Spontaneous polymer thin film assembly and organization using mutually immiscible side chains

Article abstract of DOI:10.1021/ja952225t

Polymer ultrathin film self-assembly and organization on solid substrates has been directed using grafted siloxane copolymers bearing mutually immiscible alkyl and perfluoroalkyl side chains. Polysiloxanes grafted with both alkyl disulfide and perfluoroalkyl side chains have been synthesized and characterized. These terpolymer systems assemble spontaneously on gold surfaces, forming bound polymeric monolayers organized by intramolecular phase separation. Interfacially bound polymer monolayer fabrication is driven by multipoint alkyl disulfide side chain chemisorption to gold surfaces from dilute organic solution. Immiscible perfluoroalkyl side chains of low interfacial energy enrich the ambient-exposed outer regions of these monolayers, yielding a novel bound polymer monolayer with an anisotropic, layered structure and perfluorinated surface properties. Ellipsometry indicates that these polymer films have thicknesses ranging from 22 to 32 ?, depending on solution conditions and chemistry. Angular-dependent X-ray photoelectron spectroscopy has provided a depth profile of the bound polymer films, detailing the anisotropic composition resulting from perfluoroalkyl surface enrichment. Static secondary ion mass spectrometry measurements support the enrichment of perfluoroalkyl groups in the outer atomic levels of these films. Cyclic voltammetry using the redox probes Fe(CN)₆^3- and methylviologen with film-coated gold electrodes evaluated film-attenuated electron transfer. Time-of-flight secondary ion mass spectrometry has been used to image micropatterned polymer surfaces lithographed at high resolution both before and after organic monolayer assembly. Qualitative and quantitative information on film spatial organization and surface chemistry distribution on microstructures was obtained.

Full text of DOI:10.1021/ja952225t

1
856
J. Am. Chem. Soc. 1996, 118, 1856-1866
Spontaneous Polymer Thin Film Assembly and Organization
Using Mutually Immiscible Side Chains
|
‡
†
†
§
,†
F. Sun, D. G. Castner, G. Mao, W. Wang, P. McKeown, and D. W. Grainger*
Contribution from the Department of Chemistry, Colorado State UniVersity,
Fort Collins, Colorado 80523-1872, National ESCA and Surface Analysis Center for
Biomedical Problems and Department of Chemical Engineering, UniVersity of Washington,
Seattle, Washington 98195, Physical Electronics Inc., 6509 Flying Cloud DriVe,
Eden Prairie, Minnesota 55344
X
ReceiVed July 6, 1995
Abstract: Polymer ultrathin film self-assembly and organization on solid substrates has been directed using grafted
siloxane copolymers bearing mutually immiscible alkyl and perfluoroalkyl side chains. Polysiloxanes grafted with
both alkyl disulfide and perfluoroalkyl side chains have been synthesized and characterized. These terpolymer systems
assemble spontaneously on gold surfaces, forming bound polymeric monolayers organized by intramolecular phase
separation. Interfacially bound polymer monolayer fabrication is driven by multipoint alkyl disulfide side chain
chemisorption to gold surfaces from dilute organic solution. Immiscible perfluoroalkyl side chains of low interfacial
energy enrich the ambient-exposed outer regions of these monolayers, yielding a novel bound polymer monolayer
with an anisotropic, layered structure and perfluorinated surface properties. Ellipsometry indicates that these polymer
films have thicknesses ranging from 22 to 32 Å, depending on solution conditions and chemistry. Angular-dependent
X-ray photoelectron spectroscopy has provided a depth profile of the bound polymer films, detailing the anisotropic
composition resulting from perfluoroalkyl surface enrichment. Static secondary ion mass spectrometry measurements
support the enrichment of perfluoroalkyl groups in the outer atomic levels of these films. Cyclic voltammetry using
the redox probes Fe(CN)63 and methylviologen with film-coated gold electrodes evaluated film-attenuated electron
transfer. Time-of-flight secondary ion mass spectrometry has been used to image micropatterned polymer surfaces
lithographed at high resolution both before and after organic monolayer assembly. Qualitative and quantitative
information on film spatial organization and surface chemistry distribution on microstructures was obtained.
-
Introduction
been reported.6 Self-assembly strategies using organothiols,
disulfides, and alkyl silanes have also utilized in situ polym-
erization methods to stabilize interfacial order and stability.7
Chiral recognition elements also drive monomer assembly in
two dimensions and subsequently facilitate polymerization of
Organization of molecules in two dimensions has long been
sought as a method to provide ultrathin, durable materials with
novel barrier or microstructural properties. Polymerizations
are commonly pursued in fabricating stable ultrathin organic
films in order to increase film cohesion, durability, and
functional properties. Methods involving Langmuir-Blodgett
1
8
these ordered arrays. Redox-active polymers have also been
synthesized, attaching to substrates via specific anchoring
2
9
groups. Additionally, diacetylene monomers have been an-
(L-B) mono- and multilayers, including post-transfer-thermal
chored on gold surfaces and polymerized topochemically using
3
2,4
10
polymerization and photopolymerization, electrostatic poly-
mer adsorption, and prepolymerized amphiphile assembly have
photopolymerization.
5
As an alternative, we use a strategy (Figure 1) that hybridizes
6
the concept of prepolymerized L-B organization together with
†
11
Colorado State University.
University of Washington.
Physical Electronics Inc.
Current address: Nalco Chemical Co., Naperville, IL 60563-1198
organic self-assembled film architecture on solid supports. Two
‡
architectural requirements are proposed for successful fabrication
of prepolymerized, anchored self-assembled films: a flexible
polymer backbone readily facilitating side chain rearrangements
and exposed anchor groups facilitating specific interfacial
binding. Multipoint attachment of soluble polymer coils by
numerous anchoring polymer side chains drives an irreversible
§
|
*
Address correspondence to this author.
Abstract published in AdVance ACS Abstracts, February 1, 1996.
X
(1) Swalen, J. D.; Allara, D. L.; Andrade, J. D.; Chandross, E. A.; Garoff,
S.; Israelachvilli, J.; McCarthy, T. J.; Murray, R.; Pease, R. F.; Rabolt, J.
F.; Wynne, K. J.; Yu, H. Langmuir 1987, 3, 932.
(
2) (a) Ringsdorf, H.; Schlarb, B.; Venzmer, J. Angew. Chem. Int., Ed.
Engl. 1988, 27, 113. (b) Arslanov, V. V. AdV. Colloid Interface Sci. 1992,
0, 307.
(6) (a) Elbert, R.; Laschewsky, A.; Ringsdorf, H. J. Am. Chem. Soc. 1985,
107, 4134. (b) Davis, F.; Hodge, P.; Towns, C. R.; Ali-Adib, A.
Macromolecules 1991, 24, 5695. (c) Biddle, M. B.; Lando, J. B.; Ringsdorf,
H.; Schmidt, G.; Schneider, J. Colloid Polym. Sci. 1988, 266, 806.
(7) Li, D.; Ratner, M. A.; Marks, T. J.; Zhang, C.; Yang, J.; Wong, G.
K. J. Am. Chem. Soc. 1990, 112, 7389.
4
(3) (a) Engel, A. K.; Yoden, T.; Sanui, K.; Ogata, N. J. Am. Chem. Soc.
1
985, 107, 8308. (b) Kakimoto, M.; Morikawa, A.; Nishikata, Y.; Suzuki,
M.; Imai, Y. J. Colloid Interface Sci. 1988, 121, 599. (c) Rabe, J. P.; Rabolt,
J. F.; Brown, C. A.; Swalen, J. D. J. Chem. Phys. 1986, 84, 4096.
(
4) (a) Nakanishi, F.; Okada, S.; Nakanishi, H. Polymer 1989, 30, 1959.
b) Uchida, M.; Tanizaki, T.; Oda, T.; Kajiyama, T. Macromolecules 1991,
4, 3238. (c) Akiomoto, A.; Dorn, K.; Gros, L.; Ringsdorf, H.; Schupp, H.
