Biomacromolecules 2008, 9, 505–509
505
Carbon Nanotubes as Structural Nanofibers for Hyaluronic Acid Hydrogel Scaffolds Sanjib Bhattacharyya,*,†,§ Samuel Guillot,† Hinda Dabboue,† Jean-François Tranchant,‡ and Jean-Paul Salvetat*,† Centre de Recherche sur la Matière Divisée, UMR 6619, CNRS-Université d’Orléans, 1b rue de la Férollerie, 45071 Orléans Cedex 2, France, and Laboratoire de Physicochimie, Parfums Christian Dior, F-45804 St Jean de Braye, France Received September 7, 2007; Revised Manuscript Received November 22, 2007
We have successfully dispersed functionalized single-walled carbon nanotubes (SWNTs) within hyaluronic acid–water solutions. Hybrid hyaluronic acid (HA) hydrogels with SWNTs were then formed by cross-linking with divinyl sulfone. We have found a considerable change in the morphology of the lyophilized hybrid hydrogels compared to HA hydrogels. The high water uptake capacity, an important property of HA hydrogels, remained almost unchanged after 2 wt % SWNT (vs HA) incorporation, despite a dramatic enhancement in the dynamic mechanical properties of the hybrid hydrogels compared to native ones. We have found a 300% enhancement in the storage modulus of hybrid hydrogel with only 2 wt % of SWNTs vs HA (0.06 wt % vs total weight including water content). This apparent contradiction can be explained by a networking effect between SWNTs, mediated by HA chains. As in biological tissue, HA plays a dual role of matrix and linker for the rigid reinforcing nanofibers.
Introduction In recent years carbon nanotubes (CNTs) have found a place in a variety of fields including electronics and nanocomposites as well as in biomedical research due to their remarkable electrical and mechanical properties and chemical stability.1 One of the most exciting areas of research to emerge around carbon nanotubes is in the biomedical field.2–5 There are a number of publications showing the capability of nanotubes to deliver biomolecules and therapeutic molecules,6–8 to sense the presence of DNA and proteins in solution9–11 and to enhance cell proliferation and growth.12–14 However, due to the chemical inertness of CNTs, it is always difficult to manipulate and integrate them with other materials such as polymers. One of the main reasons is the poor dispersion of CNTs within water or organic solvents. There has been much progress in the functionalization chemistry of CNTs for solubilization and dispersion within different solvents.15–18 Biopolymers are often studied for polymeric scaffolds in the field of tissue engineering,19,20 and nanotechnology has been used recently in the field.21,22 Polysaccharides are of particular interest in joint-related repairs based on their role as glycosaminoglycans in mammalian extracellular matrices. Hyaluronic acid, alginates, and chitosan hydrogels are the most intensively studied, due to their ease of processing into useful structures.23–31 Among the required properties for successful application of these hydrogels in the tissue scaffolds are mechanical stability along with porous structure and water intake capacity. Hydrogels with improved mechanical properties are required to provide mechanical support during new tissue formation in vitro and in vivo and to provide mechanical signaling to cells in vitro and * Corresponding authors S. Bhattacharyya, e-mail
[email protected], and J.-P. Salvetat, e-mail
[email protected]. † Centre de Recherche sur la Matière Divisée, UMR 6619, CNRSUniversité d’Orléans. ‡ Laboratoire de Physicochimie, Parfums Christian Dior. § Present address: Center for Bioactive Materials & Tissue Engineering, Bioengineering Department, University of Pennsylvania, Philadelphia, PA 19104.
