pubs.acs.org/Langmuir
© 2010 American Chemical Society
Surface Structures of an Amphiphilic Tri-Block Copolymer in Air and in
Water Probed Using Sum Frequency Generation Vibrational Spectroscopy
)
Cornelius B. Kristalyn,† Xiaolin Lu,†,§ Craig J. Weinman,‡ Christopher K. Ober,‡
Edward J. Kramer, ,^ and Zhan Chen*,†
†
Department of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109,
Department of Materials Science & Engineering, Cornell University, Ithaca, New York 14853, §Department of
Chemistry, Key Laboratory of Advanced Textile Materials and Manufacturing Technology of Education
Ministry, Zhejiang Sci-Tech University, Hangzhou 310018, China, Materials Department, and ^Department of
Chemical Engineering, University of California, Santa Barbara, Santa Barbara, California 93106
)
‡
Received February 16, 2010. Revised Manuscript Received April 13, 2010
Sum frequency generation (SFG) vibrational spectroscopy has been applied to investigate surface structures of an
amphiphilic surface-active block copolymer (SABC) film deposited on a CaF2 substrate, in air and in water in situ.
Developed as a surface-active component of an antifouling coating for marine applications, this amphiphilic triblock
copolymer contains both hydrophobic fluorinated alkyl groups as well as hydrophilic ethoxy groups. It was found that
surface structures of the copolymer film in air and in water cannot be probed directly using the SFG experimental
geometry we adopted because SFG signals can be contributed from the polymer/air (or polymer/water) interface as well
as the buried polymer/CaF2 substrate interface. Using polymer films with varied thicknesses, structural information
about the polymer surfaces in air and in water can be deduced from the detected SFG signals. With SFG, surface
restructuring of this polymer has been observed in water, especially the methyl and methylene groups change
orientations upon contact with water. However, the hydrophobic fluoroalkyl group was present on the surface in both
air and water, and we believe that it was held near the surface in water by its neighboring ethoxy groups.
Introduction
1
Biofouling has long been a problem with marine vessels. The
accumulation of biofoulants on ship hulls results in a decrease in
operational speed as well as an increase in fuel costs due to
increased hydrodynamic drag. Traditional antifouling coatings
have included biocides that are released into the marine environment, which can have a negative impact on nontarget organisms.2
Recently, more environmentally friendly antifouling coatings
have become a focus in research.2-8 The new coatings are usually
designed to satisfy one of two requirements: antifouling (reduce or
minimize biofouling that occurs) or fouling-release behavior
(biofouling does occur, but the foulants can be easily released
from the coating surface). Antifouling and fouling-release performance of a coating is determined by both the surface properties
(e.g., surface chemical, topographic, and biological features) and
bulk properties (e.g., Young’s modulus) of the coating.
*To whom all correspondence should be addressed. E-mail Address:
zhanc@umich.edu. Fax: 734-647-4865.
(1) Schultz, M. P. Biofouling 2007, 23, 331–341.
(2) Yebra, D. M.; Kiil, S.; Dam-Johansen, K. Prog. Org. Coat. 2004, 50, 75–104.
(3) Krishnan, S.; Weinman, C. J.; Ober, C. K. J. Mater. Chem. 2008, 18, 3405–
3413.
(4) Genzer, J.; Efimenko, K. Biofouling 2006, 22, 339–360.
(5) Webster, D. C.; Chisholm, B. J.; Stafslien, S. J. Biofouling 2007, 23, 179–192.
(6) Joshi, R. G.; Goel, A.; Mannari, V. M.; Finlay, J. A.; Callow, M. E.; Callow,
J. A. J. Appl. Polym. Sci. 2009, 114, 3693–3703.
(7) Chambers, L. D.; Stokes, K. R.; Walsh, F. C.; Wood, R. J. K. Surf. Coat.
Technol. 2006, 201, 3642–3652.
(8) Chen., S.; Jiang, S. Adv. Mater. 2008, 20, 335–338.
(9) Weinman, C. J.; Finlay, J. A.; Park, D.; Paik, M. Y.; Krishnan, S.; Fletcher,
B. R.; Callow, M. E.; Callow, J. A.; Handlin, D. L.; Willis, C. L.; Fischer, D. A.;
Sohn, K. E.; Kramer, E. J.; Ober, C. K. Polym. Mater.: Sci. Eng. Prepr. 2008, 98,
639–640.
(10) Weinman, C. J.; Finlay, J. A.; Park, D.; Paik, M. Y.; Krishnan, S.;
Sundaram, H. S.; Dimitriou, M.; Sohn, K. E.; Callow, M. E.; Callow, J. A.;
Handlin, D. L.; Willis, C. L.; Kramer, E. J.; Ober, C. K. Langmuir 2009, 25, 12266–
12274.
