Xia Adv Mater 2000

Monodispersed Colloidal Spheres:
Old Materials with New Applications**

By Younan Xia,* Byron Gates, Yadong Yin, and Yu Lu
This article presents an overview of current research activities that center on monodispersed colloidal spheres whose diam-
eter falls anywhere in the range of 10 nm to 1 mm. It is organized into three parts: The first part briefly discusses several useful
methods that have been developed for producing monodispersed colloidal spheres with tightly controlled sizes and well-de-
fined properties (both surface and bulk). The second part surveys some techniques that have been demonstrated for organiz-
ing these colloidal spheres into two- and three-dimensionally ordered lattices. The third part highlights a number of unique
applications of these crystalline assemblies, such as their uses as photonic bandgap (PBG) crystals; as removable templates to
fabricate macroporous materials with highly ordered and three-dimensionally interconnected porous structures; as physical
masks in lithographic patterning; and as diffractive elements to fabricate new types of optical sensors. Finally, we conclude
with some personal perspectives on the directions towards which future research in this area might be directed.
1. Introduction semiconductor nanocrystallites, silica colloids, or polymer

latexes in chemistry; and proteins, viruses, bacteria, or cells
The subject of colloid science covers an extremely broad in biology. Figure 1 gives a partial list of these colloidal sys-
range of seemingly different systems. What these systems tems, together with their typical range of critical dimen-
have in common are their constituent units: small objects
that have at least one characteristic dimension in the range
of 1 nm to 1 mm.[1] These objects are usually referred to as
colloidal particles, and their range of size is more or less de-
fined by the importance of Brownian motionÐthe endless
translational diffusion of the particles resulting from the
not completely averaged-out bombardment of the molecu-
lar species in the dispersion medium. In broad terms,
colloidal particles can be considered as effective molecules,
and be treated, to some degree, according to the theories of
statistical mechanics.
Colloidal particles have long been used as the major
components of industrial products such as foods, inks,
paints, coatings, papers, cosmetics, photographic films, and
rheological fluids.[1] They are also frequently seen and stud- Fig. 1. A list of some representative colloidal systems, together with their
ied in materials science, chemistry, and biology. Notable ex- typical ranges of dimensions [2]. In this chart, the upper limit of the critical
dimension for colloids has been extended from 1 mm to 100 mm.
amples include slurries, clays, minerals, aerosols, or foams
in materials science; macromolecules (including dendri-
sions.[2] Since a number of reviews have been devoted to
mers), aggregates of surfactant molecules, Au or Ag sols,
colloidal particles with dimensions less than 10 nm,[3±5] the
scope of this article will be limited to monodispersed
± colloidal spheres whose diameters are in the range of
[*] Prof. Y. Xia, B. Gates 10 nm to 1 mm. The most studied and best established ex-
Department of Chemistry amples of such colloidal materials are inorganic silica col-
University of Washington
Seattle, WA 98195-1700 (USA)
loids,[6] and polymer latexes.[7] Some of the unique aspects
Y. Yin, Y. Lu and niche applications of these colloidal spheres have re-
Department of Materials Science and Engineering cently been reviewed by a number of authors.[8,9] Here we
University of Washington intend to present an overview of the most recent activities
Seattle, WA 98195-2120 (USA)
[**] This work has been supported in part by a New Faculty Award from
in this area, with focus being placed on the effort that has
the Dreyfus Foundation, a Career Award from the National Science been directed towards the fabrication of photonic bandgap
Foundation (DMR-9983893), a subcontract from the AFOSR MURI (PBG) crystals and three-dimensional (3D) porous materi-
Center at the University of Southern California, and start-up funds
from the University of Washington. B.G. thanks the Center for Nano- als by employing monodispersed colloidal spheres as the
technology at the UW for a fellowship. building blocks.
Adv. Mater. 2000, 12, No. 10 Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 2000 0935-9648/00/1005-0693 $ 17.50+.50/0 693
Y. Xia et al./Monodispersed Colloidal Spheres
The production of monodispersed colloidal systems has tures of colloidal particles has been an intriguing subject of
been a major goal of colloid science ever since the begin- research over the past several decades.[11]
ning of the 20th century. Many advances in this area were Significant progress has been made with regard to the
brought on by the elaboration of simple, convenient, and re- formation and utilization of colloidal arrays.[9] Many types
producible methods that were able to generate monodis- of fascinating applications have also been proposed or
persed colloidal samples in relatively large quantities.[8] For demonstrated for this new class of materials, and some of
example, the availability of colloidal particles that are uni- them may have progressed past the demonstration phase.
form in size and shape plays a very important role in eluci- For example, 2D hexagonal lattices of colloidal spheres
dating and understanding the electronic, optical, magnetic, have been successfully demonstrated as ordered arrays of
and electrokinetic properties of these materials. In many optical microlenses in image processing;[12] as physical
cases, the employment of a monodispersed sample is also masks for evaporation or reactive ion etching to fabricate
central to the self-assembly of these particles into homoge- regular arrays of micro- or nanostructures (see Sec. 7.1);
neous, crystalline arrays with large domain sizes.[9] Thanks and as patterned arrays of relief structures to cast elasto-
to many years of continuous efforts, a variety of colloids can meric stamps for use in soft lithographic techniques.[13] On
now be synthesized as truly monodispersed systems in the other hand, 3D opaline lattices of colloidal spheres
which the size and shape of the particles and the net charges have recently been exploited as removable templates to
that are chemically fixed on their surfaces are all identical generate highly ordered, macroporous materials (see
to within 1±2 %.[6±9] When used as stable suspensions in liq- Sec. 5); as diffractive elements to fabricate sensors, filters,
uids, these colloidal particles have already found a broad switches, PBG crystals, or other types of optical and elec-
range of fascinating applications in fields such as drug deliv- trooptical devices (see Sec. 6 and 7.2); and as a directly ob-
ery, biodiagnostics, and combinatorial synthesis. In addition, servable (in 3D real space) model system to study a wide
the ability to assemble these colloidal particles into crystal- variety of fundamental phenomena such as crystallization,
line arrays allows one to obtain interesting and useful func- phase transition, melting, and fracture mechanics.[14] The
tionalities not only from the constituent materials but also success of all these applications strongly depends on the
from the long-range, mesoscopic order that characterizes availability of colloidal spheres with tightly controlled sizes
periodic structures.[9] A natural opal, for instance, is beauti- and surface properties, and on the ability to self-assemble
fully iridescent in color because silica colloids (colorless by them into ordered arrays with well-defined structures and
themselves!) have been organized into a three-dimension- sufficiently large domain sizes. It has also been shown that
ally ordered array with a lattice constant that is comparable a tight control over the degree of perfection on the 3D pe-
to the wavelength of visible light (400±800 nm).[10] As a re- riodic structure is as necessary to the photonic exploitation
sult, the formation and utilization of highly ordered struc- of crystalline arrays of colloidal spheres as has been the
Younan Xia (second from the left) was born in Jiangsu,