in vivo for optimization of the structural and mechanical functions of the new extracellular matrix being formed. In this context, functionalized water-soluble CNTs can be very good candidates to reinforce the biopolymer hydrogels without altering their chemical properties. The interplay of the physical properties of CNTs with a biopolymer can give birth to a new generation of tissue-engineered implants to reconstruct natural tissues that are subjected to high mechanical stress or require electrical and thermal conductivity at physiological state. Unfortunately there are very few reports on incorporation of CNTs within biocompatible polymer hydrogels.32–34 Li et al.35 have shown the successful incorporation of multiwalled carbon nanotubes (MWNTs) within gelatin hydrogel without changing the swelling characteristics of native hydrogel. Recently Wang et al.36 have reported the synthesis of supramolecular Pluronic hydrogels hybridized with single-wall carbon nanotubes (SWNTs). They have shown that the resultant hybrid hydrogels retained the basic characters of supramolecular hydrogel and pristine SWNTs, especially the shear-thinning property, which is very important in drug delivery and controlled release. Surprisingly they have observed a decline in the mechanical properties of the hybridized hydrogel, which they explain as being due to the mechanism of hydrogel formation. We know only of two reports addressing the incorporation of CNTs within biopolymer hydrogels and both are with alginate.37,38 In this article we have demonstrated a successful way to disperse SWNTs within hyaluronic acid (HA) solution and induce the formation of reinforced hydrogel in the presence of cross-linking reagent divinyl sulfone (DVS).39 We have observed a 4-fold increase in the storage modulus of the hydrogel by incorporation of only 0.06 wt % SWNTs (vs total weight including water) without significant change in the water intake capacity and shear thinning behavior of the native hydrogel.
Materials and Methods High molecular weight (1000 kDa) hyaluronic acid was a generous gift from LVMH, Orléans, France. Divinyl sulfone (purum g98%) was
10.1021/bm7009976 CCC: $40.75 2008 American Chemical Society Published on Web 01/11/2008
506 Biomacromolecules, Vol. 9, No. 2, 2008
Bhattacharyya et al.
Figure 1. Photograph of HA solution (transparent, on the left) and dispersion of 0.06% SWNTs by mass within HA solution (dark, on the right). Arrow leads to photograph showing invertible HA gels, native (white) and hybridized (dark), formed after 30 min of reaction with DVS. Scheme 1. Cross-Linking Reaction between HA and DVS
purchased from Fluka, France, and used without further purification. SWNTs were purchased from Carbon Nanotechnology Inc., USA. They were purified and carboxylated prior to use. First they were heat treated at 350 °C in air for 2 h to remove the amorphous carbon and then refluxed with 3 M nitric acid (100 mL of acid per gram of SWNTs) at 130 °C for 36 h to remove the metallic catalytic particles and to create the carboxyl functionalities on the surface. Formation of carboxyl groups has been verified by Fourier transformed infrared spectroscopy and X-ray photoelectron spectroscopy (data not shown). The carboxylated SWNTs prepared via above method was used to prepare the gel with different amounts of SWNTs. To prepare HA gel hybridized with 2 wt % (vs HA) functionalized SWNTs, first 6 mg of SWNTs was dispersed in 10 mL of double distilled water in a bath sonicator for 30 min and then 300 mg of HA was dissolved in it by magnetic stirring. At this stage, the concentration of SWNTs vs water is approximately 0.06 wt %. The pH of the solution was maintained at 9–10 by adding an appropriate amount of 2 M NaOH for efficient crosslinking. DVS was used to cross-link the HA and hybrid HA/SWNTs solution keeping the ratio HA:DVS 1:1. After 2 h the gel was thoroughly washed to get rid of any unreacted DVS and NaOH while keeping the total water amount constant. Three hydrogel samples with 0, 0.03, and 0.06 wt % of SWNT were prepared to compare the results. Success of the cross-linking of HA with and without SWNTs was verified by Fourier transform infrared (FTIR) spectroscopy on a Nicolet 710 FT-IR spectrophotometer with a resolution of 8 cm-1 by using the KBr method. For morphological characterization by scanning electron microscopy (SEM), hydrogels were frozen to rupture and the transverse sections were observed in Hitachi 4200 operated at 1 kV. The swelling behavior of the hydrogels was measured in phosphate buffer solution (pH 7.4). The starting materials were not fully dried and contained the same amount of water (10 mL for 300 mg of HA). Approximately 4 g of hydrogel samples was immersed in 50 mL of phosphate buffer solution for different time intervals. Then the hydrated gels were removed from the buffer solution and quickly dried on blotting paper to eliminate residual water on the surface. Results are expressed as percentage of swelling (S%) and were calculated by using
keep this dispersion stable up to the gelation point after a crosslinker is added. From optical observations (see photo in Figure 1 as an illustration) it was evident that the acid-treated SWNTs were well dispersed within the HA solution as it gave a homogeneous black color of the solution without any flocculation-induced turbidity. (Note that our acid-treated SWNTs were also well dispersed in pure water due to hydrophilic chemical groups grafted at the surface.) Atomic force microscopy observations showed that individual nanotubes coexisted with small bundles in HA water solutions. We do not have quantified, however, to what extend exfoliation of SWNT bundles occurred. (Additional experiments would be needed to address that point.) After 30 min of the addition of the cross-linking agent DVS, HA solutions both with and without SWNTs form the gel shown in Figure 1. It should be noted that after gelification with SWNTs color is still homogeneous indicating no macroscopic phase separation between cross-linked HA and SWNTs. With DVS, the cross-
S % ) 100 × [(Wht - W0) ⁄ W0] where W0 is the initial weight of the gel at time 0 and Wht is the weight of the hydrated gel at time t. Anton Paar Physica MRC 300 modular compact rheometer was used in the parallel plate geometry, with a 25 mm plate and a gap size of 1 mm. Constant values of deformation, around 2 mrad, were maintained throughout each frequency sweep of 0.2–15 Hz. The steady flow behaviors were measured at shear rate range from 0.1 to 100 s-1. All the dynamic tests were performed at a constant strain of 5%.
