You are on page 1of 5

COMMUNICATIONS

[8] a) R. Mahajan, C. M. Zimmerman, W. J. Koros, in Polymer Mem- Clearly, silk is affected by the ambient conditions both dur-
branes for Gas and Vapor Separation, Chemistry and Materials ing spinning and during testing. Thus the mechanical proper-
Science (Eds: B. D. Freeman, I. Pinnau), American Chemical So-
ties of silks produced and tested under very specific
ciety, Washington DC 1999, pp. 277286. b) D. R. Paul, D. R. Kemp,
J. Polym. Sci., Part C, Polym. Symp. 1973, 41, 79. conditions can be used as windows to look into the interac-
[9] V. P. Shantarovich, I. B. Kevdina, Y. P.Yampolskii, A. Y. Alentiev, tion of production conditions and silk mechanics (studied by
Macromolecules 2000, 33, 7453. varying the extrusion conditions[8]) as well as the details of the
[10] C. M. Zimmerman, A. Singh, W. J. Koros, J. Membr. Sci. 1997, 137, silk fine structure (examined by varying the testing condi-
145. tions[9]). Both are not without some importance to the spiders
[11] a) T. M. Gr, J. Membr. Sci. 1994, 93, 283. b) M. Jia, K. V. Peine-
as they can affect the performance of the web. Spiders are ec-
mann, R. D. Behling, J. Membr. Sci. 1991, 57, 289. c) J. M. Duval,
B. Folkers, M. H. V. Mulder, G. Desgrandchamps, C. A. Smolders, J. totherms, and ambient temperature has a strong effect on spi-
Membr. Sci. 1993, 80, 189. d) I. F. J. Vankelecom, E. Merckx, der behavior and life history,[10] in addition to affecting the
M. Luts, J. B. Uytterhoeven, J. Phys. Chem. 1995, 99, 13 187. e) J. material spun under those conditions. Climate further affects
Won, Y. Yoon, Y. S. Kang, Macromol. Res. 2002, 10, 9980. f) T. C. the material properties of silk and the engineering of the
Merkel, B. D. Freeman, R. J. Spontak, Z. He, I. Pinnau, P. Meakin,
whole web,[2a,11] because webs are typically spun in the early
A. J. Hill, Science 2002, 296, 519.
[12] J. P. Boom, I. G. M. Punt, H. Zwijnenberg, R. de Boer, D. Barge-
morning, when humidity is high while temperature is low.
man, C. A. Smolders, H. Strathmann, J. Membr. Sci. 1998, 138, 237. Hence we may assume that evolution has led to a set of design
[13] D. Q. Vu, W. J. Koros, S. J. Miller, J. Membr. Sci. 2003, 211, 311. criteria for silk that allow the material to function well over a
range of temperatures (from ca. 0 C to 30 C), humidities
(from ca. 20 to 100 RH (relative humidity)), and spinning
speeds (from ca. 1 mm s1 to ca. 1000 mm s1). However, what
about the properties outside that range? Moreover, as men-
Toughness of Spider Silk at High tioned above, the inter-relation of environmental conditions
and Low Temperatures** and silk behavior can provide insights into the structurefunc-
tion relationship.
By Yong Yang, Xin Chen, Zhengzhong Shao,* Hence we posed, firstly, the question of whether a major
Ping Zhou, David Porter, David P. Knight, spider silk, i.e., a silk that is normally used for lifelines and
Fritz Vollrath important supporting web-components, can perform outside
the range of temperatures under which it evolved. Secondly,
The outstanding strength and toughness of certain spider we were interested in the light that mechanical performance
