fluoropolymers with very low surface energy characteristics
TRANSCRIPT
Fluoropolymers with very low surface energy characteristics
Paul Graham, Maureen Stone, Adrian Thorpe, Thomas G. Nevell, John Tsibouklis*
School of Pharmacy and Biomedical Sciences, University of Portsmouth, St. Michael's Building, White Swan Road, Portsmouth, PO1 2DT, UK
Abstract
The synthesis and characterisation of three homologous series of polymeric materials is reported and the surface energies of ®lm
structures prepared from these are evaluated. Dependent upon the molecular design features of these materials, the initial surface energies
of ®lm structures were within the range 6±35 mJ mÿ2 but in many cases these were found to change over time due to the penetration of the
surfaces by the liquids utilised for the contact angle measurements. # 2000 Elsevier Science S.A. All rights reserved.
Keywords: Poly(methylpropenoxy¯uoroalkylsiloxane)s; Poly(per¯uoroalkylacrylate)s; Surface energy; Contact angle goniometry
1. Introduction
The commercial implications of low-surface-energy poly-
meric materials with good ®lm-forming characteristics
become apparent when the lack of universally applicable
and environmentally friendly protection against biological
fouling is considered [1,2].
It has been suggested that low-surface-energy polymers
must possess a ¯exible linear backbone onto which side-
chains with low intermolecular interactions are attached via
suitable linking groups [3]. In this work we consider two
classes of compounds that conform to these molecular
design requirements, namely: the poly(methylpropenoxy-
¯uoroalkylsiloxane)s and the poly(per¯uoroalkylacrylate)s.
The non-¯uorinated counterparts of the former class of
materials are also considered for comparison.
2. Experimental
2.1. Materials
Divinyltetramethyldisiloxane platinum catalyst was
obtained from Fluorochem and used as received. Poly-
(methylhydrosiloxane) (Mn�2270), acrylic acid, HPLC
grade tetrahydrofuran and all alcohols were obtained from
Aldrich and used as supplied. Toluene, used in the hydro-
silylation reaction, was distilled (nitrogen atmosphere) from
benzophenone and sodium metal.
2.2. Instrumental
Solution 1H-NMR and 13C-NMR spectra were recorded
using a Jeol GSX spectrometer operating at 270.05 MHz and
67.80 MHz, respectively, using CDCl3 as solvent and TMS
as reference. IR spectra (neat; NaCl plates) were obtained
using a Perkin Elmer Paragon 1000 FT-IR spectropho-
tometer. Electron-impact mass spectra were recorded on a
Jeol JMS-DX303/DA5000 double-focusing mass spectro-
meter with Matsuda optics. The purity of all the synthesised
low-molecular-mass compounds was checked by thin layer
chromatography, TLC, (silica, 1:1 v/v n-hexane:diethyl
ether) and temperature-ramped GC using a Hewlett Packard
5890 J gas chromatograph equipped with a 25 m DB225
quartz capillary column (internal diameter 0.22 mm, ®lm
thickness 25 mm); a Jeol JMS-DX303/DA5000 mass spec-
trometric detector was employed. Experiments were per-
formed under helium (12.4 mm3 minÿ1) using an ionising
voltage of 70 eV with a current of 100 mA (temperature
pro®le: 288C, 0.5 min; 10 K minÿ1 to 80 8C, 15 min; 10
K minÿ1 to 2408C, 5 min). Molecular weight distribution
studies of the polymers were performed by gel permeation
chromatography, GPC, in tetrahydrofuran using a Waters
instrument equipped with a PHENOGEL 5 linear column
(30 cm) and a differential refractometric detector. Measure-
ments were carried out at room temperature using a solvent
¯ow rate of 1 cm3 minÿ1. A GPC calibration graph was
obtained using a series of polystyrene standards. Studies of
thermal properties were carried out using a Perkin Elmer
DSC7 differential scanning calorimeter (DSC), a Stanton
Redcroft 671 differential thermal analyser (DTA) and a
Stanton Redcroft TG-750 thermogravimetric analyser
Journal of Fluorine Chemistry 104 (2000) 29±36
* Corresponding author. Fax: �44-1705843565.