(8) Stupp, S. I.; San, S.; Lin, H. C.; Li, L. S. Science 1993, 259, 59.
(9) Albagli, D.; Bazan, G. C.; Schrock, R. R.; Wrighton, M. S. J. Am.
Chem. Soc. 1993, 115, 7328.
(
2
Angew. Chem., Int. Ed. Engl. 1981, 20, 90. (d) Lopez, E.; O’Brien, D. F.;
Whitesides, T. H. J. Am. Chem. Soc. 1982, 104, 305.
(10) (a) Batchelder, D. N.; Evans, S. D.; Freeman, T. L.; H a¨ uâling, L.;
Ringsdorf, H.; Wolfe, H. J. Am. Chem. Soc. 1994, 116, 1050. (b) Kim, T.;
Crooks, R. M.; Tsen, M.; Sun, L. J. Am. Chem. Soc. 1995, 117, 3963.
(11) For a review, see: Ullman, A. Ultrathin Films; Academic Press:
New York, 1990.
(
5) (a) Decher, G.; Hong, J. D.; Schmitt, J. Thin Solid Films 1992, 210/
11, 831. (b) Shimomura, M.; Fujii, K.; Karg, P.; Frey, W.; Sackmann, E.;
Meller, P.; Ringsdorf, H. Jpn. J. Appl. Phys. 1988, 27, L1761.
2
0
002-7863/96/1518-1856$12.00/0 © 1996 American Chemical Society

Spontaneous Polymer Thin Film Assembly and Organization
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1857
Figure 1. Conceptual representation of polymer solution coil transition to anchored two-dimensional monolayer coating. Grafted terpolymers
self-organize into a three-tiered, layered structure upon interfacial binding, enriching the film’s surface region with lowest energy polymer
perfluorocarbon components. Functional units may be left along the polymer backbone for further in-situ derivatization and chemical modification.
from each other16 as well as from polysiloxane phases,15 the
copolymer architecture unfolds once bound to the solid surface.
Intrachain immiscibility compels the siloxane backbone to
position itself upon an underlying layer of chemisorbed alkyl
disulfide side chains anchored on the gold surface. Lowest
energy perfluoroalkyl chains remain layered above a region
enriched in siloxane backbone, exposed at the ambient interface
by a thermodynamic drive to minimize film interfacial tension
17
with air. The synthesis and characterization of these perfluo-
roalkyl-enriched polymer films using surface-sensitive and
electrochemical techniques are reported.
Experimental Section
Materials. l-Octadecanethiol (98%), allyl chloride (98%), allyl
alcohol (99%), n-pentanethiol (98%), thioacetic acid (96%), potassium
ferricyanide (99+%), and tetramethylammonium hydrogen sulfate
monohydrate (TAMHS) (98%) were purchased from Aldrich and used
as received. 11-Bromo-1-undecene (Lancaster, 98%), pentadecafluoro-
1
-octanol (PCR Florida, 97%), poly(hydrogenmethylsiloxane) (PHMS)
l
(
H u¨ ls America, average molecular weight 4750 as determined by H-
NMR in this laboratory), and platinum divinyltetramethyldisiloxane
H u¨ ls America, 3% in xylene) were all used as supplied without further
purification. Toluene (Chempure) was refluxed over sodium metal for
-6 h and distilled prior to use. Chloroform for final polymer sample
storage use was dried over P and then distilled. Other solvents
Figure 2. Chemical architecture for self-assembling polymers based
on poly(methylsiloxane) main chains, perfluoroalkyl side chains, and
alkyl disulfide anchoring chains.
(
4
2 5
O
adsorption process resulting in a tightly bound polymer mono-
layer. Acrylate-grafted copolymers containing alkyl disulfide
or thioether anchor chains demonstrate this behavior on gold
including reagent-grade acetone (Chempure), ethyl acetate (Chempure),
methylene chloride (Chempure), and ethanol (EM Science) were used
as received. Chromatographically pure 4,4′-dimethylbipyridinium ion
1
2,13
2+
substrates,
as do siloxane oligomers bearing propyl mer-
(methylviologen, MV ) was a gift from Professor J. K. Hurst
captan side chains.14 Assemblies of these oligomers on gold
create monolayers stabilized by thiolate-gold anchors. Propyl
mercaptan chains left unbound and surface-exposed are readily
derivatized using photoaddition reactions as schematically shown
in Figure 1.14
We describe here a new polymer (shown in Figure 2)
designed such that chemical incompatibility between bound
polymer units drives organization within the anchored polymer
assembly. Polysiloxanes are highly flexible, mobile polymers,
exhibiting very low glass transition temperatures. These
polymers also demonstrate very limited miscibility with both
hydrocarbon and fluorinated compounds.15 Because perfluo-
rocarbon and hydrocarbon side chain components phase separate
(Washington State University). Water was fed from a reverse osmosis
unit to a Millipore unit prior to all experiments (18 mΩ resistivity).
Polymer Synthesis. Polymers used for studies of spontaneous thin
film assembly feature a flexible siloxane backbone and grafted mutually
immiscible hydrocarbon and fluorocarbon side chains (Figure 2). These
polymers were prepared by co-hydrosilylation of linear poly(hydro-
genmethylsiloxane) (PHMS) with both side-chain-forming olefins as
described in the supporting information for this article.
Synthesis of n-Pentadecafluorooctyl Allyl Ether (5). Compound
18
5
was synthesized according to a reported method with minor
modifications. n-Pentadecafluoro-1-octanol (8.50 g, 21.25 mmol) was
added to a 50% caustic soda solution (60 mL) containing 0.76 g (4.0
mmol) of tetramethylammonium hydrogen sulfate (TMAHS) as a phase
transfer catalyst. The mixture was vigorously stirred for 10 min. After
20.0 g (261.3 mmol) of allyl chloride was added, the mixture was stirred
(12) (a) Sun, F.; Grainger, D. W. J. Polym. Sci. A, Polym. Chem. 1993,
3
1, 1729. (b) Sun, F.; Grainger, D. W.; Castner, D. G. Langmuir 1993, 9,
200. (c) Sun, F.; Grainger, D. W.; Castner, D. G.; Leach-Scampavia, D.
(15) (a) Yilgor, I.; Steckle, W. P., Jr.; Yilgor, E.; Freelin, R. G. J. Polym.
Sci. A, Polym. Chem. 1989, 27, 3673. (b) Kobayashi, H.; Owen, M. J.
Macromolecules 1990, 23, 4929.
(16) (a) Mukerjee, P. J. Am. Oil Chem. Soc. 1982, 59, 573. (b) Kunitake,
T.; Higoshi, N. J. Am. Chem. Soc. 1985, 107, 692. (c) Elbert, R.; Folda, T.;
Ringsdorf, H. J. Am. Chem. Soc. 1984, 106, 7687.
(17) Doeff, M. M.; Lindner, E. Macromolecules 1989, 22, 2951.
(18) Boutin, B.; Youssef, B. J. Fluorine Chem. 1987, 39, 399.
3
J. Vac. Sci. Technol. 1994, 12, 2499. (d) Sun, F.; Lei, Y.; Grainger, D. W.
Colloids Surf. 1994, 93, 191.
(
13) Lenk, T. J.; Hallmark, V. M.; Rabolt, J. F.; H a¨ ussling, L.; Ringsdorf,
H. Macromolecules 1993, 26, 1230.
14) Sun, F.; Grainger, D. W.; Castner, D. G., Leach-Scampavia, D.
Macromolecules 1994, 27, 3053.
(

1
858 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Sun et al.
