Professional Documents
Culture Documents
Surface Morphology of
Poly(caprolactone)-b-poly(dimethylsiloxane)-b-poly(caprolactone)
Copolymers: Effects on Protein Adsorption
M. A. Childs, D. D. Matlock, and J. R. Dorgan*
Chemical Engineering Department, Colorado School of Mines, Golden, Colorado 80401
T. R. Ohno
Physics Department, Colorado School of Mines, Golden, Colorado 80401 USA
Received January 2, 2001; Revised Manuscript Received March 5, 2001
Certain triblock copolymer surfaces have beneficial blood contacting properties that remain unexplained
from a mechanistic perspective. In this study, poly(caprolactone-block-dimethylsiloxane-block-caprolactone)
(PCL-b-PDMS-b-PCL) surfaces are characterized by dynamic contact angle analysis, angle-resolved X-ray
photoelectron spectroscopy (XPS), and phase detection imaging atomic force microscopy (AFM). Surface
morphology of films cast from 10 wt % MEK solutions are found to be semicrystalline possessing spherulites
on the micron scale and alternating semicrystalline PCL-rich and amorphous PDMS-rich lamellae on the
nanometer scale. Surface enrichment of the lower surface free energy block, PDMS, is observed using
angle-resolved XPS but the surface composition still consists of both copolymer blocks. Films cast from 1
wt % solutions showed similar morphologies but incomplete surface coverage. Different textural features
of adsorbed fibrinogen layers on coated and uncoated polypropylene are observed. The hypothesis that
patterned block copolymer surfaces can affect protein adsorption and thus influence compatibility is partially
supported by the findings of this study.
Figure 2. DCA results for variously treated polypropylene substrates coated with 1 wt % polymer solutions.
Results and Discussion reduces the reproducibility of DCA data. Annealing may
serve to decrease this variability.
The extent of surface enrichment of the PDMS block due A more convincing understanding of surface morphology
to its low surface free energy is easily investigated by DCA is obtained from the 10 wt % polymer solution-coated
analysis. If blooming of the PDMS block completely covers surfaces. AFM shows that full coverage is achieved at this
the surface, the contact angles of SMA and Tegomer surfaces
concentration. Figure 3 shows results for untreated and
should be the same as that of a PDMS homopolymer surface.
annealed 10 wt % homopolymer and copolymer solution-
Alternatively, the copolymer values will be intermediate
coated surfaces. Values for the polypropylene control plaques
between those characteristic of PDMS and PCL homopoly-
are also shown. Data for the annealed 10 wt % PCL solution-
mer surfaces if blooming does not fill the surface. DCA
coated surfaces are not available, because upon heating these
analysis performed on annealed samples detects changes in
coatings dewet and bead up, making DCA analysis impos-
near surface composition.
sible. These results imply that it is PDMS, and not PCL,
Figure 2 shows the contact angles for surfaces made by that wets the polypropylene substrate.
solution casting 1 wt % homopolymer and copolymer
Figure 3 shows that contact angles for the Tegomer
solutions; results for both as cast and annealed surfaces are
surfaces are lower than those for the SMA surfaces indicating
shown. In addition, measured contact angles for polypro-
a greater amount of PDMS at the SMA solution-coated
pylene control plaques are presented. As expected, PDMS
is more hydrophobic than PCL. The data reveal that contact surface. This result agrees with stoichiometric calculations
angles for copolymer solution-coated substrates lie between of mass percentages of PDMS in the copolymer based on
those coated with homopolymer solutions. It is important to AR-XPS that give 43 and 38 wt % PDMS for SMA and
remark, however, that 1 wt % solution coatings do not always Tegomer, respectively. XPS demonstrates that PDMS is
fully cover the polypropylene substrate, as shown in the enhanced at the air-copolymer interface and this enhance-
PDI-AFM images and by the AR-XPS data presented ment is apparent in the simple DCA experiments.
