Vol 460 | 23 July 2009 | doi:10.1038/nature08173
LETTERS
Near-field focusing and magnification through
self-assembled nanoscale spherical lenses
Ju Young Lee1*, Byung Hee Hong1,2*, Woo Youn Kim1, Seung Kyu Min1, Yukyung Kim1, Mikhail V. Jouravlev1,
Ranojoy Bose3, Keun Soo Kim2, In-Chul Hwang1, Laura J. Kaufman4, Chee Wei Wong3, Philip Kim5 & Kwang S. Kim1
released CHQ molecules re-assemble in three dimensions. These
are intermediate structures of anisotropically growing spheres (see
Supplementary Information A for the mechanism). These ‘planospherical convex’ (PSC) structures—with a spherical face on one side
and a flat face on the other side—can be isolated, and these CHQ
lenses are stable in air. The negative electron beam resist properties of
calixarene-based structures19 are used for fabrication and positioning
of optical devices using these CHQ lenses (Supplementary
Information D). Although these experiments are technically sophisticated, the intricately self-assembled nanolenses enable us to study the
physics of nanolens optics.
The size distribution of CHQ lenses can be controlled by the time
and temperature of the self-assembly process. Typically, PSC lenses
a
100 nm
2 µm
200 nm
100
00 nm
Fractures
b
100 nm
CHQ lens
CHQ film
CHQ crystal
1. Nucleation
2. 2D growth
3. 3D growth
1 µm
1 µm
4. Separation
Film
Crystal
1 µm
Film
c
d
Perfect sphere
1 µm
Lens
e
0.8
z (µm)
It is well known that a lens-based far-field optical microscope
cannot resolve two objects beyond Abbe’s diffraction limit.
Recently, it has been demonstrated that this limit can be overcome
by lensing effects driven by surface-plasmon excitation1–3, and by
fluorescence microscopy driven by molecular excitation4.
However, the resolution obtained using geometrical lens-based
optics without such excitation schemes remains limited by
Abbe’s law even when using the immersion technique5, which
enhances the resolution by increasing the refractive indices of
immersion liquids. As for submicrometre-scale or nanoscale
objects, standard geometrical optics fails for visible light because
the interactions of such objects with light waves are described
inevitably by near-field optics6. Here we report near-field high
resolution by nanoscale spherical lenses that are self-assembled
by bottom-up integration7 of organic molecules. These nanolenses, in contrast to geometrical optics lenses, exhibit curvilinear
trajectories of light, resulting in remarkably short near-field focal
lengths. This in turn results in near-field magnification that is able
to resolve features beyond the diffraction limit. Such spherical
nanolenses provide new pathways for lens-based near-field focusing and high-resolution optical imaging at very low intensities,
which are useful for bio-imaging, near-field lithography, optical
memory storage, light harvesting, spectral signal enhancing, and
optical nano-sensing.
Miniaturized lenses are often found in biological systems8,9, and have
been widely used for optical microelectromechanical systems10. Despite
numerous studies of miniaturized lenses11–14, no serious studies have
been undertaken of lenses for subwavelength nano-optics. To this end,
we fabricate well-defined nanoscale lenses of calix[4]hydroquinone
(CHQ), which is composed of four p-hydroquinone subunits with
eight hydroxyl groups15. Intermolecular short hydrogen-bonding16
and p2p stacking interactions17 are very useful forces for selfassembling supramolecular CHQ nanostructures18. One class of such
self-assembled nanostructures is represented by sphere-derived shapes
of diameter 50 nm to 3 mm, as seen in scanning electron microscopy
(SEM) images (Fig. 1a–c). Dissolving the CHQ monomers in 1:1
water–acetone solution leads to the formation of needle-like CHQ
nanotube crystals with infinitely long hydrogen-bonded arrays. As
the crystals grown at –14 uC are heated at 40 uC in aqueous environments for a day, CHQ molecules released from the crystals re-assemble
into nanospheres.