Table 1. Bulk Characterization Co-hydrosilylated Siloxane
Terpolymer PHMS
thorough sequential rinsing with CHCl
and drying under pure N
3
, EtOH, and Millipore water
2
.
MWa
ss (mol %)b
(mol %)c
d
e
Ellipsometry. Monolayer thicknesses were measured using a
Gaertner L117 ellipsometer (70° incident angle, 632.8 nm laser source,
three samples, 1-2 spots/sample) by assuming a refractive index of
polymer
C
C
f
N
ss
N
f
A
B
C
15 150
28 900
24 780
38
64
21
0
19
41
29.3
49.3
16.2
0
14.6
31.6
1
.41 for all monolayers (taken by averaging refractive indices for bulk
poly[methyl(trifluoropropyl)siloxane] (n ) 1.381) and poly(methyloc-
tylsiloxane)-poly(hydrogenmethylsiloxane) copolymer (0.54/0.46) (n
) 1.435)). Arbitrarily changing the n values from 1.35 (highly
fluorinated) to 1.47 (primarily hydrocarbon) resulted in less than (6
Å variation in most calculated monolayer thicknesses.
Contact Angle Analysis. Sessile drop contact angle analysis (Ram e´ -
Hart 100 apparatus) used purified (Millipore 18 MΩ‚cm resistivity)
water drops (2 µL) on three separate spots on each film surface in a
controlled environment (100% RH). Measurements were taken on both
sides of water drops at ambient temperature 30-40 s after drops were
applied to surfaces. Contact angle data report the average of three drops
at different surface locations.
a
MW ) average molecular weight of product polymers (calculated
on the basis of starting PHMS molecular weight of 4750 and polymer
composition assuming that no chain degradation occurs during co-
b
hydrosilylation).
Css (mol %) ) disulfide sided chain content in
1 c
polymer as measured by H-NMR.
chain content in polymer as measured by H-NMR.
number of disulfide side chains per polysiloxane backbone.
C (mol %) ) perfluoroalkyl side
f
1 d
N
ss ) absolute
e
N )
f
absolute number of perfluoroalkyl side chains per polysiloxane
backbone.
at 40 °C for another 12 h. The reaction suspension was then diluted
with 20 mL of water, and crude product 5 was extracted with 30 mL
of CH
2
Cl
2
.
After the extract was dried over anhydrous MgSO
4
,
Cyclic Voltammetry. Cyclic voltammograms were measured using
an EG&G Model 273 potentiostat/galvanostat and recorded on a Goerz
SE790 XY chart recorder. A single compartment, three-electrode glass
cell containing a Pt gauze counterelectrode, a calomel (saturated KCl,
Radiometer 401) reference electrode (SCE), and a polymer film-coated
methylene chloride was removed at 0 °C under reduced pressure and
a colorless liquid was collected. Final distillation of crude product
gives a clear liquid of bp 81.5-82.5 (24 mmHg). Yield: 8.98 g (96%).
1
H-NMR (CDCl
m, CH dCH), 5.89 (m, CH
Co-hydrosilylation of PHMS with Hydrocarbon/Perfluorocarbon
Side Chains (6). To a 8 mL predried glass vial were added 0.114 g
1.89 mmol of Si-H units) of poly(hydrogenmethylsiloxane) (PHMS),
.30 g (0.87 mmol) of 11-(n-pentyldithio)undecyl allyl ether (4, see
3
): δ 3.85 (t, -OCH
2
-), 4.18 (t, -CHCH
-).
2
O), 5.34
1
2d
(
2
2
dCHCH
2
gold working electrode was used as described previously.
Spin-
coated (CHCl solution, Headco vacuum chuck, 600-800 rpm) thick
3
polymer films were also prepared. Electrical leads were attached using
conducting silver epoxy, and the upper portion of the electrode was
masked with Teflon tape,19 leaving a 0.8 × 0.4 cm electrode area
exposed to the solution. Supporting electrolyte was 1 M KCl (99+%
(
0
2
supporting information and ref 12), and 4 mL of anhydrous toluene.
After the solution was purged with nitrogen for 3 min, 0.53 g (1.2
mmol) of n-pentadecafluorooctyl allyl ether (5) and platinum divi-
nyltetramethyldisiloxane-xylene (3%) (catalyst, 3 drops, 0.011 mmol)
were added and the mixed solution was stirred at 80 °C for 18 h. Crude
polymer product was purified by precipitation-redissolution in methanol/
acetone (1:2) and chloroform three times. The purification process
was monitored by TLC (chloroform/ethyl acetate, 7:3) until no
3 6
EM Science); concentration of the electroactive species (K Fe(CN)
or methylviologen) was 3 mM in Millipore-filtered water purged with
argon for 15 min prior to measurement.
Film Thermal Stability. Gold substrates covered with bound
polymer monolayer films were immersed in 1,2,3-trichlorobenzene at
various elevated temperatures for fixed time periods. At each time
point a sample was removed from the solvent, rinsed with chloroform,
detectable residual monomer (dithio-containing ether 4, R
f
) 0.74) was
2
and dried under N . Film was analyzed by aqueous wetting and
ellipsometry as described above.
found in the precipitating solution. Purified polymer 6 was then dried
in a nitrogen stream and immediately dissolved in anhydrous chloroform
to afford a 3% (by weight) stock solution (stored in the refrigerator)
Patterned Monolayer Films. Two types of micropatterned organic
thin films of polymers are reported: conventional photolithographic
methods fabricated micron-scale gold lines on semiconductor-grade
silicon oxide (thermally grown in steam) wafers in a commercial
integrated circuit fabrication process in a Class 100 clean room.
Additionally, patterns were etched into homogeneous polymer thin films
assembled on unpatterned gold surfaces using an automated commercial
ion mill (610 Series focused ion beam workstation, FEI Co., Hillsboro,
1
for thin film assembly (after further dilution). Yield: 0.35 g. H-
NMR (CDCl
3
): δ 0.075 (br, Si-CH
), 1.62 (br, S-C-CH , O-C-CH
.55, 3.90 (br, br, br, O-CH ).
Other PHMS polymer-grafted compositions were prepared by the
3
), 0.58 (br, Si-CH
2
), 0.92 (t, CH
3
),
1
3
.30 (br, CH
2
2
2
), 2.70 (t, S-CH
2
), 3.35,
2
same procedure except that initial PHMS/olefin feed ratios were
changed to alter the ratio of grafted alkyl and perfluoroalkyl chains.
Co-hydrosilylation of poly(hydrogenmethylsiloxane) (PHMS) with
both dithio- and perfluoroalkyl allyl ethers affords directly the product
polymer bearing both pendent fluorocarbon and anchoring disulfide
hydrocarbon side chains (Figure 2). Table 1 presents characterization
+
OR). A beam of Ga ions, extracted by field emission from a liquid
metal source, accelerated through 25 kV, is passed through focusing
lenses and an adjustable aperture and then digitally rastered over a
selected target area (see supporting information).
Fourier Transform Infrared Spectroscopy. Bulk transmission and
grazing angle spectra were collected on a Nicolet 60SX instrument with
1
results for the grafted PHMS polymers based on H-NMR analysis.
Unlike hydrosilylation of PHMS using terminal hydrocarbon olefins
directly adjacent to the fluorocarbon chain (e.g., 1H,1H,2H-perfluoro-
decene), hydrosilylation of the same PHMS by the perfluoroalkyl allyl
ether produces the polymer product having the desired perfluoroalkyl
side chain attachment features (absence of fluorine transfer from
2
a liquid N -cooled MCT detector. Grazing angle external reflection
spectra used a Spectratech variable angle external reflection accessory
-1
(
incident angle ) 78°, 4 cm resolution, 1024 scans). Reference
subtraction (fresh, bare gold) and flattening were achieved using
Spectracalc software (Galactic Industries), with no curve smoothing
or other alterations.