0022-1139/00/$ ± see front matter # 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ( 0 0 ) 0 0 2 2 4 - 4
(TGA). Atomic force microscopy (AFM) studies were
performed in air under ambient conditions, using a Dis-
coverer TopoMetrix TMX2000 scanning probe microscope
(SPM) which was mounted on a custom-built mass/spring
anti-vibration rig with a lateral natural frequency of 0.40 Hz
and a vertical natural frequency of 0.52 Hz.
2.3. Synthesis
All classes of materials are readily accessible; the
poly(per¯uoroacrylate)s are prepared by the bulk polymer-
isation of acrylate esters [4] whereas the synthesis of the
poly(methyl-propenoxy¯uoroalkylsiloxane)s [5] and the
poly(methylpropenoxyalkylsiloxane)s [6] involves a hydro-
silylation reaction.
2.3.1. Poly(perfluoroalkylacrylate)s, PFA
The readily obtainable monomers were polymerised, in
the bulk (no solvent; 1% w/w AIBN; 608C), to give the
corresponding poly(per¯uoroalkylacrylate)s, Fig. 1. The
polymers were puri®ed by repeated washings in methanol
and acetone and dried under reduced pressure.
2.3.2. Poly(methylpropenoxyfluoroalkylsiloxane)s, PFE
The synthesis of the poly(methylpropenoxy¯uoroalkyl-
siloxane)s, PFE, is outlined in Fig. 2.
Per¯uoroalkyl allyl ethers (CF3±(CF2)n±(CH2)2±O±CH2±
CH=CH2) were prepared from the corresponding alcohols
and allyl bromide using phase transfer catalysis [7].
For the grafting reaction [8,9], this derivative (1.0 g;
10 mol% excess against the number of Si±H groups in
the poly(methylhydrosiloxane)) was dissolved in dry,
freshly distilled toluene (100 cm3) and the appropriate
amount of poly(methylhydrosiloxane) was added. The
reaction mixture was heated to re¯ux (1108C) under
nitrogen and the divinyltetramethyldisiloxane platinum
catalyst (3% w/v solution in xylene; 100 mg equivalent
of platinum) was injected using a syringe. The reaction
mixture was re¯uxed (nitrogen atmosphere) for a further
24 h, after which time IR analysis showed that the hydro-
silylation reaction was complete. After removal of the
toluene, the polymers were puri®ed by several washes with
methanol.
2.3.3. Poly(methylpropenoxyalkylsiloxane)s, PES
The synthesis of the modi®ed polysiloxanes is outlined in
Fig. 3.
Sodium metal (1.15 g, 0.05 mol) was added to a solution
of the appropriate n-alcohol (0.06 mol) in anhydrous tetra-
hydrofuran (100 cm3) contained in a clean, dry round
bottom ¯ask (250 cm3) equipped with a double surface
condenser. After the sodium had completely dissolved
(ca. 24 h), allyl bromide (6.05 g, 0.05 mol) was added
dropwise and the mixture was allowed to re¯ux for a further
24 h. The cooled mixture was ®ltered (to remove the sodium
bromide) and the solvent was removed by rotary evapora-
tion to yield a clear yellow liquid which was further puri®ed
by column chromatography (silica, 1:1 v/v n-hexane :
diethyl ether). In each case, the pure product was obtained
as a colourless liquid. The propenoxyalkane derivative
(1.0 g; 10 mol% excess against the number of Si±H groups
in the poly(methylhydrosiloxane)) was dissolved in dry,
freshly distilled toluene (100 cm3) and the appropriate
amount of poly(methylhydrosiloxane) was added. The reac-
tion mixture was heated to re¯ux (1108C) under nitrogen
and the divinyltetramethyldisiloxane platinum catalyst
(3% w/v solution in xylene; 100 mg equivalent of platinum)
was injected with a syringe. The reaction mixture was
re¯uxed (N2 atmosphere) for a further 24 h after which
time IR analysis showed that the hydrosilylation reaction
Fig. 1. Polymer synthesis: n�3 (PFA3); n�5 (PFA5); n�7 (PFA7); n�9 (PFA9).
Fig. 2. Polymer synthesis: n�3 (PFE3); n�5 (PFE5); n�7 (PFE7); n�9
(PFE9).
