Surface-initiated polymerization of superhydrophobic polymethylene

Article abstract of DOI:10.1021/ja9086193

We report a new surface-initiated polymerization strategy that yields superhydrophobic polymethylene (PM) films from initially smooth substrates of gold and silicon. The films are prepared by assembling a vinyl-terminated self-assembled monolayer, followed by exposure of the surface to a 0.1 M solution of borane, and polymerizing from the borane sites upon exposure to a solution of diazomethane at -17 °C. Surface-initiated polymethylenation (SIPM) presents rapid growth in relation to other surface-initiated reactions, producing PM films thicker than 500 nm after 2 min of reaction and 3 μm after 24 h of reaction. AFM and SEM images show the presence of micro- and nanoscale features that enable the entrapment of air when exposed to water. Consistent with this result, these films exhibit advancing water contact angles greater than 160°, dramatically different than 103° measured for smooth PM films, and hysteresis values ranging from 2° to 40°, depending on the substrate and polymerization time. The superhydrophobic character of the films results in the entrapment of air at the polymer/solution interface to provide remarkable resistances greater than 10¹⁰ ω·cm ² against the transport of aqueous redox probes and cause the film to behave as a perfect capacitor.

Full text of DOI:10.1021/ja9086193

Published on Web 04/01/2010
Surface-Initiated Polymerization of Superhydrophobic
Polymethylene
Juan C. Tuberquia, Nabijan Nizamidin, Robert R. Harl, Jake Albert, Jason Hunter,
Bridget R. Rogers, and G. Kane Jennings*
Department of Chemical and Biomolecular Engineering, Vanderbilt UniVersity,
NashVille, Tennessee 37235
Received October 9, 2009; E-mail: kane.g.jennings@vanderbilt.edu
Abstract: We report a new surface-initiated polymerization strategy that yields superhydrophobic polym-
ethylene (PM) films from initially smooth substrates of gold and silicon. The films are prepared by assembling
a vinyl-terminated self-assembled monolayer, followed by exposure of the surface to a 0.1 M solution of
borane, and polymerizing from the borane sites upon exposure to a solution of diazomethane at -17 °C.
Surface-initiated polymethylenation (SIPM) presents rapid growth in relation to other surface-initiated
reactions, producing PM films thicker than 500 nm after 2 min of reaction and 3 µm after 24 h of reaction.
AFM and SEM images show the presence of micro- and nanoscale features that enable the entrapment of
air when exposed to water. Consistent with this result, these films exhibit advancing water contact angles
greater than 160°, dramatically different than 103° measured for smooth PM films, and hysteresis values
ranging from 2° to 40°, depending on the substrate and polymerization time. The superhydrophobic character
of the films results in the entrapment of air at the polymer/solution interface to provide remarkable resistances
greater than 1010 Ω·cm2 against the transport of aqueous redox probes and cause the film to behave as
a “perfect” capacitor.
monolayer film to graft the initiator.12 A completely different
SIP approach takes advantage of the catalytic properties of some
Introduction
Many innovations in polymer synthesis have been extended
to surfaces to create versatile methods for tailoring composition
(homopolymers, copolymers),2,3 modifying architecture (func-
tionalization, hyperbranching),3,4 and optimizing properties
(thickness, density, morphology)5 of surface-bound polymer
films. Some examples of this strategy include the adaptation of
solution-phase free radical polymerization,6,7 ATRP,8 ROMP,9
anionic,10 and cationic11 reactions into surface-initiated poly-
merizations. In surface-initiated polymerization (SIP), or “graft-
ing from” methods,12 an initiator is first immobilized on a
surface followed by exposure of the activated surface to an
appropriate monomer solution. Activation of the surface is
typically attained by self-assembling adsorbates that have the
initiator incorporated as the terminus or by modifying a
metal surfaces, since they induce spontaneous polymerization
upon exposure to reactive precursors. As an example of this
latter method, diazomethane (DM) spontaneously decomposes
at gold surfaces to produce linear polymethylene (PM).13,14
Advantages of using SIP include the preparation of robust,
bound coatings over objects of any shape with tailored structure,
thicknesses,12 grafting densities,12,15 and compositions, including
random and block copolymer films.12,15,16 Here, we adapt the
synthetic approach of polyhomologation, developed by Shea
and co-workers,17-20 to prepare surface-initiated polymethylene
films with thicknesses exceeding 1 µm on silicon and gold
supports.
In their approach, Shea et al.17-20 react a trialkyl borane
initiator in solution with an appropriate ylide monomer, such
as dimethylsulfoxonium methylide (CH2S(O)Me2), to produce
end-functionalized polymethylene (R(CH2)nX where R and X
(1) Lee, J. K.; Lee, K. B.; Kim, D. J.; Choi, I. S. Langmuir 2003, 19,
8141–8143.
(2) Bai, D.; Jennings, G. K. J. Am. Chem. Soc. 2005, 127, 3048–3056.
(3) Bai, D. S.; Ibrahim, Z.; Jennings, G. K. J. Phys. Chem. C 2007, 111,
461–466.
(13) Guo, W. F.; Jennings, G. K. Langmuir 2002, 18, 3123–3126.
(14) Seshadri, K.; Atre, S. V.; Tao, Y. T.; Lee, M. T.; Allara, D. L. J. Am.
Chem. Soc. 1997, 119, 4698–4711.
(4) Zhao, M. Q.; Zhou, Y. F.; Bruening, M. L.; Bergbreiter, D. E.; Crooks,
R. M. Langmuir 1997, 13, 1388–1391.
(15) Ruckenstein, E.; Li, Z. F. AdV. Colloid Interface Sci. 2005, 113, 43–
63.
(5) Huang, W. X.; Kim, J. B.; Bruening, M. L.; Baker, G. L. Macromol-
ecules 2002, 35, 1175–1179.
(16) Edmondson, S.; Osborne, V. L.; Huck, W. T. S. Chem. Soc. ReV. 2004,
33, 14–22.
(6) Prucker, O.; Ruhe, J. Macromolecules 1998, 31, 602–613.
(7) Prucker, O.; Ruhe, J. Macromolecules 1998, 31, 592–601.
(8) Boyes, S. G.; Brittain, W. J.; Weng, X.; Cheng, S. Z. D. Macromol-
ecules 2002, 35, 4960–4967.
(17) Shea, K. J.; Staiger, C. L.; Lee, S. Y. Macromolecules 1999, 32, 3157–
3158.
(18) Shea, K. J.; Walker, J. W.; Zhu, H.; Paz, M.; Greaves, J. J. Am. Chem.
Soc. 1997, 119, 9049–9050.
(9) Weck, M.; Jackiw, J. J.; Rossi, R. R.; Weiss, P. S.; Grubbs, R. H.