Langmuir 2010, 26(13), 11337–11343
One such type of coating is based on an amphiphilic block
copolymer system developed by Ober and co-workers.9,10 Ober et al.
were able to create a polymer with antifouling and fouling release
properties by combining both grafted hydrophobic fluoroalkyl and
grafted hydrophilic poly(ethylene glycol) groups. This approach
produced a polymer with a dynamic, responsive surface, capable
of both resisting and easily releasing fouling organisms.10,11 Another important feature enabling antifouling/fouling release is a
low Young’s modulus.12 In order to control the modulus of the
surface coating, a thick layer of polystyrene-block-poly(ethyleneran-butylene)-block-polystyrene (SEBS) was used as an anchoring
layer for a amphiphilic surface active triblock copolymer (SABC)
containing a polystyrene-b-poly(ethylene-ran-butylene) sequence
matched to the SEBS. This enables the fluorinated and PEGylated
surface active groups of the SABC to be present at the surface while
maintaining a low Young’s modulus.
In order to fully characterize and better develop marine
antifouling and fouling release coatings, an understanding of
the surface features of the coating is needed. Extensive research
has been performed to achieve in-depth understanding of surface
biological and topographical characteristics in aqueous environments using various methods such as marine field tests,13 lab
biological assays,14 and AFM.15 Determining chemical information about the coating is more challenging. Most surface-sensitive
(11) Krishnan, S.; Ayothi, R.; Hexemer, A.; Finlay, J. A.; Sohn, K. E.; Perry, R.;
Ober, C. K.; Kramer, E. J.; Callow, M. E.; Callow, J. A.; Fischer, D. A. Langmuir
2006, 22, 5075–5086.
(12) Brady, R. F.; Singer, I. L. Biofouling 2000, 15, 73–81.
(13) Swain, G. W.; Anil, A. C.; Baier, R. E.; Chia, F.; Conte, E.; Cook, A.;
Hadfield, M.; Haslbeck, E.; Holm, E.; Kavanagh, C.; Kohrs, D.; Kovach, B.; Lee,
C.; Mazzella, L.; Meyer, A. E.; Qian, P.; Sawant, S. S.; Schultz, M.; Sigurdsson, J.;
Smith, C.; Soo, L.; Terlizzi, A.; Wagh, A.; Zimmerman, R.; Zupo, V. Biofouling
2000, 16, 331–344.
(14) Finlay, J. A.; Krishnan, S.; Callow, M. E.; Callow, J. A.; Dong, R.; Asgill,
N.; Wong, K.; Kramer, E. J.; Ober, C. K. Langmuir 2008, 24, 503–510.
Published on Web 05/13/2010
DOI: 10.1021/la100701b
11337
Article
approaches have disadvantages when looking at aqueous environments. For example, X-ray photoelectron spectroscopy (XPS)
and near-edge X-ray absorption fine structure (NEXAFS) spectroscopy require an ultrahigh-vacuum environment and thus
cannot be applied to study surfaces in water. In a previous study,
Weinman et al. used XPS to investigate the surface of a SABC film
on a silicon wafer.10 Important surface structural information can
be detected using XPS in vacuum. For instance, it was found for
this SABC that both the fluorinated and PEG portion segregated
to the surface. Nevertheless, it is difficult to directly relate such a
surface structure probed in vacuum to the antifouling and
fouling-release behaviors of such a material, because it is believed
that such behaviors are mediated by the surface structure of the
material in an aqueous environment, which is very likely different
from that in vacuum. XPS is unable to characterize whether the
hydrophobic segments are still present at the surface when the
polymer film is placed in an aqueous environment. Attenuated
total reflection infrared spectroscopy (ATR-FTIR) can be
applied to study solid/liquid interfaces.16 However, the surface
sensitivity of ATR-FTIR is limited, and the signals may suffer
from water absorption bands that can obscure vital information. Recently, sum frequency generation vibrational spectroscopy (SFG) has been applied to investigate molecular surface
structures in various chemical environments,17-25 including
those in an aqueous environment,26-44 and solid/solid inter(15) Bartels, J. W.; Cheng, C.; Powell, K. T.; Xu, J.; Wooley, K. L. Macromol.
Chem. Phys. 2007, 208, 1676–1687.
(16) Wang, J.; Paszti, Z.; Clarke, M. L.; Chen, X.; Chen, Z. J. Phys. Chem. B
2007, 111, 6088–6095.
(17) Shen, Y. R.; Ostroverkhov, V. Chem. Rev. 2006, 106, 1140–1154.
(18) Eisenthal, K. B. Chem. Rev. 1996, 96, 1343–1360.
(19) Bain, C. D. J. Chem. Soc. Faraday. Trans. 1995, 91, 1281–1296.
(20) Richmond, G. L. Annu. Rev. Phys. Chem. 2001, 52, 357–389.
(21) Tadjeddine, A.; Peremans, A. Surf. Sci. 1996, 368, 377–383.
(22) Shultz, M. J.; Schnitzer, C.; Simonelli, D.; Baldelli, S. Int. Rev. Phys. Chem.
2000, 19, 123–153.
(23) Williams, C. T.; Beattie, D. A. Surf. Sci. 2002, 500, 545–576.
(24) Perry, A.; Neipert, C.; Space, B.; Moore, P. B. Chem. Rev. 2006, 106, 1234–
1258.