X-ray Photoelectron Spectroscopy. X-ray photoelectron spectros-
copy (XPS) experiments were performed on a Surface Science SSX-
2
4
perfluorocarbon R to the terminal olefin).
Meanwhile, disulfide
species present in the dithioalkyl allyl ether do not cause any significant
poisoning of the hydrosilylation catalyst, though a relatively slower
reaction compared to conventional hydrosilylation without disulfide
100 spectrometer (Mountain View, CA) equipped with a monochromatic
2
5
1
moieties is observed. The H-NMR spectrum shows that the grafted
polymers possess ether-oxygen-adjacent methylene groups derived from
both the dithioalkyl allyl ether (δ 3.38) and perfluoroalkyl allyl ether
Al KR source, hemispherical analyzer, and a multichannel detector.
Conditions and methods for these studies were identical to those
12,14
published previously.
Loss of fluorine from these samples, resulting
(δ 3.51 and 3.88), illustrating the coexistence of both side chains along
from x-ray exposure, was negligible: the fluorine F1s signal varied
less than 1% between the start and end of experimental analysis. A
Teflon standard was used to calibrate fluorine:carbon ratios. For
calculation of XPS elemental composition, the analyzer transmission
function was assumed not to vary with photoelectron kinetic energy
the PHMS siloxane backbone.
Polymer Monolayer Film Assembly. Gold substrates for (dithio-
alkyl)polysiloxane self-assembly were freshly prepared by evaporative
deposition of a layer of 1500-2000 Å gold (99.999%, Johnson-Mathey)
1
2,14
onto Pd-coated (200 Å) silicon wafer surfaces as described.
coated substrates (cut into 3 × 0.8 cm pieces) were immersed into
polymer-CHCl solutions (0.1-1.5 mM) for up to 72 h, followed by
Gold-
(19) Martin, C. R.; Rubenstein, I.; Bard, A. J. J. Am. Chem. Soc. 1982,
3
104, 4817.

Spontaneous Polymer Thin Film Assembly and Organization
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1859
KE),20 the photoelectron escape depth was assumed to vary as KE
0.7 20
,
Table 2. Polymer Ultrathin Film Properties before and after
(
2
1
Treatment at Elevated Temperature in Trichlorobenzenea
and Scofield’s photoionization cross sections were used.
Angle-dependent XPS data were collected at nominal photoelectron
take-off angles of 0°, 55°, and 80°. The take-off angle was defined as
the angle between the surface normal and the axis of the analyzer lens
system. Using mean free paths calculated from the equations given
by Seah and Dench, the sampling depth (three times the mean free
path) should decrease from 90 to 15 Å as take-off angle increases from
film
thickness
(Å)
aqueous
contact
angle (deg)
monolayer film
octadecanethiol (initially)
octadecanethiol (after 30 min,
23
7
97
63
2
2
9
0 °C in solvent)
0
° to 80°.
polymer A (initially)
polymer B (initially)
polymer C (initially)
polymer C (after 60 min, 90 °C in solvent)
polymer C (after 120 min, 120 °C in solvent)
polymer C (after 60 min, 150 °C in solvent)
22
28
42
28
18
4
89
92
107
107
102
92
Quadrupole Static Secondary Ion Mass Spectrometry. Static
secondary ion mass spectrometry (SIMS) experiments were performed
with a Physical Electronics 3700 SIMS system (PHI Electronics, Eden
Prairie, MN) mounted on a custom ultra-high-vacuum (UHV) system
(see supporting information for details). The ion beam was rastered
over a 5 × 5 mm area, and the total exposure time of the sample to the
a
Monolayers on gold substrates; solvent ) 1,2,3-trichlorobenzene.
ion beam, including setup and data acquisition, was less than 7 min.
1
2
2
Corresponding total ion doses per sample (<5 × 10 ions/cm ) are
within the generally accepted limit for static SIMS conditions for
tion times (>72 h). Multilayers presumably result from side
chain interlayer associations which increase polymer adsorption
and deposition.
23-25
organic surfaces.
Both positive and negative secondary ions were
collected over a m/z range of 0-300 with a nominal mass resolution
of unity. Data acquisition and control of the energy filter and
quadrupole used the Physical Electronics SIMS software package.
Sample Imaging Using Time-of-Flight Secondary Ion Mass
Spectrometry (ToF-SIMS). Imaging of patterned polymer thin films
was performed on a Model 7200 Physical Electronics ToF-SIMS
instrument (PHI Electronics, Eden Prairie, MN). A gallium liquid metal
ion source (LMI Ga), operating at a beam energy of 25 keV, was utilized
for all analysis presented in this study (see supporting information for
details).
The aqueous contact angle shows that the fluorocarbon side
chains in the polymers increase film hydrophobicity compared
to that of the pure dithioalkyl side chain analog (polymer A)
and that surface hydrophobicity increases with increasing the
perfluoroalkyl side chain content in the polymer (Table 2). This
is attributed to the relative enrichment of fluorocarbon chains
in the outermost regions of the monolayer (see XPS film
characterization section below). However, compared to other
fluorinated surfaces such as Teflon, plasma-deposited perfluo-
ropropane films, and self-assembled monolayers from organic
perfluoro compounds,
monolayers derived from spontaneous adsorption are 7-20°
lower, suggesting incomplete coverage of the underlying silox-
ane backbone and dithio-anchoring side chains by the overlaying
perfluoroalkyl chains (Figure 2).
Electrochemical Characterization of Polymer Films on
Gold Electrodes. Soluble redox probes, including Ru(NH3)6
Fe(CN)6
used for electrochemical measurements on organothiol mono-
layers in aqueous electrolyte.
function of a number of variables including alkanethiol length,
probe chemistry, and film defects (pinholes, grain boundaries).
Generally, little Faradaic current is observed through alkanethiol
monolayers of dodecanethiol and longer alkyl lengths when
Results and Discussion
Three-component, grafted polysiloxanes bearing both per-
fluoroalkyl and alkyl disulfide chains were synthesized in
various compositions, with three examples shown in Table 1:
polymer A bearing only alkyl disulfide chains; polymer B
bearing 19 mol % perfluoroalkyl chains and 64 mol % alkyl
disulfide side chains; and polymer C bearing 41 mol %
perfluoroalkyl and 21 mol % alkyl disulfide side chains.
Polymer Film Thickness and Wettability on Gold Sub-
strates. Perfluoroalkyl and dithioalkyl co-hydrosilylated poly-
siloxanes were self-adsorbed onto fresh gold-deposited wafers
from dilute chloroform solution. Ellipsometric thicknesses of
the fabricated polymer films range from 22 to 42 Å (Table 2),
consistent with measurements of alkyl disulfide side chain
15,26,27
contact angles for the polymeric
3+/2+
,
3-/4-
2+/+/0
, and methylviologen (MV
) have often been
28-31
Current attenuation is a
12
monolayers alone (14 Å), plus the addition of a siloxane layer
4 Å) and a perfluoroalkyl layer (14 Å). Film thicknesses are
(
28
assembled on carefully prepared gold substrates.
Cyclic voltammograms for Fe(CN)6 on clean, bare gold
electrodes show the well-characterized, reversible, one-electron
transfer process (data not shown, ∆Ep ) 60 mV, E1/2 ) 190 (
mV).
on gold electrodes are able to completely attenuate electron
transfer under the same conditions (not shown). Thinner
multilayer films of polymer C on gold electrodes (64 Å) are
also very effective at attenuating electron transfer to/from the
consistent for each polymer from “batch to batch” after 24 h of
dilute solution adsorption (standard deviations of (4 Å) and
do not increase with extended adsorption times up to 1 week,
demonstrating only stable monolayer formation and absence of
any significant interchain cross-linking by disulfide reduction.