J. Am. Chem. Soc. 1999, 121, 4088–4089.
(19) Busch, B. B.; Paz, M. M.; Shea, K. J.; Staiger, C. L.; Stoddard, J. M.;
Walker, J. R.; Zhou, X. Z.; Zhu, H. D. J. Am. Chem. Soc. 2002, 124,
3636–3646.
(10) Jordan, R.; Ulman, A.; Kang, J. F.; Rafailovich, M. H.; Sokolov, J.
J. Am. Chem. Soc. 1999, 121, 1016–1022.
(11) Jordan, R.; Ulman, A. J. Am. Chem. Soc. 1998, 120, 243–247.
(12) Jennings, G. K.; Brantley, E. L. AdV. Mater. 2004, 16, 1983–1994.
(20) Busch, B. B.; Staiger, C. L.; Stoddard, J. M.; Shea, K. J. Macromol-
ecules 2002, 35, 8330–8337.
9
10.1021/ja9086193  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 5725–5734 5725

A R T I C L E S
Tuberquia et al.
Scheme 1. Mechanism of Reaction between a Trialkylborane and
Dimethylsulfoxonium Methylide to Prepare Polymethylene Chains
Scheme 2. Strategy for the Polymethylenation of Diazomethane at
Heterogeneous Borane Sites on Gold with and without a
Monolayer of Silver, and Silicon Surfacesa
are various terminal groups). Scheme 1 shows the polyhomolo-
gation mechanism. In this mechanism the three alkyl groups of
the initiator have the same probability to undergo a 1,2
migration, resulting in a low polydispersity product that reacts
with additional ylide to extend the chain length.17 Oxidation
with basic peroxide18 or trimethylamine N-oxide dehydrate21
produces an alcohol-terminated polymethylene that can be
converted to various functional groups (X).18 The terminal
R-group at the opposite end of the chain is introduced by
reacting an R-terminated olefin with diborane (B2H6) prior to
reaction with the ylide in Scheme 1. The primary advantages
of polyhomologation18,20 are the low polydispersity of the
polymer (1.04-1.17), the rapid kinetics of polymerization, and
the tight control over the molecular weight by varying the ratio
of ylide to trialkylborane.
a A vinyl-terminated monolayer is exposed to a 0.1 M solution of borane
in THF, followed by exposure to a DM solution in ether at -17 °C. The
silver layer on gold prevents the spontaneous polymerization of DM into
PM that is observed on gold.
monolayer on the gold surface to greatly inhibit the spontaneous
growth of PM films.13,25
The spontaneous growth of ultrathin PM films on gold has
been the subject of previous studies2,14,25 that take advantage
of the low cost, chemically inert character, and low dielectric
constant of PM26 to prepare ultrathin polymer films and
overcome limitations of PM processability due to poor solubility
in organic solvents27 and poor adhesion to metal substrates28
that limit spin coating and solution casting. Seshadri et al.14
grew ∼50 nm PM films by exposing gold surfaces to a DM
solution in ether. Our group13,25 developed approaches to
enhance or prevent PM growth by atomically modifying gold
surfaces with copper or silver, respectively. Different surfaces29-31
also favor DM decomposition to form PM. However, polym-
erization by these routes relies on the inherent catalytic
properties of the substrate. Using SIPM, we intend to expand
the initiation capabilities of boranes to noncatalytic surfaces,
thus diversifying the number of substrates from which PM films
can be grown.
Here, we develop a surface-initiated polymethylenation
(SIPM) to grow polymethylene films from borane-modified gold
or silicon surfaces by exposure to dilute diazomethane in ether
at -17 °C. Scheme 2 presents our strategy for polymerizing
over immobilized monoalkylborane sites on gold and silicon
substrates. First, we prepare a vinyl-terminated monolayer on
the substrate.22 Then, the initiator is immobilized on the surface
by hydroborating the vinyl termini in a solution of borane (BH3)
in tetrahydrofuran (THF) under inert atmosphere. Immersion
of the functionalized surface into a dilute DM solution enables
polymerization from the borane centers, yielding a PM film.23,24
Exposure of unmodified gold surfaces to DM also produces a
PM film due to spontaneous polymerization at gold surface
sites.14 To isolate the effect of the monoalkylborane initiation
sites on the gold surface in our process, we predeposit a silver
(25) Guo, W. F.; Jennings, G. K. AdV. Mater. 2003, 15, 588–591.
(26) James, M. Polymer Data Handbook; Oxford University Press: New
York, 1999.
(21) Wagner, C. E.; Rodriguez, A. A.; Shea, K. J. Macromolecules 2005,
38, 7286–7291.
(27) Brandrup, J.; Immergut, E. H.; Grulke, E. A.; Abe, A.; Bloch, D. R.
In Polymer Handbook; Wiley & Sons: New York, 1999.
(28) Kong, J.-S.; Lee, D.-J.; Kim, H.-D. J. Appl. Polym. Sci. 2001, 82,
1677–1690.
(22) Peanasky, J. S.; McCarley, R. L. Langmuir 1998, 14, 113–123.
(23) Davies, A. G.; Hare, D. G.; Khan, O. R.; Sikora, J. J. Am. Chem. Soc.
1963, 4461–4471.
(24) Davies, A. G.; Hare, D. G.; Khan, O. R.; Sikora, J. Proc. Chem. Soc.
1961, 172.
(29) Lie, L. H.; Patole, S. N.; Hart, E. R.; Houlton, A.; Horrocks, B. R. J.
Phys. Chem. B 2002, 106, 113–120.
9
5726 J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010

Surface-Initiated Polymerization of Polymethylene
A R T I C L E S
In this article, we report the first surface-initiated polymeth-
ylenation for the growth of PM films from gold and silicon
surfaces. This surface reaction is rapid in comparison to many
surface-initiated polymerizations, producing PM films thicker
than 500 nm after 2 min of reaction and 3 µm after 24 h of
reaction. Extremely high contact angles (>160°) and low
hystereses (∼2°) indicate that PM films exhibit superhydro-
phobic behavior due to their low surface energy and their
combination of micrometer- and nanoscale morphology. Thus,
our approach modifies a smooth substrate into a rough,
superhydrophobic surface through a single polymerization step.
Due to the presence of entrapped air at the water/polymer
interface, these superhydrophobic films present large resistances
on the order of 1 × 1010 Ω·cm2 against the transport of aqueous
redox probes.
