Corrosion-based synthesis of single-crystal Pd nanoboxes and nanocages and their surface plasmon properties

Article abstract of DOI:10.1002/anie.200502722

(Figure Presented) Inside out: Corrosive pitting and etching can be combined in a simple, elegant, and powerful method to transform single-crystal Pd nanocubes into nanoboxes (see SEM image) and nanocages in a one-pot synthesis without the involvement of exotic templates. Emptying the interiors enables the surface plasmon resonance peaks of the Pd nanostructures to be tuned from 410 to 520 nm.

Full text of DOI:10.1002/anie.200502722

Angewandte
Chemie
Nanostructures
controllable void size, wall porosity, and shell thickness.[4]
More interestingly, the surface plasmon resonance (SPR)
peaks of the resultant nanostructures could be continuously
shifted from the blue (400 nm) to the near infrared (1200 nm)
region by controlling the molar ratio of Ag to HAuCl4 and
thus the extent of replacement, alloying, and dealloying.
Ideally, one would prefer a one-step synthesis that is able to
tune the SPR properties of metal nanostructures without the
requirement of templates made of a different metal. Here we
demonstrate that corrosive etching of single-crystal Pd nano-
cubes could be used to transform them into nanoboxes and
nanocages by controlling the reaction time.
DOI: 10.1002/anie.200502722
Corrosion-Based Synthesis of Single-Crystal Pd
Nanoboxes and Nanocages and Their Surface
Plasmon Properties**
Yujie Xiong, Benjamin Wiley, Jingyi Chen, Zhi-Yuan Li,
Yadong Yin, and Younan Xia*
Corrosion can be broadly defined as “the damage to metal
caused by reactions with its environment,” and may take place
in many different forms such as pitting, crevice etching,
intergranular exchange, galvanic replacement, and dealloy-
ing.[1] Although corrosion is generally undesirable (automo-
bile corrosion alone costs $16 billion annually!), it can be
exploited as a versatile route to porous structures such as
metal membranes.[2] While many solution-phase methods
have been developed for synthesizing nanoparticles with well-
controlled shapes,[3] it is still a great challenge to generate
hollow nanostructures without involving exotic templates. We
recently found that the galvanic replacement reaction
between Ag templates and HAuCl4 in an aqueous medium
could lead to the formation of hollow nanostructures with
As demonstrated in previous studies, Pd nanoparticles
play an important role in hydrogen storage,[5] serve as the
primary catalyst for the reduction of pollutants emitted from
automobiles,[6] and facilitate organic reactions such as Suzuki,
Heck, and Stille couplings.[7] Their performance could be
tuned or improved for most of these applications by process-
ing Pd nanostructures into hollow ones. For example,
although solid Pd nanoparticles usually lose their catalytic
activity after one round of Suzuki coupling reactions,[7a]
hollow Pd spheres remained highly active for at least 24
rounds of reactions.[7b] While the SPR properties of Au and
Ag have enabled many applications such as colorimetric
sensing, nanoscale wave-guiding, and enhancement of elec-
tromagnetic fields and light transmission,[8] the SPR proper-
ties of Pd nanoparticles remain largely unexplored since Pd
particles less than 10 nm in size give no resonance peak in the
visible and near infrared regions.[9] Furthermore, the specific
sensitivity of Pd towards hydrogen makes it especially
interesting for fabricating SPR-based sensors for hydrogen
gas.[10] We found by using discrete dipole approximation
(DDA) calculations[11] that the resonant peak of Pd nano-
shells with outer diameters of 48 nm could be tuned from 530
to 870 nm by tailoring their wall thickness from 6 to 3 nm. To
demonstrate that Pd nanostructures with hollow interiors
could indeed exhibit tunable SPR extinction peaks in the
visible region we have synthesized single-crystal Pd nano-
boxes and nanocages by a self-templating, corrosive pitting
process.