Given the uncertainty in the polysiloxane contribution film,
thickness differences between films of polymers A and B are
not significant. Thicker multilayer films can be prepared from
concentrated polymer solutions (20 mM) and extended adsorp-
3-
28-30
5
Thick, spin-cast films of polymer C (330 Å thick)
(26) (a) Chau, L.-K.; Porter, M. D. Chem. Phys. Lett. 1990, 167, 198.
(
20) Application note from Surface Science Instruments, Mountain View,
(b) Arduengo, A. J.; Moran, J. R.; Rodriguez-Parada, J. R.; Ward, M. D. J.
Am. Chem. Soc. 1990, 112, 6153. (c) Chidsey, C. E. D.; Loiacono, D.
Langmuir 1990, 6, 682. (d) Alves, C. A.; Porter, M. D. Langmuir 1993, 9,
3507. (e) Lenk, T. J.; Hallmark, V. M.; Hoffmann, C. L.; Rabolt, J. F.;
Castner, D. G.; Erdelen, C.; Ringsdorf, H. Langmuir 1994, 10, 4610.
(27) (a) Haque, Y.; Ratner, B. D. J. Appl. Polym. Sci. 1986, 32, 4369.
(b) Dann, J. R. J. Colloid Interface Sci. 1970, 32, 302.
(28) (a) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J.
Am. Chem. Soc. 1987 109, 3559. (b) Fabinowski, W.; Coyle, L. C.; Weber,
B. A.; Granata, R. D.; Castner, D. G.; Sadownik, A.; Regen, S. L. Langmuir
1989 5, 35.
CA (1987).
(
(
(
(
21) Scofield, J. H. J. Electron Spectrosc. Relat. Phenom. 1976, 8, 129.
22) Seah, M. P.; Dench, W. A. Surf. Interface Anal. 1979, 1, 2.
23) Briggs, D.; Hearn, M. J. Vacuum 1986, 36, 1005.
24) We have also studied hydrosilylation of PHMS by (1H,1H,2H-
perfluoroalkyl)olefins (C10 and C6) and found in both circumstances an
unexpected transfer of fluorine from the -CF2- group in the R-position to
the olefin to the polymer backbone Si atoms. Resultant polymers had less-
than-desired side chain density (unpublished result). This undesirable
reaction was not observed with compound 5, where the terminal olefin is
moved away from the perfluoroalkyl moieties.
(29) Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990,
112, 558.
(
25) A model study conducted under similar reaction conditions where
n-pentyl disulfide was used as a “poisoning” additive (replacing dithioalkyl
(30) Niwa, M.; Mori, T.; Higashi, N. J. Mater. Chem. 1992, 2, 245;
Macromolecules 1993, 26, 1936.
(31) Chailapakul, O.; Crooks, R. M. Langmuir 1993, 9, 884.
allyl ether) in hydrosilylation of PHMS with 1-octene shows over 90-
1
9
5% coupling conversion proven by both H-NMR and FTIR.

1
860 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Sun et al.
Figure 3. Cyclic voltammograms for MV2 on polymer-coated gold electrodes (probe concentration, 3 mM; scan rate, 200 mV s-1; 1 M KCl
supporting electrolyte). Large plot: bare gold electrode (solid line); gold electrode spun-coat (330 Å thick) with polymer C (dashed line). Inset plot:
electrode with multilayer film (inner voltammogram) of polymer C (64 Å thick); electrode with monolayer (outer voltammogram) of polymer C (30
Å thick).
+
2
+
redox probe, showing little Faradaic current (data not shown).
Monolayer films of polymer C on gold electrodes (30 Å thick)
by contrast show a considerably reduced ability to effectively
attenuate electron transfer (data not shown), indicating solvent-
accessible film defects which permit transport of Fe(CN)6
through the polymer barrier to allow electron transfer with the
underlying gold electrode.
to those for bare gold electrodes when methylviologen (MV )
is the redox probe. Despite the fact that peak current (ip) is
reduced by ∼80% relative to bare gold, two voltammetric waves
(similar to those observed at the bare electrode) are clearly seen
at the monolayer-coated electrode. Hence, the voltammetric
3
-
2
+
response of the monolayer-coated film to the MV probe is
quite different from that of the Fe(CN)6 probe. This suggests
that the more lipophilic methylviologen penetrates the film more
3-
Because probe chemistry has been implicated in the electron
transfer blocking abilities of organothiol monolayers,31 a meth-
ylviologen redox couple was also investigated. The redox
properties of viologens in aqueous solutions have been studied
3-
readily than Fe(CN)6 . Analogous results were observed for
1
2d
monolayers of acrylate polymers on gold.
Polymer Film Thermal Stability. Film stability differences
between monomer and polymer assemblies at elevated temper-
ature in 1,2,4-trichlorobenzene are reflected by temperature-
dependent data shown in Table 2. Reductions in aqueous
contact angles are correlated to reductions in film thicknesses
for all films as a function of solvent temperature, indicating
film destruction at elevated temperatures is linked to film
stability and film chemistry. Noteworthy are temperature
stability differences between monomeric and polymer films.
Octadecanethiol films on gold (desorption energy of 35 kcal/
mol ) are removed within 30 min under these extreme
conditions, while polymer films are stable for substantially
longer times. This is consistent with previous results reported
for polyacrylate-based films on gold. Films of octadecanethiol
on gold are readily removed by moderate heating in hexadecane,
while dithioalkyl-derivatized polyacrylates are stable over an
extended time and elevated temperatures without noticeable
3
2-35
extensively.
Two separate, reversible, one-electron transfer
processes are observed for MV2 at -750 and -1000 mV in
+
+
0
aqueous solution, with both reduced species, MV and MV ,
retaining water solubility.36
Cyclic voltammograms for MV2+ on clean, bare gold
electrodes show the two characteristic reversible oxidation and
reduction waves for this probe (Figure 3, voltammogram A).
Again, thick, spin-cast films of polymer C (330 Å thick)
completely attentuate electron transfer with this redox couple
3
7
(Figure 3, voltammogram B). Thinner multilayer films of
polymer C (64 Å thick) on gold electrodes are also effective in
blocking a substantial amount of Faradaic current as well (Figure
3
, voltammogram C). Nevertheless, monolayer assemblies of
polymer C on gold electrodes are unable to attenuate electron
transfer effectively. The peak-to-peak separation (∆E) and E1/2
values obtained for polymer C monolayer-coated electrodes
demonstrate results (∆E ) 50 mV, E1/2 ) -750 mV) identical
1
2
effects. Significantly, polyacrylates bearing long alkyl side
chains but no sulfur anchoring groups were not able to adsorb
to gold to any measurable extent (<5 Å thick), indicating that
sulfur-mediated attachment promotes film formation and sub-
(
32) Gomez, M.; Li, J.; Kaifer, A. E. Langmuir 1991, 7, 1797.
(33) Bae, I. T.; Huang, H.; Yeager, E. B.; Scherson, D. A. Langmuir
1
991, 7, 1558.
34) Willner, I.; Katz, E.; Riklin, A.; Kasher, R. J. Am. Chem. Soc. 1992,
14, 10965.
(
12
sequent stability.
1
(
(
35) Li, J.; Kaifer, A. E. Langmuir 1993, 9, 591.
36) Lei, Y.; Hurst, J. K. J. Phys. Chem. 1991, 95, 7918.
(37) Nuzzo, R. G.; Fusco, F. A.; Allara, D. L. J. Am. Chem. Soc. 1987,
109, 2358.

Spontaneous Polymer Thin Film Assembly and Organization
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1861
Figure 4. High-resolution XPS Cls spectrum for a polymer ultrathin
film of a siloxane terpolymer assembled on a gold substrate. The
siloxane terpolymer contained 70% perfluorocarbon and 30% alkyl
disulfide side chains.