China, in 1965. He received a B.S. degree in chemical phys-
ics from the University of Science and Technology of China
(USTC) in 1987, and then worked as a graduate student for
four years at the Fujian Institute of Research on the Struc-
ture of Matter, Academia Sinica. He came to the United
States in 1991, received an M.S. degree in polymer chemis-
try from the University of Pennsylvania (with Prof. A. G.
MacDiarmid) in 1993, and a Ph.D. degree in physical
chemistry from Harvard University (with Prof. G. M.
Whitesides) in 1996. He has been an Assistant Professor of
Chemistry at the University of Washington in Seattle since
1997. His research interests include three-dimensional
microfabrication, nanostructured materials, self-assembled
monolayers, inorganic functional materials, conducting polymers, microfluidic and microanalytical systems, microelectrome-
chanical systems (MEMS), and novel devices for optics, optoelectronics, and displays. Byron Gates (first from the left) was born
in Spokane, Washington (USA), in 1975. He obtained a B.S. degree in chemistry from the Western Washington University in
1997, and an M.S. degree in analytical chemistry from the University of Washington. Yadong Yin (second from the right) was
born in Jiangsu, China, in 1973. He obtained his B.S. and M.S. degrees in chemistry from the USTC in 1996 and 1998, respec-
tively. Yu Lu (first from the right) was born in Guizhou, China, in 1974. She obtained her B.S. and M.S. degrees in chemistry
from the USTC in 1997 and 1999, respectively. All three are currently pursuing their Ph.D. degrees under the supervision of
Prof. Y. Xia.
694 Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 2000 0935-9648/00/1005-0694 $ 17.50+.50/0 Adv. Mater. 2000, 12, No. 10
case in the microelectronic usage of a semiconductor. As