[*] Dr. Y. Xiong, J. Chen, Prof. Y. Xia
Department of Chemistry
University of Washington
Seattle, WA 98195 (USA)
Fax: (+1)206-685-8665
E-mail: xia@chem.washington.edu
B. Wiley
Department of Chemical Engineering
University of Washington
Seattle, WA 98195 (USA)
Prof. Z.-Y. Li
Laboratory of Optical Physics
Institute of Physics
Chinese Academy of Sciences
Beijing 100080 (P.R. China)
The synthesis was conducted in a solution containing
ethylene glycol, water, and poly(vinyl pyrrolidone) (PVP)
with the addition of Na2PdCl4 as a precursor to Pd. The
reaction can be divided into six distinctive stages based on
both microscopic and spectroscopic observations. Reduction
of [PdCl4]2À to Pd atoms dominated up until t = 18 h (Fig-
ure 1a), with the Pd atoms nucleating and growing into
nanocubes (with slightly truncated corners) with edge lengths
of approximately 48 nm. Oxidative etching of the Pd nano-
cubes by O2 (that is, corrosion) started to surpass the
reduction process at t = 20 h. Small pits were observed in
the surfaces of individual Pd nanocubes (Figure 1b), thus
implying that the corrosion was initiated locally rather than
uniformly over the entire surface of each cube. Over the next
hour the interior of each cube became increasingly empty
while the size of the hole in the surface started to shrink to
form a box with relatively thick walls (Figure 1c). After
another hour the void was further increased in size and each
Pd cube evolved into a completely empty box with somewhat
Dr. Y. Yin
The Molecular Foundry
Lawrence Berkeley National Laboratory
Berkeley, CA 94720 (USA)
[**] This work was supported in part by a grant fromthe NSF (DMR-
0451788), a DARPA-DURINT subcontract fromHarvard University,
and a fellowship fromthe David and Lucile Packard Foundation. Y.X.
is a Camille Dreyfus Teacher Scholar (2002). B.W. and J.C. thank the
Center for Nanotechnology at the University of Washington for an
IGERT (funded by the NSF, DGE-9987620) and a Nanotech
fellowship, respectively. Z.-Y.L. was supported by the National Key
Basic Research Special Foundation of China (Grant no.
2004CB719804). Y.Y. was supported by the Director, Office of
Science, U.S. Department of Energy, under Contract DE-AC03-
76SF00098. This work used the Nanotech User Facility (NTUF), a
member of the National Nanotechnology Infrastructure Network
(NNIN) funded by the NSF. We thank the Molecular Foundry at the
Lawrence Berkeley National Laboratory for HRTEM analysis.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or fromthe author.
Angew. Chem. Int. Ed. 2005, 44, 7913 –7917
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7913

Communications
Figure 1. TEM images of the products at different reaction stages:
a) t=18 h; b) t=20 h; c) t=21 h; d) t=22 h; e) t=24 h; and f) t=28 h.
Note that the 100-nmbar applies to all images. The upper right and lower
left insets show enlarged TEM and SEM images of an individual particle
taken fromeach product, respectively.
Figure 2. Surface plasmon properties of the Pd nanostructures shown
in Figure 1: a) UV/Vis extinction spectra recorded fromaqueous
suspensions of the as-synthesized samples (see the inset, all spectra
are normalized against the intensities of their strongest peaks); and
b) DDA calculation of extinction, scattering, and absorption cross-
sections of the Pd nanocage depicted in the inset.
truncated corners (Figure 1d). Note that the hole in the
surface had by now disappeared so that the nanobox was
enclosed by a smooth, thin wall with a thickness of about
8 nm. Small holes appeared at the corners of each nanobox if
the reaction was continued beyond this point. Holes of about
10 nm in diameter punctured each corner at t = 24 h (Fig-
ure 1e), thus turning the nanobox into a cage. The Pd
nanocage had an edge length of approximately 48 nm and was
characterized by a uniform wall thickness of about 6 nm. It is
worth noting that the outer diameter of the nanobox and
nanocage exhibited no significant change relative to the initial
Pd nanocube. The nanocage finally evolved into a ring
(probably through additional etching and Ostwald ripening),
thus causing the wall thickness to increase slightly from 6 to
10–20 nm (Figure 1 f).
The Pd nanostructures displayed distinctive colors since
SPR is strongly dependent on wall thickness. As shown by the
inset in Figure 2a, the aqueous suspensions changed from
yellow to orange to purple as the solid Pd nanocubes were
transformed into hollow boxes and then porous cages.