below. Interpretation of the DCA data is further complicated A crucial finding from the DCA data is that both blocks
since polypropylene control plaques have contact angles that appear to be present at the surface in both the 1 and 10 wt
also fall intermediate between those coated with homopoly- % solution coatings. This is consistent with the existence of
mer solutions; contact angle values for 1 wt % copolymer a microdomain structure at the copolymer interface. Certainly
solution-coated surfaces may be intermediate between the conditions exist that permit, for example, the PDMS block
homopolymer values simply because polypropylene is ex- to cover the entire surface while still producing a micro-
posed at the surface. domained structure in the bulk. However, crystallization of
Annealing of the 1 wt % solution-coated surfaces results one of the blocks in block copolymers is known to lead to
in contact angles within the standard deviations of the the rapid formation of robust microstructure (one in the so-
respective untreated samples. This indicates that all surfaces called strong segregation regime).14 The block copolymer
are stable and remain unaltered throughout the annealing samples did not change upon annealing significantly above
process. It should be noted that the large standard deviations the Tg’s of both blocks thus indicating considerable stability
observed for untreated substrates coated with PCL solutions as a result of the PCL crystallinity.
are likely due to surface roughness. Nonhomogeneous AR-XPS is an effective method to measure depth profiles
coatings of this semicrystalline polymer produce a large of atomic composition. This type of analysis is germane for
amount of visible surface roughness, a factor that strongly block copolymer systems since they can produce a dissimilar
Effects on Protein Adsorption Biomacromolecules, Vol. 2, No. 2, 2001 529
Figure 3. DCA results for variously treated polypropylene substrates coated with 10 wt % polymer solutions.
Figure 7. Series of low magnification height (left) and phase (right) PDI-AFM images of a 10 wt % SMA solution-coated surface: (a) 50.0 µm;
(b) 13.0 µm; (c) 5.00 µm.
tems.18 Regarding copolymers, Magonov and Heaton em- advantage of such a powerful mode of analysis to reveal the
ployed PDI-AFM to verify TEM and XPS studies of wide variety of surface morphologies and properties dem-
poly(styrene)-b-poly(butadiene)-b-poly(styrene) (PS-b-PBD- onstrated by block copolymer systems.16,20-22
b-PS) triblock copolymers that showed an enhancement of The existence of phase-separated triblock copolymer
lower surface energy PBD blocks at the surface.19 AFM not surfaces is directly revealed with PDI-AFM. Height and
only was able to verify the previous studies, but also was phase images of a 10 wt % SMA solution-coated surface
able to generate more detailed information about this are shown for one sample at low magnification in Figure 7
copolymer system, revealing a wormlike pattern of stiff PS and another sample at high magnification in Figure 8; the
within an amorphous sea of PBD. Many others have taken morphologies shown are typical of 10 wt % SMA solution-
532 Biomacromolecules, Vol. 2, No. 2, 2001 Childs et al.
Figure 8. Series of high magnification height (left) and phase (right) PDI-AFM images of a 10 wt % SMA solution-coated surface: (a) 1.88 µm;
(b) 1.00 µm.
coated surfaces. Bright areas in the height images presumably
indicate elevated regions and dark regions represent de-
pressed material. However, as discussed in ref 19, height
images of block copolymer morphology in tapping mode are
not reliable because the relative contrast of the blocks
depends sensitively on the driving frequency in height
images. For this reason, no scale is provided. Such sensitivity
does not exist in phase images, and the contrast between
rubbery PDMS-rich and crystalline PCL-rich areas is excel-
lent. However assignment of each phase to a particular
contrast cannot be definitively stated. The low magnification
scans reveal semicrystalline spherulites on the order of 5-10
µm. Such spherulitic morphology is consistent with the
studies by Lovinger et al. of PCL-b-PDMS-b-PCL triblock Figure 9. Schematic models of possible molecular arrangements in
copolymers with similar molecular weights and linker the lamellar crystals of SMA and Tegomer copolymers (after Lov-
inger): (a) extended-type PCL blocks and (b) once-folded PCL blocks.
groups.8 High magnification shows a lamellar morphology
is present in which fibrous, PCL-rich semicrystalline spheru- the conclusion that PCL chains are either folded once within
lites are seen to be completely space filling with amorphous the semicrystalline core or not folded at all, as shown in
PDMS located in the interlamellar regions. Exact evaluation Figure 9. A once-folded conformation corresponds to a
of interlayer thicknesses are made difficult by the lack of semicrystalline thickness of approximately 5 nm, whereas
distinct, sharp boundaries that separate such layers; however, extended PCL chains are plausible for a 10 nm lamellar
thicknesses measured using a ruler overlaid on the 1.00 µm thickness. This is again in agreement with Lovinger et al.