For polymeric films with disulfide anchoring chains, a time-
dependent, thermally enhanced, restructuring phenomenon is
plausible, allowing single chemisorbed side chains opportunities
to detach but remain surface-localized and to reanchor into
ordered side chain clusters while each polymer remains attached
to the surface by multiple anchors. Side chain multiplicity
promotes film stability and layer permanence while allowing
side chain rearrangements toward equilibrium structures. Ad-
ditionally, the pentanethiol cleavage product from disulfide
reduction would likely desorb from the surface under these
competitive conditions.38
Angular-Dependent X-ray Photoelectron Spectroscopy
(
1
ADXPS). Figure 4 shows a typical high-resolution XPS carbon
s spectrum, detailing the four types of carbon in monolayer
films of siloxane terpolymers on gold. The peak centered at
85 eV is characteristic of the hydrocarbon species (e.g., CH2
present in the alkyl side chains while the peak at 287 eV results
2
39
from the C-O (ether) carbon present in the side chain linkages.
Fluorocarbon species present in the perfluoroalkyl side chains
give rise to the peaks at 292 eV (CF2 groups) and 294 eV (CF3
groups).39 Quantitation of high-resolution spectra from the
various chemical species in these systems can be compared as
a function of take-off angle to yield depth-dependent compari-
sons of film composition.
Figure 5 presents depth-dependent XPS analyses for mono-
layer films of four different siloxane polymers on gold.
Quantitation for the fluorine:carbon ratio as a function of film
depth is collated in Table 3. The most evident point is that all
four surfaces have widely different compositions, characteristic
of both the differences in bulk side chain stoichiometry and
surface enrichment of particular chemical components. The
relative abundance of the fluorine signal at all take-off angles
changes in direct relation to bulk fluorine content of the samples.
Another prominent trend is that as fluorine content increases
across the four samples (regardless of sampling depth), carbon
signal decreases. The increasing F:C ratio with decreased
sampling depth (increased XPS take-off angle) is consistent with
the presence of a perfluorocarbon overlayer at the films’
outermost surfaces attenuating the carbon signals from film
layers underneath. The calculated XPS F:C atomic ratios
expected for the polymeric thin films are summarized in Table
Figure 5. Angular-dependent XPS compositional analysis for polymer
ultrathin films of siloxane terpolymers assembled on gold substrates
as a function of side chain composition: (a) 70% perfluorocarbon, 30%
alkyl disulfide side chains; (b) 41% perfluorocarbon, 21% alkyl disulfide
side chains; (c) 23% perfluorocarbon, 77% alkyl disulfide side chains;
(d) 5% perfluorocarbon, 95% alkyl disulfide side chains.
The fluorocarbon layer was the outermost layer and assigned a
thickness of 14 Å. The siloxane backbone was directly below
the fluorocarbon layer and assigned a thickness of 4 Å. The
alkyl disulfide layer was the deepest layer and assigned a
thickness of 14 Å. The composition of each layer was
determined from the stoichiometry of each component (fluo-
rocarbon chains, siloxane backbone, alkyl disulfide chains) and
the grafting percentages of the fluorocarbon and alkyl disulfide
chains. The calculated values in Table 4 compare favorably
with the experimentally determined values in Table 3, consider-
ing that quantities such as elastic scattering and surface
roughness are not accounted for in the XPS calculations. The
main effect of elastic scattering and surface roughness is to
increase the relative contribution of deeper layers (e.g., the alkyl
disulfide chains) and is the most pronounced at the glancing
take-off angle (smallest XPS sampling depth). This probably
accounts, at least in part, for the reason the calculated ratios
4
. These values were obtained using the standard expression
for calculating XPS intensities and the three layer model
depicted in Figure 1 (see ref 40). Mean free paths of 32 and
2
2
2
6 Å were used for the C 1s and F 1s photoelectrons. The
composition within each layer was assumed to be homogeneous.
(
(
38) Biebuyk, H. A.; Whitesides, G. M. Langmuir 1993, 9, 1766.
39) Dilks, A. In Electron Spectroscopy: Theory, Techniques and
Applications; Baker, A. O., Brundle, C. R., Eds.; Academic Press: London,
981; p 277.
40) Paynter, R. W. Surf. Interface Anal. 1981, 3, 186.
1
(

1
862 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Sun et al.
Table 3. XPS Analysis of Surface Perfluorocarbon Enrichment in Polymeric Perfluorosiloxane Monolayers on Golda
fluorine:carbon elemental ratios
depth-dependent F:C ratio
F:C ratio (80° take-off angle)
polymer
monolayer
b
0° (90 Å)
55° (50 Å)
80° (15 Å)
F:C ratio (stoichiometric bulk)
F:C ratio (stoichiometric bulk)
7
4
2
5
0%FC/30%SS
1%FC/21%SS
3%FC/77%SS
%FC/95%SS
0.83
0.84
0.29
0.09
0.94
0.91
0.40
0.10
1.08
1.15
0.56
0.14
0.73
0.64
0.17
0.03
1.5
1.8
3.3
4.7
a
XPS 0°, 55°, and 80° take-off angles represent XPS sampling depths from 90 Å (deep) to 15 Å (shallow). b FC represents perfluoroalkyl graft;
SS represents alkyl disulfide graft.
Table 4. Caalculated XPS F:C Atomic Ratios for Perfluorosiloxane
Monolayers
calcd F:C atomic ratio
polymer monolayer
0°
55°
80°
7
4
2
5
0%FC/30%SS
1%FC/21%SS
3%FC/77%SS
%FC/95%SS
0.72
0.60
0.21
0.04
0.82
0.69
0.28
0.06
1.13
1.06
0.81
0.32
a
Based on ref 40.
show a larger increase in the F:C ratio from 0 to 80° take-off
angles than the experimental ratios do.
Data from Table 3 show that relative perfluorocarbon
enrichment (compare 80° take-off angle data against stoichio-
metric bulk data) is significantly larger for samples with lower
bulk perfluorocarbon content. As perfluorocarbon content in
the polymer increases, relative surface enrichment decreases.
Additionally, as a function of increasing take-off angle (decreas-
ing depth) for each sample, fluorine signal increases beyond
the stoichiometric bulk content, further supporting the presence
of an enriched perfluorocarbon overlayer in these polymer
monolayer films. This is paralleled by the predicted opposite
trend in carbon signal with take-off angle; that is, a consistent
decrease in carbon below its bulk content with increasing take-
off angle (decreasing depth).
Sulfur content steadily decreases in all film surfaces with
decreasing sampling depth, reaching concentrations significantly
below the stoichiometric bulk content at the shallowest sampling
depth. This observation is consistent with the sulfur anchor
groups located under the film adjacent to the gold interface.
Furthermore, the silicon signal remains roughly constant despite
the various changes in other species with sampling depth,
indicating a relatively diffuse partitioning of the polysiloxane
backbone between a perfluorinated overlayer and an alkyl
disulfide underlayer. Some samples show a clear maximum in
the silicon signal at intermediate depths, consistent with a
polysiloxane-enriched layer between two adjacent, composi-
tionally distinct zones. The ratios of fluorine:silicon, carbon:
silicon, and sulfer:silicon are further evidence for a layered
polymer monolayer structure: perfluoroalkyl-enriched outer
surface layer, siloxane-enriched intermediate stratum, and a
basal, anchoring layer containing sulfur (disulfide) anchoring
chains.
Figure 6. Statis SIMS results for a monolayer film of polymer B on
gold: (a) positive ion spectrum (m/z ) 0 to 75); (b) positive ion
spectrum (m/z ) 75 to 150); (c) negative ion spectrum (F- counts )
1900).
enriched by perfluorocarbon chains at its outer surface (Figure
5
, A-D). Static SIMS results for the terpolymer monolayer
film of polymer B on Au, shown in Figure 6, provide yet further
support for this model. Prominent positive ion peaks (Figure
Static Secondary Ion Mass Spectrometry (SIMS).
