subject of ongoing investigation. Nevertheless, the images
support the notion that adsorption of excess material is occurring
3d

Contact Angle Studies of Fluorinated Alkyltrichlorosilanes
J. Phys. Chem. B, Vol. 106, No. 18, 2002 4751
in this self-assembly process and show that such adsorption does
not involve continuous buildup of smooth films. At this point,
we believe that the surface roughness introduced by these
particulates is responsible for the significant hysteresis in the
water dynamic contact angles, since the hysteresis and the
density of adsorbed particles both increase with dip time. One
might also argue that the increase in hysteresis can be viewed
within the framework of “molecular scale roughness” advanced
by Fadeev and McCarthy.62a In this case, the roughness would
be caused by adsorbed or grafted oligomeric segments. How-
ever, their work showed water contact angles relatively inde-
pendent of molecular scale roughness,6 whereas Table 2 shows
a large increase in hysteresis at longer dip times. Therefore, we
favor the interpretation that the contact angle hysteresis in Figure
2b
3
originates from larger-scale (tens of nanometers) roughness
induced by adsorption of oligomeric particulate species.
Comparison of Water and Oil Contact Angles for Films
from Silanes 1 and 2. Some interesting features emerge when
comparing dynamic contact angles for the four compounds 1
(n ) 6, 8, 10) and 2 (Table 1). For instance, water receding
contact angles for 1 passed through a maximum for n ) 8, while
n ) 6 also gave lower advancing angles than its homologues.
Thus, among 1 (n ) 6, 8, 10), n ) 8 gave the highest-quality
3
8
films for depositions at room temperature.
Dynamic contact angle behavior of silanes 1 (n ) 8) and 2
also makes for an interesting comparison. Although the water
advancing angles were essentially identical, we obtained higher
receding angles for 2 under all conditions investigated, including
many side-by-side runs with both compounds under identical
conditions. This is surprising in view of the mixture of linear
and branched perfluoroheptyl segments as well as the linking
group ether oxygen in 2, both of which might be expected to
adversely affect contact angle behavior. Differences also ap-
peared in the heptane data, with 2 giving near zero hysteresis
and lower angles. Rondelez and co-workers reported low
hysteresis for films prepared from 1 (n ) 8, 10) by dip coating
at subambient temperatures and stated that “good grafting with
Figure 5. Wilhelmy water dynamic contact angle data on two quartz
rods, bearing self-assembled films from silane 2, after storage under
conditions indicated. These two rod samples were prepared using
identical conditions and both gave adv/rec ) 120/104 initially after
stabilization (see text.)
(n ) 8) and give more uniform film coverage. The branched
isomer content in 2 is perhaps the origin of the lower alkane
contact angles relative to those for 1 (n ) 8). Stated differently,
hysteresis may be more related to substrate coverage and film
uniformity, while chain structure differences show up in alkane
2
2
fluorinated silanes requires low temperatures of reaction.”
(
but not water) advancing angles. Results supporting this notion
Silane 2 is an example of a fluorinated silane that can assemble
at room temperature into films showing almost no contact angle
hysteresis with heptane.
4
0
were also obtained by Fadeev and co-workers in dynamic
contact angle studies on a series of monofunctional fluoroalkyl-
silanes bearing linear and branched terminal fluorochemical
segments. They reported higher water advancing and receding
contact angles and lower pentadecane contact angles for
branched fluoroalkyl segments relative to their linear counter-
parts. Such a picture is also consistent with the assertion put
In their same 1994 study, Rondelez and co-workers22
proposed the existence of a critical temperature Tc for silane
film formation, below which optimum packing of the monolayer
can be achieved. This behavior was thought to be related to the
transition between liquid expanded and liquid compressed states
in Langmuir monolayers. The value of Tc is a function of silane
structure; for hydrocarbon silanes, Tc increases with increasing
alkyl chain length. In this work, both hydrocarbon and fluoro-
carbon silanes (specifically, 1 (n ) 8, 10)) showed lower alkane
contact angle hysteresis when coated below Tc (which is below
62a
forth by Fadeev and McCarthy that organic wetting liquids
can penetrate some types of hydrocarbon silane monolayers and
sample other structural features of the silane alkyl chain, whereas
water does not. Traditionally, this is thought not to be the case
31,62a
with crystalline monolayers exhibiting high packing density,
but these shorter-chain fluorinated silane films may fall short
22
room temperature for the fluorinated silanes ).