Results and Discussion To obtain a proper integration of nanofillers within the biopolymer gels, it is very important to stabilize a good dispersion of the fillers within the polymer–water solution and
Figure 2. FTIR spectra of starting HA, oxidized SWNT, and DVS cross-linked HA gels with 2 wt % (vs HA) SWNTs. (Spectra are shifted vertically for a better view.)
CNTs as Structural Nanofibers for Scaffolds
Biomacromolecules, Vol. 9, No. 2, 2008 507
Figure 3. SEM images of the morphologies of native hydrogels (a, b) and hydrogels hybridized with 0.06 wt % SWNTs (c, d). Scale bars in panels a and c are 30 µm, and those in panels b and d are 10 µm.
linking occurs via the hydroxyl groups of HA forming an ether bond as shown in Scheme 1. Due to cross-linking through the hydroxyl group, the intensity of the peak of the OH group in FTIR spectra (Figure 2, at around 3400 cm-1) decreased for the cross-linked HA gel with and without SWNTs, which indicates the presence of SWNTs did not hinder the cross-linking reaction. There is also an appearance of two enhanced peaks in the case of cross-linked gels at around 1290 and 1120 cm-1 which are coming both from C-O stretch modes involved in the HA-DVS bonding (1260 cm-1 and 1110 cm-1) and from SdO stretch modes of DVS (1316 and 1124 cm-1). A contribution from C-H bend modes should also be present around 1270 cm-1. We did not noticed any modification of the 1640 cm-1 peak, corresponding to the symmetric CdO stretch mode of HA, which further confirmed that the main reaction between HA and DVS was between the hydroxyl groups. These data confirm the successful cross-linking of HA despite the presence of SWNTs. It was not possible, however, to assess by FTIR whether covalent reactions occurred between oxidized SWNTs and HA via DVS. Morphology of Freeze-Dried Hydrogels. Panels a and b of Figure 3 show the typical interconnected porous morphology for the native hydrogel having uniform pore sizes ranging from 6 to 9 µm. We observed a considerable change in the morphology of the hybrid hydrogels. In panels c and d of Figure 3, the material constituting the original honeycombed or spongelike structure seems to have ”relaxed” and separated into separate fibers or slivers of material, although the chambered structure can still be recognized. Pores seem, surprisingly, larger
than those in the pristine structure but are filled with filamentous material. Mechanically rigid SWNTs are probably involved in the backbone of the fibrous substructure, which suggests that they interact strongly with HA during the cross-linking reaction with DVS. We may wonder whether the spontaneous liquid crystal phase separation of nanotubes in HA reported by Moulton can be involved in the filamentous structure formation.40 However, the ratio between SWNT and HA concentration (0.06:3) wt % is not in favor of a nematic phase according to the phase diagram established by the authors.40 Swelling Behavior. From Figure 4, it can be clearly seen that the hybrid and native hydrogels show the same general swelling trend. After 24 h of water absorption we have observed only 12% difference in the equilibrium water uptake capacity between native (190%) and hybrid (170%) hydrogel with 2% (wt/wt HA) of SWNTs. These results demonstrate that the presence of SWNTs has very little influence on the water uptake capacity of the gel. Rheological Properties of the Native and Hybrid Hydrogels. To examine the influence of SWNTs on the dynamical viscosity and moduli of the hydrogels, two samples with different concentrations of SWNTs were employed for the rheology test. The steady flow results showed a considerable increase in the viscosity due to the incorporation of SWNTs (Figure 5a). Both the native and hybrid hydrogels exhibited a shear-thinning behavior (decrease of the viscosity with increasing the shear rate), which is a typical character of supramolecular hydrogels.26 In Figure 5a, we plotted the viscosity as a function
508 Biomacromolecules, Vol. 9, No. 2, 2008
Figure 4. Swelling behavior of native (O) and hybridized hydrogel with 1 wt % ([) and 2 wt % (b) of SWNTs vs HA, upon incubation with PBS buffer (pH 7.4) at 25 °C.