At the beginning of this process, film-like structures of CHQ cover
the surface of the crystals. CHQ molecules released from the surface
accumulate in a small volume under the film, leading to the nucleation and growth of two-dimensional disk-shaped structures.
Spherical curvatures are then gradually formed, as more of the
ι
0.6
1 µm
0.2
ι′
D = 1.610 µm
H = 0.590 µm
R = 0.844 µm
0.4
–0.5
0.0
x (µm)
0.5
500 nm
Figure 1 | CHQ plano-spherical convex lenses. a, SEM images of growing
CHQ nanospheres and their intermediate structures. b, Schematic diagrams
and SEM images showing the self-assembly of CHQ lenses (see text for
details). c, SEM image showing various sizes of CHQ lenses separated as an
aqueous suspension and drop-dried on a substrate. d, AFM profile showing
the near perfect spherical face of the lens. Inset, corresponding SEM image.
e, Optical microscope image of CHQ lenses on a CHQ nanotube crystal,
showing the magnification by the lens. The line spacing (l) behind the lens is
considerably increased (l9).
1
Center for Superfunctional Materials, Department of Chemistry, Pohang University of Science and Technology, Hyojadong, Namgu, Pohang 790-784, Korea. 2Department of
Chemistry and SKKU Advanced Institute of Nanotechnology, Sungkyunkwan University, Suwon 440-746, Korea. 3Department of Mechanical Engineering, 4Department of Chemistry,
5
Department of Physics, Columbia University, New York, New York 10027, USA.
*These authors contributed equally to this work.
498
©2009 Macmillan Publishers Limited. All rights reserved
LETTERS
NATURE | Vol 460 | 23 July 2009
a
a
b
~µm images
(objective lens)
b
c
~400 nm
In focus
Far-field path
–1.0
–0.5
0
+0.5
+1.0
+1.5
+2.0
+2.5
+3.0
+3.5
+4.0
+4.5
+5.0
+5.5
+6.0
+6.5
+7.0
+7.5
+8.0
+8.5
Observed path
Face -up
2 µm
c
Image inversion
(focal point)
2 µm
30º tilted
d
250 nm
CHQ nanolens
250 nm
~470 nm
1×1 cm2 object
(condenser lens)
d
0
+1.0
+2.0
+3.0
+4.0
+5.0
e
3
Face -down
1 µm
30º tilted
e
f
~380 nm
220 nm
Image radius (µm)
2
1 µm
rved
Obse
1
paths
aths
eld p
Far-fi
Incident
beams
0
–1
–2
–3
1 µm
–5
1 µm
Figure 2 | Optical microscope/SEM images of CHQ lenses on patterned
substrates. a, Optical microscope image of a face-up lens placed on a glass
substrate with Pd stripe patterns (see Supplementary Information C for
more images taken with different focus and magnification). b, SEM image
corresponding to a. Inset, image of tilted lens. c, Optical microscope image of
a face-down lens. d, SEM image corresponding to c. e, Optical microscope
image of a face-up lens. Inset, light intensity profile taken from the blue
dotted line. f, SEM image corresponding to e. The sub-diffraction-limit
patterns cannot be resolved in conventional optical microscopy, but the
magnifying effect through the lens allows the stripe patterns of 250/220 nm
spacing to be resolved.
with nanoscale thickness H , 800 nm and diameter D 5 0.05–3 mm
can be synthesized and separated from the aqueous suspension for
further experiments. Surface roughness of the lenses deposited on a
SiO/Si substrate is determined by an atomic force microscope (AFM).
The round surface exhibits a typical deviation from a spherical surface
of less than 3% with a surface roughness of ,1 nm (Fig. 1d). The
nearly perfect PSC structure demonstrates that self-assembled CHQ
lenses are high-quality optical elements.
Figure 1e shows an optical microscope image of a CHQ lens
(D 5 970 nm, H 5 220 nm) on top of CHQ tubule bundles under
filtered light (lmax 5 472 nm) from a halogen lamp. The image indicates that the lens magnifies the underlying object with a magnification factor M (5l9/l 5 ,1.6). The paraxial focal length estimated
from the observed magnification is F 5 HM/(M 2 1) 5 590 nm,
much shorter than that expected from geometric optics,
Fgeo 5 R/(n 2 1) 5 1.3 mm (R, lens radius; n 5 1.5, refractive index).