Results and Discussion
Characterization of Surface Borane. Hydroboration of olefins
in solution results in trialkylboranes, except in the case of
sterically hindered olefins, in which hydroboration yields
dialkylboranes or monoalkylboranes.32,33 Here, we exposed a
dense vinyl-terminated monolayer to a solution of borane in
THF. The immobilization of the chains within the monolayer
prevents the formation of trialkylboranes, and further, the
formation of dialkylboranes that would require chain termini
to be within ∼3 Å based on the C-B bond length34 and the
geometry of the binding is unlikely on average, given the ∼5
Å center-to-center distance between chains in an alkanethiolate
monolayer.35 Therefore, we deduce that the exposure of a vinyl-
terminated monolayer to borane results predominately in a
monoalkylborane and a -BH2 terminus. However, characteriza-
tion of the monoalkylborane self-assembled monolayer (SAM)
is challenging because our primary surface characterization
methods (PM-IRRAS, XPS, etc.) require the surfaces to be
exposed to air, and alkyl boranes are known to react rapidly
with oxygen in the air to yield a variety of oxidized species.36-38
Our approach was to preoxidize39 the surface prior to charac-
terization. Therein, we exposed a freshly hydroborated vinyl-
terminated monolayer on gold to O2-saturated THF for 1 h and
water for 30 min (denoted as HBO) before exposure to air for
subsequent characterization with PM-IRRAS and XPS.
Figure 1. (a) PM-IRRAS spectrum of a hydroborated SAM after its
oxidation with pure oxygen (HBO) for 1 h and immersion in water for 30
min. For comparison a spectrum for a vinyl-terminated SAM exposed to
the same conditions (CdC/Au) is also displayed. Modulation was focused
in the lower (1129 cm-1) region of the spectra. (b) XPS spectra for a
hydroborated (HBO) and a vinyl-terminated (CdC/Au) SAM after exposure
to oxidation conditions. Spectrum of a hydroxyl-terminated monolayer (OH/
Au) is presented for comparison. The suvey XPS spectra (not shown)
confirm the presence of Au, C, O, and S on all surfaces except for the
absence of O on the vinyl-terminated monolayer. Spectra have been offset
for clarity.
CdC modes, consistent with hydroboration of the vinyl
functionality at the surface of the monolayer, and the appearance
of a peak at 1264 cm-1 that we assign to symmetric B-O
stretching,42,43 suggesting the subsequent oxidation of B-H2
bonds to B-(OH)2 (boronic acid).32,34,38,39 Orientation of the
B-OH groups at the outer surface would result in a dominant
symmetric B-O stretch, based on surface selection rules,44 as
compared to the more commonly observed asymmetric B-O
stretch,45,46 which we do not observe here. Higher reactivity of
the H-B bond toward O2 as compared to the B-C bond could
result from the confinement of the latter within the SAM, which
lowers its exposure and increases the steric hindrance toward
oxidation.47,48
Figure 1 shows the IR and the XPS spectra of an HBO surface
and a vinyl-terminated SAM (CdC/Au) that were each exposed
to the same oxidizing conditions. The IR spectrum of the vinyl-
terminated SAM displays the C-H out-of-plane deformation
(912 cm-1) and the out-of-plane deformation of the CdC (993
cm-1) groups characteristic of the vinylic bond.22,40,41 The
spectrum for the HBO monolayer shows a disappearance of the
(30) Kubota, H.; Morawetz, H. J. Polym. Sci. Part A-1: Polym. Chem. 1967,
5, 585–591.
(31) Bart, J. C.; Cariati, F.; Erre, L.; Gessa, C.; Micera, G.; Piu, P. Clays
Clay Miner. 1979, 27, 429–432.
(32) Brown, H. Boranes in Organic Chemistry; Cornell University Press:
Ithaca, 1972.
(33) Brown, H. C.; Rao, B. C. S. J. Am. Chem. Soc. 1959, 81, 6428–6434.
(34) Onak, T. Organoborane Chemistry; Academic Press: New York, 1975.
(35) Ulman, A. Chem. ReV. 1996, 96, 1533–1554.
(36) Mirviss, S. B. J. Am. Chem. Soc. 1961, 83, 3051–3056.
(37) Mirviss, S. B. J. Org. Chem. 1967, 32, 1713.
(38) Davies, A. G. J. Chem. Res.-Synop. 2008, 361–375.
(39) Brown, H. C.; Midland, M. M. Tetrahedron 1987, 43, 4059–4070.
(40) Snyder, R. G.; Strauss, H. L.; Elliger, C. A. J. Phys. Chem. 1982, 86,
5145–5150.
(42) Shurvell, H. F.; Faniran, J. A. Can. J. Chem. 1968, 46, 2082.
(43) Faniran, J. A.; Shurvell, H. F. Can. J. Chem. 1968, 46, 2089.
(44) Bradshaw, A. M.; Richardson, N. V. Pure Appl. Chem. 1996, 68, 457–
467.
(45) Barriet, D.; Yam, C. M.; Shmakova, O. E.; Jamison, A. C.; Lee, T. R.
Langmuir 2007, 23, 8866–8875.
(46) Brewer, S. H.; Allen, A. M.; Lappi, S. E.; Chasse, T. L.; Briggman,
K. A.; Gorman, C. B.; Franzen, S. Langmuir 2004, 20, 5512–5520.
(47) Brown, H. C.; Dodson, V. H. J. Am. Chem. Soc. 1957, 79, 2302–
2306.
(41) Sulc, R.; Zhou, X. Z.; Shea, K. J. Macromolecules 2006, 39, 4948–
4952.
9
J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010 5727

A R T I C L E S
Tuberquia et al.
Table 1. Thickness Measurements for PM Films Grown from the
Indicated Surfaces upon Exposure to a 13 mM Solution of DM in
Ether at -17 °C for 24 h
To provide an estimate of boron surface coverage on the HBO
monolayer, we used X-ray photoelectron spectroscopy (XPS).
Figure 1b shows the C(1s) and O(1s) high-resolution spectra
for the HBO monolayer, the vinyl-terminated monolayer, and
a hydroxyl-terminated monolayer (OH/Au) prepared from
HS(CH2)11OH on Au. These spectra show that all monolayers
exhibit the characteristic C(1s) peak for aliphatic hydrocarbons
at 284.7 eV. The spectrum for the hydroxyl-terminated mono-
layer exhibits a shoulder at 286.8 eV corresponding to the
outermost carbon that is covalently bound to the oxygen of the
hydroxyl group.49 The atomic concentration ratio of these two
peaks is 10:1, consistent with the chain length of the hydroxyl
monolayer. Appearance of a single and sharp peak for the HBO
monolayer indicates the absence of the C-O bond, suggesting
that the oxygen present in the system is not bonded to carbon,
but most likely to boron. Direct detection of boron was not
observed due to its low atomic concentration (e1 B:11 C) and
small sensitivity factor, ∼20% of that for oxygen.50 Instead,
we used the presence of oxygen in the spectra as indirect
evidence of an oxidized boron functionality. Importantly, high-
resolution spectra of the O(1s) region indicate the presence of
oxygen on the HBO surface but not on the vinyl-terminated
control. On the basis of a comparison of the atomic ratios of O
to C in both the HBO and OH-terminated monolayers, we
estimate that the oxygen level on the HBO surface is 70 ( 9%
of that on a dense HO-terminated SAM. Since boronic acids
dehydrate in vacuum to form boroxine glasses with a stoichi-
ometry of one oxygen per boron,51 we assume that our oxidized
species has one oxygen per boron per chain, yielding an estimate
of boron surface coverage of 70 ( 9% from XPS. As further
evidence of the presence of boron, exposure of the HBO surface
on silicon to diazomethane in ether results in a PM film (vide
infra) that is ∼35% as thick as a film grown from an unoxidized
HB monolayer, consistent with the presence of a boron
functionality within the monolayer after oxidation since a
nonhydroborated vinyl monolayer on silicon yields no polymer.