phase image are estimated to be between 5 and 10 nm for who found, by estimating crystallinity from X-ray diffrac-
bright regions and between 5 and 15 nm for dark regions. tometry and assuming the density of amorphous phase to be
Assigning a value to the fully extended PCL chain length 0.85 times that of the crystalline phase, a semicrystalline
of 15 nm and to the fully extended PDMS of 13 nm leads to thickness of 6 nm corresponding to a once-folded PCL
Effects on Protein Adsorption Biomacromolecules, Vol. 2, No. 2, 2001 533
Figure 10. Series of low magnification height (left) and phase (right) PDI-AFM images of a 10 wt % Tegomer solution-coated surface: (a) 50.0
µm; (b) 15.0 µm; (c) 7.73 µm.
morphology. It is again worth emphasizing that the rapid SMA. Measurement by hand of lamellar thicknesses result
crystallization of the PCL block must lead to a pinning of in approximately 10 nm for bright regions and range from 5
the PDMS thus preventing the formation of an overlayer. to 20 nm for the dark regions.
A series of height and phase PDI-AFM images of a 10 The only significant compositional difference between the
wt % Tegomer solution-coated surface, shown in Figures copolymers is that the middle block has a molecular weight
10 (low magnification) and 11 (high magnification), dem- of 2700 in Tegomer as opposed to 3200 in SMA. Lamellar
onstrate spherulitic structures similar to those revealed in thickness measurements of 10 wt % Tegomer solution-coated
10 wt % SMA solution-coated surfaces. However Tegomer surfaces do not elucidate the conformation of the PCL chains
spherulites range from approximately 5 to 25 µm. No other since it is not known if the crystalline regions are bright or
significant differences are observed in images of 10 wt % dark in the phase images. If the dark regions represented
Tegomer solution-coated surfaces compared to those of PCL-rich material, they would allow for both extended and
534 Biomacromolecules, Vol. 2, No. 2, 2001 Childs et al.
Figure 11. Series of high magnification height (left) and phase (right) PDI-AFM images of a 10 wt % Tegomer solution-coated surface: (a)
3.78 µm; (b) 1.42 µm. The lamellar spacings of the domains are plainly visible in the high magnification PDI-AFM scans.
once-folded arrangements since they measure 5-20 nm wide. SAXS, respectively.8 Clearly the present results are in
If, however, the bright regions corresponded to PCL-rich excellent agreement with these previous findings for bulk
material, the measured thickness of 10 nm would indicate spacings.
extended PCL chains with chain ends incorporated into the AFM demonstrates that the 1 wt % SMA and Tegomer
amorphous PDMS-rich region. solution-coated surfaces exhibit a variety of morphologies
Image analysis tools may be used to find the lamellar as a result of different surface coverages. A small fraction
spacing from the AFM results. Power spectrum density of 1 wt % copolymer solution-coated surfaces analyzed have
analysis calculates an average spacing, or a length defined
morphologies identical to those coated with 10 wt %
as one-half the length of the repeating unit (a repeating unit
copolymer solutions, demonstrating the same characteristic
in the present case corresponds to the lamellar spacing). If
space-filling crystalline spherulites at low magnification and
one considers the present system as a clumpy mixture in
fibrous lamellar phase separation at higher magnification.
which the repeating structures are statistically similar in
However, most 1 wt % copolymer solution-coated surfaces
shape, yet have no regular arrangement, then the linear scale
of segregation, SL, can be calculated as the average spacing.23 analyzed exhibit some exposed polypropylene regions. Figure
The 10 wt % SMA solution-coated surfaces are not analyzed 12 is a series of height and phase images of a 1 wt % SMA
because the high magnification images do not have sharp solution-coated surface in which semicrystalline spherulitic
enough boundaries between bright and dark regions. How- structures are evident in low magnification scans, yet
ever, a 10 wt % Tegomer solution-coated surface is easily randomly located copolymer-deficient areas are observed at
analyzed by this method, and the average spacing is higher magnification. These areas, indicated in the height
calculated to be approximately 8 nm corresponding to a images as dark spots, are consistent with low copolymer
lamellar spacing of 16 nm. Utilizing a Fourier transform solution concentrations forming films that are too thin. In
method, the average lamellar spacing (distance from one light of the AR-XPS data and the additional AFM results
bright domain to the next) is calculated to be approximately not presented, it is evident that full surface coverage is not
18 nm. Lovinger et al. determined average lamellar spacings guaranteed using a 1 wt % copolymer solution dip coating.
of 15 and 18 nm by analysis of TEM micrographs and by Similar observations were made for the Tegomer triblock.