The deviations from geometrical optics for the subwavelength-size
CHQ lenses are the key signature of near-field focusing. The reduction of the focal length in the CHQ nanolens implies enhanced magnifying effects through the nanolens.
To demonstrate enhanced spatial resolution, we further investigate the optical properties of CHQ lenses on pre-fabricated
–4
–4
–3
–2
–1
0
1
2
zip (µm)
3
4
5
6
7
8
Figure 3 | Optical images and beam trajectories of alphabetical characters
projected through CHQ lenses and PMMA disks. a, Schematic illustration
of the imaging process. The light sources transmitting a patterned glass are
collimated by a condenser lens. As the nanolens focuses the beam, inverted
images are formed after passing through the focal point. Each slice of the
images is obtained at a distance zip from the lens/disk bottom. b, Optical
microscope images of various inverted alphabetical characters (A–F)
projected through the lens. The upright non-inverted characters are imaged
due to the image inversion at the image plane ,3.5 mm away from the
nanolens. c, Optical microscope images of the alphabetical character ‘‘E’’
projected through a CHQ lens (D 5 1.7 mm, H 5 480 nm) on vertical image
planes at varying distances zip. The number in each slice denotes the value of
zip. ‘In-focus’ denotes the image focused on the plane of the Pd stripe
patterns. d, Projected optical microscope images through a 1.7-mm-wide and
400-nm-thick PMMA disk, which shows no clear alphabetical character
image formation. e, Beam trajectory with reduced focal length in the nearfield PSC lens. Small insets on the left, AFM images of the CHQ lens (upper)
and the PMMA disk (lower). Large insets on the left, FDTD simulation
results of the radial component of the electric field (Ex) of the PSC lens
(upper) and the PMMA disk (lower) (l 5 472nm). All scale bars, 2 mm.
subwavelength objects as follows (Fig. 2). First, 250- and 220-nm
pitch metallic (Pd/Cr, 120/3 nm thick) stripe arrays are fabricated
by electron beam lithography on glass substrates. The CHQ lenses
isolated in aqueous suspensions (Fig. 1c) are randomly placed on the
patterned substrate by spinning (Fig. 2b, d, f). We use a high resolution optical microscope to take reflection mode images of the same
part of the sample (Fig. 2a, c, e). The optical images were obtained
through a 1003 objective lens with a numerical aperture (NA) of 0.9.
The Rayleigh resolution limit for point objects (r 5 0.61l/NA) is
320 nm, while that for line objects (r 5 0.5l/NA) is 262 nm. A more
stringent Sparrow resolution limit (r 5 0.475l/NA)20 is 249 nm.
Indeed, the optical images outside the CHQ lenses do not resolve
the underlying metallic stripes, as the stripe spacings of d 5 220 and
250 nm are narrower than, or very similar to, the stringent resolution
limit. On the other hand, resolved individual metallic stripes are
clearly imaged through the CHQ lens (Fig. 2a, c for 250 nm and
Fig. 2e for 220 nm). The image magnification increases as the distance
499
©2009 Macmillan Publishers Limited. All rights reserved
LETTERS
NATURE | Vol 460 | 23 July 2009
a
b
D = 0.8 µm
1.0
c
1
D = 2 µm
0.5
z (µm)
0.0
H
0
H
0
–1
–0.5
–1.0
D = 4 µm
1
H
Fmax = 0.85H
–1
Fmax = 1.46H
–2
Fmax =1.68*H
–1.5
–1.5 –1.0 –0.5 0.0 0.5 1.0 1.5
x (µm)
Fmax = 1.68H
–3
–2
–2.0
–1.5 –1.0 –0.5 0.0 0.5 1.0 1.5
x (µm)
–2
–1
0
x (µm)
1
2
Figure 4 | Focal length changes for various sizes of CHQ lenses (fixed H/D 5 0.35). a, D 5 0.8 mm; b, D 5 2 mm; c, D 5 4 mm. Data were obtained from
FDTD simulation results of | Ex | 2 (l 5 472 nm).