and lepidopteran silks[1] has aroused considerable interest in would shed on the structure of the material if we tested our
recent years,[2] with research focusing primarily on the rela- silks under as wide a temperature regime as possible (i.e., 60
tionship between molecular structure and mechanical proper- to +150 C). Clearly, this would not only provide data of
ties.[3] Environmental conditions such as ambient humidity,[4] purely scientific value but also data of considerable commer-
acidity,[5] and UV radiation[6] all affect the mechanical proper- cial interest, as spider silk is often billed as a material with a
ties of native silks to some degree.[7] Pronounced differences great future. To study these questions (i.e., to examine the
in mechanical properties were also observed when conditions mechanical properties of a major spider silk over a wide tem-
such as the speed or temperature at spinning were varied[8] or perature range), a dynamic mechanical thermal analyzer
when the silk was (or had been) submerged in solvents such (DMTA) was used to measure both static and dynamic
as water, urea solution, or a range of alcohols, where it con- mechanical properties of single spider-silk threads. We mea-
tracts in varying degrees.[9] sured the strain response of single-thread filaments (ca. 6 lm
in diameter) of a principal dragline fiber from the large
orb-weaving spider that has become the benchmark of spider-
silk studies, the golden silk spider Nephila edulis. All silk had
[*] Prof. Z. Shao, Y. Yang, Dr. X. Chen, Dr. P. Zhou
Department of Macromolecular Science and been collected under the same control silking conditions of
The Key Laboratory of Molecular Engineering of Polymers 20 mm s1 at 25 C. Figure 1 shows the data thus collected.
of the Ministry of Education Figure 1a clearly demonstrates that, at 15 C, the DMTA
Fudan University
Shanghai 200433 (P. R. China) used in this experiment produces stressstrain curves for the
E-mail: zzshao@fudan.edu.cn single and fine silk filaments that closely resemble curves col-
Dr. D. Porter, Dr. D. P. Knight, Prof. F. Vollrath lected on our extremely sensitive, custom-built, rapid-re-
Department of Zoology, University of Oxford sponse stressstrain gauge from the same kind of silk.[8,9a] This
South Parks Road, Oxford OX1 3PS (UK)
control experiment ensured that the machine was adequate to
[**] This work is supported by the National Natural Science Foundation
of China (NSFC, No. 20244005), the Science and Technology Devel- our requirements, although this DMTA is normally used to
opment Foundation of Shanghai, the China-UK Science and Tech- study the dynamic mechanics of more robust polymer fi-
nology Research Foundation. For funding FV, DK and DP thank the bers.[12] We concluded that this method is an accurate and
British EPSRC (grant GR/NO1538/01) and BBSRC (S12778) as well
as the European Commission (grant G5RD-CT-2002-00738) and the powerful tool to survey the stretching properties of silk fibers
AFSOR of the United States of America (grant F49620-03-1-0111). under a regime of well-controlled temperatures.

