NEWS & VIEWS

nature materials | VOL 7 | JANUARY 2008 | www.nature.com/naturematerials �

Gennady Shvetsis in the Department of Physics, University of Texas at Austin, 1 University Station, Austin, Texas 78712, USA.

e-mail: gena@physics.utexas.edu

S ince time immemorial people have tried to create materials whose optical properties are not available in natural

media. For example, the capability of some gold-containing colourless glasses to dramatically change their optical properties was known as early as the fourth century ad. Romans took full advantage of this ability when producing beautiful vases such as the famous Lycurgus Cup — now on display at the British Museum, London — which changes its colour depending on the illumination angle (Fig. 1). Similarly, medieval red-stained glass used in church windows also owes its remarkable colours to ‘nano-engineered’ materials containing metallic nanoparticles.

In their quest to design such optical properties, the ancients used a trial-and-error-based approach to achieve the desired colour effect — it was the nineteenth century before a systematic experimental study was made, by Michael Faraday, of the optical properties of colloidal gold1. Satisfactory quantitative understanding of the resonant scattering and absorption of light by metallic nanoparticles that is responsible for these exotic colours was only achieved after the pioneering theoretical work of Gustav Mie2. However, full control over key aspects of light propagation in a medium at optical frequencies has so far been missing. On page 31 of this issue, Na Liu and her colleagues now present a nanofabrication technique that enables the realization of fully three-dimensional artificial ‘metamaterials’ at optical frequencies close to the visible region3.

Because light is an electromagnetic wave that is composed of electric and magnetic fields, its propagation in matter is determined by both the electric and magnetic responses of the medium. For isotropic media, those responses are characterized by the dielectric permittivity,

ε, and magnetic permeability, µ. For most natural materials both ε and µ are of order unity, however, adding resonant inclusions such as metallic nanoparticles can dramatically change both parameters. Naturally, the more tightly packed the resonant particles are the stronger is the modification. Although in all the historic examples above, light propagation is modified by affecting only the dielectric permittivity, it has been suggested by Soviet physicist Victor Veselago that even more drastic effects can be expected when both ε and µ are so strongly modified as to become negative4. Artificial materials with negative ε and µ (often referred to as metamaterials because their properties represent a step beyond what is naturally available) have been predicted to have the remarkable property of negative refractive index: a light ray entering such a negative index material (NIM) bends ‘the wrong’ way from the surface normal. Even more surprisingly, a flat lens made out of NIM material can overcome the diffraction-focusing limit of conventional curved lenses resulting in a ‘perfect’ lens5.

Achieving negative permeability is a particularly challenging task because the magnetic response of most natural materials at optical frequencies is negligible. To simplify this task, the first experiments were conducted in the microwave part of the

spectrum, where precision-engineered loops shaped as so-called split-ring resonators (SRRs) were used for maximizing the magnetic response6. Further enhancement of the magnetic response is accomplished by dense packing of SRR in a crystal-like three-dimensional arrangement. Not surprisingly, as soon as the microwave NIM had been accomplished, the race towards optical realization started. However, the first metamaterials exhibiting magnetic activity in the optical part of the spectrum7,8 tended to be only one layer thick: monolayers as opposed to bulk materials that are required for developing actual devices and imaging tools.

In their successful demonstration of a fabrication technique that enables fully three-dimensional optical NIMs to be constructed, Na Liu and colleagues use a layer-by-layer assembly technique to create densely packed arrays of metallic nanoscale SRRs that exhibit, according to the researchers’ numerical calculations, both electric and magnetic activity in the near- to mid-infrared parts of the electromagnetic spectrum. The calculations indicate that spectral regions of negative ε and µ can be found, albeit not overlapping as would be desirable for making a NIM. The fabrication process involves the deposition of an array of metallic SRRs using electron-beam lithography. A polymer film is then coated

In a major departure from their humble origins as ultrathin monolayers, optical metamaterials have now advanced to three-dimensional bulk media exhibiting both electric and magnetic activity.

PhotonicS

Metamaterials add an extra dimension

Figure 1 Evolution of three-dimensional metamaterials: from gold-containing glass (left) to negative index microwave metamaterials10 (centre) to the multilayer nanostructured optical metamaterials fabricated by Liu et al.3 (right).

© 2

004

AAAS

© t

he t

ruSt

eeS

of t

he b

ritiS

h m

uSeu

m

© 2008 Nature Publishing Group

NEWS & VIEWS

�� nature materials | VOL 7 | JANUARY 2008 | www.nature.com/naturematerials

over them to make the structure planar and to separate the layers. These steps are repeated as many times as the number of desired layers.