between nanolens and image is increased with micromanipulators
and piezo-controlled nano-positioning stages. The magnifying effect
enhances the resolution substantially (by as much as 2.5 times). The
magnified images of the face-up lenses show pin-cushion distortion
(Supplementary Information C), whereas no notable distortion
appears for the face-down lens. This difference arises because the
near-field image of the face-up lens is formed by the interference of
the secondary Fresnel’s waves on the flat and convex surfaces of the
lens, while the near-field image of the face-down lens is formed by the
secondary surface waves due to the convex surface.
To analyse the imaging and focusing behaviours through the CHQ
lenses, we projected a series of alphabetical character images into a
CHQ lens (D 5 1.7 mm, H 5 0.48 mm) via a condenser lens
(NA 5 0.8) in the far field (Fig. 3a–c). The images of ‘‘E’’ formed
by the PSC lens (Fig. 3c) are compared with those formed by a flat
disk (Fig. 3d). The transmitted images through the PSC lens or flat
disk were recorded by an optical microscope with a CCD camera
focused on different image planes at a distance zip from the lens/disk
bottom (Supplementary Information F). In Fig. 3e, the solid lines are
guides to the eye, following the square dots (which are the measured
optical beam trajectories of the top and bottom edges of the ‘‘E’’
image along zip). The far-field optical paths calculated from geometrical ray optics (dotted lines) are also presented for comparison. As
the image plane moves away from the CHQ lens, the images are
inverted and magnified. In Fig. 3c, we observe clear magnified and
inverted images of ‘‘E’’ along the axial positions for zip $ 1.5–2.0 mm,
and the letter shape is still seen at zip 5 1.0 mm. At zip , 0.5 mm, we
observe that the bright spot with a dark annular ring at zip 5 0.5 mm
changes to the grey/dark spot with a bright annular ring surrounding
the dark spot. We did not observe the direct non-inverted image for
zip , 0.5 mm. However, since the ‘‘E’’ image is clearly inverted for
zip $ 1.5 mm, we deduce that the image inversion should occur at
the focal point between the two adjacent points at zip 5 0.5 mm (lens
top) and zip 5 1.0 mm, which have the first and second smallest bright
spot radii (0.38 and 0.42 mm, respectively) among many images taken
at every 0.5-mm step.
As the lens height is 0.48 mm and the bright spot size at zip 5 0.5 mm
is slightly smaller than that at zip 5 1.0 mm, the focal point is expected
to be located at 0.48 mm , zip , 0.75 mm (then the focal length is not
more than 0.27 mm). Thus, this focal length is in agreement with the
theoretically estimated focal length (0.3 mm, or zip 5 ,0.8 mm)
assumed as the first Fresnel zone focal length (F 5 4l/(pNA)2).
Given the axial resolution limit (Rayleigh range zR 5 (4/p)l/NA2)
of 0.9 mm according to Gaussian beam optics, the measured focal
point is at zip 5 ,0.7 6 0.9 mm, or F 5 ,0.2 6 0.9 mm, much smaller
than the geometrical focal length Fgeo (2.0 mm; zip 5 ,2.5 mm). This
drastically shortened focal length (F = Fgeo), which should yield the
curvilinear trajectories for zip , 1.0 mm, is supported in Fig. 3e by
tracing the image size of the ‘‘E’’ in Fig. 3c down to zip 5 1.0 mm. It is
further clearly seen in accurate finite-difference time domain
(FDTD) simulations that the inversion occurs at zip 5 0.8 mm
(Fig. 3e top left inset). Here, the images between zip 5 0.0 mm and
the focal point are in the Fresnel diffraction region, appearing as the
Fourier transformed image. In this region, the asymptotic beam path
is curvilinear so that the image can be inverted at the short focal
point. This result explains the experiment showing the inverted
image of ‘‘E’’ consistently obtained for zip $ 1.0 mm in Fig. 3c.