we proceed, we will address most of these issues, which, we
believe, must be addressed before colloidal arrays can
reach their potential as a new class of industrial materials.
The objectives of this review are: i) to briefly discuss cur-
rent approaches to the production of colloidal spheres with
monodispersed sizes and desired properties; ii) to address
some experimental issues related to the self-assembly of
colloidal spheres into highly ordered 2D or 3D lattices; and
iii) to highlight and assess a number of intriguing applica-
tions based on the intrinsic periodic structure of colloidal
arrays.
2. Monodispersed Colloidal Spheres
There are a number of reasons for choosing monodis-

persed colloidal spheres as the primary target for scientific
research. Theoretical treatments of heterogeneous systems
frequently utilize the spherical symmetry. Most theoretical
models (e.g., the Mie theory[15]) that deal with the proper-
ties of colloidal particles and the interactions between them
Fig. 2. Schematic illustrations of three types of representative colloidal
are usually based on the spherical shape. In principle, the spheres: A) a solid sphere; B) a core±shell sphere; and C) a hollow sphere.
intrinsic properties of monodispersed colloidal spheres can The size of the surface groups, X, has been exaggerated in (A). The polarity
be tightly controlled by changing the parameters illustrated and density of charges on a colloidal sphere are mainly determined by the
surface group. The surface of silica spheres is usually terminated in silanol
in Figure 2: i) the diameter (D); ii) the chemical composi- groups (±Si±OH). For polymer latexes, the surface groups can span over a
tion; iii) the bulk substructure; iv) the crystallinity (poly- diverse range that includes ±NH2, ±COOH, ±SO4H, ±SO3H, ±OH,
±CONH2, ±CH2NH2, ±CH2Cl, and epoxy groups. In some cases, the bulk
crystalline, single crystalline, or amorphous); and v) the sur- structure (chemical composition and substructure) and surface morphology
face functional group (thus the interfacial free energy and of colloidal spheres may also play an important role, although the effect of
surface charge density). The sphere may also represent the these parameters is usually neglected in most discussions.
simplest form that a colloidal particle can easily adopt dur-
ing the nucleation or growth process, as driven by minimi-
zation of interfacial energy.[6±8]
2.1. Chemical Synthesis of Monodispersed Colloidal