However, many challenges need to be overcome before some of the spectacular applications of metamaterials demonstrated at microwave frequencies, such as perfect lensing and optical cloaking, are realized. Furthermore, it will be a challenge to extend the optical resonance of these structures from the present infrared region into near-infrared and visible parts of the spectrum. Simply scaling down the size of SRRs and therefore reducing the resonance wavelength does not seem to be viable from both fundamental and fabrication standpoints

and, therefore, new geometries may be needed. Another challenge is to combine the electric and magnetic responses in such a way that negative ε and µ are obtained at the same optical frequency range, resulting in a true negative-index material.

Although some two-layer structures in the near-infrared indeed have such NIM properties7, they tend not to show strong metamaterials properties and are difficult to fabricate in more than two layers. The fabrication technique developed by Na Liu and colleagues offers the freedom to fabricate multilayer metamaterials that, for example, could be used to construct some of the recently designed subwavelength NIM structures9. Regardless of what shape

the future designs of nanoengineered metamaterials are going to take, this powerful fabrication technique is likely to influence the way three-dimensional metamaterials are going to be produced.

References1. Faraday, M. Philos. Trans. Royal Soc. Lond. 147, 145–181 (1857).2. Mie, G. Ann. Phys. 330, 377–445 (1908).3. Liu, N. et al. Nature Mater. 7, 31–37 (2008).4. Veselago, V. G. Sov. Phys. USPEKHI 10, 509–514 (1968).5. Pendry, J. B. Phys. Rev. Lett. 85, 3966–3969 (2000).6. Smith, D. R., Padilla, W. J., Vier, D. C., Nemat-Nasser, S. C. &

Schultz, S. Phys. Rev. Lett. 84, 4184–4187 (2000).7. Zhang, S. et al. J. Phys. Rev. Lett. 95, 137404 (2005).8. Shalaev, V. M. et. al. Opt. Lett. 30, 3356–3358 (2005).9. Lomakin, V., Fainman, Y., Urzhumov, Y. & Shvets, G. Opt. Expr.

14, 11164–11177 (2006).10. Smith, D. R., Pendry, J. B. & Wiltshire, M. C. K. Science

305, 788–792 (2004).

SURFAcE chEMiStRY

Mussel power

J. herbert Waiteis in the Departments of Molecular Cell & Developmental Biology and Chemistry & Biochemistry, University of California, Santa Barbara, California 93106, USA.

e-mail: waite@lifesci.ucsb.edu

What would you use to join two different materials to form a rigid structure — nails, rivets, screws

or glue? In general, the experts advocate adhesion because the interface in such bonded materials contains a myriad of molecular contacts and, as a result, there is better load transfer and lower stress concentration. A great frustration in science and technology, however, is that adhesives do not always work as desired. First, they may not find anything of consequence to bind to on a given surface; second, they may be prevented from binding because of moisture or other contaminants, and third, they may not wear or weather well. The latter two are of particular concern for adhesives and coatings in biomedical applications. One need only think of loose hip implants or dental fillings. Despite the development of numerous pre-adhesive treatments in an attempt to put a more reactive face on uncooperative surfaces, no general strategy has been shown to improve adhesion on all surfaces. But now,

this sticky problem could be solved by a newly developed surface modification method published recently in Science by Phillip Messersmith and colleagues1. The method draws inspiration from the adhesive proteins found in mussels (Fig. 1a) and the outcome is similar — the adhesive polymer film sticks to almost anything, coating many

different types of surfaces. In addition to the modification of a wide variety of inorganic and organic materials, the resultant surface can react further, and hence the scope for applications widens dramatically.

The adhesive film is formed from dopamine, a catecholic (1,2-dihydroxybenzene) compound with a

The adhesive proteins secreted by mussels are the inspiration behind a versatile approach to the surface modification of a wide range of inorganic and organic materials, resulting in the fabrication of multifunctional coatings for a variety of applications.

NH2

CH2

CH2

OH

OH

Dopamine

Figure 1 Dopamine is a small-molecule mimic of the adhesive proteins found in the ‘footprints’ of mussels. a, in their natural environment, mussels show strong adhesion to marine surfaces. b, Dopamine has a chemical structure comprising both alkylamine (yellow, lysine-like) and catechol (blue, DoPA-like) functionalities.

© 2008 Nature Publishing Group