Bhattacharyya et al.
Figure 6. Shear modulus vs weight fraction of SWNTs. Data points are linked by a cubic spline curve.
frequency range and a G′′ which is considerably smaller than the storage moduli (3 orders of magnitude). This feature indicates the formation of strong and rigid gels for both the cases. We have observed a 3- and 4-fold increase in the storage modulus for the hybrid hydrogels compared to native hydrogel with 0.03 wt % and 0.06 wt % of SWNTs, respectively. Origin of the Reinforcement. A rule-of-mixture approximation used for polymer matrices in their glassy state is certainly nonadapted to HA hydrogels, since it predicts that 0.03 wt % of SWNT would increase the modulus by 300 MPa. The elasticity of HA hydrogels, whether they are native or reinforced by SWNTs, is better described by entropy considerations. The presence of nanofillers in a gel can bring various causes of reinforcement. The first is hydrodynamical in origin and affects the polymer chain movement. The associated increase of shear modulus can be roughly estimated by the Guth law taking into account the filler aspect ratio,f:41
G ) G0(1 + 0.67fΦ + 1.62f2Φ2)
Figure 5. (a) Shear-thinning behaviors of the viscosity of the native (O) and hybridized hydrogels with 0.03 wt % ([) and 0.06 wt % (b) of SWNTs. (b) Dynamical rheological behavior of the native and hybridized hydrogels.
of the shear rate; the shear-thinning behavior can be well fitted by the Ostwald-de Waele or “power law” model given by
η ) K(dγ ⁄ dt)n-1 with K a constant and n the flow behavior index (0 < n < 1) which is characteristic of the deviation from the Newtonian behavior. Viscoelastic properties were explored by frequency sweep experiments. Storage and loss moduli G′ and G′′ were measured as a function of angular frequency at fixed strain of 5%. Figure 5b shows the variation of equilibrium moduli G′ and G′′ for the hybrid hydrogels with angular frequency compared to the native hydrogel. One of the general features observed for both types of hydrogels is the pronounced plateau of G′ in the full
G0 being the initial modulus of the matrix and Φ the volume fraction. We cannot explain in our case the large increase of modulus by such a law with realistic values of f: transmission electron microscopy observations showed that SWNTs are not completely debundled and f ∼ 100. Moreover the variation of G vs Φ is not compatible with the Guth formula (Figure 6). A second effect appears if relatively strong bonds (covalent or not) are created between polymer chains and nanofillers, the latter playing the role of cross-linkers. Lastly, similarly to what occurs with elastomers,42 it is expected that networking of nanofillers should bring a third efficient reinforcing effect in hydrogels. Depending on the filler concentration, filler aspect ratio, and filler-polymer bond density, networking effect can dominate over cross-linking effect. We suggest that in our case SWNTs are linked by segments of HA polymer chains noncovalently (HA may wrap SWNTs) or covalently via DVS that may react with carboxyl or hydroxyl groups at the SWNT surface. The fact that rheological properties did not changed after SWNTs were added to HA solution without DVS suggests, however, that the SWNT network cannot form just through noncovalent interaction with HA molecules. Without involving the formation of a SWNT network, it is difficult to explain the large increase in the storage modulus along with the small decrease of swelling capacity. In the classical theory of gels and elastomers, modulus and swelling are interdependent via the interchain cross-link density.43,44 The higher the modulus, the smaller the swelling ratio. In our case, the SWNT network provides enough rigidity in shear mode but cannot prevent the HA matrix from swelling with water. This
CNTs as Structural Nanofibers for Scaffolds
result is significant for biomedical applications where both mechanical properties and water uptake are important.