30 P. Graham et al. / Journal of Fluorine Chemistry 104 (2000) 29±36
was complete. The polymers were separated and puri®ed by
several reprecipitations from tetrahydrofuran solution into
methanol.
2.4. Film formation
Films of the synthesised polymers were deposited on
glass and/or poly(methylmethacrylate) supporting sub-
strates (10 mm�10 mm�1 mm). The acrylates were depos-
ited from the melt or from CF2ClCFCl2 (0.1% w/w) solution
(dipping speed: 10 mm sÿ1) without further treatment. The,
liquid, silicone samples formed into glassy structures after
thermal crosslinking at 1058C for 16 h; it was found that a
slow cooling rate (<108C hÿ1) was important in determining
®lm quality.
2.5. Goniometry
The surface free energies of polymer samples were
determined by contact-angle goniometry using a manual
Kernco instrument or a Kruss G10 goniometer interfaced to
image capture software; both instruments were ®tted with an
enclosed thermostated cell. Relaxed ®lm structures that had
been annealed at 258C for at least 3 weeks were utilised
for the experiments. For both advancing (yA) and receding
(yR) contact angle experiments, measurements of droplets
(2±10 ml) were recorded at 25�18C using doubly distilled
water (surface tension gL�73.4 mN mÿ1 at 18.758C, litera-
ture value�73.05 mN mÿ1 at 188C [10,11]), diiodomethane
(>99%; gL�48.7 mN mÿ1 at 18.88C, literature value�50.76 mN mÿ1 at 208C [10,11]) and 1,1-ethanediol (ethy-
lene glycol, >99%; gL�47.7 mN mÿ1 at 18.88C, literature
value�48.40 mN mÿ1 at 208C [10,11]).
3. Results and discussion
3.1. Materials
The chemistry of the polyacrylates (Fig. 1) is thoroughly
documented in standard textbooks whereas that of poly(-
methylpropenoxy¯uoroalkylsiloxane)s (Fig. 2) has been
described elsewhere [5].
The preparation of the propenoxyalkanes by phase trans-
fer catalysis resulted in yields in the range 84±90% whereas
a synthetic procedure utilising the SN2 reaction of allyl
bromide with the corresponding alcohol (Williamson Synth-
esis) gave signi®cantly lower yields (61±37%). The phase
transfer reaction procedure proved to be a less complicated
route in that: (i) no potentially hazardous metallic sodium is
required, (ii) the long period of time necessary to form the
alkoxide is eliminated and (iii) in contrast to the Williamson
synthesis, the method may be employed for the preparation
of ¯uorinated ethers.
The infrared spectra of all propenoxyalkanes were char-
acterised by absorptions at 1647 cmÿ1, ethylenic n(C=C),
and at 1105 cmÿ1, n(C±O); 1H-NMR and 13C-NMR spectra
were consistent with the expected structures [6]. The purity
of all the synthesised allylic ethers was checked by TLC and
further con®rmed by GC-MS; a single component was
identi®ed in all cases. The molecular ion, [CH3(CH2)n±
O±CH2±CH=CH2]�, and characteristic fragments due to the
long hydrocarbon chain as well as those due to the allylic
system (m/z 41) were observed in all cases.
The poly(methylpropenoxyalkylsiloxane)s were synthe-
sised by the reaction of the above propenoxyalkylsiloxanes
with poly(methylhydrosiloxane) as shown in Fig. 3. The
utilisation of an excess amount of the propenoxyalkane
derivative allowed the reaction to go to completion. Any
excess propenoxyalkane was removed by repeated precipi-
tation from tetrahydrofuran solution with methanol.
The preparation of the polymers by the grafting reaction
with Karstedt catalyst requires ultra-dry conditions as
moisture causes deactivation; sodium dried toluene and a
nitrogen atmosphere are employed. An excess amount of
the allylic ether ensures the reaction goes to completion.
The reaction can be followed spectroscopically: the strong
Si±H absorption at �2165 cmÿ1 and the ±C=C stretch at
�1650 cmÿ1 are seen to progressively decrease in intensity
and eventually disappear as the reaction moves to comple-
tion. In addition, the IR spectra of the poly(methylhydro-
siloxane)s are characterised by strong absorptions in the
regions 1000±1200 cmÿ1 [n(Si±O±Si)] and 800 cmÿ1±
900 cmÿ1 [n(Si±C)]; the Si±CH3 deformation mode is at
1258 cmÿ1 (Fig. 4).