(25) Chen, X.; Chen, Z. Biochim. Biophys. Acta 2006, 1758, 1257–1273.
(26) Shi, Q.; Ye, S.; Kristalyn, C.; Su, Y.; Jiang, Z.; Chen, Z. Langmuir 2008, 24,
7939–7946.
(27) Ye, S.; McClelland, A.; Majumdar, P.; Stafslien, S. J.; Daniels, J.;
Chisholm, B.; Chen, Z. Langmuir 2008, 24, 9686–9694.
(28) Chen, Z.; Shen, Y. R.; Somorjai, G. A. Annu. Rev. Phys. Chem. 2002, 53,
437–465.
(29) Chen, Z. Polym. Int. 2006, 56, 577–587.
(30) Zhang, D.; Ward, R. S.; Shen, Y. R.; Somorjai, G. A. J. Phys. Chem. B
1997, 101, 9060–9064.
(31) Chen, C. Y.; Clarke, M. L.; Wang, J.; Chen, Z. Phys. Chem. Chem. Phys.
2005, 7, 2357–2363.
(32) Chen., Q.; Zhang, D.; Somorjai, G. A.; Bertozzi, C. R. J. Am. Chem. Soc.
1999, 121, 446–47.
(33) Li, G.; Shen, Y.; Morita, S.; Nishida, T.; Osawa, M. J. Am. Chem. Soc.
2004, 126, 12198–12199.
(34) Clarke, M. L.; Wang, J.; Chen, Z. Anal. Chem. 2003, 75, 3275–3280.
(35) Clarke, M. L.; Wang, J.; Chen, Z. J. Phys. Chem. B 2005, 109, 22027–22035.
(36) Rangwalla, H.; Schwab, A. D.; Yurdumakan, B.; Yablon, D. G.; Yeganeh,
M. S.; Dhinojwala, A. Langmuir 2004, 20, 8625–8633.
(37) Yang, C. S. C.; Wilson, P. T.; Richter, L. J. Macromolecules 2004, 37, 7742–
7746.
(38) Chen, C. Y.; Wang, J.; Chen, Z. Langmuir 2004, 20, 10186–10193.
(39) Shi, Q.; Ye, S.; Spanninga, S. A.; Su, Y.; Jiang, Z.; Chen, Z. Soft Matter
2009, 5, 3487–3494.
(40) Chen, X.; Clarke, M. L.; Wang, J.; Chen, Z. Int. J. Mod. Phys. B 2005, 19,
691–713.
(41) Lambert, A. G.; Davies, P. B.; Neivandt, D. J. Appl. Spectrosc. Rev. 2005,
40, 103–145.
(42) Casford, M. T. L.; Davies, P. B. J. Phys. Chem. B 2008, 112, 2616–2621.
(43) Wang, J.; Paszti, Z.; Even, M. A.; Chen, Z. J. Am. Chem. Soc. 2002, 124,
7016–7023.
(44) Wang, J.; Woodcock, S. E.; Buck, S. M.; Chen, C.; Chen, Z. J. Am. Chem.
Soc. 2001, 123, 9470–9471.
(45) McGall, S. J.; Davies, P. B.; Neivandt, D. J. J. Phys. Chem. B 2004, 108,
16030–16039.
(46) Gautam, K. S.; Schwab, A. D.; Dhinojwala, A.; Zhang, D.; Dougal, S. M.;
Yeganeh, M. S. Phys. Rev. Lett. 2000, 85, 3854–3857.
11338 DOI: 10.1021/la100701b
Kristalyn et al.
Figure 1. Molecular formula for SABC (reproduced with permission from Langmuir 2009, 25 12266-12274. Copyright 2009 Am.
Chem. Soc.). The polystyrene block has a molecular weight of ∼8
kDa. The random ethylene-butylene block has a molecular weight
of ∼25 kDa. The isoprene block has a molecular weight of ∼10
kDa with an R group with x = 3.5 and y = 5.
faces involving polymers.45-48 In this research, we will apply
SFG to investigate surface structures of the SABC developed
previously as a marine coating.
SFG is a second-order nonlinear spectroscopy technique ideally
suited to looking at the chemical structure of a surface in many
different environments.17-44 For example, SFG has been used to
examine the surface restructuring behavior of polymers in water. It
was observed using this technique that surface dominating methyl
groups of the poly(n-butyl methacrylate) (PBMA) stand up with a
broad distribution in air, but upon contact with water, they lay
down with a much narrower distribution.43 In contrast, the surface-dominating ester methyl groups of poly(methyl methacrylate)
(PMMA) exhibit no detectable structural change upon PMMA
contact with water.44 SFG has also been used to examine reversible
changes occurring in polymers as they are taken from air to water
and back to air. Uncured poly(dimethylsiloxane) (PDMS) surfaces
have been observed to undergo an irreversible orientation change
upon contact to water. Once the PDMS surface is fully cured,
however, the reorganization that occurs in water is reversible when
the surface is dried and examined again.49 The above-discussed
PBMA surface restructuring in water is also reversible: the surface
structure recovered after the sample was removed from water and
exposed to air again.43 However, a different polymethacrylate,
poly(n-octyl methacrylate), exhibits irreversible surface structural
change in water.44 SFG results demonstrated that it is important to
investigate polymer surface structure in situ to understand the
polymer surface behavior.