in fulfilling these requirements. Indeed, 1 (n ) 8) has yielded
The above observation that 2 gives lower hysteresis than 1
21
amorphous monolayers under some conditions.
(n ) 8) when coated at room temperature seems inconsistent
Effects of Storage on Water Dynamic Contact Angles in
Self-Assembled Films from 2. The contact angles shown in
Table 1 for stabilized films from trichlorosilane 2 and 1 (n )
with the Rondelez picture. Given the correlation between Tc
and chain length for hydrocarbon silanes, we expect silane 2,
with its branched perfluoroalkyl chain content and the flexi-
bilizing ether oxygen, to have a lower Tc than the linear and
more highly fluorinated 1 (n ) 8). Thus, at room temperature,
8
) on quartz rods were not stable to storage at room temperature
37
under high humidity. Bunker and co-workers and de Boer and
co-workers have also pointed out that high humidity storage
67
2
should be further above Tc than 1 (n ) 8); with respect to
results in degradation of wetting and lubricative properties of
silicon wafers treated with self-assembled films derived from
silane 1 (n ) 8.) Figure 5 illustrates this behavior for films
from silane 2. It shows water dynamic contact angle data on
two stabilized film samples, taken after storage in 0 or 100%
film properties, one would predict a result opposite that
observed.
We account for this behavior by proposing that the more
flexible molecule, 2, can assemble and pack more easily than 1

4752 J. Phys. Chem. B, Vol. 106, No. 18, 2002
Pellerite et al.
relative humidity (RH) for 12 weeks. These two samples were
prepared under identical conditions, and when fresh, both gave
water adv/rec contact angles of 120/104. While 0% RH storage
did essentially nothing to these values, 100% RH induced
profound changes. The first-cycle advancing angle was largely
unchanged from that for the fresh sample, but all subsequent
advancing and receding angles were significantly lower. This
pattern appeared consistently, although the exact magnitude of
the contact angle changes and the storage time needed to
produce them could vary from sample to sample. Second and
third cycle advancing angles would typically drop 5-15°, while
the receding angles could drop 10-20°, relative to those for
the fresh coating. Some samples exhibited these changes after
only 1-2 weeks’ storage at high humidity, while others took
longer. The large drop in advancing angles between the first
and second cycle indicates significant changes occurring in the
film upon exposure to liquid water.
Proposed Mechanism for Self-Assembly of Fluorinated
Alkyltrichlorosilane Films. Much of the literature on alkyl-
trichlorosilane self-assembly on silica utilizes a model in which
condensation of silanols leads to formation of cross-linked
siloxane networks, covalently bound to the substrate through
2
,10,25
occasional condensation with surface silanol groups.
3
6
Recently, Stevens emphasized that long-chain alkyl groups
are too bulky to allow silanol condensation and construction of
well-ordered, two-dimensionally cross-linked films, because
their van der Waals cross-sectional radii are larger than the Si-O
bond length. One of the basic assumptions that has guided
research in this area for more than 20 years is thus called into
question.
This situation can only be worse for fluorinated alkylsilane-
derived films, since perfluoroalkyl segments exhibit larger cross-
19,21
sectional areas than their hydrocarbon analogues.
Thus, it
appears extremely unlikely that these systems, including those
of the present study, can be forming the classical two-
dimensional cross-linked siloxane network with vertically
oriented, ordered fluorochemical chains.