On silicon substrates, the abundance of oxygen in the native
oxide layer hindered our ability to use XPS to detect oxygen
levels in the monolayer as an estimate of boron coverage.
Instead, we incorporated the approach of Wasserman et al.52 to
completely oxidize the surface boranes to hydroxyl groups by
exposure to a 0.2 N solution of NaOH in H2O2 (30%) for 12 h
under inert atmosphere.53 Wasserman et al.52 used this approach
to estimate the conversion of vinyl groups to a hydroxyl
functionality. We observed that this harsh oxidation destroys
thiolate SAMs but does not damage the more robust siloxane-
linked monolayers on silicon. Exposure of the H2B-termini to
the oxidizing conditions resulted in an advancing water contact
angle (θHB) of 53°, whereas a vinyl-terminated control was not
affected by the conditions as the contact angle (θCdC) remained
95°. Calculations of yield were performed using the Cassie
equation,52
substrate
ellipsometry, nm
QCM, nm
profilometry, nm
HB+Au
3608
238
3006
3500 ( 200
344 ( 15
SAM+Au
HB+Ag/Au
SAM+Ag/Au
Ag/Au
363 ( 12
1 ( 0.1
6 ( 3
224 ( 1
Au
HB+Si
SAM+Si
182
277 ( 5
1667 ( 190
1 ( 0.2
where φCdC and φOH are the surface area fractions of the vinyl-
and hydroxyl-terminations in contact with water, respectively.
Assuming θOH ≈ 0°, the final monolayer exhibits a higher
surface energy as compared to the initial vinyl-terminated
monolayer and indicates that ∼35% of the surface area presents
the vinyl termini (φCdC) and 65% represents hydroxyl termini
(φOH) from the direct conversion of surface boranes. This result
on silicon is in good agreement with XPS results on gold and
suggests that after hydroboration ∼65% of the vinyl-termini are
converted to H2B-terminations through the hydroboration step.
Effect of Substrate. To investigate the effect of hydroboration
on the surface-initiated growth of PM, we measured the
thicknesses of PM films grown from hydroborated surfaces and
control substrates upon exposure to a 13 mM solution of DM
in ether at -17 °C for 24 h. Table 1 shows the results of the
thickness measurements for PM films grown from six surfaces,
including two hydroborated SAMs, one on gold (HB+Au) and
another on Ag/Au UPD (HB+Ag/Au); two vinyl-terminated
SAMs not exposed to borane (SAM+Au, SAM+Ag/Au), as
well as bare Au and Ag/Au surfaces. In principle, ellipsometric
measurements were performed on all the films, assuming that
they are uniform and isotropic and have the same refractive
index as polyethylene (n ) 1.5).27 All the control films could
be fit with a Cauchy model to determine their thickness, but
PM films grown from hydroborated SAMs could not be fit with
this simple model, perhaps due to their rough morphology (vide
infra). To assess the rough morphology of the surface, we used
an effective medium approximation54 model that combined the
optical constants for PM and air. However, this approach did
not provide a good fitting of the experimental data. This
characteristic behavior suggests that PM films grown from SIPM
are distinct in comparison to the control films. To provide a
measure of thickness for these rougher PM films, we performed
profilometry and ex situ quartz crystal microgravimetry (QCM)
measurements.14
To verify the validity of profilometry and QCM, we first
compared the results for a PM film on Au with those from
ellipsometry to determine if the techniques are in relative
agreement. Comparison of the thickness values for PM growth
on Au and SAM+Au surfaces obtained by QCM, ellipsometry,
and profilometry shows that QCM-estimated thicknesses are
∼25% lower than the average of the three techniques. However,
the fact that, in all cases, PM films grown from SAM+Au
present much higher thicknesses as compared to those grown
cos θHB ) φCdC cos θCdC + φOH cos θOH
(1)
(48) Mikhailov, B. M.; Bubnov, Y. Organoboron Compounds in Organic
Synthesis; Harwood Academic: New York, 1984.
(49) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides, G. M.;
Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321–335.
(50) Ulvac-PHI, Inc. Operator’s PHI Multipak software manual, version
8.2C; Ulvac-PHI, Inc.: Chigasaki City, Kanagawa Prefecture, Japan,
2007.
(52) Wasserman, S. R.; Tao, Y. T.; Whitesides, G. M. Langmuir 1989, 5,
1074–1087.
(53) Faucheux, N.; Schweiss, R.; Lutzow, K.; Werner, C.; Groth, T.
Biomaterials 2004, 25, 2721–2730.
(54) Guide to using WVASE32 Software for spectroscopic ellipsometry
data acquisition and analysis, Version 3.335; J.A. Woollam Co., Inc.,
2001.
(51) Carey, R. I.; Folkers, J. P.; Whitesides, G. M. Langmuir 1994, 10,
2228–2234.
9
5728 J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010

Surface-Initiated Polymerization of Polymethylene
A R T I C L E S
from Au indicates that the techniques certainly agree with
respect to the overall trend.
Table 1 shows that SIPM allows the growth of films that are
10- to 1000-fold thicker than films grown from non-hydrobo-
rated substrates. In fact, the ∼3 µm thick films produced by
SIPM are among the thickest films yet reported by surface-
initiated routes. From the evaluation of the non-hydroborated
surfaces, we notice that silver deposition on gold surfaces (Ag/
Au and SAM+Ag/Au) inhibits spontaneous PM formation, as
previously described by our group.13,25 This inhibition indicates
that the PM film grown on the HB+Ag/Au surface results from
SIPM and not from spontaneous polymerization. In the case of
the HB+Au surface, SIPM produces a film about 10 times
thicker than the thickest non-hydroborated PM film (SAM+Au),
supporting the dominance of SIPM toward film growth when
both SIPM and spontaneous14 modes are present.