Effects on Protein Adsorption Biomacromolecules, Vol. 2, No. 2, 2001 535
Figure 12. Series of height (left) and phase (right) PDI-AFM images of a 1 wt % SMA solution-coated surface: (a) 10.0 µm; (b) 4.11 µm; (c)
1.27 µm. Incomplete coverage is evident in the form of small elongated voids.
These findings may have important consequences for indus- image, it can be seen that the underlying lamellar morphology
trial coating processes utilizing such triblock copolymers. appears to template the adsorption of the fibrinogen. That
Knowing the intimate details of the surface morphology is, the fibrinogen molecules or clusters lie with their long
is a first step in understanding the mechanistic functioning axis along the lamellae (see the upper right quadrant of the
of such surfaces in improving biocompatibility. Figure 13 phase image). This finding supports the idea that protein
shows an AFM result for a block copolymer surface that adsorption can be mediated, in a conformational sense,
has been exposed to a 0.1 mg/mL human fibrinogen surface through the interaction with block copolymers. However, this
for 30 s. Again height and phase mode images are presented image represents scant experimental evidence for such an
and it is seen that for this short exposure time, incomplete effect. Dewetting effects during sample drying could influ-
surface coverage is obtained. Particularly in the phase mode ence fibrinogen deposition. An in situ AFM experiment at
536 Biomacromolecules, Vol. 2, No. 2, 2001 Childs et al.
Figure 13. Height (left) and phase (right) PDI-AFM images of a 1 µm2 area of a 10 wt % SMA solution-coated surface exposed to a 0.1 mg/mL
human fibrinogen solution for 30 s. The fibrinogen adsorbs along the lamellae, showing a templating action of the surface on adsorption.
Figure 14. Comparison of the morphology of adsorbed fibrinogen layers over a 100 µm2 area. The sample on the left is untreated polypropylene,
and the one to the right is polypropylene coated with block copolymer; both surfaces are covered with a layer of fibrinogen protein but exhibit
different textures (see text).
the copolymer liquid interface is needed but facilities for present study and those of other groups,7,17 this interpretation
such an experiment were not available to the authors. appears to be a rather simplistic view of the action of these
Further evidence of the effects of the block copolymer block copolymer compatibilizers. Certainly, changes brought
surface on protein adsorption is presented in Figure 14. Here, about by micropatterning lead to surfaces possessing fun-
phase mode images of polypropylene surfaces exposed to a damentally different blood contacting properties.
0.1 mg/mL fibrinogen solution for 24 h at 37 °C are shown.
The surface to the right is coated with a 10 wt % SMA Conclusions
solution. The morphologies of the two fully established
fibrinogen layers are distinct; the copolymer treated surface This study investigates the surface morphology of triblock
exhibits a finer, grained texture. The untreated surface copolymer solution coatings on polypropylene substrates and
possesses coarser features (a morphology reminiscent of its affect on the adsorption of fibrinogen. DCA analysis
cauliflower blooms). These findings more directly support shows that copolymer solution-coated surfaces have contact
the premise that some type of templating action may exist. angles that lie within the range of the respective homopoly-
An alternative explanation is that the exposure of both mers and provides perhaps the most compelling indication
copolymer blocks at the surface simply produces an inter- of the presence of both PDMS and PCL at the surfaces. AR-
mediate contact angle. Given the combined findings of the XPS studies show enhancement of the lower surface free
Effects on Protein Adsorption Biomacromolecules, Vol. 2, No. 2, 2001 537
energy block, PDMS, at the interface. However, AR-XPS (2) Singer, S. J.; Nicolson, G. L. Science 1972, 175, 720-731.
results also indicate the presence of both blocks at the surface. (3) Deppisch, R.; Gohl, H.; Smeby, L. Nephrol. Dial. Transplant. 1998,
13, 1354-1359.
PDI-AFM images are also revealing, showing completely (4) Bates, F. S.; Fredrickson, G. H. Phys. Today 1999, 52, 32-38.
space-filling semicrystalline spherulites for all 10 wt % (5) Okano, T.; Aoyagi, T.; Kataoka, K.; Abe, K.; Sakurai, Y.; Shimada,
solution-coated surfaces. On the nanometer scale, such M.; Shinohara, I. J. Biomed. Mater. Res. 1986, 20, 919-927.
spherulites show alternating bright and dark lamellae cor- (6) Yui, N.; Sanui, K.; Ogata, N.; Kataoka, K.; Okano, T.; Sakurai, Y.
responding to semicrystalline PCL-rich domains and amor- J. Biomed. Mater. Res. 1986, 20, 929-943.