We propose a mechanistic picture (Figure 6) for formation
of self-assembled films from fluorinated alkyltrichlorosilanes
on quartz that takes this view into account while also rational-
izing much of the experimental results described here. In this
mechanism, the active amphiphilic species is the silanetriol
RfSi(OH)3, formed by hydrolysis of the trichlorosilane with
surface-adsorbed water.56 The silanetriol can self-assemble into
immobilized monolayer films stabilized by hydrogen bonding
with the silica substrate and van der Waals interactions between
the fluorochemical groups. It is this construction that we believe
gives rise to the films displaying the minimum contact angle
Figure 6. Schematic illustration of proposed mechanism for humidity-
induced degradation of water dynamic contact angles for self-assembled
films from fluorinated trichlorosilanes such as 2.
under dry conditions. We suggest that the contact angle
71
degradation that occurs upon exposure of the low-hysteresis
films to high humidity is due to disruption, by gaseous water,
of the hydrogen bonding between silanetriol and substrate. Under
humid conditions, water can penetrate the monolayer film and
displace the silanetriol from the surface; these species are then
free to condense with other neighboring silanetriols, generating
regions of disordered material and leading to greatly reduced
receding contact angles and unstable advancing values (Figures
hysteresis discussed earlier. This mechanism is similar to that
which has appeared in the literature1
7,20,22,28,29,37,44,55
many times,
except without the extensive condensation and cross-linking
generally assumed to occur. Within this framework, and as
37
proposed by Bunker and co-workers, the use of water-saturated
heptane probably facilitates coating formation by preventing
dissolution of the adsorbed water film into an anhydrous coating
medium. Simultaneously, and consistent with other observa-
5
b and 6). An alternative mechanism, suggested by a referee,
involves an initial monolayer with disordered headgroups and
some siloxane cross-links; upon exposure to humidity, these
3
7,68
tions,
the trichlorosilane can react in a competitive process
with water in the solvent to eventually yield condensed
oligomeric species that can also adsorb to the surface and to
the monolayer as contact times with the solution become long
69
hydrolyze, leading to film degradation. In either case, the
71
changes in advancing angles from first to second cycle may
suggest further hydration of the quartz surface and disruption
of the coating upon immersion in liquid water.
(>1 min).
Our experimental data also indicate that, if formed, this
arrangement of self-assembled silanetriols in low-hysteresis
films is not stable to exposure to 100% RH, although it can be
This mechanistic hypothesis raises a number of basic ques-
tions that we are not able to answer at this point. Issues of chain
packing density, siloxane bond formation, and the role of the

Contact Angle Studies of Fluorinated Alkyltrichlorosilanes
J. Phys. Chem. B, Vol. 106, No. 18, 2002 4753
substrate (quartz and silicon may not behave identically due to
different levels of surface silanol and adsorbed water; these
could affect film properties in as yet undertermined ways) have
yet to be sorted out and can provide the basis for future
investigations. For instance, detailed AFM analysis could in
principle address questions of packing density and chain order,
as well as providing a tool for examining structural effects of
film exposure to high humidity. The humidity-induced changes
we are proposing may be detectable by AFM. Infrared spec-
ples the advancing and receding values were perfectly stable
over three Wilhelmy cycles, after aging in a 100% RH environ-
ment the dynamic contact angle behavior changed completely.
The first cycle advancing angle was largely unchanged, but
subsequent advancing and all receding cycles showed contact
angles 10-20° lower than the same sample before aging.71
Finally, we propose a mechanism that can account qualita-
tively for the bulk of these results. The central feature of this
mechanism is reaction of the fluorinated alkyltrichlorosilane with
surface-adsorbed water to yield a self-assembled thin film
consisting primarily of alkylsilanetriol molecules hydrogen-
bonded to the substrate. Our results with high humidity aging
of these films suggest that the film is at most only weakly cross-
linked, although we cannot rule out covalent bonding of some
fraction of the adsorbed molecules to each other and the
substrate by silanol condensation. Additional work is needed
to address the role of siloxane bonding in these films. Of course,
this mechanism and the work reported here still leave many
questions unanswered. Nevertheless, we hope that it can serve
as a guide for additional efforts that will lead to further
fundamental understanding of these interesting systems, as well
as provide insight into routes to films with enhanced perfor-
mance in areas such as surface protection and lubrication.