Shape-dependent focusing and imaging was studied by performing a similar experiment through a patterned poly(methylmethacrylate) (PMMA) disk formed by electron beam lithography. The disk
has flat surfaces on both sides, but its thickness and size are similar to
those of the CHQ lens. In contrast to the clear images created through
the PSC lens, the light propagating through the disk shows no clear
images (Fig. 3d). These experimental images are consistent with the
electromagnetic simulation (Fig. 3e bottom left inset); at
zip 5 ,0 mm along the axis, the image has a small bright spot; at
zip 5 ,1 mm, the image has a small dark spot at the centre and an
annular bright spot; and the images for zip $ ,2 mm again have
bright spots at the centre. The difference between the images created
through the PSC lens and the flat disk is thus clearly confirmed by
electromagnetic simulations, and the image formation depends on
the shape (surface curvature) of the optical elements. We note that
the focal point of a nanolens originates mainly from surface waves at
the spherical interface of the PSC lens, and the near-field focal length
is drastically shortened by the interference of the propagating waves
from the lens edges because the nanolens length-scale is comparable
to the wavelength (Supplementary Information B). The resulting
curvilinear ray path with a wave propagation to match the amplitude
and phase inside and outside the lens forms a small focused spot at a
very short near-field focal distance (solid lines in Fig. 3e).
To investigate the size-dependent diffraction/refraction phenomena
in the near-field regime, we performed FDTD simulations21 for subwavelength-size lenses of different diameters (D 5 0.8–4 mm) with
fixed ratio of H/D 5 0.35 (Fig. 4). The incident plane waves are
polarized along the x-direction, and Fig. 4 shows the spatial distribution of jExj2 where Ex is the x-component of the electric field. As
the lens size approaches the wavelength, the near-field focal length
(Fmax) showing the maximum light intensity decreases remarkably
towards the first Fresnel zone. This near-field focusing is due to superposition of the diffraction on the flat aperture and the interference of
secondary Fresnel waves on the spherical surface of the PSC lens, whose
diameter is comparable to the wavelength of light.
Given that high resolution imaging beyond or near the diffraction
limit has been achieved by relying on a stimulated-emission-depletion
method, a lens-less near-field optical method, and novel materials
approaches22–26, the near-field focusing and magnification discussed
here represent a complementary approach to obtaining lens-based high
resolution beyond the diffraction limit at low intensities. This magnification can be further increased by using immersion lensing techniques. Such a combination would lead to hyper-refraction phenomena
due to surface waves on the interfaces. Based on the merit of spherical
lens-based optics, the near-field focusing and magnification phenomena in nanoscale lenses would have wide applications, including
super-resolution by a nanolens array and by a nanolens on an AFM
500
©2009 Macmillan Publishers Limited. All rights reserved
LETTERS
NATURE | Vol 460 | 23 July 2009
or aperture-less near-field scanning optical microscope (ANSOM). This
is possible because nanolenses at arbitrary locations can successfully be
transferred to pre-determined locations by using micromanipulators
with the assistance of an SEM, and fabricated in array in order to
increase the area over which high spatial resolution is achieved
(Supplementary Information E). In particular, the focusing and magnification effect of the nanolens would have useful applications in signal
enhancements in spectroscopy—in micro-photoluminescence intensity with quantum dots for single (near-infrared) photon spectroscopy
(Supplementary Information H), and in Raman intensity on substrates
such as graphene27,28 (Supplementary Information I). Application to
deep ultraviolet lithography is also possible (Supplementary
Information G), because regardless of the visible or ultraviolet wavelength, nanolenses can be used to obtain high resolution beyond the
diffraction limit, as long as the wavelength is comparable to the lens size.
METHODS SUMMARY
CHQ nanoscale lenses are self-assembled in 1:1 acetone–water solution by
evaporating the solution slowly for a few days in ambient conditions. After
dissolving 10 mg CHQ monomers in 2 ml acetone solvent, 2 ml water is added.