Fig. 3. Polymer synthesis: n�4 (PES4); n�6 (PES6); n�8 (PES8); n�10 (PES10); n�12 (PES12); n�14 (PES14); n�16 (PES16).
P. Graham et al. / Journal of Fluorine Chemistry 104 (2000) 29±36 31
All polymers (except for n�14 and n�16) were obtained
as viscous liquids. It was found that polymers substituted
with lower propenoxyalkanes (up to n�12) had a tendency
to trap organic solvents. Attempts to remove these by rotary
evaporation or distillation resulted in the materials cross-
linking to form clear, glassy solids. The n�14 and n�16 poly-
mers were obtained as low-melting white solids; recoveries
were 96 and 97%, respectively. In all cases, 1H-NMR spectro-
scopy was used to follow the disappearance of the Si±H bond.
GPC results (Fig. 5) demonstrated an increase in average
molar mass from 2270 to ca. 11,800 for the polymer n�16.
In accordance with expectation, this corresponds with the
addition of about 30 ether units to each poly(methylhydro-
siloxane) chain. Analogous GPC analysis results were
obtained with all other synthesised polymers.
It was found that all polymer samples could be cross-
linked on thermal annealing at 1108C. In order to examine
the effect of partial hydrosilylation on the cross-linking
reaction, a series of poly(methylpropenoxyalkylsiloxane)s
was prepared using equimolar quantities of reactants. Even
after a 24 h reaction time at 1108C, IR [n(Si±H), 2166 cmÿ1]
and 1H-NMR [d(Si±H), 4.7 ppm] spectroscopies con®rmed
the presence of unreacted Si±H moieties (1±2 sites per
average polymer chain). Film structures prepared from
partially hydrosilylated materials were found to be parti-
cularly susceptible to thermal cross-linking.
Differential scanning calorimetry showed the presence of
a melting endotherm (448C) for the higher-melting polymer
(n�16), suggesting a degree of crystallinity within the
sample; thermally cross-linked samples of the same poly-
mer did not show this endotherm. In accordance with the
NMR experiments, thermogravimetric analysis (Fig. 6),
combined with DTA (Fig. 7) and DSC experiments, con-
®rmed the presence of trapped solvent within all liquid
polymer samples. A common feature characterising the
dynamic DTA and DSC thermograms (®rst heating) of
Fig. 4. FT-IR spectrum of poly(methylpropenoxyoctylsiloxane).
Fig. 5. GPC analysis results presented as log (molecular weight) versus retention time; polystyrene provides the calibration.
32 P. Graham et al. / Journal of Fluorine Chemistry 104 (2000) 29±36
the solid poly(methylpropenoxyalkylsiloxane)s (n�14 and
n�16) is the presence of a two-step exotherm. The onset of
the ®rst exotherm, for solid samples heated at 10 K minÿ1, is
at ca. 1108C. Since this is not observed in subsequent
heating cycles, it is assumed to be associated with the
cross-linking process characterising these materials.
Liquid samples exhibited a similar thermal behaviour
which, was nonetheless, complicated by the presence of an
evaporation exotherm, the position and intensity of which
were found to be a function of the nitrogen ¯ow rate as well
as the heating rate. The onset of degradation, for all poly-
mers, is at ca. 2208C.
3.2. Surface energies
Surface energies were evaluated using the surface-ten-
sion-component theory [12±15].
3.2.1. Poly(1H,1H,2H,2H-perfluoroalkyl acrylate)s
The surface energy of the poly(1H,1H,2H,2H-per¯uorodo-
decyl acrylate) ®lm structures has been evaluated as 6 mJ mÿ2
(Table 1); the material combines good hydrophobicity with
unusual incompatibility with the organic liquids used.
The large difference between the initial surface free
energies of poly(1H,1H,2H,2H-per¯uorooctyl acrylate)
(7.0�0.4 mJ mÿ2) and poly(octyl acrylate) (33.1�0.9
mJ mÿ2) demonstrate the signi®cance of ¯uorine substitution
in the molecular design requirements for such materials.