troscopy may provide data with which to assess the role of
siloxane bond formation,1
7,18,44,70
both between molecules and
with the substrate, in these self-assembled films, and perhaps
allow the above two mechanistic hypotheses for humidity-
induced film degradation to be distinguished. At present we have
no information on the importance of condensation of silanetriols
with each other (cross-linking), or with surface silanols to yield
covalently surface-bound fluorinated alkylsilanes.70 These reac-
tions are potential stabilizing mechanisms for the films. If they
occur to varying extents in different samples, this could account
for the differences we have seen in humidity-induced contact
angle changes and the rates at which they occur.
Summary and Conclusions
Dynamic contact angle studies on self-assembled thin films
from fluorinated alkyltrichlorosilane 2 reveal a range of interest-
ing behavior. Conditions have been identified that allow
preparation of what are believed to be essentially monolayer
films on quartz rods exhibiting water advancing/receding contact
angles of 119/104 ( 2°; this receding angle is one of the highest
yet reported for fluorinated trichlorosilane films on smooth
surfaces. Use of dilute (1.2 mM) solutions of silane, water-
saturated heptane as the solvent, and short dip times (1-3 min)
were critical to the most consistent preparation of films with
minimum hysteresis. These films gave even lower contact angle
hysteresis with hydrocarbon wetting liquids. Silane 2 is an
example of a fluorinated trichlorosilane that is able to deliver
low-hysteresis films by deposition at room temperature.
Measurements on films prepared on silicon wafers as a
function of dip time showed that advancing angles continued
to increase and receding angles decreased as dip times were
increased up to 60 min. This behavior appears to arise from
adsorption of oligomeric material from solution to yield thicker,
less ordered, and rougher films. Ellipsometric data confirmed
that film thicknesses increased with dip time, while AFM
imaging showed that this excess material was deposited in the
form of particulates. Solvent rinsing removed a fraction of this
excess material, with this fraction decreasing as dip times
increased. Adsorption of oligomeric siloxane particulates has
been noted in previous work with other self-assembled tri-
Acknowledgment. Thanks go to Dr. T. D. Dunbar (3M
Advanced Materials Technology Center) for ellipsometric
assistance; Drs. D. A. Weil and K. A. Thakur (3M Corporate
Analytical Technology Center) for mass spectrometric and NMR
work, respectively; and Drs. Dunbar, L. D. Boardman, H. E.
Johnson, C. R. Kessel, and R. R. Shah (3M Advanced Materials
Technology Center) for helpful discussions and critical reading
of the manuscript.
References and Notes
(1) A portion of this work was first presented at the 216th National
ACS Meeting, Boston, MA, August 1998.
(2) For leading references, see, for instance: (a) Sagiv, J. J. Am Chem.
Soc. 1980, 102, 92-98. (b) Maoz, R.; Sagiv, J. J. Colloid Interface Sci.
1
984, 100, 465-496.
(3) Ulman, A. An Introduction to Ultrathin Organic Films; Academic
Press: San Diego, CA, 1991.
4) Brunner, H.; Vallant, T.; Mayer, U.; Hoffmann, H.; Basnar, B.;
(
Vallant, M.; Friedbacher, G. Langmuir 1999, 15, 1899-1901.
(5) Fontaine, P.; Goldmann, M.; Rondelez, F. Langmuir 1999, 15,
1
348-1352.
(6) Tian, F.; Xiao, X.; Loy, M. M. T.; Wang, C.; Bai, C. Langmuir
1
999, 15, 244-249.