Table 1 also shows the thicknesses obtained after exposure
of DM at the above conditions to a silicon substrate modified
by 5-hexenyltrichlorosilane (SAM+Si) and one that was further
modified by borane (HB+Si). While PM does not grow on
SAM-modified silicon, as evidenced by a thickness that is
approximately that of the SAM, it does grow from HB+Si to
achieve a thickness that is 55% of that for SIPM from HB+Ag/
Au after 24 h of polymerization. We did observe that HB+Si
requires longer times of polymerization to obtain comparable
thicknesses to those obtained for shorter times on HB+Ag/Au
and HB+Au. The slower polymerization on silicon could result
from a greater extent of oxidation of this monolayer during
transfer of the substrate into the DM solution for polymerization.
As the 6-carbon vinyl-terminated monolayer on silicon is thinner
and less crystalline than the 11-carbon monolayer on Au and
Ag/Au,55 the C-B bond within the monolayer on silicon is more
exposed and, thus, more susceptible to attack by oxygen.
Insertion of oxygen into the C-B bond would result in a lower
number of C-B bonds that enable the polymerization. Ap-
preciable differences in the reactivity of various boranes toward
DM was reported by Davies et al.23 in which the polymerization
of PM by alkylboranes is faster as compared to that of oxidized
states: R3B > (RBO)3 > (R2B)2O > RB(OR)2 > (RO)3B.
Nonetheless, growth of PM films from HB+Si demonstrates
the ability to extend SIPM to other vinyl-terminated substrates.
For PM films grown from HB+Ag/Au and HB+Au surfaces,
discrepancies in their thicknesses result from the catalytic
activity of Au toward DM, resulting in slightly thicker films.
Effect of Concentration and Time. Figure 2 shows the effect
of the time of polymerization and the concentration of DM on
the PM film thickness on HB+Ag/Au substrates. Figure 2a
shows the variation of film thickness measured by profilometry
at various times of polymerization after exposure to a 15 mM
solution of DM at -17 °C. The films achieve thicknesses of
∼1 µm after 10 min of reaction, presenting rapid kinetics as
compared to other SI routes.13,14 The growth rate slows as the
reaction proceeds, suggesting occlusion of monomers to the
active borane sites. Figure 2b shows the effect of DM
concentration on the growth of PM films after 24 h of exposure
to DM solutions with concentrations between 0.2 and 15 mM
at -17 °C. Film thickness increases sharply with concentration
from 0 to 1 mM with a moderate dependence for higher
concentrations. Figure 2b shows that films with thicknesses
greater than 1 µm can be grown in 24 h if the concentration of
DM is greater than 1 mM.
Figure 2. Thickness of PM films grown from HB+Ag/Au surfaces as
affected by (a) time of polymerization after exposure to a 15 mM solution
of DM at -17 °C and (b) DM concentration (0.2 to 15 mM) after 24 h of
reaction at -17 °C. The data points and error bars represent the averages
and standard deviations, respectively, of measurements obtained on two
samples prepared independently.
Figure 3. RAIR spectrum for a PM film after 1 h of polymerization in a
8.5 mM solution of DM in ether for 1 h.
Film Composition and Structure. We used RAIRS to assess
the composition and structure of polymer films grown by SIPM
and enable comparison with PM films formed by spontaneous
polymerization (controls) that have been characterized previ-
ously.14,25 We first exposed a HB+Au surface to an 8.5 mM
solution of DM in ether for 1 h. RAIR spectra of the film (Figure
3) show the relevant regions of PM: CH2 stretching40,41 (νas
)
2918 cm-1, νs ) 2850 cm-1), CH2 bending13,14,56 (νas ) 1473
cm-1, νs ) 1463 cm-1), and CH2 rocking modes13,14,56,57 (split:
νas ) 730 cm-1 and νs ) 720 cm-1). The positions of these
modes are consistent with crystalline PM chains, and the absence
of methyl modes in the IR spectra at 2960 and 2880 cm-1
indicates the linear character of the PM chains.2
To further evaluate the structure of PM films, we exposed
the same six substrates listed in Table 1 to a 13 mM solution
of DM in ether for 24 h and collected RAIR spectra. Figure 4
shows the relevant regions of the RAIR spectra for CH2
bending13,14,56 and CH2 rocking modes.13,14,56,57 Comparison
(56) Colthup, N.; Daly, L.; Wiberley, S. Introduction to Infrared and Raman
Spectroscopy, 3rd ed.; Academic Press: Boston, 1990.
(57) Snyder, R. G. J. Chem. Phys. 1979, 71, 3229–3235.
(55) Brunner, H.; Vallant, T.; Mayer, U.; Hoffmann, H. Surf. Sci. 1996,
368, 279–291.
9
J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010 5729

A R T I C L E S
Tuberquia et al.
Figure 5. Scanning electron microscopy of PM films grown from HB+Si
(a, b), HB+Ag/Au (c), and HB+Au (d) upon exposure to a 13 mM solution
of DM in ether at -17 °C for 24 h.
HB+Ag/Au surfaces after 24 h of reaction in a 13 mM solution
of DM in ether at -17 °C. Figure 5a presents a PM film grown
from HB+Si showing nanometer-scale ridges and valleys and
a micrometer-scale cluster that is characteristic of PM films
grown from SIPM. Comparison of PM films grown from
HB+Si, HB+Ag/Au, and HB+Au surfaces (Figure 5b, c, and
d) shows that the main topographical difference of the three
substrates is related to the concentration of the micrometer-
scale clusters, with films grown from HB+Si providing the least
concentrated clusters on the surface and those on HB+Au the
most concentrated clusters. To confirm that PM clusters result
from the SIPM, we compared SEM images for films with no
additional treatment after extraction from the DM solution and
films that were vigorously rinsed with trichlorobenzene and
ethanol after polymerization. Rinsing the films removed some
large agglomerates (>5 µm in diameter), but the ∼1 µm PM
clusters remained firmly attached to the surface. These clusters
survive AFM scanning (see Figure 6), and their mountain-like
morphology (Figure 5a or 6a) suggests they are produced from
the surface-initiated polymerization.
Figure 4. RAIR spectra for PM films grown by exposure to a 13 mM
solution of DM in ether for 24 h. (a) CH bending region of PM films grown
from the same six substrates listed in Table 1, with four of the spectra
multiplied by 10 to enable comparison with spectra for thicker PM films
grown from HB+Au and HB+Ag/Au. (b) CH rocking region for PM films
grown from Ag/Au and Au surfaces. The spectra have been offset vertically
for clarity.
of the RAIR peak absorbance (Figure 4a) for the various surfaces
supports thickness measurements and shows that SIPM enables
the formation of PM films. Figure 4 also provides information
on the structure and conformation of PM films on HB+Au and
HB+Ag/Au surfaces. Figure 4a shows the CH bending region
for PM films on HB+Ag/Au and HB+Au. The split character
of the peak into asymmetric (νas ) 1473 cm-1) and symmetric
(νs ) 1463 cm-1) modes indicates that there are two types of
chain configurations in the unit subcell,58 a prime characteristic
of orthorhombic chain packing. In agreement with the orthor-
hombic conformation, the CH rocking mode also splits into in-
phase and out-of-phase modes14 at 731 and 719 cm-1 (Figure
4b), respectively. If we assume that the orthorhombic conforma-
tion is the only crystalline structure present and the rest of the
chains have no conformational order,14 we can estimate an
orthorhombic crystalline content of ∼70% for the HB+Au and
HB+Au/Ag surfaces. This calculation uses the theoretical
consideration for paraffins where the ratio of the rocking
intensities I720/I730 ) 1.233.57 Spectra (not shown) obtained after
exposure of a HB+Si surface to a 13 mM solution of DM in
ether at -17 °C for 24 h are similar to those shown in Figure
4.