For all liquids employed, the advancing contact angles
observed for the two higher esters, poly(1H,1H,2H,2H-
per¯uorodecylacrylate) and poly(1H,1H,2H,2H-per¯uoro-
dodecyl-acrylate), were time-independent (monitored over
12 months) whereas those for the lower analogues and for
the non-¯uorinated material were found to decrease with
time. The extremely low surface energies associated with
the higher-¯uorocarbon-based materials indicate very little
af®nity for the liquids used; the time-independent behaviour
con®rms that the liquids do not become absorbed.
The length of the per¯uorocarbon chain is important in
determining the time-dependence of the wetting behaviour;
the surface free energy of poly(1H,1H,2H,2H-per¯uorooc-
tyl acrylate) increased, over time, at a much lower rate than
that of poly(1H,1H,2H,2H-per¯uorohexyl acrylate). From
observations of the dimensions of liquid drops, it appeared
that penetration of the surface by the liquids was responsible
for the observed reduction in contact angles; with the polar
protic liquids, hydrogen bonding with oxygen atoms of the
ester-linking group may be responsible for this behaviour
whereas with diiodomethane, interactions with the hydro-
carbon backbone may provide the driving force behind the
observed phenomena.The n-hexadecane contact angle is
Fig. 6. Thermogram of poly(methylpropenoxydodecylsiloxane).
Fig. 7. DTA thermogram of poly(methylpropenoxyhexadecylsiloxane).
P. Graham et al. / Journal of Fluorine Chemistry 104 (2000) 29±36 33
generally accepted as the index of oleophobicity [1]. For
poly(1H,1H,2H,2H-per¯uorododecyl acrylate) this contact
angle was found to be 838. For comparison, it is worth
noting that ¯uoropolymers have been identi®ed [16,17]
with n-hexadecane values of 858, but it is the combined
effect of the hydrophobic (1258) and oleophobic nature
of poly(1H,1H,2H,2H-per¯uorododecyl acrylate) that is
responsible for the unusually low surface-energy value
associated with this material; surface energies of the same
order of magnitude (6±8 mJ mÿ2) have only been reported
for, thermodynamically unstable, Langmuir±Blodgett ®lm
structures of per¯uorocarboxylic acids [18±21].
Receding contact angles were also measured for the
two materials that exhibited time-independent behaviour.
To this end, a slightly larger drop (3±6 ml) was placed on
the surface of the sample and allowed to stand for 30 min.
Using a microcapillary pipette, a portion of liquid (1±2 ml)
was then removed, care being taken to avoid disturbing
the edge of the drop; the contact angle was measured
after ca. 30 s. Following small changes which occurred
rapidly after the drop size was reduced, receding contact
angles, yR, remained almost constant and were lower than
advancing angles, yA. For both samples, hysteresis
(H�yAÿyR, Table 1) was greatest for ethylene glycol.
The small hysteresis effects shown with water imply that
the surfaces of poly(1H,1H,2H,2H-per¯uorodecyl acrylate)
and poly(1H,1H,2H,2H-per¯uorododecyl acrylate) were
mechanically smooth and homogeneous with respect to
van der Waals' and/or hydrogen bonded interactions;
surface roughness pro®ling using AFM con®rmed the
smoothness of the surface (Ra 0.49 nm) [22,23].
3.2.2. Poly(methylpropenoxyfluoroalkylsiloxane)s
Depending on the length of the pendent ¯uorocarbon
chain, the calculated initial surface energies of the poly-
(methylpropenoxy¯uoroalkylsiloxane) ®lm structures, as
determined from advancing contact angle measurements,
range between 17 and 9 mJ mÿ2 (Table 2).