(7) Vallant, T.; Brunner, H.; Mayer, U.; Hoffmann, H.; Leitner, T.;
Resch, R.; Friedbacher, G. J. Phys. Chem. B 1998, 102, 7190-7197.
(
(
8) Vidon, S.; Leblanc, R. J. Phys. Chem. B 1998, 102, 1279-1286.
9) Richter, A. G.; Durbin, M. K.; Yu, C.-J.; Dutta, P. Langmuir 1998,
1
4, 5980-5983.
(10) Rye, R. R.; Nelson, G. C.; Dugger, M. T. Langmuir 1997, 13,
965-2972.
2
(11) Huang, J. Y.; Song, K. J.; Lagoutchev, A.; Yang, P. K.; Chuang,
37,68
chlorosilane films;
these results and ours showed qualitative
T. J. Langmuir 1997, 13, 58-64.
differences that were attributed to different silane solubilities
and film preparation time scales.
(12) Banga, R.; Yarwood, J.; Morgan, A. M.; Evans, B.; Kells, J. Thin
Solid Films 1996, 284-285, 261-266.
(13) Lobau, J.; Rumphorst, A.; Galla, K.; Seeger, S.; Wolfrum, K. Thin
Comparison of dynamic contact angles for films prepared
from silane 2 with those prepared from silane 1 (n ) 8) revealed
some interesting structure-property trends. Films from silane
Solid Films 1996, 289, 272-281.
14) Takahara, A.; Kojio, K.; Ge, S.-R.; Kajiyama, T. J. Vac. Sci.
Technol. A 1996, 14, 1747-1754.
15) Calistri-Yeh, M.; Kramer, E. J.; Sharma, R.; Zhao, W.; Rafailovich,
M. H.; Sokolov, J.; Brock, J. D. Langmuir 1996, 12, 2747-2755.
16) Bierbaum, K.; Grunze, M. Langmuir 1995, 11, 2143-2150.
(
(
2
always yielded lower hysteresis under all conditions studied.
Water advancing contact angles were similar, but 1 (n ) 8)
gave higher contact angles with hydrocarbon liquids. We
propose that these differences arise from enhanced coverage of
monolayer films from silane 2 due to the flexibilizing ether
linkage, and the branched perfluoroheptyl group content in the
trichlorosilane film precursor 2.
Water dynamic contact angles on self-assembled films from
fluorinated trichlorosilanes on quartz were found to be unstable
to storage in high humidity conditions. Whereas for fresh sam-
(
(17) Allara, D. L.; Parikh, A. N.; Rondelez, F. Langmuir 1995, 11,
357-2360.
2
(
18) Parikh, A. N.; Liedberg, B.; Atre, S. V.; Ho, M.; Allara, D. L. J.
Phys. Chem. B 1995, 99, 9996-10008.
19) Banga, R.; Yarwood, J.; Morgan, A. M.; Evans, B.; Kells, J.
Langmuir 1995, 11, 4393-4399.
20) Nakagawa, T.; Ogawa, K.; Kurumizawa, T. Langmuir 1994, 10,
25-529.
21) Ge, S.; Takahara, A.; Kajiyama, T. J. Vac. Sci. Technol. A 1994,
12, 2530-2536.
(
(
5
(

4754 J. Phys. Chem. B, Vol. 106, No. 18, 2002
Pellerite et al.
(
22) Brzoska, J. B.; Ben Azouz, I.; Rondelez, F. Langmuir 1994, 10,
367-4373.
23) Fujii, M.; Sugisawa, S.; Fukada, K.; Kato, T.; Shirakawa, T.;
Seimiya, T. Langmuir 1994, 10, 984-987.
24) Wei, M.; Bowman, R. S.; Wilson, J. L.; Morrow, N. R. J. Colloid
Interface Sci. 1993, 157, 154-159.