Conclusions Our objective was to apply carbon nanotubes to enhance the mechanical properties of biopolymer HA hydrogels without changing their water uptake capacity. We have demonstrated a successful method to incorporate functionalized SWNTs within HA hydrogels and shown that these carbon nanotubes are extremely efficient to enhance the viscoelastic properties of the hydrogels. The most important thing to be noted is that the swelling behavior, which is a key factor in the field of biomedical applications, remains almost unchanged despite considerable change in the gel morphology after freeze-drying. Thus we believe this novel hybrid hydrogel may find suitable applications in biomedical fields such as drug delivery, controlled release, and tissue engineering due to the unique properties of both HA and SWNTs. Acknowledgment. Financial support by the French National Research Agency (ANR), project “Bionanocomp”, is gratefully acknowledged. We thank Annie Richard (Université d’Orléans) for cryo-SEM observations and Dominique Langevin (Lab. Physique des Solides, Orsay) for access to the rheometer.
References and Notes (1) Understanding carbon nanotubes; Lecture Notes in Physics 677; Springer: Berlin and Heidelberg, 2006. (2) Lacerda, L.; Bianco, A.; Prato, M.; Kostarelos, K. AdV. Drug DeliVery ReV. 2006, 58, 1460–1470. (3) Polizu, S.; Savadogo, O.; Poulin, P.; Yahia, L. J. Nanosci. Nanotechnol. 2006, 6, 1883–1904. (4) Murugesan, S.; Park, T. J.; Yang, H. C.; Mousa, S.; Linhardt, R. J. Langmuir 2006, 22, 3461–3463. (5) Shim, M.; Kam, N. W. S.; Chen, R. J.; Li, Y. M.; Dai, H. J. Nano Lett. 2002, 2, 285–288. (6) Ajima, K.; Yudasaka, M.; Murakami, T.; Maigne, A.; Shiba, K.; Iijima, S. Mol. Pharmaceutics 2005, 2, 475–480. (7) Kam, N. W. S.; Jessop, T. C.; Wender, P. A.; Dai, H. J. J. Am. Chem. Soc. 2004, 126, 6850–6851. (8) Kam, N. W. S.; Dai, H. J. J. Am. Chem. Soc. 2005, 127, 6021–6026. (9) Heller, D. A.; Jeng, E. S.; Yeung, T. K.; Martinez, B. M.; Moll, A. E.; Gastala, J. B.; Strano, M. S. Science 2006, 311, 508–511. (10) Star, A.; Gabriel, J. C. P.; Bradley, K.; Gruner, G. Nano Lett. 2003, 3, 459–463. (11) Chen, R. J.; Choi, H. C.; Bangsaruntip, S.; Yenilmez, E.; Tang, X. W.; Wang, Q.; Chang, Y. L.; Dai, H. J. J. Am. Chem. Soc. 2004, 126, 1563–1568. (12) Harrison, B. S.; Atala, A. Biomaterials 2007, 28, 344–353. (13) Zhao, B.; Hu, H.; Mandal, S. K.; Haddon, R. C. Chem. Mater. 2005, 17, 3235–3241. (14) Meng, J.; Song, L.; Kong, H.; Zhu, G. J.; Wang, C. Y.; Xu, L. H.; Xie, S. S.; Xu, H. Y. J. Biomed. Mater. Res., Part A 2006, 79A, 298– 306.