Figure 2a shows the UV/Vis extinction spectra recorded
from the six aqueous suspensions. The Pd nanocubes exhib-
ited an SPR peak at 410 nm, which was consistent with the
results of DDA calculations (see the Supporting Informa-
tion).[11] The SPR peak was red-shifted to 430 nm as holes
appeared in their surfaces. The formation of hollow structures
with thick walls (and relatively rough inner surfaces) resulted
in the sample exhibiting an SPR peak at 443 nm. The
extinction peak was further shifted to 452 nm as the wall of
the nanoboxes was thinned to 8 nm. Once the product had
evolved into nanocages with an edge length of 48 nm and a
wall thickness of 6 nm, the SPR peak was shifted to 520 nm,
which matched the calculated spectra (Figure 2b). The
extinction (Cext), absorption (Cabs), and scattering (Csca
cross-sections are related by the following equation: Cext
abs + Csca
Figure 3a and b show typical SEM and TEM images of the
)
=
C
.
Pd nanoboxes at relatively low magnifications which indicate
that over 95% of the hollow structures inherited the
truncated, cubic shape of the initial cubes. The fringes in the
HRTEM images (Figure 3c and d) showed a period of 2.0 Š,
which was consistent with the {200} lattice spacing of face-
centered-cubic Pd. These images also displayed well-resolved,
continuous fringes with the same orientation, thus implying
that the Pd nanobox, like the starting Pd nanocube, was a
piece of single crystal. The inset of Figure 3c shows a
7914
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 7913 –7917

Angewandte
Chemie
expedite the pitting process.[13] On the other hand, the
presence of a small amount of water can enhance the
solubility of O2 and profoundly increase the corrosion
rate.[1] In addition to the augmentation by O2, ClÀ ions, and
H2O, the amount of PVP was also found to be critical to the
formation of hollow Pd nanostructures through a corrosion
process. A large amount of PVP (with the ratio of PVP to Pd
being 15:1) could ensure that the surface of each Pd nanocube
was effectively covered by this polymer. In addition, the
surface of each Pd nanocube is likely covered by a thin layer
of palladium oxide.[13a] Both of these factors could make the
surface of a Pd nanocube less reactive than the freshly
exposed interior once the pitting process has started.[13b]
Figure 4 summarizes all the major steps involved in the
corrosive etching process. Compared with the system without
water (which produced Pd particles ca. 8 nm in size),[12b] the
presence of water significantly accelerated the oxidative
etching of Pd, and thus reduced the number of seeds formed
in the nucleation step. As we have recently demonstrated in a
Figure 3. Electron microscopy characterization of the Pd nanoboxes
obtained at t=22 h: a) SEM image; b) TEM image; and c, d) high-
resolution TEM images. The inset of (c) shows an electron diffraction
pattern taken fromthe side surface of an individual Pd nanobox.
case where FeIII ions were added as the oxidative etchant,[14]
a
decrease in the number of seeds resulted in formation of Pd
nanocubes with larger sizes when the concentration of the Pd
precursor was maintained. As in the corrosion of bulk metals,
the presence of water also facilitated the pitting process. Once
a pit was formed at a specific site on the surface of a Pd cube
(step a), further corrosion would occur preferentially inside
nanoprobe electron diffraction (ED) pattern obtained by
aligning the beam perpendicular to one of the side faces of a
cubic box. The spots could be indexed to the {200} diffraction
of face-centered-cubic Pd. This diffraction pattern also
suggests that the Pd nanobox was a single crystal with its
symmetry similar to that of the initial Pd nanocube, which was
enclosed by {100} facets. Both powder XRD and microprobe
ED patterns (see the Supporting Information) further con-
firms the high phase purity and crystallinity of the Pd
nanoboxes.