25) Le Grange, J. D.; Markham, J. L.; Kurkjian, C. R. Langmuir 1993,
, 1749-1753.
(60) Nishino, T.; Meguro, M.; Nakamae, K.; Matsushita, M.; Ueda, Y.
Langmuir 1999, 15, 4321-4323.
(61) Fukushima, H.; Seki, S.; Nishikawa, T.; Takiguchi, H.; Tamada,
K.; Abe, K.; Colorado, R.; Graupe, M.; Shmakova, O. E.; Lee, T. R. J.
Phys. Chem. B 2000, 104, 7417-7423.
4
(
(
(62) (a) Fadeev, A. Y.; McCarthy, T. J. Langmuir 1999, 15, 3759-
3766. (b) We are grateful to a referee for bringing this point to our attention.
The referee also noted the disparate effects of molecular scale roughness
on water and hexadecane contact angles for the series of hydrocarbon silane
films in ref 62a. While water values were largely insensitive to increases
in molecular scale roughness, those for hexadecane increased. We have
characterized effects of particulate-based roughness on water contact angles
in our fluorinated system (Table 2); it would be informative to probe the
response of contact angles for other liquids such as hexadecane, to compare
the behavior with that studied in ref 62a. We predict that, like water, the
response of hexadecane contact angles would also be different for the two
systems. Such experiments will be the subject of future work.
(63) Adamson, A. W. Physical Chemistry of Surfaces, 4th Ed.; Wiley-
Interscience: New York, 1982; pp 350-352.
(
9
7
5
(26) Tripp, C. P.; Hair, M. L. Langmuir 1992, 8, 1961-1967.
(27) Tripp, C. P.; Hair, M. L. Langmuir 1992, 8, 1120-1126.
(28) Silberzan, P.; Leger, L.; Ausserre, D.; Benattar, J. J. Langmuir 1991,
, 1647-1651.
(29) Angst, D. L.; Simmons, G. W. Langmuir 1991, 7, 2236-2242.
(30) Kessel, C. R.; Granick, S. Langmuir 1991, 7, 532-538.
(31) Wasserman, S. R.; Tao, Y.-T.; Whitesides, G. W. Langmuir 1989,
, 1074-1087.
(
32) DePalma, V.; Tillman, N. Langmuir 1989, 5, 868-872.
(33) Wasserman, S. R.; Whitesides, G. W.; Tidswell, I. M.; Ocko, B.
M.; Pershan, P. S.; Axe, J. D. J. Am. Chem. Soc. 1989, 111, 5852-5863.
34) Maoz, R.; Netzer, L.; Gun, J.; Sagiv, J. J. Chim Phys. 1988, 85,
059-1065.
35) Finklea, H. O.; Robinson, L. R.; Blackburn, A.; Richter, B.; Allara,
(
(64) Menawat, A.; Henry, J.; Siriwardane, R. J. Colloid Interface Sci.
1984, 101, 110-119.
1
(
(65) Molecular modeling calculations (G. Caldwell, unpublished results)
using the Dreiding force field suggest an extended length of 16 Å (silicon
to trifluoromethyl fluorines) for silane 2. Given the estimated uncertainty
of (2 Å (ref 43) for ellipsometric thicknesses measured by fixing the
refractive index, we can say only that the fluorinated silane chains appear
to be oriented largely normal to the surface, with a relatively small tilt
angle.
D.; Bright, T. Langmuir 1986, 2, 239-244.
(
36) Stevens, M. Langmuir 1999, 15, 2773-2778.
(37) Bunker, B. C.; Carpick, R. W.; Assink, R. A.; Thomas, M. L.;
Hankins, M. G.; Voigt, J. A.; Sipola, D.; de Boer, M. P.; Gulley, G. L.
Langmuir 2000, 16, 7742-7751.
(38) Hozumi, A.; Ushiyama, K.; Sugimura, H.; Takai, O. Langmuir 1999,
1
5, 7600-7604.