In this paper, we discuss the surface structures of the SABC
developed by the Ober group derived from polystyrene8K-blockpoly(ethylene-ran-butylene)25K-block-polyisoprene10K (PS8K-bP(E/B)25K-b-PI10K) triblock copolymer precursor functionalized
with ethoxylated fluoroalkyl groups on the polyisoprene block
(Figure 1).10 The surface structures of the SABC in air and water
were examined by SFG through the use of films with varying
thicknesses, as was done in previous studies involving buried
polymer/metal interfaces.50,51 It was found that the SFG spectra
(47) Wilson, P. T.; Briggman, K. A.; Wallace, W. E.; Stephenson, J. C.; Richter,
L. J. Appl. Phys. Lett. 2002, 80, 3084–3086.
(48) Chen, C. Y.; Wang, J.; Even, M. A.; Chen, Z. Macromolecules 2002, 35,
8093–8097.
(49) Chen, C.; Wang, J.; Chen, Z. Langmuir 2004, 20, 10186–10193.
(50) Lu, X.; Shephard, N.; Han, J.; Xue, G.; Chen, Z. Macromolecules 2008, 41,
8770–8777.
Langmuir 2010, 26(13), 11337–11343
Kristalyn et al.
Article
generated in the alkyl stretching region are a combination of both
the buried substrate/SABC interface as well as the SABC at the air
or aqueous interface. We detected the presence of fluorinated
groups in air as well as in the presence of D2O, and this is the first
case of fluorinated signal observation by SFG of a polymer
material in contact with an aqueous environment. As fluorinated
materials are of increasing interest to the polymer community, it is
important to be able to observe their behavior at a water interface.
Experimental Section
Figure 2. SFG experimental geometry.
Sample Preparation. The synthesis of the SABC has been
given in detail previously.10 The SABC triblock copolymer was
dissolved in chloroform (Sigma) in a 1% w/w solution. The
samples were then prepared by spin coating (Specialty Coatings
Systems) at 500 rpm intervals from 500 to 3000 rpm onto one leg of
a CaF2 right angle prism (Altos) to create a series of polymer films
with different thicknesses. Deuterated water (D2O) was obtained
from Sigma Aldrich and used to avoid water bands overlapping
the primary C-H stretching signals. The film thicknesses were
measured with a Dektek 3 profilometer from Veeco. For each
sample, 3-4 measurements were made and then averaged.
SFG. Sum frequency generation (SFG) is a second-order
nonlinear spectroscopy technique that has proven powerful in
probing surfaces and interfaces. The full theory has been discussed
in previous publications and will not be repeated here.17-44 A
frequency tunable infrared (IR) beam and a frequency fixed
visible beam are overlapped spatially and temporally on the
surface/interface. A third beam is generated at the sum of the
two input beam frequencies. The signal of the sum signal is
proportional to the square of the effective second-order nonlinear
susceptibility tensor, which is an intrinsic property of a material,
as follows
I SFG jχeff ð2Þ j2
ð1Þ
Here, ISFG is the intensity of the SFG signal beam and χeff(2) is the
effective second order nonlinear susceptibility tensor. The effective second-order nonlinear susceptibility tensor components can
then be related to the second-order nonlinear susceptibility tensor
components defined in the lab coordinate system. For example,
ð2Þ
χeff ssp ¼ F ssp χð2Þ
yyz
ð2Þ
where χeff(2)ssp is the effective second-order nonlinear susceptibility tensor component probed using the ssp polarization combination (s polarized sum, s polarized visible, p polarized IR) of the
input and output laser beams. Fssp is the Fresnel coefficient and
χyyz(2) is the second-order nonlinear susceptibility tensor component. Different components of the tensor can be probed by
different polarization combinations, which can be used to examine orientation of the molecules or functional groups. χeff(2) can
then be further broken into resonant and nonresonant terms as
follows
ð2Þ
χeff ¼ χnr þ
X
Aq
ωIR - ωq þ iΓq
ð3Þ
where χnr(2) is the nonresonant background and Aq, ωq, and Γq are
the strength, resonant frequency and damping coefficient (width)
for the vibrational mode q.
Under the dipole approximation, the second-order nonlinear
susceptibility vanishes for materials with inversion symmetry.
Most bulk materials have an inversion center, resulting in SFG
being surface or interface sensitive, as the inversion symmetry
must be broken at the surface or interface. For the SFG experiments in this paper, the SABC coated CaF2 prisms were mounted
as seen in Figure 2. The polarization combination used was ssp.
Langmuir 2010, 26(13), 11337–11343
Figure 3. SFG spectra of SABC films with varying thicknesses
in the ssp polarization in contact with air. Spectra are offset for
clarity.