Table 1
Contact angles and hysteresisa
Sample t (minÿ1) Contact angle, (y/8); (H/8) Surface energy (mJ mÿ2)
H2O DIM EG gLWS gS
� gSÿ gS
PFDDA 0 125 (4) 112 (16) 120 (25) 5.0 0.1 1.6 5.6
30 123 (4) 112 (18) 119 (26) 5.0 0.1 2.8 6.1
PFDA 0 117 (8) 108 (12) 108 (17) 6.2 0.1 2.2 6.9
30 106 (2) 107 (15) 103 (13) 6.3 0.0 7.3 7.3
PFOA 0 114 105 108 6.9 0.0 3.7 7.0
30 89 100 89 8.7 0.2 18.0 12.1
PFHA 0 113 96 110 10.2 0.4 4.3 12.9
30 86 69 89 23.3 1.2 18.9 32.6
POA 0 96 52 70 33.0 0.0 1.5 33.1
30 54 42 63 38.6 1.3 46.5 54.2
a (H) for water, diiodomethane (DIM) and ethylene glycol (EG) on poly(1H,1H,2H,2H-perfluorododecyl acrylate) PFDDA, poly(1H,1H,2H,2H-
perfluorodecyl acrylate) PFDA, poly(1H,1H,2H,2H-perfluorooctyl acrylate) PFOA, poly(1H,1H,2H,2H-perfluorohexylacrylate) PFHA, poly(octyl acrylate)
POA and surface energies of these materials. Each contact angle value is the mean of six drops on two independently prepared polymer samples; standard
deviations were in the range 0.6±1.1.
Table 2
Contact angles for watera
Sample Time Contact angle, (yA/8) Surface energy (mJ mÿ2)
n (Ra nmÿ1)(t minÿ1)
H2O DIM EG gLWS gS
� gSÿ gS
PFE3 0.5 0 105.0 89.5 79.2 12.9 0.1 4.6 14.2
30 84.3 72.9 67.3 21.3 6.4 1.3 26.9
PFE5 1.0 0 102.4 86.9 82.6 14.1 0.8 1.7 16.4
30 89.6 82.2 80.8 16.4 9.9 0.1 17.8
PFE7 2.3 0 106.7 95.6 91.3 10.4 1.3 1.1 12.6
30 99.5 89.8 92.0 12.8 7.2 0.0 13.2
PFE9 3.1 0 109.3 94.8 94.3 10.7 1.3 0.5 12.2
30 100 94.0 94.5 11.0 7.8 0.0 11.8
PFE9u 133.3 0 117.1 102.2 102.1 7.8 0.5 0.5 8.8
30 109.5 102.7 100.9 7.7 3.1 0.2 9.4
a Diiodomethane (DIM) and ethylene glycol (EG) on poly(methylpropenoxyfluoroalkylsiloxane)s and surface energies of these materials. Each contact
angle value is the mean of six drops on two independently prepared polymer samples (standard deviations were in the range 0.6±1.9). PFE9u has not been
subjected to thermal treatment. (Ra�surface roughness as determined by AFM measurements).
34 P. Graham et al. / Journal of Fluorine Chemistry 104 (2000) 29±36
All materials were found to combine good hydrophobi-
city with unusual incompatibility with the organic liquids
used. However, from observations of the dimensions of
water drops, it appeared that penetration of the surface
by this liquid was responsible for the observed reduction
in contact angles. This may be brought about by hydrogen
bonding with oxygen atoms of the ether-linking group.
Analogous behaviour was observed with diiodomethane
for all polymers except PFE9. The lower surface energy
associated with the non-crosslinked polymer (PFE9u), rela-
tive to its crosslinked counterpart (PFE9), can be readily
explained in terms of the `¯exible-backbone' molecular
design requirement [1].
Since on surfaces of PFE9 contact angles for ethylene
glycol and diiodomethane remained constant for 30 min
(Table 2), receding contact angle experiments were per-
formed in order to assess hysteresis effects. To this end, a
diiodomethane or ethylene glycol drop (8±10 ml) was placed
on the surface of the sample. The liquid was then removed in
small increments (0.5 ml) until the drop edge spontaneously
contracted to a new stationary position at which the receding
contact angle was measured. In the case of PFE9 ®lm-
structures. Hysteresis was of the order of 88 for ethylene
glycol and 128 for diiodomethane, indicating considerable
surface roughness. Surface roughness pro®ling using AFM
con®rmed these ®ndings (Ra 3.24 nm). The surface rough-
ness of the PFE ®lm-structures was found to be a function of
the length of the propenoxy¯uoroalkyl side-chain. Thus,
®lm structures prepared from the short side-chain analogue,
PFE3, exhibited hysteresis values of the order of 38; AFM
imaging further con®rmed the relative smoothness of this
surface (Ra 0.49 nm). The small hysteresis effects associated
with this material imply that the surface of PFE3 was
mechanically smooth and homogeneous with respect to
van der Waals' and/or hydrogen bonded interactions
[16,17].