A
6
a) are observed for fragments at m/z ) 31, 69, 100, 119, and
significant amount of recent work characterizing self-assembled
4
2
organic films has used SIMS surface analysis.41 Static SIMS
131, corresponding to perfluoroalkyl fragments.
Peaks rep-
resenting positive ion fragments of the allyl ether chain terminus
linking the perfluoroalkyl and disulfide side chains to the
siloxane backbone are observed at m/z ) 27, 29, 39, 41, 43,
measurements complement XPS data by virtue of the high
surface sensitivity of static SIMS (∼20 Å) which allows a more
complete examination of chemistry at the outermost film surface.
ADXPS measurements support a stratified monolayer assembly
+
and 45. The peak at m/z ) 28 can be attributed to Si from
the siloxane backbone. The negative secondary ion spectrum
(Figure 6b) also exhibits prominent peaks corresponding to
(
41) (a) Huang, J.; Hemminger, J. C. J. Am. Chem. Soc. 1993, 115, 3342.
(
b) Frisbie, C. D.; Martin, J. R.; Duff, R. R.; Wrighton, M. S. J. Am. Chem.
Soc. 1992, 114, 7142. (c) Tarlov, M. J.; Burgess, D. R. F.; Gillen, G. J.
Am. Chem. Soc. 1993, 115, 5305.
(42) Castner, D. G.; Lewis, K. B.; Fischer, D. A.; Ratner, B. D.; Gland,
J. L. Langmuir 1993, 9, 537.

Spontaneous Polymer Thin Film Assembly and Organization
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1863
indicative of structure in the polymeric monolayer. So-called
-
1
-1
“
axial CF2” bands at 1373 and 1346 cm and the 1257 cm
band have substantially increased intensity in the reflection
spectrum (Figure 7A) compared to the bulk spectrum (Figure
7
B). Recently, Lenk et al. have interpreted this effect to be
due to alignment of the axial CF2 stretch dipole moment change
2
6e
along helical -CF2- chains parallel with the surface normal.
-1
Enhancement of the 1257 cm band in (Figure 7A), tentatively
assigned to ν(C-C) and δ(C-C-C) components,2 also
supports orientation of perfluoroalkyl groups normal to the
surface. Additionally, changes of relative intensities for axial
CF2 stretching bands at 1373 and 1346 cm in grazing angle
spectra are consistent with spectra for well-ordered monolayer
6e
-
1
2
6
films of perfluoroalkanethiols on gold. In fact, reversal of
these intensity ratios between bulk (Figure 7B) and reflection
(
Figure 7A) spectra is consistent with a randomly ordered to
2
6c
organized transition.
Lastly, asymmetric CF2 stretching bands at 1221, 1205, and
-
1
1
151 cm decrease their relative intensity in the grazing angle
reflection spectrum compared to bulk. This, again, is consistent
with attenuated monolayer vibrations from dipole moment
changes perpendicular to a perfluorocarbon helical axis oriented
normal to the surface. A band observed at 725 cm , unique
to the reflection spectra and characteristic of the CF3 end group
stretching vibration, would be expected to be intensified in the
grazing angle spectrum for chains oriented normal to the
Figure 7. FTIR spectra for a polysiloxane derivatized with 74%
perfluoroalkyl side chains and 19% dithioalkyl side chains: (A) grazing
incidence reflection spectra (78°, E ) for a monolayer assembled on
⊥
gold and (B) bulk transmission spectra. Scale bars indicate respective
absorbance units for each spectra.
-1
2
6d
surface.
This generates an overall orientation of the C-F
perfluoroalkyl fragments (m/z ) 19, 31, 38, 69, 93, etc.) and
peaks corresponding to fragments from the siloxane backbone
bond in each CF3 group relatively perpendicular to the substrate
45
surface. Polysiloxanes with varied perfluoroalkyl contents all
exhibit these spectral features (data not shown). Changes in
the FTIR vibrational modes in the reflection spectra of the
polymer monolayers compared to their bulk transmission spectra
support a structural model where polymer fluorocarbon side
chain packing in the ultrathin films is more organized aniso-
tropically toward the surface normal than that in bulk coat-
(
m/z ) 16 and 43). Most importantly, the intensity of negative
-
-
ions originating from the sulfur anchor groups (S and SH at
m/z ) 32 and 33) are very weak. In contrast, assembled
monolayers of siloxane oligomers14 that have a nearly homo-
geneous distribution of thiol groups throughout the film produce
-
-
43
very strong S and SH signals.
Thus, the prominent
perfluoroalkyl fragments and very weak sulfur fragments of
Figure 6 are consistent with the stratified monolayer assembly
depicted in Figures 1 and 2.
2
6,45
ings.
Polymer Film Structure. On the basis of this extensive set
of surface analytical data, the structural picture supported by
FTIR, XPS, and SIMS data on these monolayer films is one
where perfluoroalkyl side chains extend away from the substrate-
anchored siloxane backbone to enrich the outermost film region
in an orientation more vertical than horizontal to the substrate
plane. Such an approximation is, in fact, quite close to the
structure proposed in Figure 1 for these ultrathin films. The
only specific chemical interactions mediating this organization
are the gold-thiolate bonds formed by anchoring side chains.
This ensures that some fraction of the alkyl disulfide grafted
chains are directly adjacent to the gold surface. Intrinsic
immiscibility (unfavorable enthalpy of mixing) between this
anchored alkyl component and more mobile siloxane and
perfluoroalkyl substituents likely promotes formation of a
stratified organic film, rich in perfluorinated groups near its
surface. However, an alternative explanation is that specific
alkyl side chain attachment to gold excludes the rest of the
polymer from this interface from entropic reasons, promoting
perfluoroalkyl surface enrichment as a consequence of polymer
interface exclusion. It is not unreasonable to assume some
degree of surface perfluoroalkyl self-association to form an
incomplete yet significantly organized overlayer atop the
polysiloxane superstructure: several precedents for such self-
organization in perfluoroalkane systems have been reported.46
Infrared Spectroscopy. Comparisons of vibrational FTIR
spectra acquired from these same ultrathin polymer films at
grazing incidence vs bulk coated polymer films also support
film organization through differences regarding both vibrational
peak position and relative intensity.44,26e Figure 7 presents the
FTIR spectral region with various C-F-active IR bands for a
perfluoroalkyl-grafted polysiloxane monolayer on gold contain-
ing 74% grafted perfluoroalkyl chains. The transmission spectra
of the bulk perfluoroalkyl-grafted polysiloxane film (Figure 7B)
is nearly identical to that calculated for a randomly oriented
perfluoroalkyl system.2
6d
Bands at 1373 and 1344 cm
-1
represent the symmetric CF2 stretch progression, or what has
2
6e
been recently called the “axial CF2 stretching vibrations.”
-
1
Bands at 1236, 1205 and 1151 cm represent fluorocarbon
asymmetric stretches [νa(CF2)]. Other bands characteristic of
-
1
the polysiloxane backbone [e.g., ν(Si-O), 1105 cm ] and
various hydrocarbon assignments are also present, but the
dichroism of the bands representing fluorocarbon-grafted side
chains, in particular, are important in understanding monolayer
organization.
The upper spectrum (A) in Figure 7 was recorded at a grazing
incidence (78°, E⊥ ) for a monolayer of the same polysiloxane
(74% perfluoroalkyl side chain content) on gold. Dramatic
changes in this spectra compared to the bulk spectra (B) as well
as polarization effects reflected in grazing incidence spectra are
(45) Bertilsson, L.; Liedberg, B. Langmuir 1993, 9, 141.
(
43) Castner, D. G.; Grainger, D. W. Unpublished results.