Biomacromolecules, Vol. 9, No. 2, 2008 509 (15) Karajanagi, S. S.; Yang, H. C.; Asuri, P.; Sellitto, E.; Dordick, J. S.; Kane, R. S. Langmuir 2006, 22, 1392–1395. (16) Sabba, Y.; Thomas, E. L. Macromolecules 2004, 37, 4815–4820. (17) Sun, Y. P.; Fu, K. F.; Lin, Y.; Huang, W. J. Acc. Chem. Res. 2002, 35, 1096–1104. (18) Fei, B.; Lu, H. F.; Hu, Z. G.; Xin, J. H. Nanotechnology 2006, 17, 1589–1593. (19) Hollander, A. P.; Hatton, P. V. Biopolymer methods in tissue engineering; Methods in Molecular Biology 238; Humana Press: Totowa, NJ, 2004. (20) Ma, P. X.; Elisseeff, J. H. Scaffolding in tissue engineering; Taylor & Francis: Boca Raton, FL, 2006. (21) Murugan, R.; Ramakrishna, S. Tissue Eng. 2006, 12, 435–447. (22) Thomas, V.; Dean, D. R.; Jose, M. V.; Mathew, B.; Chowdhury, S.; Vohra, Y. K. Biomacromolecules 2007, 8, 631–637. (23) Lee, K. Y.; Mooney, D. J. Chem. ReV. 2001, 101, 1869–1879. (24) Crescenzi, V.; Cornelio, L.; Di Meo, C.; Nardecchia, S.; Lamanna, R. Biomacromolecules 2007, 8, 1844–1850. (25) Ghosh, K.; Shu, X. Z.; Mou, R.; Lombardi, J.; Prestwich, G. D.; Rafailovich, M. H.; Clark, R. A. F. Biomacromolecules 2005, 6, 2857– 2865. (26) Lee, K. Y.; Alsberg, E.; Hsiong, S.; Comisar, W.; Linderman, J.; Ziff, R.; Mooney, D. Nano Lett. 2004, 4, 1501–1506. (27) Li, X. Q.; Liu, T. Q.; Song, K. D.; Yao, L. S.; Ge, D.; Bao, C. Y.; Ma, X. H.; Cui, Z. F. Biotechnol. Prog. 2006, 22, 1683–1689. (28) Hannouche, D.; Terai, H.; Fuchs, J. R.; Terada, S.; Zand, S.; Nasseri, B. A.; Petite, H.; Sedel, L.; Vacanti, J. P. Tissue Eng. 2007, 13, 87– 99. (29) Iwasaki, N.; Yamane, S. T.; Majima, T.; Kasahara, Y.; Minami, A.; Harada, K.; Nonaka, S.; Maekawa, N.; Tamura, H.; Tokura, S.; Shiono, M.; Monde, K.; Nishimura, S. I. Biomacromolecules 2004, 5, 828– 833. (30) Crompton, K. E.; Goud, J. D.; Bellamkonda, R. V.; Gengenbach, T. R.; Finkelstein, D. I.; Horne, M. K.; Forsythe, J. S. Biomaterials 2007, 28, 441–449. (31) Kumar, M.; Muzzarelli, R. A. A.; Muzzarelli, C.; Sashiwa, H.; Domb, A. J. Chem. ReV. 2004, 104, 6017–6084. (32) Liu, C.; Zhang, J.; He, J.; Hu, G. Polymer 2003, 44, 7529–7532. (33) Nabeta, M.; Sano, M. Langmuir 2005, 21, 1706–1708. (34) Tong, X.; Zheng, J.; Lu, Y.; Zhang, Z.; Cheng, H. Mater. Lett. 2007, 61, 1704–1706. (35) Li, H.; Wang, D. Q.; Liu, B. L.; Gao, L. Z. Colloids Surf., B 2004, 33, 85–88. (36) Wang, Z. M.; Chen, Y. M. Macromolecules 2007, 40, 3402–3407. (37) Kawaguchi, M.; Fukushima, T.; Hayakawa, T.; Nakashima, N.; Inoue, Y.; Takeda, S.; Okamura, K.; Taniguchi, K. Dental Materials Journal 2006, 25, 719–725. (38) Jiang, Z. Y.; Xu, S. W.; Lu, Y.; Yuan, W. K.; Wu, H.; Lv, C. Q. J. Biomater. Sci., Polym. Ed. 2006, 17, 21–35. (39) Larsen, N. E.; Pollak, C. T.; Reiner, K.; Leshchiner, E.; Balazs, E. A. J. Biomed. Mater. Res. 1993, 27, 1129–1134. (40) Moulton, S. E.; Maugey, M.; Poulin, P.; Wallace, G. G. J. Am. Chem. Soc. 2007, 129, 9452–9457. (41) Guth, E. J. Appl. Phys. 1945, 16, 20–25. (42) Ehrburger-Dolle, F.; Bley, F.; Geissler, E.; Livet, F.; Morfin, I.; Rochas, C. Macromol. Symp. 2003, 200, 157–167. (43) Treloar, L. R. G. The physics of rubber elasticity, 3rd ed.; Clarendon Press: Oxford, 1975. (44) Rubinstein, M.; Colby, R. H.; Dobrynin, A. V.; Joanny, J. F. Macromolecules 1996, 29, 398–406.
BM7009976