3.2.3. Poly(methylpropenoxyalkylsiloxane)s
Depending on the length of the pendent hydrocarbon
chain, the calculated initial surface energies of the poly(-
methylpropenoxyalkylsiloxane) ®lm structures, as deter-
mined from advancing contact angle measurements,
range between 35 and 21 mJ mÿ2 (Table 3). It must be
noted that, because of the observed variation in surface
roughness between samples and the absence of receding
contact angle data, the reported values do not represent the
true surface energy of the materials and are presented for
reasons of comparison only. In particular, the dramatic
decrease in the calculated initial surface energy of the
n�10 polymer as compared to the n�8 analogue
(Table 3) may be a re¯ection of the large differences in
average surface roughness associated with ®lm structures
prepared using these materials whereas the corresponding
values for the n�14 and n�16 polymers are those of very
rough surfaces and, hence, of little value.
From observations of the dimensions of drops of the three
liquids under consideration, it appeared that penetration of
the surface by the liquids was responsible for the observed
reduction in contact angles and consequent time dependence
of the surface energy [14]. In the case of water, hydrogen
bonding with oxygen atoms of the ether-linking group may
provide the driving force behind the observed phenomenon
whereas with diiodomethane and ethylene glycol, enhanced
van der Waals interactions due to the hydrocarbon nature of
the side-chains may account for this effect. This hypothesis
is supported by the observation that the ethylene glycol
contact angle associated with the shortest hydrocarbon
analogue (n�4) exhibited a time-independent behaviour
Table 3
Surface roughness dataa
Sample Time Contact angle (yA/8) Surface energy (mJ mÿ2)
(n) (Ra nmÿ1) (t minÿ1) H2O DIM EG gLWS gS
� gSÿ gS
4 1.9 0 79.6 55.2 70.9 31.3 0.3 14.9 35.4
30 61.6 51.0 70.8 33.7 1.5 42.1 49.5
6 3.5 0 91.0 62.1 82.0 27.3 0.5 8.7 31.6
30 77.3 57.6 81.2 29.9 1.6 25.1 42.7
8 8.2 0 72.2 51.6 61.2 33.3 0.1 18.5 35.2
30 63.1 46.4 54.2 36.2 0.1 26.2 38.4
10 5.6 0 80.5 66.8 73.8 24.6 0.1 16.4 26.6
30 72.8 63.2 66.1 26.7 0.0 21.6 27.5
12 0.9 0 85.2 70.0 75.4 22.9 0.0 12.0 23.4
30 73.4 62.4 67.7 27.2 0.0 21.7 29.2
14 115.5 0 84.7 72.1 75.6 21.7 0.0 12.8 21.8
30 70.0 63.6 70.3 26.5 0.2 28.7 31.2
16 99.9 0 106.9 73.4 81.8 21.0 0.1 0.3 21.3
30 99.7 64.6 82.8 25.9 0.2 2.9 27.3
a (Ra value; mean of at least six samples) as determined by AFM and contact angles (mean of six drops on two independently prepared samples) for water,
diiodomethane (DIM) and ethylene glycol (EG) on poly(methylpropenoxyalkylsiloxane)s (n�4, 6, 8, 10, 12, 14, 16). The surface energies of these materials
are also given. Standard deviations were in the range 0.4±3.4.
P. Graham et al. / Journal of Fluorine Chemistry 104 (2000) 29±36 35
and that for the next member of the series (n�6) was only
weakly time-dependent.
Since contact angles on the surfaces of ®lm structures
from the polymers under consideration were found to
change with time, receding contact angle experiments were
not performed. However, surface roughness pro®ling using
AFM (performed on at least six, independently prepared,
®lm samples) showed that all ®lms were relatively rough
(Table 3). The n�10 polymer was found to produce parti-
cularly rough surfaces (Ra 8.15 nm) with the greatest varia-
tion between samples whereas the n�12 analogue
reproducibly yielded the smoothest ®lm structures (Ra
0.90 nm). Structures derived from the two solid polymers,
n�14 and n�16, were very rough (Ra 115.5 and 99.9 nm,
respectively) rendering the reported surface energy values
essentially meaningless.