84  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/adma.200400344 Adv. Mater. 2005, 17, No. 1, January 6
COMMUNICATIONS
a) b)

c) d)

Figure 1. The tensile properties of single filaments of major ampullate silk of the golden silk spider Nephila edulis measured by a dynamic mechanical
thermal analyzer at different temperatures. a) Stressstrain curves of spider silk at selected temperatures. Influence of the temperature on the mechan-
ical properties: b) initial modulus, c) breaking strength, and d) elongation at break. The measurement at each condition was performed at least 5
times, and the data were averaged.

We observed that increasing the temperature up to 150 C mer/silk protein blends,[14] or thick bundles of silk fila-
produced remarkably little change in the tensile behavior. ments.[15] To analyze in detail the effect of temperature on the
Surprisingly, cooling down to 60 C from room temperature mechanical behaviors of natural silk we tested single filaments
produced an increase of strength while the elongation to frac- of the Nephila spider dragline silk on the DMTA. A represen-
ture doubled. This means that at low temperatures our spider tative scan of the dynamic mechanical properties of such silk
silk can outperform the toughest (i.e., most energy absorbent) is displayed in Figure 2.
synthetic polymer fibers. Figures 1b1d show detailed mea-
surements on the effect of temperature on the tensile proper-
ties of single spider-silk fiber. Tensile strength decreased with
increasing temperature, comparable to most polymeric mate-
rials as well as even high-performance polymeric fibers such
as PPTA (Kevlar) and HBA/PET (hydroxybenzoic acid/poly-
(ethyleneterephthalate)) copolyester fibers.[13] However, (and
this is highly unusual for any of these fibers) in our silk, the
elongation at break decreased with increasing temperature up
to 70 C. Accordingly, our static mechanical measurements
suggest that the combination of high tensile strength and high
extensibility give the natural spider silk remarkable toughness
at low temperature. Gosline et al. demonstrated that the
strength and toughness of spider silk would be enhanced at
high strain rates.[3h] Their observation might arise from a simi-
lar phenomenon to ours, since for viscoelastic processes low Figure 2. Typical dynamic mechanical curves of Nephila edulis spider
dragline silk. From 66 C to 60 C, the phenomenon of E generally in-
temperatures are equivalent to high strain rates. creasing with temperature may be an experimental artifact and the slop-
Dynamic mechanical analysis has previously been used to ing baseline of tan d between 0 C and 360 C can be explained by the dif-
study either films of regenerated silk protein or films of poly- ferent creep abilities of spider silk with changing temperature.

Adv. Mater. 2005, 17, No. 1, January 6 http://www.advmat.de  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 85
COMMUNICATIONS