Spheres
A rich variety of chemical approaches are available for

producing colloidal spheres that are monodispersed in
size.[8] The best established and most commonly used meth-
ods seem to be controlled precipitation for inorganic (hy-
drous) oxides and emulsion polymerization for polymer la- Fig. 3. The TEM images of two representative colloidal systems that can be
readily prepared as monodispersed samples at large quantities: A) ~400 nm
texes, respectively. Using these methods, inorganic oxides
silica spheres; and B) ~200 nm polystyrene beads. Both samples were ob-
such as amorphous silica have been readily prepared as tained from Polysciences.
uniform spheres with diameter ranging from a few nm to
1 mm; polymer latexes of 20 nm to 100 mm in size have also steps: nucleation and growth of the nuclei. To achieve
been routinely produced as uniform beads. Figure 3 shows monodispersity, these two stages must be strictly separated
two transmission electron microscopy (TEM) images of and nucleation should be avoided during the period of
such colloidal samples: ~400 nm silica spheres and ~200 nm growth. In a closed system, the monomer (usually exists as
polystyrene beads. a complex or a solid precursor) must be added or released
slowly at a well-controlled rate in order to keep it from
2.1.1. Inorganic Colloidal Spheres by Controlled passing the critical supersaturation levels during the growth
Precipitation period. This criteria for obtaining monodispersed colloidal
particles was summarized by LaMer and co-workers as a
Inorganic colloids are usually prepared via precipitation general rule (the LaMer diagram) in their study of sulfur
reactions, a process that often involves two sequential colloids prepared by acidifying thiosulfate solutions.[16] Ac-
cording to this model, all the sulfur particles started at the chromophores have also been incorporated into silica col-
same time and subsequently grew at the same rate until the loids during the synthetic step to functionalize these parti-
final size was attained. cles with fluorescence or other useful optical properties.[22]
In 1968, Stöber and Fink applied this strategy to other
systems and demonstrated an extremely useful procedure 2.1.2. Polymer Latexes by Emulsion Polymerization
for preparing monodispersed silica colloids.[17] They hydro-
lyzed a dilute solution of tetraethylorthosilicate (TEOS) in Polymer colloids of different chemical compositions can
ethanol at high pH and obtained uniform spheres of amor- be produced as exceedingly uniform spheres by a process
phous silica whose sizes could be varied from 50 nm to called emulsion polymerization.[7] The major components
2 mm simply by changing the concentrations of the reac- of this process include a monomer, a dispersion medium
tants. This method was later improved by many others, and (in most cases, water), an emulsifier (surfactant), and an
now seems to be the simplest and most effective route to initiator (usually water-soluble). The monomer is dispersed
monodispersed silica spheres.[6] Matijestic and co-workers as an aqueous emulsion (~1±100 mm in diameter) with the
extensively investigated and further developed this ap- help of the emulsifier. According to the proposed mecha-
proach.[8] They have successfully applied this strategy to nism, most surfactant molecules exist as micelles (~10 nm
the production of monodispersed colloidal spheres, cubes, in diameter), and the majority of these micelles have been
rods, and ellipsoids from a broad range of materials such as swollen by the monomer. The formation of polymer latexes
metal oxides and carbonates. As a general rule, it is neces- begins with the decomposition of the water-soluble initiator
sary to precisely control the reaction conditionsÐfor exam- during which a burst of primary free radicals are generated.
ple, the temperature, the pH, the method for mixing of re- These radicals polymerize the small amount of monomer
actants, the concentration of reactants, and the that is dissolved in the aqueous phase to form the nuclei-
concentration of counterionsÐto generate a single, short oligomers in the form of tiny particles. These nuclei subse-
burst of nuclei and then let these nuclei grow uniformly. quently enter the micelles and eventually grow into larger
It is worth noting that the LaMer model may not be valid particles until all the monomer dissolved in each micelle
for the formation of some monodispersed colloidal parti- has been consumed. At the same time, the monomer en-
cles.[18] For instance, a number of recent studies suggested capsulated in emulsion droplets acts as reservoir to provide
that several types of micrometer-sized colloidal particles a supply of repeating units to the growing polymer chains
prepared by the precipitation method were, in fact, the re- through diffusion. The growth of polymer latexes will stop
sult of aggregation of much smaller subunits (or nano- at the point when all the monomer has been depleted.[7]
meter-sized primary particles) rather than continuous For a polymer latex that is 100 nm in size, there are ap-
growth by diffusion of species from the solution towards proximately 1000 macromolecular chains entangled as coils
the surfaces of nuclei.[19] In some cases, a broad range of in the sphere; each chain starts and ends with a functional
size distribution was also observed during the period of group formed by the decomposition of the radical initiator.
growth, indicating the occurrence of multiple nucleation Monodispersed polymer colloids such as poly(methyl
events.[20] The uniformity in size for the final product could methacrylate) (PMMA) and polystyrene (PS) have been
have been achieved through a self-sharpening growth pro- produced in large quantities by using this technique.[7] A
cess during which small particles grew more rapidly than similar approach has also been successfully applied to the
larger ones. synthesis of monodispersed samples of inorganic colloidal
Silica colloids represent one of the best characterized in- particles.[23] The size of polymer latex spheres can be con-
organic systems that have been manufactured as monodis- trolled at will to span a broad range from 20 nm to ~1 mm.
persed samples, as well as in large quantities. The surface of Ugelstad et al. and El-Asser et al. developed a two-step
the as-synthesized silica colloids is often terminated with si- method for preparing monodispersed polymer spheres with
lanol groups (±Si±OH), which can ionize to generate a neg- diameters larger than 1 mm.[24] In their approach, submicro-
atively charged interface at pH values higher than 7.[6] Pris- meter-sized polystyrene beads were synthesized using the
tine samples of silica colloids will undergo a series of classical emulsion polymerization method, followed by
changes when they are thermally treated at elevated tem- swelling with another monomer (or the same monomer) in
peratures: The absorbed water (~5 wt.-%) will be released the presence of an aprotic solvent miscible with water. Sub-
first at ~150 C; the silanol groups will be crosslinked via de- sequent polymerization of the encapsulated monomer
hydration in the temperature range of 400±700 C; and leads to the formation of monodispersed latex spheres with
these particles will start to fuse into aggregates when the diameters up to several hundred micrometers. When an im-
temperature is raised above the glass transition temperature miscible monomer is introduced into the polymer beads,
of amorphous silica (~800 C). The surface properties of sil- polymerization in the following stage may result in com-
ica colloids can be changed in a controllable way by using si- plex, non-spherical particles due to phase separation (see
loxane chemistry to form self-assembled monolayers Sec. 8).
(SAMs) with the silanol groups.[21] Other materials such as When potassium persulfate is used as the water-soluble
semiconductor nanocrystallites, metal sols, and organic initiator, the surface of polymer latexes prepared by emul-
sion polymerization is usually terminated in the negatively It is worth noting that free samples of silica colloids can
charged sulfate group.[7] Other acidic (e.g., ±COOH) and be obtained from Nissan Chemical Industries, a company
basic (e.g., ±NH2) groups have also been introduced into that has produced silica colloids for many years and whose
the surface layer by adding the appropriate component to monodispersed products are marketed under the trade
the monomer. The polarity of these surface groups can be name Snowtex. In addition to its monodispersed research
changed by controlling the pH value of the dispersion me- microspheres, Duke Scientific also provides samples that
dium. In some cases, both positively and negatively charged can serve as certified particle size standards. The standard
groups can also be placed on a single polymer latex parti- deviation in diameter for these colloidal spheres is usually
cle. At a particular pH, the negative and positive charges well-controlled in the range of 1±2 %. Dyno Particles AS
on such latex surface are balanced and these particles are seems to be the only company that provides polymer la-
usually referred to as amphoteric or zwitterionic colloidal texes with porous morphologies or core±shell structures.
systems.[1]
2.3. Monodispersed Colloidal Spheres with Core±Shell