Slow evaporation of solution leads to the self-assemblies of CHQ nanostructures
including CHQ lenses. Pd/Cr (120/3 nm thick, 220 and 250 nm pitch) line
patterns are fabricated on a 0.13–0.16 mm thick glass substrate with e-beam
lithography. After depositing chromium to a thickness of 3 nm using a thermal
evaporator (BOC Auto 306, Edwards), positive electron-beam resist PMMA is
spin-coated on a SiO2 substrate at 4,000 r.p.m. The PMMA film is exposed to
30 kV electron beam and developed with MIBK:IPA 5 1:3 developer for patterning. Deposition of 120 nm palladium is performed with an electron-beam
evaporator (SC2000, SEMICORE), and then the patterns are accomplished
through a lift-off process. We obtained the magnified image through the CHQ
lens using an optical microscope equipped with micromanipulators and piezocontrolled nano-positioners (E-610, Physik Instrumente). FDTD simulations
were performed using the FullWAVE 4.0 program (RSoft Design Group). In
the Supplementary Information we describe relocation of CHQ lenses with a
dual focused ion beam microscope (Helios Nanolab 600, FEI) and an SEM
(JSM6390, Jeol). CHQ lenses are attached to the end of a tungsten tip (which
is adhesive after contact with an adhesive carbon tape), and then transferred to
pre-determined positions. The enhancement of Raman intensity of graphene
through a CHQ lens is observed by a micro-Raman microscope (InVia Raman
microscope, Renishaw; laser wavelength 633 nm, power 4 mW).
Received 30 December 2008; accepted 28 May 2009.
1.
Smolyaninov, I. I., Hung, Y.-J. & Davis, C. C. Magnifying superlens in the visible
frequency range. Science 315, 1699–1701 (2007).
2. Liu, Z., Lee, H., Xiong, Y., Sun, C. & Zhang, X. Far-field optical hyperlens magnifying
sub-diffraction-limited objects. Science 315, 1686 (2007).
3. Fang, N., Lee, H., Sun, C. & Zhang, X. Sub-diffraction-limited optical imaging with a
silver superlens. Science 308, 534–537 (2005).
4. Hell, S. W. Far-field optical nanoscopy. Science 316, 1153–1158 (2007).
5. Yano, T., Shibata, S. & Kishi, T. Fabrication of micrometer-size glass solid
immersion lens. Appl. Phys. B 83, 167–170 (2006).
6. Merlin, R. Radiationless electromagnetic interference: evanescent-field lenses
and perfect focusing. Science 317, 927–929 (2007).
7. Whitesides, G. M. & Grzybowski, B. Self-assembly at all scales. Science 295,
2418–2421 (2002).
8. Aizenberg, J., Tkachenko, A., Weiner, S., Addadi, L. & Hendler, G. Calcitic
microlenses as part of the photoreceptor system in brittlestars. Nature 412,
819–822 (2001).
9. Lee, L. P. & Szema, R. Inspirations from biological optics for advanced photonic
systems. Science 310, 1148–1150 (2005).
10. Dong, L., Agarwal, A. K., Beebe, D. J. & Jiang, H. Adaptive liquid microlenses
activated by stimuli-responsive hydrogels. Nature 442, 551–554 (2006).
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
Yang, S. K. U., Chaitanya, T. E. L., Chen, G. & Aizenberg, J. Microlens arrays with
integrated pores as a multipattern photomask. Appl. Phys. Lett. 86, 201121 (2005).
Jeong, K.-H., Kim, J. & Lee, L. P. Biologically inspired artificial compound eyes.
Science 312, 557–561 (2006).
Fletcher, D. A., Goodson, K. E. & Kino, G. S. Focusing in microlenses close to a
wavelength in diameter. Opt. Lett. 26, 399–401 (2001).
Wu, H., Odom, T. W. & Whitesides, G. M. Connectivity of features in microlens
array reduction photolithography: generation of various patterns with a single
photomask. J. Am. Chem. Soc. 124, 7288–7289 (2002).