For the spectrum collected with the SABC in contact with D2O,
the time of contact was approximately 30 min. No changes in the
spectrum as a function of time in contact with D2O were observed.
Results and Discussion
SFG ssp spectra of a series of SABC samples with different film
thicknesses were collected in air and are displayed in Figure 3. The
peak assignments of C-H stretching modes have been extensively
studied.43,52 Here, two dominant peaks, each with a shoulder and
one weaker peak, can be observed in the spectra. The peak at
∼2880 cm-1 is attributed to the symmetric stretch of the CH3
group in the amphiphilic polyisoprene block. The polymer chain
end methyl groups have a much lower overall concentration;
therefore, they are not likely to be the dominant groups to
contribute this signal. The other dominant peak at ∼2950 cm-1
is attributed to the CH3 Fermi resonance. The shoulder found
between these two peaks is a result of the asymmetric CH2 stretch
centered at ∼2910 cm-1. The CH3 asymmetric stretching signal
centered at ∼2960 cm-1 appears as a shoulder or as a tail of
the Fermi resonance. Lastly, there is a weak peak found at
∼2840 cm-1 that arises from the symmetric stretch of the CH2
group. Both methylene signals can be contributed from all the
methylene groups in the molecule. As the thickness of the thin film
increases, the intensity of the SFG signal decreases. If the
polymer/air interface dominates, then there should be no SFG
signal dependence on thickness. Thus, there exists a second
interface where the inversion symmetry can be broken, the
CaF2/polymer buried interface. For many polymers, SFG signals
generated from this interface are negligible,53 but in our system,
(51) Lu, X.; Li, D.; Kristalyn, C.; Han, J.; Shephard, N.; Rhodes, S.; Xue, G.;
Chen, Z. Macromolecules 2009, 42, 9052–9057.
(52) Snyder, R. G.; Strauss, H. L.; Elliger, C. A. J. Phys. Chem. 1982, 86, 5145–
5150.
(53) Wang, J.; Chen, C.; Buck, S. M.; Chen, Z. J. Phys. Chem. B 2001, 105,
12118–12125.
DOI: 10.1021/la100701b
11339
Article
Figure 4. Fresnel coefficients as a function of SABC film thickness for
the CaF2/SABC interface and the SABC/Air interface at 2880 cm-1.
this interface generates some SFG signals which cannot be ignored.
The SFG signals detected in Figure 3 might be from a contribution
from two interfaces (the polymer/air and the CaF2/polymer
interfaces), or even solely from the CaF2/polymer interface.
In order to determine where the SFG signal is coming from, we
compared the Fresnel coefficients of both interfaces. As stated
before, the SFG signal is related to χeff(2) and, therefore, to the
Fresnel coefficients. The Fresnel coefficients may vary for the two
interfaces when the sample thickness changes. By examining the
Fresnel coefficients of both interfaces, we hope to determine the
source(s) of the SFG signal. Using the detailed analysis presented
in the Supporting Information, we can generate thickness dependent curves of the Fresnel coefficients of the two interfaces.
For the case of the polymer/air interface, it can be seen that
there is little change in the value of the Fresnel coefficients in the
thickness range under study (Figure 4). The Fresnel coefficients
grow from 0.683 at 24 nm to 0.811 at 130 nm, an increase of
∼18%. This means as the thickness of the film increases, the
Fresnel coefficients increase. This should lead to an increase in the
observed SFG signal, which is not what was observed. If we take
the strength over the width of the symmetric stretching peak from
the fit parameters (Supporting Information, Table S1) we get an
indication of the relative value of the symmetric stretch for each
thickness. As the thickness increases, the value goes from 7.73 at
24 nm to 2.31 for 130 nm, a drop of ∼77% percent. So, not only is
the change in Fresnel coefficients trending in the opposite way, but
the magnitude of the calculated trend (18% change) is almost four
times less than the experimentally observed results (77% change).
As the thickness-dependent SFG signal strength from the
polymer/air interface does not adequately explain our results,
we then calculated the Fresnel coefficients of the CaF2/polymer
interface. For the CaF2/polymer interface, the Fresnel coefficients
are seen to change to a much greater degree than the polymer/air
interface (Figure 4). At 24 nm, the CaF2/polymer Fresnel coefficient is 0.755, which increases to 2.50 at 130 nm. This is an increase
of >300%, which would indicate for a film of 130 nm the buried
interface signal strength should be over 3 times stronger (intensity
will be 9 times stronger) than that found for a film of 24 nm. This
value is much closer to the magnitude of change seen in the
collected spectra, but once again the trend is in the opposite
direction. According to the Fresnel coefficients for both interfaces, the signal should increase from both interfaces as the
thickness increases.
The SFG data follow neither trend when considering only one
interface at a time; thus, it may be possible that SFG signals are
11340 DOI: 10.1021/la100701b
Kristalyn et al.
Figure 5. Lines: Plot of the signal strength for the symmetric
stretch of CH3 functional group of SABC films as a function of
thickness using known Fresnel coefficients of the SABC/Air interface and the CaF2/SABC interface, and trial values of a and b.