(7) Kawamoto, N.; Mori, H.; Terano, M.; Yui, N. J. Biomater. Sci. Polym.
phous PDMS-rich regions. The 1 wt % copolymer solution- Ed. 1997, 8, 859-877.
coated surfaces vary in morphology but commonly show (8) Lovinger, A.; Han, B. J.; Padden, F. J., Jr.; Mirau, P. A. J. Polym.
semicrystalline weblike fibers. Utilization of the present Sci., B: Polym. Phys. 1993, 31, 115-123.
powerful suite of analytical tools provides a detailed picture (9) Tsai, C.; Deppisch, R. M.; Lloyd, J. F.; Ritzau, G. H.; Orem, A. D.;
of the intimate nature of the surface morphology formed from Gohl, H. J.; Voorhees, M. E. ASAIO J. 1994, 40, 619-624.
(10) Spijker, H. T.; Haan, J.; Graaff, R.; Boonstra, P. W.; van Oeveren,
solvent casting such films. W. Cells Mater. 1997, 7.
Observation of the effects of the surface morphology on (11) Gu, Y. J.; Boonstra, P. W.; Rignsburger, A. A.; Haan, J.; van Oeveren,
protein adsorption supports a novel approach to understand- W. Thorac. Surg. 1998, 65.
ing the mechanism of improved blood compatibility on block (12) Rubens, F. D.; Labow, R. S.; Lavallée, G. R.; Watson, M. I.; Robblee,
copolymer surfaces. Namely, that microdomain patterned J. A.; Voorhees, M. E.; Nathan, H. J. Ann. Thorac. Surg. 1999, 67,
689-698.
synthetic surfaces profoundly alter the nature of the adsorbed (13) Jalbert, C. J.; Koberstein, J. T.; Balaji, R.; Bhatia, Q.; Salvati, L.,
protein layer. This study reports some evidence for such Jr.; Yilgor, I. Macromolecules 1994, 27, 2409-2413.
effects by showing modified adsorbed protein morphology (14) Rangarajan, P.; Register, R. A.; Adamson, D. H.; Fetters, L. J.; Bras,
brought about by the micropatterned copolymer coating. W.; Naylor, S.; Ryan, A. J. Macromolecules 1995, 28, 1422-1428.
(15) Gorelova, M. M.; Pertsin, A. J.; Muzafarov, A. M.; Gritsenko, O.
Future design of biomedical materials should include a direct
T.; Vasilenko, N. G. J. Appl. Polym. Sci. 1995, 55, 1131-1135.
recognition of the surface as a two-dimensional object in (16) Chen, J. T.; Thomas, E. L. J. Mater. Sci. 1996, 31, 2531-2538.
which lateral compositional variations are of significant (17) Wen, J.; Somorjai, G.; Lim, F.; Ward, R. Macromolecules 1997, 30,
importance. 7206-7213.
(18) Magonov, S.; Heaton, M. G. Am. Lab. 1998, 9-16.
Acknowledgment. This work was supported by the U.S. (19) Magonov, S. N.; Cleveland, J.; Elings, V.; Denley, D.; Whangbo,
National Science Foundation, Division of Biological and M. H. Surf. Sci. 1997, 389, 201-211.
(20) Leclere, P.; Lazzaroni, R.; Bredas, J. L.; Yu, J. M.; Dubois, P.;
Environmental Sciences, through Grant BES 9709959. The
Jerome, R. Langmuir 1996, 12, 4317-4320.
authors are grateful to the reviewers for a number of (21) Motomatsu, M.; Mizutani, W.; Tokumoto, H. Polymer 1997, 38,
suggestions leading to the improvement of the presentation 1779-1785.
of the findings and for pointing out an error in the original (22) Bar, G.; Thomann, Y.; Whangbo, M. H. Langmuir 1998, 14, 1219-
treatment of the XPS data. 1226.
(23) Tucker, C. L. In Principles of Mixing Measurements; Rauwendaal,
References and Notes C., Ed.; Marcel Dekker: New York, 1991.
(1) Okano, T.; Nishiyama, S.; Shinohara, I.; Akaike, T.; Sakurai, Y.
Polym. J. 1978, 10, 223-228. BM0100054