Usually, silk protein is regarded as a biological representa- as the temperature increased above 290 C, probably as the
tive of the nylon family, Nylon 2.[2b] Hence the thermal me- result of an annealing process that improved the hitherto im-
chanical relaxation behavior of spider silk can be addressed perfect[19] crystalline structure before its destruction. There-
similarly to those of commercial polyamides.[16] Figure 2 fore, it is reasonable to consider that the relaxation peaks of
shows that the tan d curve (giving the ratio of loss modulus E spider silk at 198 C and 309 C can be attributed to the break-
to storage modulus E) displayed a small peak at about ing down of hydrogen bonds in the disordered glassy phase
70 C. This suggests a structural sub-transition at this point[15a] and the subsequent partial crystallization of the silk,
probably due to the relaxation of hydrocarbonhydrocarbon respectively.
interactions.[16] A slight fluctuation was also seen at approxi- It will be interesting to examine the mechanisms of failure
mately 20 C; one reason for this may be related to the phase at low temperatures. Fractographic studies of spider silks had
transition of water. This interpretation is supported by the ob- previously been performed on the spiders Nephila clavipes
servation that taking a sample to about 110 C in the DMTA and Argiope trifasciata but only at room temperature and
resulted in a slight diminution of the fluctuation when a sec- after stretching.[15a,20] Un-stretched spider silk broken at very
ond warming scan was carried out after re-cooling to 100 C. low temperatures had not been studied before. Figure 3 shows
The increase of storage modulus up to a maximum occur- the fractured surface of a Nephila edulis spider-silk fiber bro-
ring at approximately 60 C correlated with the data obtained ken in liquid nitrogen. The fractured part frequently showed
by statistical measurements. A transition of initial modulus evidence that it stretched during the fracture process
and elongation at break values was found to occur around (Figs. 3a,b). This leads to the remarkable conclusion that,
70 C (Figs. 1b,d). This could indicate an increase in the intrin- unlike most other polymeric materials, the fracture of spider
sic stiffness of individual molecular chains or, alternatively, a silk is ductile and therefore tough rather than brittle, even at
stronger interaction between chains with an increase in tem- liquid-nitrogen temperatures. In contrast, silkworm silk gave a
perature from 66 to 60 C, or a combination of both. Tsukada very coarse transverse break in liquid nitrogen, clearly reveal-
et al.[17] also observed a slight increase of E from 30 to 60 C ing its microfibril structure (Fig. 3d).
in tussah silk fibers (albeit in bundles not single fibers) and After spider silk had fractured in liquid nitrogen, the outer
suggested that at that temperature there might be some rear- layer of the broken silk often showed wrinkles (Fig. 3c).
rangement of the molecules in the amorphous regions during These might arise from the different degrees of contraction of
the heating process, resulting in a strengthening of the inter- skin and core of the silk, comparable to the behavior during
chain interactions. Nevertheless, this phenomenon may simply the super-contraction when silk is immersed in water or other
be caused by artifacts during the experiment since a single organic solvents.[9] During the period when the broken silk fi-
thread of spider silk is normally thinner than 6 lm. Still, the ber is taken from the liquid nitrogen and is being mounted on
exact mechanism remains to be understood. the scanning electron microscopy (SEM) stubs, the moisture
There were two main peaks on the tan d curve: a large one in the air would have condensed on the fiber, which might
at 198 C and a small at 309 C. The large one we assigned to have led to super-contraction. Spider silk has a skincore
the breakdown of intermolecular hydrogen bonds in the non- structure,[3e] and a difference in the contraction in these com-
crystalline section of spider-silk fiber, indicating that this frac- ponents may contribute to this effect.
tion of the material had begun to enter the rubber state. It has Although both spider and silkworm silk appear to be large-
been suggested that the lattice of hydrogen bonds holding ly constructed from nanofibrils,[21] and both show a similar
together the b-sheet crystalline component of spider silk b-sheet secondary structure,[22] only silkworm silk exhibited a
breaks down at ~200 C.[15a] However, we think it more likely microfibrillar structure when fractured in liquid nitrogen
that the crystal phase breakdown or melting occurs at a much (Fig. 3d). This difference may correlate with a greater
higher temperature, and imparts the high temperature stabil- tendency of silkworm silk to fasciculate into bundles of nano-
ity of silk fibers; Nephila dragline silk shows excellent heat re- fibrils when homogenized (Knight, unpublished observa-
sistance, failing mechanically only at 371 C. tions). Both observations may indicate a weaker bonding of
The mechanical properties of spider dragline silk have been nanofibrils to the inter-nanofibrillary matrix in Bombyx (silk-
modelled by Termonia[3a] as a matrix composed of poorly worm) compared to spider silk.
ordered amorphous flexible rubber-like chains reinforced by Future work will attempt to link the dynamic mechanical
hard crystals. As a revision of this model, we suggest that, be- characteristics of silks to their low-temperature toughness.
low 198 C, the non-crystalline segments are orientated with The tan d peak at 70 C is large enough and broad enough to
respect to the thread[8,18] and held together by hydrogen bonds suggest that the associated relaxation process may encourage
that are broken when this critical temperature is reached. a local yield mechanism that promotes a tough failure
Thus, the molecular chains separate and become disordered mechanism.
to give the rubber state. However, in such a state, the crystal- In summary, major ampullate silk from the spider Nephila
line part of the silk still exists and provides multivalent cross- edulis showed remarkable toughness at very low temperatures
links. This interpretation explains two observations: firstly, (60 to 0 C) that had not been observed in most synthetic
spider silk showed a considerable storage modulus above fibers. Below 0 C, it had a high initial modulus (~14 GPa),
198 C; and secondly, the storage modulus increased slightly good tensile strength (~1.5 GPa) and high extensibility

86  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim http://www.advmat.de Adv. Mater. 2005, 17, No. 1, January 6
COMMUNICATIONS
a) c)

b) d)

Figure 3. The morphology of single filaments of silks fractured in liquid nitrogen. a) Profile, b) front view of cross-section, c) inclined plane of broken
Nephila edulis spider silk; d) front view of broken Bombyx mori silkworm silk, which was drawn back into the embedding epoxy resin.