39) Zybill, C. E.; Ang, H. G.; Lan, L.; Choy, W. Y.; Meng, E. F. K. J.
Organomet. Chem. 1997, 547, 167-172.
40) Fadeev, A. Y.; Soboleva, O. A.; Summ, B. D. Colloid J. 1997, 59,
(66) Everett, D. H. Basic Principles of Colloid Science; Royal Society
of Chemistry: London, 1988.
(
(67) De Boer, M. P.; Knapp, J. A.; Mayer, T. M.; Michalske, T. A.
Proc. SPIE-Int. Soc. Opt. Eng. 1999, 3825, 2-15.
(
2
22-225.
(68) Very recent results also suggest a similar mechanism, involving
(41) Hoffmann, P. W.; Stelzle, M.; Rabolt, J. F. Langmuir 1997, 13,
deposition of both monomer and aggregate structures, for octadecyltri-
chlorosilane self-assembled monolayers on fused silica: Liu, Y.; Wolf, L.
K.; Messmer, M. C. Langmuir 2001, 17, 4329-4335.
1
877-1880.
(
42) Tada, H.; Nagayama, H. Langmuir 1994, 10, 1472-1476.
(43) Geer, R. E.; Stenger, D. A.; Chen, M. S.; Calvert, J. M.; Shashidhar,
(69) (a) This referee questioned the mechanistic hypothesis illustrated
R.; Jeong, Y. H.; Pershan, P. S. Langmuir 1994, 10, 1171-1176.
in Figure 6, noting that equilibrium considerations argued against a scenario
in which addition of gaseous water to the system induced condensation,
which is formally a dehydration process. However, we note that silanol
condensations can occur even in aqueous solution (ref 69b), and silanol/
siloxane equilibria can lie well toward siloxane (ref 69c). On this basis we
prefer our interpretation, in which the silanetriols are stabilized by hydrogen
bonding in the initial monolayer (see also ref 70). However, at this point
we cannot distinguish experimentally between the two alternatives, nor do
we claim that the initial monolayer is free of intermolecular siloxane
bonding. (b) Plueddemann, E. P. Silane Coupling Agents, 2nd ed.;
Plenum: New York, 1991; p 57. (c) Wilczek, L.; Chojnowski, J. Die
Makromol. Chem. 1983, 184, 77.
(44) Tripp, C. P.; Veregin, R. P. N.; Hair, M. L. Langmuir 1993, 9,
3
518-3522.
(
45) Lindner, E.; Arias, E. Langmuir 1992, 8, 1195-1198.
(46) Yanazawa, H.; Matsuzawa, T.; Hashimoto, N. J. Adhes. Sci.
Technol. 1990, 4, 145-153.
(
(
(
(
47) Bascom, W. D. J. Colloid Interface Sci. 1968, 27, 789-796.
48) See refs 12, 14, 19, 21-23, 31, 32, 34, 37-47.
49) U. S. Patent 5,851,674 (1999).
50) For another example where a fluorinated material with this type of
isomeric chain distribution has been used in preparation of organic thin
films, see: Schmidt, D. L.; DeKoven, B. M.; Coburn, C. E.; Potter, G. E.;
Meyers, G. F.; Fischer, D. A. Langmuir 1996, 12, 518-529.
(70) Tripp, C. P.; Hair, M. L. Langmuir 1995, 11, 1215-1219. This
work demonstrates that, for octadecyltrichlorosilane self-assembled films
deposited from wet carbon tetrachloride, few if any siloxane bonds form
between a fumed silica substrate and adsorbed silanols.
(
51) U. S. Patent 3,012,006 (1961).
(52) Yoshino, N.; Yamamoto, Y.; Hamano, K.; Kawase, T. Bull. Chem.
Soc. Jpn. 1993, 66, 1754-1758.