Finally, to more accurately determine the dimensions of the
topographical features of PM films via SIPM, Figure 6 presents
an atomic force microscopy (AFM) 3D representation of the
HB+Au surface after 24 h of reaction in a 13 mM solution of
DM in ether at -17 °C. Figure 6a shows that the PM film
contains microscale features as high as ∼1 µm rising above a
base with nanoscale features. For the latter type of features,
Figure 6b presents a clearer representation to verify that the
differences in height are below 200 nm. Both SEM and AFM
images of the PM films show that its topography is rough on
multiple scales, and as shown below, the rich topographic
characteristics of the surface yield extreme behavior when the
surface is in contact with water.
Surface Wettability. Table 2 shows the advancing (θA) and
receding (θR) water contact angles for the same PM films grown
for 24 h on HB+Au, HB+Ag/Au, and HB+Si and used in SEM
characterizations. PM films on all substrates present advancing
contact angles greater than 160° and differences between θA
and θR, known as hysteresis, of 2°, 13°, and 39° for the HB+Au,
HB+Ag/Au, and HB+Si surfaces, respectively. In the case of
PM films grown from HB+Au, the film presents the charac-
Surface Topography. Figure 5 shows scanning electron
microscopy (SEM) images to reveal the distinctive topographical
features for PM films grown from HB+Si, HB+Au, and
(58) Snyder, R. G. J. Mol. Spectrosc. 1961, 7, 116–144.
teristics of a superhydrophobic film (θA > 150°, θA - θR <
9
5730 J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010

Surface-Initiated Polymerization of Polymethylene
A R T I C L E S
comparedtosmoothPMsurfaces(∼103°).14 Severalauthors59,61,62
have shown that the film topography is responsible for the
elevation of contact angles and the diminution of hysteresis,
and surface roughness has been directly or indirectly included
in the Wenzel and the Cassie contact angle models. In the
Wenzel model63-65 the water is assumed to wet the entire solid
surface area, and the film topography of the surface is accounted
by defining the roughness ratio (r),59,63
actual area wetting the solid
r )
(2)
geometrical area
The predicted or Wenzel contact angle (θW) can be related to
the contact angle on a smooth PM surface (θPM) by
cos θW ) r cos θPM
(3)
In the Cassie model,59,66-68 the water is assumed to not
completely wet the solid surface area, enabling the presence of
entrapped air in the grooves under the drop. Therefore, the
composite interface is comprised of liquid-solid, liquid-air,
and solid-air interfaces. In the Cassie model, the resulting
contact angle (θC) is estimated as a contribution between the
contact angle of the drop on a flat PM surface (θPM ) 103°)
and the contact angle of the liquid in air (θAir ) 180°).
cos θC ) φPM(cos θPM + 1) - 1
φPM + φAir ) 1
(4)
(5)
where φPM and φAir are the fractions of the projected area of
the surface for polymer and air, respectively. The Cassie and
Wenzel models are conceived as different wetting regimes.64,65
Bico et al. proposed that both states are delimited by a critical
roughness (rc), in which surfaces with roughness ratios lower
than rc favor the formation of drops in the Wenzel state.64
φPM - 1
cos θPM
rc )
+ φPM
(6)
Figure 6. Tapping mode AFM images that show (a) microscale and (b)
nanoscale features of PM films prepared by exposure of HB+Au to a 13
mM solution of DM in ether at -17 °C for 24 h. (b) Image acquired using
a higher definition scan within the area of (a). Line scans represent the
topographical features for a diagonal line across the image from the bottom
left to the top right in the image.
To analyze the wetting behavior of PM films grown by SIPM,
we classify them depending on whether they follow the Cassie
or the Wenzel state. Films in the Cassie state exhibit large
contact angles and low hystereses,59,60 which is the case for
contact angles on PM films grown from HB+Au and HB+Ag/
Au surfaces. For PM films grown from HB+Si, we observe
that θA is similar to contact angles for PM films grown from
HB+Au and HB+Ag/Au surfaces, and hence, the water drops
advance in the Cassie state for all three films. For the case of
θR for PM films grown from HB+Si, we first use eq 6 to
calculate an rc value of 2.9, assuming that φPM , 1.60 Then,
we calculate the roughness ratio of the surface as 2.3 with eq
3, which, as compared69 to rc, suggests that after advancing in
the Cassie state the drop collapses into the Wenzel state upon
receding. After using eq 4 and the θR for PM films grown from
Table 2. Water Contact Angles for PM Films Grown from HB+Au,
HB+Ag/Au, and HB+Si Surfaces after 24 h of Polymerization in a
13 mM Solution of DM in Ether at -17 °C for 24 h
substrate
θA (deg)
θR (deg)
HB+Au
HB+Au/Ag
HB+Si
167 ( 2
162 ( 3
160 ( 2
165 ( 1
149 ( 6
121 ( 3
10°)59,60 and enables impinging water drops to easily rebound
from the surface without suffering noticeable shape deformation.
HB+Ag/Au and HB+Si surfaces also present superhydrophobic
behavior after 72 and 120 h of reaction, respectively, consistent
with an increase in the number of microscale features as
polymerization proceeds.
(61) Nosonovsky, M. Langmuir 2007, 23, 3157–3161.
(62) Gao, L. C.; McCarthy, T. J. Langmuir 2006, 22, 2966–2967.
(63) Ulman, A. Characterization of Organic Thin Films; Butterworth-
Heinemann: Boston, 1994.
The behavior of the advancing contact angle (θA) for PM
films grown from SIPM represents a substantial difference as
(64) Bico, J.; Thiele, U.; Que´re´, D. Colloids Surf., A 2002, 206, 41–46.
(65) Marmur, A. Langmuir 2003, 19, 8343–8348.
(66) Gao, L. C.; McCarthy, T. J. Langmuir 2006, 22, 6234–6237.
(67) Roach, P.; Shirtcliffe, N. J.; Newton, M. I. Soft Matter 2008, 4, 224–
240.