Dots: experimental data. When a is set at 17, and b is varied,
experimental values for χeff(2)ssp fit best when b = -4.4.
generated from both interfaces and they are somehow interfering
with each other. For the case of interference between the two
interfaces, the SFG intensity can be related as follows
I SFG ja F Polymer=Air þ b F CaF2=Polymer j2
ð4Þ
This means that some values of a and b may fit the trend of the
signal strength dependence on film thickness of the experimental data. It is clear that we can consider a and b as χPolymer/Air
and χCaF2/Polymer, respectively. Apparently, χPolymer/Air or a and
χCaF2/Polymer or b would not change as a function of sample
thickness, but FPolymer/Air and FCaF2/Polymer are different when
the sample thicknesses are varied, as shown in Figure 4. Here,
we fit the SFG spectra for samples with different thicknesses.
From the fitting, we can detect values of a FPolymer/Air þ b
FCaF2/Polymer at different sample thicknesses, which we plot as
dots in Figure 5. We then calculated the values of a FPolymer/Air þ
b FCaF2/Polymer using known FPolymer/Air and FCaF2/Polymer values
(plotted in Figure 4) of samples with different thicknesses and
different trial values of a and b. A plot of curves using various
a and b values along with the experimental data generated from
the fit parameters can be seen in Figure 5. When a = 17 and b =
-4.4, the thickness-dependent relationship between the Fresnel
coefficients for both interfaces approaches the experimental data.
The different signs (positive and negative) of a and b indicate a
difference in the absolute orientation of the methyl groups at the
two interfaces. We believe that, for the polymer/air interface, the
methyl groups adopt an orientation pointing away from the
surface toward the air, similar to those on other polymer (e.g.,
PBMA) surfaces in air, because both methyl groups and air are
hydrophobic. At the CaF2/polymer buried interface, the methyl
groups point away from the bulk of the film toward the substrate.
This results in methyl groups pointing in opposite directions and
their SFG signals destructively interfering with each other. The
difference in magnitude of the two constants is an indication of the
different amplitudes of the second-order nonlinear optical susceptibility components (χPolymer/Air vs χCaF2/Polymer). The polymer/
air interface is almost four times stronger than the buried CaF2/
polymer interface, indicating that the amplitude of the secondorder nonlinear optical susceptibility component at the polymer/
air interface is larger. This observation could be due to an increase
in the surface coverage of the methyl groups, greater ordering
(a narrower orientation distribution) of the methyl groups, or the
Langmuir 2010, 26(13), 11337–11343
Kristalyn et al.
Figure 6. SFG spectra of SABC films of varying thickness contacted to D2O in the ssp polarization. Spectra are offset for clarity.
Figure 7. Fresnel coefficients as a function of SABC film thickness for
the CaF2/SABC interface and the SABC/D2O interface at 2880 cm-1.
polymer/air methyl groups having a smaller angle to the surface
normal (stand up). As the sample thickness increases, the buried
interface signal strength begins to increase due to the increase of
the Fresnel coefficient (Figure 4), which results in the decrease in
the observed SFG signals due to the destructive interference
between signals from the two interfaces.
Upon contacting the SABC films with D2O, a conformational
change can be observed. As seen in Figure 6, the symmetric CH3
stretching and Fermi resonance signals are no longer the dominant
peaks. The asymmetric CH2 (∼2930 cm-1) signal is the strongest
peak, which is an indication of gauche defects developing in the
polymer chains at the surface or more ethylene glycol groups
migrating to the surface because they are more hydrophilic. The
second peak (centered at ∼2955 cm-1) that overlaps with the
asymmetric CH2 stretching peak is from the asymmetric CH3
stretch. The symmetric CH2 stretching signal (∼2845 cm-1) is still
present, as is the CH3 symmetric stretching signal (∼2880 cm-1),
as a shoulder to the asymmetric CH2. The change in signal
intensities of the different methyl stretches is most likely a result
of the methyl groups adopting a large angle to the surface normal
to minimize interactions with the hydrophilic D2O, which will be
confirmed by the data analysis below. It is also apparent that there
is a much smaller variation of the signal intensity between the
different thickness samples.
For the Fresnel coefficients of the polymer/D2O interface, there
is very little variation in the values for films with different
thicknesses. At 24 nm, the Fresnel coefficient is 3.27, which
Langmuir 2010, 26(13), 11337–11343
Article
Figure 8. Lines: Plot of the signal strength for the symmetric
stretch of CH3 functional group of SABC films as a function of
thickness using known Fresnel coefficients of the SABC/D2O
interface and the CaF2/SABC interface and trial values of a and
b. Dots: experimental data. When a is set at 2.6 and b is varied,
experimental values for χeff(2)ssp fit best when b = -2.4.
decreases by ∼5% to 3.12 at 103 nm (Figure 7). Fitting the
spectra, we can find the signal strength and width of each
component stretch, with parameters shown in the Supporting
Information (Table S2). Using the ratio of strength over width of
both the symmetric and asymmetric CH3 stretches, we develop a
gauge of how the SFG intensity changes with thickness. For the
symmetric CH3 stretch, there is a ∼123% change between 24 and
103 nm. For the asymmetric CH3, the difference is less pronounced
with a ∼20% difference. Therefore, the experimentally observed
thickness-dependent signal strength does not match the calculated one when only signals from the polymer/D2O interface are
considered. Furthermore, the trends of the thickness-dependent
signal observed and the Fresnel coefficients are opposite. The
observed signals are strongest for the thickest and thinnest
samples and drop in intensity for intermediate values of film
thickness. The Fresnel coefficients indicate that the opposite
should be true with the films of intermediate values of thickness
exhibiting the strongest signal.