Hong, B. H. et al. Self-assembled arrays of organic nanotubes with infinitely long
one-dimensional H-bond chains. J. Am. Chem. Soc. 123, 10748–10749 (2001).
Kim, K. S. et al. Assembling phenomena of calix[4]hydroquinone nanotube
bundles by one-dimensional short hydrogen bonding and displaced p-p stacking.
J. Am. Chem. Soc. 124, 14268–14279 (2002).
Kim, K. S., Tarakeshwar, P. & Lee, J. Y. Molecular clusters of p-systems: theoretical
studies of structures, spectra and origin of interaction energies. Chem. Rev. 100,
4145–4186 (2000).
Hong, B. H., Bae, S. C., Lee, C.-W., Jeong, S. & Kim, K. S. Ultrathin single-crystalline
silver nanowire arrays formed in an ambient solution phase. Science 294, 348–351
(2001).
Fujita, J., Ohnishi, Y., Ochiai, Y. & Matsui, S. Ultrahigh resolution of calixarene
negative resist in electron beam lithography. Appl. Phys. Lett. 68, 1297–1299
(1996).
Goldstein, D. J. Resolution in light microscopy studied by computer simulations. J.
Microsc. 166, 185–197 (1992).
Taflove, A. & Hagness, S. C. Computational Electrodynamics: The Finite-difference
Time-domain Method (Artech House, 2000).
Simpson, G. J. Biological imaging: the diffraction barrier broken. Nature 440,
879–880 (2006).
Lewis, A. et al. Near-field optics: from subwavelength illumination to nanometric
shadowing. Nature Biotechnol. 21, 1378–1386 (2003).
Smith, D. R., Pendry, J. B. & Wiltshire, M. C. K. Metamaterials and negative
refractive index. Science 305, 788–792 (2004).
Parimi, P. V., Lu, W. T., Vodo, P. & Sridhar, S. Photonic crystals: imaging by flat lens
using negative refraction. Nature 426, 404 (2003).
Chatterjee, R. et al. Achieving subdiffraction imaging through bound surface
states in negative refraction photonic crystals in the near-infrared range. Phys.
Rev. Lett. 100, 187401 (2008).
Novoselov, K. S. et al. Electric field effect in atomically thin carbon films. Science
306, 666–669 (2004).
Kim, K. S. et al. Large-scale pattern growth of graphene films for stretchable
transparent electrodes. Nature 457, 706–710 (2009).
Supplementary Information is linked to the online version of the paper at
www.nature.com/nature.
Acknowledgements We thank T. F. Heinz, C. K. Hong, J. H. Lee and W. J. Kim for
discussions, and K. Cho, J. T. Han, J. W. Lee and C. S. Lee for assisting in
characterization. This work was supported by the Korea Foundation for
International Cooperation of Science and Technology (Global Research Laboratory
programme), Korea Science and Engineering Foundation grants funded by the
Korea Government (World Class University, R32-2008-000-10180-0,
R33-2008-000-10138-0; EPB Center, 2009-0063312; 2009-0062808;
2009-0060271), the Brain Korea 21 (Korea Research Foundation), the National
Science Foundation (NSF: CHE-0641523; ECCS-0747787) and the New York State
Office of Science (NYSTAR).
Author Contributions J.Y.L. and B.H.H. conducted experiments (synthesis,
characterization, optical measurements). Y.K. assisted in synthesis. R.B., B.H.H.
and C.W.W. conducted electromagnetic simulations, and W.Y.K., S.K.M. and
M.V.J. analysed the simulation results. L.J.K. assisted in the high-resolution optical
imaging analysis. I.-C.H. conducted lens transfer and lens array formation. Keun S.
Kim and J.Y.L. obtained micro-Raman spectra. P.K. supervised optical
measurements. Kwang S. Kim supervised the whole project.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. Correspondence and requests for materials should be
addressed to Kwang S. Kim (kim@postech.ac.kr) or P.K.
(pkim@phys.columbia.edu).
501
©2009 Macmillan Publishers Limited. All rights reserved