(~40 %). This correlates with the tough fracture behavior in A dynamic mechanical thermal analyzer (DMTA IV; Rheometric
liquid nitrogen inferred from our SEM observations. Further- Scientific Inc.; Piscataway, NJ, USA) was used to measure both static
and dynamic mechanical properties of single spider silk threads. Sam-
more, and highly unusual, the elongation at break decreased ples were prepared as described previously [23]. Using single threads
with increasing temperature and reached a minimum around instead of bundles of fibers for dynamic mechanical thermal analysis
70 C. Meanwhile, dynamic mechanical measurements on the prevents uncertainty in interpretation resulting from lack of unifor-
same silk (using the same instrument) have showed relaxation mity of strain and from threads slipping past each other. To determine
the diameter of silk fibers, a different part of the same thread used in
processes associated with different molecular-level interac-
dynamic mechanical thermal analysis measurement was mounted
tions in the silk: hydrocarbonhydrocarbon, wateramide, and directly on SEM stubs and viewed at a magnification of about 3000
amideamide, as well as possible crystallization at higher tem- after sputtering with gold. At least three diameter measurements were
peratures. In addition, the failure temperature of dynamic recorded and averaged for each sample. The mean diameter of major
ampullate filaments was 6.2 0.4 lm in the present study. For static
measurement of the spider silk (371 C) was higher than silk-
measurements the strain rate was 0.05 % s1, temperature range from
worm silk (308 C). Thus, remarkably, spider silk retains quite 60 C to 150 C, and temperature control 0.5 C. For dynamic mea-
exceptional mechanical properties over a temperature range surements the frequency was 1 Hz, strain 0.6 %, temperature range
from at least 66 Cand probably down to liquid nitrogen from 100 C to 370 C, and temperature ramp 5 Cmin1.
temperaturesup to about 100 C. This temperature depen- Received: March 8, 2004
dence of mechanical properties of spider silk (as opposed to Final version: June 18, 2004
that of other polymers) adds to the other remarkable proper-
[1] F. Vollrath, Int. J. Biol. Macromol. 1999, 24, 81.
ties of this accomplished material[7a] and demonstrates the po-
[2] a) F. Vollrath, D. P. Knight, Nature 2001, 401, 541. b) J. P. O'Brien,
tential usefulness of such a super-fiber in harsh environments. S. R. Fahnestock, Y. Termonia, K. C. H. Gardner, Adv. Mater. 1998,
10, 1185. c) D. Kaplan, W. W. Adams, B. Farmer, C. Viney, ACS
Symp. Ser. 1994, 544, 3.
Experimental [3] a) Y. Termonia, Macromolecules 1994, 27, 7378. b) E. L. Mayes,
F. Vollrath, S. Mann, Adv. Mater. 1998, 10, 801. c) A. H. Simmons,
Single dragline silk filaments were forcibly reeled from the major C. A. Michal, L. W. Jelinski, Science 1996, 271, 84. d) T. Y. Yu,
ampullate glands of the Australian golden silk spider Nephila edulis. S. Zhang, Z. Z. Shao, Chem. J. Chin. Univ. 1996, 17, 323. e) W. K.
The technique is described in detail elsewhere [8,9a]. For these experi- Zhang, Q. B. Xu, S. Zou, H. B. Li, W. Q. Xu, X. Zhang, Z. Z. Shao,
ments the reeling rate was 2.0 cm s1. For the fracture experiments, M. Kudera, H. E. Gaub, Langmuir 2000, 16, 4305. e) S. Frische,
spider silk threads were embedded directly in epoxy resin without A. Maunsbach, F. Vollrath, J. Microsc. (Oxford) 1998, 189, 64.
treatment. After polymerization for 1 day at room temperature the f) D. T. Grubb, L. W. Jelinski, Macromolecules 1997, 30, 2860.
resin bars were cooled in liquid nitrogen for 10 min and then frac- g) J. D. van Beek, S. Hess, F. Vollrath, B. H. Meier, Proc. Natl. Acad.
tured. A Philips XL30 scanning electron microscope was used to ob- Sci. USA 2002, 99, 10 266. h) J. M. Gosline, P. A. Guerette, C. S. Ort-
serve the fracture surfaces of silks at 2025 kV. lepp, K. N. Savage, J. Exp. Biol. 1999, 202, 3295.