Applying the same approach as what we used to study polymer
surface in air, here we continued to examine the Fresnel coefficients for the polymer/D2O and buried the CaF2/polymer interface
(Figure 8) when the polymer surface is in contact with D2O. This
time, there is not a drastic difference between the two interfaces as
was seen before. Both the buried CaF2/polymer and polymer/D2O
interfaces exhibit the same trends with the highest values being
located in the middle of the tested thicknesses. The difference is
once again larger for the buried interface, but to a much lesser
degree in comparison to the air case. The difference between 24 nm
(3.33) and 103 nm (2.67) is ∼20%. While this difference is very
close to the difference for the asymmetric stretch as stated
previously, the trends do not match up.
The differences in the thickness-dependent trends of the Fresnel
coefficients and the observed SFG results lead us to again see if we
could interpret the signals as the interference of the two interfaces.
Using eq 4, we again found a relationship that approximated our
observed SFG signals for both the symmetric CH3 and the
asymmetric CH3 stretches. In this case, a = 2.6 and b = -2.4
for the symmetric stretch and a = 3.3 and b = -2.1 for the
asymmetric stretch (Figures 8 and 9). As the signs of the two
constants are again opposite, it is likely the groups have opposite
absolute orientations. The buried interface should not be seeing
any environment changes in comparison to that probed when the
DOI: 10.1021/la100701b
11341
Article
Figure 9. Lines: Plot of the signal strength for the asymmetric
stretch of CH3 functional group of SABC films as a function of
thickness using known Fresnel coefficients of the SABC/D2O
interface and the CaF2/SABC interface and trial values of a and
b. Dots: experimental data. When a is set at 3.3 and b is varied,
experimental values for χeff(2)ssp fit best when b = -2.1.
Kristalyn et al.
Figure 11. Fresnel coefficients as a function of SABC film thickness
for the CaF2/SABC interface and the SABC/Air interface at 1370 cm-1.
Figure 12. SFG spectra of SABC films of varying thickness in
contact to D2O in the ssp polarization in the C-F spectral region.
Spectra are offset for clarity.
Figure 10. SFG spectra of SABC films of varying thickness in
air in the ssp polarization in the C-F spectral region. Spectra are
offset for clarity.
polymer is in contact with air. So, the methyl groups at the
CaF2/polymer interface will still point away from the bulk toward
the CaF2/polymer substrate. The methyl groups at the polymer/
D2O interface are then still pointing away from the bulk into the
D2O. At this point, it is likely that the methyl groups at the
polymer/water interface tilt more toward the surface, which
contributes to why the constants (susceptibility components) for
symmetric methyl stretch are much closer between the polymer/
substrate and the polymer/water interface than those between the
polymer/substrate and the polymer/air interface.
As the SABC has a short fluorinated portion on the side chain,
we also examined the SFG signals generated in the C-F stretching frequency region (Figure 10). A dominant peak appears at
1440 cm-1, which we attribute to our CaF2 substrates. We do not
know the origin of this signal, but it can be observed from the
CaF2 substrate without the polymer. There is also a shoulder that
appears centered at 1380 cm-1 assigned to the CF3 asymmetric54
(54) Tyrode, E.; Johnson, C. M.; Rutland, M. W.; Day, J. P. R.; Bain, C. D.
J. Phys. Chem. C. 2007, 111, 316–329.
(55) Lenk, T. J.; Hallmark, V. M.; Hoffman, C. L.; Rabolt, J. F.; Castner, D. G.;
Erdelen, C.; Ringsdor, H. Langmuir 1994, 10, 4610–4617.
11342 DOI: 10.1021/la100701b
or CF2 asymmetric55 mode. Upon first examination of the collected SFG spectra, it becomes readily apparent that there is not a
large change in the signals as the polymer film thickness changes.
This is different from what was seen in the C-H stretching frequency region and more in line with what is generally seen with
SFG where polymer film thickness does not affect the spectra.
Upon fitting the spectra (Supporting Information Table S3), it was
found from the ratio of the strength to width that the largest
variation was about ∼25% and the decrease from 24 nm (4.30) and
103 nm (3.60) was ∼16% (Supporting Information Table S4).