2.2. Commercial Sources for Monodispersed Colloidal Structures
Spheres
The properties of colloidal spheres can be further modi-
Some monodispersed colloidal spheres can also be com- fied by coating them with thin shells of a different chemi-
mercially obtained in relatively large quantities from a cal composition, as well as in varying thickness (Fig. 2B).
number of companies, although the diversity of materials It has been demonstrated that the structure, size, and com-
might be limited. Table 1 gives a partial list (in alphabetic position of these hybrid particles could be altered in a
order) of such companies with whom we have been inter- controllable way to tailor their optical, electrical, thermal,
acting in the past. The major products of these companies mechanical, electro-optical, magnetic, and catalytic prop-
are based on silica colloids or polystyrene latexes, and are erties over a broad range.[25±27] This modification is also
usually supplied as stabilized suspensions (>1 wt.-%) in useful for tuning the interactions between colloidal
either water or organic solvents. The polarity and density spheres, and stabilizing dispersions of these spheres in a
of the surface charges can be specified when orders are given medium. Furthermore, the cores can also be re-
placed. All of these products can be directly used without moved in a subsequent step using procedures such as sol-
further purification or separation. Several companies also vent extraction or calcination at elevated temperatures to
take orders for customized synthesis. generate hollow spheres of the coating material
Table 1. Some commercial sources of monodispersed colloidal spheres [a].
[a] Information contained here is in accordance with our knowledge of these companies at this time. [b] Custom synthesis available from
these companies.
(Fig. 2C).[26,28] In many cases, these hollow spheres may