To analyze the corrosive etching process that transformed
Pd nanocubes into nanocages we recorded UV/Vis spectra of
the solution at different stages of the reaction (see the
Supporting Information, the Pd nanoparticles were removed
by centrifugation before the spectra were taken). The
extinction peak at 242 nm (see the Supporting Information)
corresponds to [PdCl4]2À while that at 278 nm can be
attributed to water-substituted [PdCl4]2À ([PdCl4Àn
-
(H2O)n](2Àn)À) species.[12] The peak at 242 nm disappeared
quickly when [PdCl4]2À was reduced by ethylene glycol to
generate Pd atoms and then nanocubes (see the Supporting
Information). The concentration of water-substituted
[PdCl4]2À steadily increased as the nanocubes were etched
by corrosive pitting. Similar to the corrosion of bulk metals,[13]
O2, ClÀ ions, and H2O are all instrumental for the initiation of
pitting on the surfaces of Pd nanocubes. As documented in the
literature, corrosion most commonly occurs when the O2 in air
is dissolved in a thin film of water on the surface of a metal,
thereby picking up electrons from the cathode to form
hydroxide ions that then migrate towards the anode.[1] If the
anode and cathode are located at different sites on the metal
surface because of differences in the local environment (for
example, surface heterogeneity and/or concentrations of
ions), either pitting or crevice corrosion will take place.[1]
Like the corrosion of bulk metals, adsorption of chloride ions
can greatly increase the rate of surface diffusion and therefore
Figure 4. Schematic illustration summarizing all the major morpholog-
ical changes involved in the synthesis of Pd nanoboxes and nanocages
by corrosive etching: a) Pitting at a specific site on the surface of a
nanocube where the O2 dissolved in the solvent received electrons
fromthe cathode and generated hydroxide ions that migrated towards
the anode; b) formation of hollow structures after further etching of
the interior of the nanocube, and the concentration of water-substi-
tuted [PdCl4]2À increased; c) formation of a completely enclosed Pd
nanobox by reducing the water-substituted [PdCl4]2À species to form
Pd atoms at the edge of each hole; d) formation of nanocages by
dissolving Pd fromthe corners of the nanobox; and e) reconstruction
of the nanocage by relocating all the holes fromthe corners to side
faces and thickening of the wall as a result of additional reduction of
the water-substituted [PdCl4]2À species.
Angew. Chem. Int. Ed. 2005, 44, 7913 –7917
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7915

Communications
palladium(ii) sodium chloride (0.0486 g, Na2PdCl4, Aldrich) and
poly(vinyl pyrrolidone) (0.2747 g, PVP, M.W. = 55000, Aldrich)
were dissolved at room temperature in ethylene glycol (3 mL) and
water (3 mL), respectively. These two solutions (with the molar ratio
between Pd and the repeating unit of PVP being 1:15) were then
injected simultaneously into the flask with a syringe pump at a rate of
45 mLhÀ1. Heating of the reaction at 1008C was continued in air for
28 h. A set of samples were taken over the course of a synthesis with a
glass pipette. To minimize temperature perturbation during sampling,
the glass pipette was held just above the solution and preheated for
30 s before immersion. For the argon-protected synthesis, all other
experimental parameters were kept the same except for the
continuous bubbling of argon. The samples were washed with acetone
and then with ethanol several times before they were centrifuged at a
rate of 6000 rpm for 10 min to remove most of the ethylene glycol and
excess PVP. The samples were characterized by transmission electron
microscopy (TEM), high-resolution TEM, scanning electron micro-
scopy (SEM), electron diffraction (ED), and powder X-ray diffrac-
tion (PXRD).
the pit, thus leading to formation of a hollow nanostructure
(step b). As etching was continued, the concentration of
water-substituted [PdCl4]2À sufficiently increased to allow the
reduction by ethylene glycol to take place at the edge of a
hole, thus causing it to gradually shrink in size. Since the
reduction of water-substituted [PdCl4]2À by ethylene glycol
was minimal, most of these ionic species were retained in the
solution rather than being reduced and deposited onto the
outer surface of the Pd nanoboxes. This argument is
supported by the UV/Vis spectra (see the Supporting
Information), and it is for this reason why the outer
dimensions of the Pd nanocubes, nanoboxes, and nanocages
were essentially the same. It is the combination of continuous
etching of the Pd nanocube from the interior and some
reduction of the PdII species at the edge of the hole that
resulted in the formation of Pd nanobox by t = 22 h (step c). It
is worth pointing out that such a self-templating and self-
etching process has also been observed in the solution-phase
synthesis of ZnO microtubes.[15] The Pd nanoboxes were as
unstable as the nanocubes when exposed to the corrosive
environment. Under the illumination of light, the electric field
intensity should concentrate at the corners, thus increasing
the surface charge at these sites.[16] As a result, corrosive
etching of the Pd nanoboxes preferentially occurred at the
corners, thus leading to the formation of Pd nanocages with
all eight corners being truncated (step d). This is different
from the synthesis performed in the dark (see the Supporting
Information). Further corrosion caused these holes to
increase in dimension until the Pd nanocages evolved into
rings. At the same time, continuous reduction of the water-
substituted [PdCl4]2À species occurred on the exterior surface
of the nanocages, which led to the formation of nanorings
(step e). This course of materials transfer resembles the
conventional “Ostward ripening” process. Overall, the void
size, shell thickness, and wall porosity of Pd hollow nano-
structures were determined by the net difference between two
opposite reactions: the reduction of the PdII precursor to form
Pd0 atoms, and the oxidation of Pd (both atoms and nano-
structures) into PdII species.