Looking at the calculated Fresnel coefficients for polymer/air
and CaF2/polymer (Figure 11) interfaces, we proceeded to
determine if there was any interference between the signals from
the two interfaces. For the polymer/air case, an increase of ∼16%
occurs from 24 to 103 nm, while the CaF2/polymer increases by
>250% from 24 to 103 nm. As these numbers are similar to the
C-H region Fresnel coefficients, it would be expected that if there
were a significant interference between the two interfaces it should
result in a major change in the intensity of the C-F signal. As the
change in both the Fresnel coefficients for the polymer/air interface and measured results are similar, we conclude that the buried
interface contributes very little to the observed SFG signal, and
the signal mainly comes from the polymer surface in air. This
shows that fluorinated segments are present on the triblock
copolymer surface in air, as shown by XPS before (in vacuum).10
Langmuir 2010, 26(13), 11337–11343
Kristalyn et al.
Article
As the hydrophilic PEG groups move to the surface for the more
favorable interaction with the hydrophilic environment provided
by the D2O, the fluorinated portions are forced to stay on the
surface. We are then able to detect these groups even while it
would be unfavorable for them to be present.
Figure 13. Fresnel coefficients as a function of SABC film thickness for
the CaF2/SABC interface and the SABC/D2O interface at 1370 cm-1.
The C-F region was also observed from polymer films in D2O
(Figure 12). Due to the change from air to water, the 1440 cm-1
peak from the substrates has decreased allowing the C-F signal
(1375 cm-1) to be resolved cleanly. As is the case in air, in the C-F
and C-H region in D2O there is little variation with change in
thickness. Examination of the fit parameters (Supporting Information Table S5) for 24 nm (1.50) and 78 nm (1.25) (Supporting
Information Table S4) shows a decrease in the signal of ∼17% as
the thickness increases.
The change in the observed results is much larger than the
change in the Fresnel coefficients for either the polymer/D2O
interface or the CaF2/polymer buried interface (Figure 13).
There is only a 1-2% change in the Fresnel factors in the
24-78 nm range for the polymer/D2O interfaces. The buried
interface experiences a slightly larger change with a ∼5%
difference between 24-78 and a ∼9% difference between the
highest and lowest Fresnel coefficient in that range. Therefore,
we believe that the C-F signal is a contribution from both
interfaces. The overall signal intensities observed in air are much
stronger than those in water, as shown in Figures 10 and 12. We
believe that, even though the contribution from the polymer/
substrate interface is the same in two cases, it does not play a
noticeable role in the signals observed in the air case, because the
signal in air is dominating. For the water case, it will affect the
observed overall signal because the signal from the polymer/
water interface is weak.
Our observation of the C-F stretching signal from the polymer/D2O interface is surprising, as it would be expected that the
highly hydrophobic fluorinated groups would not expose to water
but bury themselves in the bulk polymer. We believe that, because
the fluorinated portion is attached to the hydrophilic PEG
portion, it is physically constrained to remain near the surface.
Langmuir 2010, 26(13), 11337–11343
Conclusions
SFG has been used to examine the surface structures of a
potential marine antifouling coating. This coating material contains hydrophobic fluoroalkyl groups as well as a hydrophilic
PEG group. In air, the alkyl groups tend to stand more toward the
surface normal; when in contact with water, they adopt a much
larger angle to the surface normal to minimize unfavorable
interactions with the D2O. The fluorinated groups are seen at
both the polymer/air and polymer/D2O cases. The fluorinated
groups appear at the surface in air due to the hydrophobic
interactions, while the hydrophilic PEG portion is likely further
from the surface. Upon contact with water, the PEG migrates to
the surface to interact with the water resulting in the increase of
the SFG asymmetric methylene stretch signal. Surprisingly,
despite the fact that the fluorinated groups are expected to retreat
into the bulk in an aqueous environment, they are still detected at
the surface. As the fluorinated groups are attached to the
hydrophilic PEGylated portion of the polymer, they may be
constrained to remain near the surface while the PEG optimizes
its interaction in the hydrophilic D2O environment.
The surface behavior of this polymer correlates well with
antifouling testing results using Ulva spores and Navicula cells,
two materials commonly used to test antifouling properties. Ulva
is known to adhere strongly to hydrophilic surfaces, while
Navicula is known to attach strongly to hydrophobic surfaces.
Both Ulva and Navicula were shown to have high removal from
the SABC, indicating the amphiphilic nature of the SABC.10 The
amphiphlic nature of the SABC surface in water is reflected in the
observed SFG spectra as discussed above.
Acknowledgment. This research was supported by the Office
of Naval Research (ONR) through award N00014-08-1-1211 for
ZC, and was supported by United States Department of Defense’s
Strategic Environmental Research and Development Program
(SERDP), grant WP #1454 with additional support from ONR
through award N00014-02-1-0170 to CKO and EJK. XL acknowledges the financial supports from Department of Education
of Zhejiang Province (Y200909780) and Key Laboratory of
Advanced Textile Materials and Manufacturing Technology of
Education Ministry, Zhejiang Sci-Tech University (2009QN06).
Supporting Information Available: More discussions on
Fresnel coefficient calculation. This material is available free
of charge via the Internet at http://pubs.acs.org.
DOI: 10.1021/la100701b
11343