4. Conclusions
Three homologous series of polymers have been synthe-
sised and characterised and the surface energies of ®lm
structures prepared from these have been evaluated. Depen-
dent upon the type of backbone as well as nature and length
of the pendant hydrocarbon chain, the initial surface ener-
gies were within the range 6±35 mJ mÿ2 but in many cases
these were found to change over time due to the penetration
of the surfaces by the liquids utilised for the contact angle
measurements.
Acknowledgements
The ®nancial support of this work by EPSRC/MTD Ltd
(research studentships to MS and AT) and NERC (research
studentship to PG) is gratefully acknowledged.
References
[1] H. Kobayashi, M.J. Owen, Trends in Polymer Science 3 (1995) 10.
[2] E.F. Cuddihy, in: K. Mittal (Ed.), Particles on Surfaces 1: Detection,
Adhesion and Removal, Plenum Press, New York, 1988, p. 91.
[3] M.J. Owen, Comments Inorg. Chem 7 (1988) 195.
[4] M. Stone, T.G. Nevell, J. Tsibouklis, Mater. Lett. 37 (1998) 102.
[5] A.A. Thorpe, S.A. Young, T.G. Nevell, J. Tsibouklis, Appl. Surf. Sci.
136 (1998) 99.
[6] A.A. Thorpe, T.G. Nevell, J. Tsibouklis, Appl. Surf. Sci. 137 (1999)
1±10.
[7] R. Dorigo, D. Teyssie, J.M. Yu, S. Boileau, Polym. Prepr, Am.
Chem. Soc. Div. Polym. Chem. 31 (1990) 420.
[8] B. Youssef, B. Boutevin, A.M. Garnault, S. Boileau, J. Fluor. Chem.
35 (1987) 399.
[9] S. Boileau, Polym. Mater. Sci. Eng. 56 (1987) 384.
[10] G.W.C. Kaye, T.H. Laby (Eds.), Table of Physical and Chemical
Constants, 15th Edition, Longman, Harlow, 1992.
[11] D.R. Lide (Ed.), Handbook of Chemistry and Physics 1995±1996,
76th Edition, CRC Press Boca, Raton, 1995.
[12] R.G. Good, C.J. van Oss, in: M.E. Schrader, G. Loeb, Modern
Approaches to Wettability: Theory and Applications, Plenum Press
New York, 1991, pp. 1±27.
[13] R.G. Good, M.K. Chaudhury, C.J. van Oss, in: L.H. Lee (Ed.),
Fundamentals of Adhesion, Plenum Press, New York, 1991, pp. 153±
172.
[14] T.G. Nevell, D.P. Edwards, A.J. Davis, R.A. Pullin, Biofouling 10
(1996) 199.
[15] C.J. Drummond, D.Y.C. Chan, Langmuir, 13 (1997) 38±90 and
references therein.
[16] H. Kobayashi, Makromol. Chem. 194 (1993) 2569.
[17] S.J. McLain, B. Sauer, L. Firment, Polym. Prepr, Am. Chem. Soc.
Div. Polym. Chem. 34 (1993) 666.
[18] H.W. Fox, W.A. Zisman, J. Colloid Interface Sci. 5 (1950) 514.
[19] M. Suzuki, Y. Saotome, M. Yanagisawa, Thin Solid Films 160
(1988) 453.
[20] H. Nakahama, S. Miyato, T. Wang, S. Tasaka, Thin Solid Films 141
(1986) 165.
[21] T. Wang, C.K. Ober, Macromolecules 30 (1997) 7560.
[22] T. Katsuragawa, E. Chiba, K. Okada, K. Tani, H. Tomoto, Jpn. J.
Appl. Phys. Part 1 34 (1995) 649.
[23] Y. Katano, H. Tomoto, T. Nakajima, Macromolecules 27 (1994)
2342.
36 P. Graham et al. / Journal of Fluorine Chemistry 104 (2000) 29±36