Adv. Mater. 2005, 17, No. 1, January 6 http://www.advmat.de  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 87
COMMUNICATIONS

[4] J. Perez-Rigueiro, C. Viney, J. Llorca, M. Elices, Polymer 2000, 41,


8433.
Effects of Carbon Nanotubes on Grain
[5] M. Kitagawa, H. Sasagawa, M. Kawagoe, J. Soc. Mater. Sci. Jpn. Boundary Sliding in Zirconia
1995, 44, 1445.
[6] M. Kitagawa, T. Kitayama, J. Mater. Sci. 1997, 32, 2005. Polycrystals**
[7] a) J. M. Gosline, M. E. MeMont, M. W. Denny, Endeavour 1986, 10,
37. b) J. Perez-Rigueiro, C. Viney, J. Llorca, M. Elices, J. Appl. By Maren Daraktchiev,* Bertrand Van de Moortle,
Polym. Sci. 1998, 70, 2439. Robert Schaller, Edina Couteau, and Lszl Forr
[8] F. Vollrath, B. Madson, Z. Z. Shao, Proc. R. Soc. London B 2001,
268, 2239.
[9] a) Z. Z. Shao, F. Vollrath, Polymer 1999, 40, 1799. b) Z. Z. Shao, Mechanical properties of zirconia polycrystals decrease
R. J. Young, F. Vollrath, Int. J. Biol. Macromol. 1999, 24, 295. drastically at high temperature. This is due to thermally acti-
[10] R. Foelix, in Biology of Spiders, Oxford University Press, Oxford, vated grain boundary (GB) sliding which leads to plastic or
UK 1996. even super-plastic deformation.[15] As GB sliding is a source
[11] a) L. H. Lin, D. T. Edmonds, F. Vollrath, Nature 1995, 373, 146.
of energy dissipation in the material, mechanical loss mea-
b) L. H. Lin, D. T. Edmonds, Environmental effects on the me-
chanical properties of silk and webs, presented at Int. Conf. on De-
surements[6,7] are well suited to study such a mechanism. They
formation, Yield and Fracture of Polymers, Cambridge, UK, April 7 reveal, in general, a mechanical loss peak, which evolves into
11, 1997. an exponential increase at higher temperatures.[811] When in-
[12] a) A. G. Simanke, G. B. Galland, L. Freitas, J. A. H. da Jornada, tergranular glassy films or/and amorphous pockets are pre-
R. Quijada, R. S. Mauler, Polymer 1999, 40, 5489. b) D. Mader, sented in polycrystalline ceramics, the mechanical loss[12,13]
Y. Thomann, J. Suhm, R. Mulhaupt, J. Appl. Polym. Sci. 1999, 74,
and creep rate are higher.[5,14,15] We show that introducing
838. c) L. C. Simon, R. F. de Souza, J. B. P. Soares, R. S. Mauler,
Polymer 2001, 42, 4885. carbon nanotubes (CNTs) in zirconia, in particular, drastically
[13] a) L. E. Nielsen, in Mechanical Properties of Polymers and Compos- reduces GB sliding and consequently the mechanical loss at
ites, Marcel Dekker, New York 1974, Ch. 2. b) J. R. Schaefgen, T. I. high temperatures. The nanotubes were observed at the grain
Bair, J. W. Ballou, S. L. Kwolek, P. W. Morgan, M. Panar, J. Zimmer- boundaries by high-resolution transmission electron microsco-
man, in Ultra High Modulus Fibres (Eds: A. Ciferri, I. M. Ward), py (HRTEM) and were related to the reduction of superplas-
Applied Science, London 1979. c) J. Jackson, J. Wu, H. F. Kuhfuss,
tic flow through the boundaries, which should improve the
J. Polym. Sci., Polym. Chem. Ed. 1976, 14, 2043.
[14] a) M. Tsukada, G. Freddi, N. Kasai, P. Monti, J. Polym. Sci., Part B: material creep resistance.