(59) Zhu, L. B.; Xiu, Y. H.; Xu, J. W.; Tamirisa, P. A.; Hess, D. W.; Wong,
C. P. Langmuir 2005, 21, 11208–11212.
(68) Bhushan, B.; Jung, Y. C. Nanotechnology 2006, 17, 2758–2772.
(69) Bittoun, E.; Marmur, A. J. Adhes. Sci. Technol. 2009, 23, 401–411.
(60) Callies, M.; Quere, D. Soft Matter 2005, 1, 55–61.
9
J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010 5731

A R T I C L E S
Tuberquia et al.
Scheme 3. Equivalent Circuit Model Used to Represent
Electrochemical Impedance Spectra for PM Filmsa
a Components of the circuit: Rs: resistance of the solution, Ci: capacitance
of the interface, Ri: resistance of the interface, Cf: capacitance of the film,
Rf: resistance of the film.
HB+Au and HB+Ag/Au surfaces, we estimate that less than
5% and 19%, respectively, of the area of the drop at the interface
is in contact with PM.
Figure 7. Electrochemical impedance spectra for PM films grown by
exposure of various surfaces to a 13 mM solution of DM in ether at -17
°C for 24 h. Surfaces include two hydroborated SAMs, one on gold
(HB+Au) and another on Ag/Au UPD (HB+Ag/Au), two vinyl-terminated
SAMs not exposed to borane (SAM+Au, SAM+Ag/Au), as well as a Au
and a Ag/Au surface. Solutions of 1 mM K3Fe(CN)6 and 1 mM K4Fe(CN)6
were used as soluble probes in a 0.1 M NaSO4(aq) solution.
The surface topography of the films influences the behavior
of water droplets on the surface. Gao and McCarthy62 showed
that introduction of nanoscale roughness to an originally
microscale rough surface decreased the surface hysteresis from
20° to ∼0°. Here, we observe from SEM images that an increase
in the concentration of microscale features (HB+Au > HB+Ag/
Au > Si+Au) decreases hysteresis (HB+Au < HB+Ag/Au <
Si+Au), thus increasing the stability of the drop in the Cassie
state59 and enabling the drops to easily rebound from the
surface.70 PM films grown from SIPM exhibit superhydrophobic
behavior after times of polymerization that enable the growth
of a sufficient number of micrometer-scale features that provide
stability of the drops in the Cassie state, with times ranging
from 24, 72, and 120 h for HB+Au, HB+Ag/Au, and HB+Si
surfaces, respectively. In summary, these results suggest that
SIPM enables the preparation of PM films with high contact
angles (>150°) due to the pervasive presence of nanoscale ridges/
valleys and that an increase of the isolated microscale features
decreases hysteresis, enabling the preparation of superhydro-
phobic PM films.
Barrier Properties. The remarkable hydrophobic character
of SIPM films due to the presence of air at the polymer/water
interface opens the door to a diverse number of applications in
which minimizing the contact between water and a substrate is
critical. One of those applications is the protection of substrates
using polymer coatings where the protective coating serves as
a barrier to substances that can react and degrade the substrate
interface. To investigate the effect that trapped air at the
polymer/water interface has on the barrier properties against
redox probes, we characterized PM films with electrochemical
impedance spectroscopy (EIS) in the presence of 1 mM
K3Fe(CN)6, 1 mM K4Fe(CN)6 ·3H2O, and 0.1 M Na2SO4(aq).
EIS enables a comparison of the effective thickness and blocking
properties of SIPM films (HB+Ag/Au, HB+Au) in relation to
smoother PM films grown from non-polymethylenated surfaces
(Au, Ag/Au, SAM+Ag/Au, SAM+Au) after 24 h of exposure
to a 13 mM solution of DM at -17 °C. Comparison of the
barrier properties depends on the ability to fit any resultant
impedance spectrum to an equivalent circuit model composed
of capacitors and resistors. Scheme 3 presents the general circuit
model of a metal surface covered by a polymer film.3,71 In this
model, both the film and the organic/metal interface contribute
to the total impedance. This model may be simplified on the
basis of the relative values of each circuit component. For
example, if the resistance of the film (Rf) is much larger than
the interfacial impedance (defined as the parallel contribution
of the interfacial capacitance (Ci) and resistance (Ri)), the
spectrum will exhibit only a single time constant that is
representative of the film (Randles model). Figure 7 shows a
Bode plot of the experimental results with fits by a Randles71
equivalent circuit model as solid lines. Table 3 lists the
capacitance and resistance of the films that were quantified by
fits to experimental data.
EIS results show that, after exposure to DM, Ag/Au surfaces
(Ag/Au, SAM+Ag/Au) exhibit Warburg behavior (W) charac-
teristic of uncoated metallic surfaces, indicating that little if any
PM is grown on the surface.72 In the case of Au surfaces (Au,
SAM+Au), their exposure to DM enables the growth of PM
films having impedance spectra that follow the Randles model
and that exhibit well-defined resistances at low frequencies as
the redox probes begin to penetrate the film. In contrast, films
grown from the SIPM process show no signs of a resistive
plateau (Rf > 1 × 1010 Ω·cm2) at low frequencies but a linear
increase of the log impedance with decreasing log frequency,
with a slope of -1, reflecting ultrahigh impedance against the
transfer of redox probes. Variation in the phase angle (φ) in
the 10-1 to 105 frequency window for the same six substrates
reported in Figure 7 (Figure S1 in Supporting Information)
shows its balance between pure capacitive (φ ) -π/2) and pure
resistive (φ ) 0) behavior and the higher sensitivity of φ versus
|Z| to measure breakdowns in the protection by the film.73 The
results agree with those of Figure 7, and more importantly, the
fact that the superhydrophobic film grown from HB+Au
displays a purely capacitive behavior (φ ) -π/2) for all
frequencies indicates perfect barrier performance. The PM film
grown from HB+Ag/Au (θA/θR ) 162°/149°) shows a slight
(70) Chen, W.; Fadeev, A. Y.; Hsieh, M. C.; Oner, D.; Youngblood, J.;
McCarthy, T. J. Langmuir 1999, 15, 3395–3399.
(72) Brantley, E. L.; Jennings, G. K. Macromolecules 2004, 37, 1476–
1483.
(71) Brantley, E. L.; Holmes, T. C.; Jennings, G. K. J. Phys. Chem. B
2004, 108, 16077–16084.
(73) Bard, A.; Faulkner, L. Electrochemical Methods, 2nd ed.; John Wiley
and Sons: New York, 2001.