53) (a) Tretinnikov, O. N.; Ikada, Y. Langmuir 1994, 10, 1606-1614.
b) Hayes, R. A.; Ralston, J. Chem. Australia 1992, 524-528. (c) Lander,
L. M.; Siewierski, L. M.; Brittain, W. J.; Vogler, E. A. Langmuir 1993, 9,
(
(71) Note Added in Proof. Close inspection of the force vs displacement
trace in Figure 5b reveals that the sample was undergoing electrostatic
charging upon withdrawal from the liquid. The effects of this charging can
be seen as a positive (attractive) force as the bottom of the sample
approached the liquid on cycles 2 and 3 (0-2 mm displacement). This
general behavior has been studied previously for silanized glass surfaces
immersed in water (Yaminsky, V. V.; Johnson, M. B. Langmuir 1995, 11,
4153-4158). These workers pointed out that this electrification produces
a reduction in apparent advancing contact angles on subsequent immersion
cycles, exactly what we see in Figure 5b. Thus, it is conceivable that the
contact angle effects seen in Figure 5b are due only to static charging, rather
than the humidity-induced film degradation proposed in the text. While we
cannot entirely rule this out at present, we believe it to be unlikely for the
following reasons. First, we have seen samples which exhibit varying
degrees of electrification yet show no changes in contact angle with cycling.
(For instance, the second set of three cycles in Figure 2 show slight static
charging and perfectly stable contact angles.) Second, Yaminsky and
Johnson note that another important result of the charging process is
pronounced nonlinearity of the receding force curves, and this phenomenon
is not observed in our data. Finally, the magnitude of the attractive force is
significantly smaller here than in the systems studied by Yaminsky and
Johnson. However, for reasons we do not currently understand, the sample
in Figure 5 stored at low humidity was considerably less susceptible to
static charging than was the sample stored at high humidity. This appeared
consistently in our work. High humidity storage seems to have some effect
on the self-assembled film itself which is not produced by low humidity
exposure. These effects are irreversible; moving a sample such as that from
Figure 5b from high- into low-humidity storage did not regenerate the
behavior shown in Figure 5a. Obviously, more work is needed here to fully
clarify the situation.
(
2
237-2239. (d) Sessile drop measurements of hexadecane contact angles
on the rinsed 1 and 3 min coated silicon wafer samples in Table 2 showed
hysteresis and contact angle values within a few degrees of those on quartz
rods (Table 1). Thus, we did not observe the effect noted by Fukushima
6
1
and co-workers in which Wilhelmy and sessile drop methods yielded
significantly different results. While this was attributed to differences
inherent in fabrication of the gold substrates (ref 61,) at least one study
(
ref 53c) has shown that systematic differences can occur in water contact
angles measured with the two techniques. This was not seen in our work,
at least to a significant extent. Also, for short dip times (1-3 min), the
static water contact angles in Table 2 are in good agreement with those
reported (ref 37) for films from 1 (n ) 8); however, we did not see values
approaching 140° at long times as reported there.
(
(
54) Jasper, J. J. J. Phys. Chem. Ref. Data 1972, 1, 841-1009.
55) Parikh, A.; Allara, D.; Azouz, I. B.; Rondelez, F. J. Phys. Chem.
1
994, 98, 7577-7590.
56) A small amount of water adsorbed on the substrate is known (refs
8, 29, and 37) to be essential for formation of high-quality alkyltri-
(
2
chlorosilane films.
57) Some workers (see, for instance, refs 22 and 31) use (cos θadv -
cos θrec) to obtain hysteresis values proportional to surface energies.
58) Values on microstructured surfaces can be much higher due to
roughness effects: (a) O¨ ner, D.; McCarthy, T. J. Langmuir 2000, 16, 7777-
782. (b) Shibuichi, S.; Yamamoto, T.; Onda, T.; Tsujii, K. J. Colloid
Interface Sci. 1998, 208, 287-294.
59) Katano, Y.; Tomono, H.; Nakajima, T. Macromolecules 1994, 27,
342-2344.
(
(
7
(
2