Polym. Phys. 1998, 36, 2717. b) H. Y. Kweon, I. C. Um, Y. H. Park, The mechanical loss, tan(u), spectrum associated with the
Polymer 2000, 41, 7361. brittleductile transition in fully tetragonal zirconia polycrys-
[15] a) P. M. Cunniff, S. A. Fossey, M. A. Auerbach, J. W. Song, ACS tals stabilized by 3 mol-% yttria (3Y-TZP) is presented in
Symp. Ser. 1994, 544, 234. b) K. B. Guess, C. Viney, Thermochim.
Figure 1. As measured as a function of temperature, the
Acta 1998, 315, 61.
[16] N. G. McCrum, B. E. Read, G. Williams, Anelastic and Dielectric
mechanical loss angle tan(u) shows an exponential increase
Effects in Polymeric Solids, Wiley, London 1967, Ch. 12. accompanied by a steep shear modulus (G) decrease above
[17] M. Tsukada, G. Freddi, M. Nagura, H. Ishikawa, N. Kasai, J. Appl. 1200 K.
Polym. Sci., 1992, 46, 1945. High-temperature plastic deformation of polycrystalline
[18] a) J. D. van Beek, L. Beaulieu, H. Schafer, M. Demura, T. Asakura, zirconia is associated with GB sliding, and consequently it
B. H. Meier, Nature 2000, 405, 1077. b) J. D. van Beek, J. Kummer-
seems reasonable to link the mechanical loss spectrum with
len, F. Vollrath, B. H. Meier, Int. J. Biol. Macromol. 1999, 24, 173.
c) C. Riekel, F. Vollrath, Int. J. Biol. Macromol. 2001, 29, 203.
d) C. Riekel, C. Branden, C. Craig, C. Ferrero, F. Heidelbach,
M. Muller, Int. J. Biol. Macromol. 1999, 24, 179.
[19] B. L. Thiel, K. B. Guess, C. Viney, Biopolymers 1997, 41, 703. [*] Dr. M. Daraktchiev[+]
[20] P. Poza, J. Perez-Rigueiro, M. Elices, J. Llorca, Eng. Frac. Mech. Institut des Matriaux (IMX)
2002, 69, 1035. Ecole Polytechnique Fdrale de Lausanne (EPFL)
[21] a) S. Putthanarat, N. Stribeck, S. A. Fossey, R. K. Eby, W. W. CH-1015 Lausanne (Switzerland)
Adams, Polymer, 2000, 41, 7735. b) L. D. Miller, S. Putthanarat, E-mail: maren.daraktchiev@epfl.ch
R. K. Eby, W. W. Adams, Int. J. Biol. Macromol. 1999, 24, 159. Dr. B. Van de Moortle, Dr. R. Schaller, Prof. L. Forr
c) D. P. Knight, F. Vollrath, Philos. Trans. R. Soc. London B 2002, Institut de Physique de la Matire Complexe (IPMC)
357, 155. Ecole Polytechnique Fdrale de Lausanne (EPFL)
[22] Z. Z. Shao, F. Vollrath, J. Sirichaisit, R. J. Young, Polymer 1999, 40, CH-1015 Lausanne (Switzerland)
2493. Dr. E. Couteau
[23] Y. Yang, X. Chen, P. Zhou, W. H. Yao, Z. Z. Shao, Chem. J. Chin. DSM Nutritional Products, R&D Chemical Process Research
Univ. 2001, 22, 1592. P.O. Box 3255, CH-4002 Basel (Switzerland)
[+] Present Address: Department of Earth Sciences, University of
Cambridge, Downing Street, Cambridge CB2 3EQ, UK.
[**] This work was supported by the Swiss National Science Foundation
and in part by the TOP NANO 21 Program. The authors would like
______________________ to thank L. Gremillard for his assistance in TEM observations,
B. Guisolan for technical help and Thomas Lagrange for critically
reading the manuscript. The CLYME (Consortium Lyonnais de Mi-
croscopie Electronique) were grateful acknowledged for the micro-
scope access.

88  2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/adma.200400598 Adv. Mater. 2005, 17, No. 1, January 6

You might also like