9
5732 J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010

Surface-Initiated Polymerization of Polymethylene
A R T I C L E S
Table 3. Film Resistance (Rf), Capacitance (Cf), and Thicknesses (d) for PM Films Grown from the Indicated Surfaces upon Exposure to a
13 mM Solution of DM in Ether at -17 °C for 24 h
Ag/Au
SAM+Ag/Au
Au
SAM+Au
HB+Ag/Au
HB+Au
d (nm)a
Rf (Ω)
Cf (F)
6
1
228
315
3006
3554
7.2 × 105
5.3 × 106
>1 × 1010
>1 × 1010
1.6 × 10-8
7.8 × 10-9
3.9 × 10-10
2.2 × 10-10
Ri (Ω)
1.7 × 102
1.1 × 10-5
3.6 × 10-3
3.0 × 102
3.3 × 10-6
3.1 × 10-3
Ci (F)
W (s1/2/Ω)
a Average thickness calculated from the values of Table 1.
the aqueous solution with reduced surface tension. This effect
on the wetting properties (θSDS ) 75 ( 5° static contact angle
of a ∼16 mM SDS(aq) solution on PM) results from the
alteration of the interfacial tension of water after the surfactant
molecules adsorb at the liquid-solid and liquid-gas inter-
faces.76 The dramatic 105 Ω·cm2 fold difference between these
spectra at low frequencies (10-1 Hz) represents the magnitude
of the superhydrophobic effect in maximizing the barrier
properties. These results agree with those reported by Yin77,78
and Liu,79 in which high water contact angles (∼150°) of a
surface coating correlated with reduced degradation of the metal
substrate in the presence of corrosive environments. Here,
however, for the first time, we have quantified the effect that
superhydrophobicity has on the barrier properties as compared
to non-superhydrophobic films of the same characteristics. In
relation to the electrochemical behavior of SDS, Wang et al.80
showed that organization of SDS at the interface decreases the
passage of current in a cyclic voltammogram for a methyl-
terminated SAM. To determine the implications of having SDS
in solution, and as a manner of control, Figure 8 shows how
the resultant spectra of a densely packed SAM before and after
exposure to the surfactant exhibit little change in the electro-
chemical properties, suggesting that the substantial decrease in
the barrier properties of the superhydrophobic film is not due
to any electrochemical effect of the surfactant, but rather to the
penetration of the solution through the grooves of the film. This
argument is supported by the fact that the resistance for the
PM film in the presence of SDS is lower than the resistance
offered by the densely packed SAM that is only 2.5 nm thick.
These results suggest that the PM film contains a topology that
enables either the entrapment of air or the penetration of aqueous
solution into the film when it is exposed to solutions of high or
low surface tensions, respectively.
Figure 8. Change in the electrochemical impedance spectra for a
superhydrophobic PM film (1.6 µm thick) grown from HB+Au after
addition of SDS (∼16 mM) to the aqueous solution of 1 mM K3Fe(CN)6
and K4Fe(CN)6 and 0.1 M NaSO4. A SAM on Au prepared from
n-octadecanethiol both with and without surfactant served as a control.
increase in φ at low frequency, consistent with the onset of
penetration by redox probes and highlighting the role of
superhydrophobicity in the ability of the film to perfectly block
redox probes from the substrate.
Tremendous differences in barrier properties between PM
films grown from SIPM versus spontaneous DM polymerization
could, at first sight, be attributed to differences in thickness.
However, to determine the effect of superhydrophobicity on the
barrier properties of PM films grown on Au from SIPM, we
collected the impedance spectrum of a PM film under super-
hydrophobic and non-superhydrophobic conditions. To achieve
superhydrophobic conditions, we used the aqueous redox solu-
tion mentioned above. For non-superhydrophobic conditions,
we collected a new spectrum after decreasing the surface tension
of the previous solution by adding sodium dodecyl sulfate (SDS,
2 × cmc ) 16 mM).74 Figure 8 shows a Bode magnitude plot
of the resulting spectra and their fits as solid lines for a PM
film grown from SIPM and, as a control, a SAM prepared from
n-octadecanethiol both with and without surfactant. We selected
a SAM as a control due to its molecularly smooth surface and
inability to entrap much air at the organic/aqueous interface.
The superhydrophobic PM film behaves as an ideal capacitor
that provides a defined separation of charge between the solution
and the metal substrate due to the combined presence of air
and PM as dielectric materials. However, addition of the
surfactant breaks the superhydrophobic behavior,74,75 favoring
the displacement of the air in the grooves of the interface by
Conclusion
We have developed a new strategy to produce polymethylene
films on gold and silicon surfaces by immobilizing borane on
a vinyl-terminated monolayer, followed by its exposure to a
DM solution in ether. This approach enables the control of film
thickness by varying the time of polymerization or the DM
concentration to produce PM films with superhydrophobic
behavior and outstanding barrier properties. This methodology
(75) Ferrari, M.; Ravera, F.; Rao, S.; Liggieri, L. Appl. Phys. Lett. 2006,
89.
(76) Feng-Ming, C.; Yu-Jane, S.; Hui, C.; Heng-Kwong, T. Appl. Phys.
Lett. 2007, 91, 094108.
(77) Liu, T.; Chen, S.; Cheng, S.; Tian, J.; Chang, X.; Yin, Y. Electrochim.
Acta 2007, 52, 8003–8007.
(78) Yin, Y. S.; Liu, T.; Chen, S. G.; Cheng, S. Appl. Surf. Sci. 2008, 255,
2978–2984.
(79) Liu, H. Q.; Szunerits, S.; Xu, W. G.; Boukherroub, R. ACS Appl.
Mater. Interfaces 2009, 1, 1150–1153.
(74) Mohammadi, R.; Wassink, J.; Amirfazli, A. Langmuir 2004, 20, 9657–
(80) Wang, J.; Zeng, B.; Fang, C.; Zhou, X. J. Electroanal. Chem. 2000,
484, 88–92.
9662.
9
J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010 5733

A R T I C L E S
Tuberquia et al.
Acknowledgment. We gratefully acknowledge the National
Science Foundation (CBET 0731168) for financial support of our
work. N.N. acknowledges support from the Vanderbilt University
School of Engineering Summer Undergraduate Research Experience
Program.
of creating a superhydrophobic surface from an originally
smooth surface has the versatility to be further extended to many
other vinyl-terminated surfaces.
PM films grown from SIPM present a micrometer- and
nanoscale morphology that is responsible for their superhydro-
phobic behavior. This morphology leads to entrapped air at the
polymer/aqueous interface, which elevates both advancing and
receding contact angles and enhances aqueous-phase barrier
properties by orders of magnitude. This remarkable improvement
in the barrier properties, combined with the precise control over
film properties and the ability to generate superhydrophobic
films, could open the door to a broad number of applications in
which highly blocking surface coatings are required.
Supporting Information Available: Experimental methods
and a Bode phase plot that serves as a companion to Figure 7
are provided and discussed. This material is available free of
charge via the Internet at http://pubs.acs.org.
JA9086193
9
5734 J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010