TEM images and microprobe ED patterns were recorded on a
Phillips 420 transmission electron microscope operated at 120 kV.
HRTEM images and nanoprobe ED patterns were taken on a
Jeol 2010 LaB6 high-resolution transmission electron microscope
operated at 200 kV. PXRD patterns were recorded on a Philips 1820
diffractometer equipped with
a CuKa radiation source (l =
1.54180 Š). SEM images were taken on a FEI field-emission scanning
electron microscope (Sirion XL) operated at an accelerating voltage
of 20 kV. UV/Vis spectra were obtained using a Hewlett–Packard
8452A diode array spectrophotometer.
The calculation used in the article is based on the DDA
method.[11] DDA is a computational method developed for studying
both the scattering and absorption of electromagnetic radiation by
particles with sizes on the order of or smaller than the wavelength of
incident light. In the calculation, the particle is divided into an array
of N polarizable point dipoles, each of which is characterized by a
polarizability of ai. When the system is excited by a monochromatic
incident plane wave Einc, each dipole of the system will be subjected to
an electric field that can be split in two contributions: 1) the incident
radiation field Ei,inc and 2) the field radiated by all of the other
induced dipoles Ei,dip. The sum of both fields is the so-called local field
at each dipole (Ei,loc = Ei,inc + Ei,dip). Each dipole can be expressed as
an oscillating polarization with the dipole moment being Pi = ai Ei,loc
.
Both absorption and scattering cross-sections (Cabs and Csca) can be
obtained directly from Pi. The refractive index of bulk Pd was used in
all calculations reported in this article, and the nanoparticle was
assumed to be surrounded by and filled with water (when it became
porous).
In summary, metal corrosion is a familiar and common
natural phenomenon that, while normally viewed as a
detrimental process that destroys the usefulness of tools and
machines, can also serve as a simple, elegant, and powerful
method for the fabrication of hollow nanostructures with
controllable optical properties. More specifically, corrosive
pitting and etching could be combined to transform single-
crystal Pd nanocubes into nanoboxes and nanocages in a one-
pot synthesis without the involvement of exotic templates.
The SPR peaks of Pd nanostructures could be tuned from 410
to 520 nm by empting their interiors. Our DDA calculations
indicated that the SPR peak of the Pd nanocages with edge
lengths of 48 nm could be further red-shifted to 870 nm by
decreasing their wall thickness to 3 nm.
Received: August 3, 2005
Published online: November 22, 2005
Keywords: corrosion · nanostructures · palladium·
.
scanning probe microscopy · surface plasmon resonance
[1]S. A. Bradford, Corrosion Control, 2nd ed., CASTI Publishing,
Edmonton, Canada, 2001, pp. 1 – 51.
[2]Y. Ding, J. Erlebacher, J. Am. Chem. Soc. 2003, 125, 7772 – 7773.
[3]a) F. Kim, S. Connor, H. Song, T. Kuykendall, P. Yang, Angew.
Chem. 2004, 116, 3759 – 3763; Angew. Chem. Int. Ed. 2004, 43,
3673 – 3677; b) V. F. Puntes, K. M. Krishnan, A. P. Alivisatos,
Science 2001, 291, 2115 – 2117; c) L. A. Gugliotti, D. L. Feld-
heim, B. E. Eaton, Science 2004, 304, 850 – 852; d) X. Teng, H.