(46) (a) Russell, T. P.; Rabolt, J. F.; Twieg, R. J.; Siemens, R. L.; Farmer,
B. L. Macromolecules 1986, 19, 1135. (b) Mahler, W.; Guillon, D.; Skoulios,
A. Mol. Cryst. Liq. Cryst., Lett. Sect. 1985, 2, 111.
(44) Sun, F.; Mao, G.; Grainger, D. W.; Castner, D. G. Thin Solid Films
1
994, 242, 106.

1
864 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Sun et al.
Figure 8. ToF-SIMS imaging results for polymer ultrathin films of siloxane terpolymers lithographed after assembly using ion milling: (a) total
ion yield; (b) fluorine ion yield; (c) CN- contaminant ion yield exposed in gold substrate after ion milling; (d) Au substrate ion yield.
Surface Imaging Using Time-of-Flight Secondary Ion
Mass Spectrometry (ToF-SIMS). Time-of-flight mass spec-
trometry (ToF-SIMS) provides yet again a further dimension
to the surface analysis of these films by yielding both high mass
resolution for surface-derived fragments as well as the capability
to conveniently map and image all possible film chemical
fragment combinations across the sample’s surface. Several
recent studies of organic thin films have used ToF-SIMS to
image patterned substrates.47 Spectroscopy and imaging were
performed on patterned polymer films in both positive and
negative SIMS modes by ToF-SIMS. Quasi-molecular ions
originating from the perfluoro alkyl chains (e.g., CF, CF3, C2F2,
C F , C F , C F ...) are observed (data not shown).
3
3
2
5
3 5
Figure 8 is a ToF-SIMS image acquired from a 50 × 50 µm
area (256 × 256 pixels) of ion-milled lines (20 µm long and
widths ranging from 2 to 0.2 µm) in homogeneous polymer
assemblies. The upper left quadrant (Figure 8a) shows the total
secondary ion image. The contrast in the total ion image is the
result of variation in the secondary ion yields for the species
present in the analysis area. For this sample, topography (i.e.,
shadowing of the primary ion beam) does not contribute to the
observed contrast as the etching process removes monolayers
of material. The upper right quadrant (Figure 8b) shows the
(monomer units) were not observed in any of the acquired
spectra from the perfluorinated polymers; however, characteristic
fragment ions were observed. In the positive SIMS spectrum
the base peak was CF (m/z 31). Many fragment ions associated
with the fluorinated chain were observed such as CF, CF2, CF3,
C3F3, C2F4, and C3F5. In addition, fragment positive ions
originating from the siloxane portion of the polymer were also
observed; Si, CH3Si, and SiC3H9. In negative SIMS, fluorine
was the base peak in the spectrum and, again, fragment ions
-
fluorine (F ) distribution and hence the distribution of the
perfluorosiloxane. Images may be generated from any of the
-
fluoro-containing fragments; however, the fluorine (F ) image
provides the highest S/N ratio (image quality) from which line
scans can easily be determined. The line scans generated from
this image (data not included) show that actual line widths from
the ion mill etched areas are 1.3, 2.1, 2.5, and 3.5 µm,
respectively, which deviate substantially from the ion beam’s
focused resolution during milling. The lower left quadrant
(
47) (a) Offord, D. A.; John, C. A.; Griffin, J. H. Langmuir 1994, 10,
61. (b) Vargo, T. G.; Thompson, P. M.; Gerenser, L. J.; Valentini, R. F.;
Aebischer, P.; Hook, D. J.; Gardella, J. A. Langmuir 1992, 8, 130.
7
-
(Figure 8c) shows the distribution of CN ion, likely a trace-

Spontaneous Polymer Thin Film Assembly and Organization
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1865
Figure 9. ToF-SIMS imaging results for siloxane terpolymers assembled from dilute solution onto photolithographed gold lines on SiO
a) total ion yield; (b) oxygen ion yield (primarily SiO substrate); (c) perfluorocarbon fragment ion yield (polymer film); (d) Au ion yield (lithographed
gold lines).
2
wafers:
(
2
level contaminant in the gold underlayer, which even at a gold
purity of 99.999% seems exposed during the ion milling of the
organic overlayer. This ion is not detected by XPS but is easily
imaged by the ToF instrument in the ion-milled, exposed gold
areas. Detecting and locating contaminants in these materials
is very important to the understanding and characterization of
these materials. The parallel analytical capabilities of both XPS
and ToF-SIMS is, therefore, significant and complimentary.
Figure 9 shows analogous ToF-SIMS images for bound
polymer films of the siloxane terpolymer C anchored specifically
on photolithographed gold lines (2 µm wide) on SiO2 wafers.
We have previously used scanning Auger mapping to demon-
strate that disulfide-derivatized acrylate polymers attach to
lithographed gold lines specifically by solution exposure of
thin polymeric film. The high mass resolution spectra ac-
companying these ion fragment maps will be discussed in a
future publication.
Conclusions
Our interpretation of these data confirms our hypothesis that
polymer coil-to-monolayer transitions can be driven by specific
anchoring events at interfaces, and together with chain-chain
immiscibility, allowing three-dimensional polymer coils in
solution to form two-dimensional networks at the expense of
entropic tendencies. Significantly, poorly miscible components
mutually tied to a common polymer main chain are compelled
to segregate, even in these two-dimensional monolayer film
assemblies. Low surface energy perfluoro side chains, therefore,
migrate to the surface of these films, away from anchoring
hydrocarbon chains, yielding a distinctly structured, self-
organized polymer monolayer with a fluoroalkyl-enriched
surface.
patterned wafers to soluble polymer.1
2c
The ToF-SIMS ion
fragment “maps” show that the siloxane terpolymer attaches
exclusively to the patterned gold lines, mediated by Au
substrate-disulfide side chain anchor interactions. Oxygen
-
negative ions (Figure 9b) map to SiO2 areas while CF ions
(
9
Figure 9c) map accurately to the gold lines (compare to Figure
d). Both sets of ToF-SIMS images provide detailed spatial
information on the distribution of surface chemistry present in
Dense, monolayer polymer films can be anchored on solid
surfaces by spontaneous assembly from dilute solution. Judi-
cious choice of polymer components leads to films which self-

1
866 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Sun et al.
organize to yield bound anisotropic structures. We have also
demonstrated selected partitioning of polymer components to
yield surface enrichment. This has practical implications,
namely, tailoring of interfacial properties by predictable anchor-
ing of polymers to surfaces in ultrathin arrays, and further routes
for chemically modifying these arrays in situ as a versatile
surface modification strategy. These methods may also prove
useful for fundamental studies of polymer chain behavior at
interfaces.
Graduate Institute) provided outstanding assistance with focused
ion beam milling to prepare lithographed polymer surfaces.
Support is gratefully acknowledged from the Whitaker Founda-
tion (D.W.G.), National Science Foundation/EPRI Grant MSS-
9212496 (D.W.G.), National Science Foundation Grant DMR-
9357439 (D.W.G.), Tektronix, Inc. (D.W.G.), and National
Institutes of Health Grant RR-01296 (D.G.C.).
Supporting Information Available: Text giving experi-
mental details (3 pages). This material is contained in many
libraries on microfiche, immediately follows this article in the
microfilm version of the journal, can be ordered from the ACS,
and can be downloaded from the Internet; see any current
masthead page for ordering information and Internet access
instructions.
Acknowledgment. We thank C. R. Martin, M. D. Porter, J.
D. Andrade, H. Wolf, J. Venzmer, M. Lerner, and T. J. Lenk
for their critique and suggestions. The authors also thank Dr.
Bob Drosd (BIT, Inc., Portland, OR) for supplying freshly
photolithographed gold lines on silicon wafers and Dr. J. K.
Hurst (Washington State University) for assistance with cyclic
voltammetry experiments. John Hunt and Jim Cser (Oregon
JA952225T