Yang, Nano Lett. 2005, 5, 885 – 891.
Experimental Section
In a typical synthesis, ethylene glycol (5 mL, J. T. Baker) was placed in
a 3-neck flask (equipped with a reflux condenser and a teflon-coated
magnetic stirring bar) and heated in air at 1008C for 1 h. Meanwhile,
[4]a) Y. Sun, Y. Xia, Science 2002, 298, 2176 – 2179; b) Y. Sun, Y.
Xia, J. Am. Chem. Soc. 2004, 126, 3892 – 3901; c) B. Wiley, Y.
7916
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 7913 –7917

Angewandte
Chemie
Sun, J. Chen, H. Cang, Z.-Y. Li, X. Li, Y. Xia, MRS Bull. 2005,
30, 356 – 361.
[5]Y. Sun, Z. Tao, J. Chen, T. Herricks, Y. Xia, J. Am. Chem. Soc.
2004, 126, 5940 – 5941.
[6]Y. Nishihata, J. Mizuki, T. Akao, H. Tanaka, M. Uenishi, M.
Kimura, T. Okamoto, N. Hamada, Nature 2002, 418, 164 – 167.
[7]a) Y. Li, X. M. Hong, D. M. Collard, M. A. El-Sayed, Org. Lett.
2000, 2, 2385 – 2388; b) S.-W. Kim, M. Kim, W. Y. Lee, T. Hyeon,
J. Am. Chem. Soc. 2002, 124, 7642 – 7643; c) M. T. Reetz, E.
Westermann, Angew. Chem. 2000, 112, 170 – 173; Angew. Chem.
Int. Ed. 2000, 39, 165 – 168; d) S. U. Son, Y. Jang, J. Park, H. B.
Na, H. M. Park, H. J. Yun, J. Lee, T. Hyeon, J. Am. Chem. Soc.
2004, 126, 5026 – 5627.
[8]Y. Xia, N. J. Halas, MRS Bull. 2005, 30, 338 – 343.
[9]J. A. Creighton, D. G. Eadon, J. Chem. Soc. Faraday Trans. 1991,
87, 3881 – 3891.
ˇ
[10]P. Tobis ka, O. Hugon, A. Trouillet, H. Gagnaire, Sens. Actuators
B 2001, 74, 168 – 172.
[11]a) W. H. Yang, G. C. Schatz, R. P. Van Duyne, J. Chem. Phys.
1995, 103, 869 – 875; b) K. L. Kelly, E. Coronado, L. L. Zhao,
G. C. Schatz, J. Phys. Chem. B 2003, 107, 668 – 677; c) I. O. Sosa,
C. Noguez, R. G. Barrera, J. Phys. Chem. B 2003, 107, 6269 –
6275.
[12]a) B. Veisz, Z. Kirꢀly, Langmuir 2003, 19, 4817 – 4824; b) Y.
Xiong, J. Chen, B. Wiley, Y. Xia, S. Aloni, Y. Yin, J. Am. Chem.
Soc. 2005, 127, 7332 – 7333.
[13]a) R. C. Newman, K. Sieradzki, Science 1994, 263, 1708 – 1709;
b) J. C. Scully, The Fundamentals of Corrosion, 3rd ed., Perga-
mon, Oxford, 1990, pp. 1 – 57.
[14]Y. Xiong, J. Chen, B. Wiley, Y. Xia, Y. Yin, Z.-Y. Li, Nano Lett.
2005, 5, 1237 – 1242.
[15]L. Vayssieres, K. Keis, A. Hagfeldt, S. E. Lindquist, Chem. Mater.
2001, 13, 4395 – 4398.
[16]a) J. P. Kottmann, O. J. F. Martin, D. R. Smith, S. Schultz, Phys.
Rev. B 2001, 64, 235402(1 – 10); b) A. J. Haes, C. L. Haynes,
A. D. McFarland, G. C. Schatz, R. P. Van Duyne, S. Zhou, MRS
Bull. 2005, 30, 368 – 375.
Angew. Chem. Int. Ed. 2005, 44, 7913 –7917
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7917