exhibit optical properties that are substantially different
from those of solid ones. When used as fillers (or pig-
ments), hollow spheres definitely offer advantages over
their solid counterparts because hollow spheres have a
much lower density. The empty interior of hollow spheres
also make them particularly useful in several other niche
applications. For example, they can serve as extremely
small containers for encapsulationÐa process that has
been extensively exploited for use in catalysis, delivery of
drugs, development of artificial cells, and protection of
biologically active agents such as proteins, enzymes, or Fig. 4. An SEM image of monodispersed TiO2 hollow spheres that were pro-
duced by templating a sol±gel precursor against a 3D crystalline array of
deoxyribonucleic acids (DNAs).[29] polystyrene beads, followed by selective etching in toluene [33]. The inset
A variety of methods have been successfully demonstrat- shows the TEM image of such a hollow sphere, with a wall thickness of
ed for coating colloidal spheres with thin shells of the de- ~50 nm.
sired material. Most of them involves the use of a con-

trolled adsorption and/or reaction (e.g., precipitation, 2.4. Interactions Between Colloidal Spheres Suspended in
grafted polymerization, or sol±gel condensation) on the Liquids
surfaces of colloidal spheres.[30] Although these methods
are straightforward to conduct, it is difficult to control the The interactions between colloidal particles and their
homogeneity and thickness of the coating, and sometimes ramifications have been investigated intensively for almost
this may lead to clumping and heterocoagulation. Two ele- two centuries due to their profound effect on the behavior
gant approaches were recently demonstrated by several of a colloidal dispersion, especially with respect to its
groups, which allowed for the generation of homogeneous, stability, crystallization, and flow.[1] For hard colloidal
dense, thin coatings of silica on various types of colloidal spheres whose surfaces are electrically neutral, the pairwise
spheres. In the first method, the surfaces of colloidal potential energy of interaction consists of two terms: the
spheres (e.g., gold or silver colloids) were grafted with an short-range, steric repulsive interaction (Fig. 5A); and the
appropriate primer that could greatly enhance the coupling
(and thus deposition) of silica monomers or oligomers to
these surfaces.[31] In the second method, electrostatic at-
tractive adsorption of polyelectrolytes and charged nano-
particles was used to build a thin shell (layer-by-layer)
around the templateÐcolloidal spheres whose surfaces had
been derivatized with certain charged functional groups.[32]
Both methods have been successfully applied to the forma-
tion of homogeneous and dense coatings of ceramic materi-
als on the surfaces of a variety of colloidal spheres. Subse-
quent removal of the core particles yielded hollow spheres
made of the ceramic material.
More recently, our group also demonstrated a procedure
for producing ordered arrays of mesoscale hollow spheres
of ceramic materials that have a well-controlled, uniform
size and homogeneous wall thickness.[33] These hollow
spheres were formed by templating an appropriate sol±gel
precursor against a 3D crystalline array of PS beads. Be-
cause the templating process was carried out under the
confinement of two glass substrates, the thickness of the
wall could be easily changed by controlling the concentra-
tion of the sol±gel precursor solution. We have demonstrat-
ed the capability and feasibility of this method by fabricat-
ing 380 and 190 nm hollow spheres of TiO2 and SnO2 with Fig. 5. A±C) The pairwise potential of interaction between two colloidal
spheres: hard spheres (A); electrically neutral spheres (B); and highly
a wall thickness that falls anywhere in the range of charged spheres or soft spheres (C). In the ideal, hard-sphere model, there
30±100 nm. Figure 4 shows the scanning electron micros- is essentially no interaction between two colloidal spheres until they ap-
copy (SEM) and TEM images of such hollow spheres that proach to a distance of Deff, at which the potential energy of repulsion is so
strong that it effectively goes to infinity. The curve in (B) is usually referred
were fabricated from a sol±gel precursor to titania.[33] The to as the Sogami potential [1]. D) An illustration of the double-layer struc-
thickness of the wall is approximately 50 nm. ture around a highly charged colloidal sphere [1].
long-range (exists over distances of ~100 nm) attractive in- liquids using a number of strategies.[38±40] Figure 6 shows
teraction, which is usually referred to as the van der Waals the schematic diagrams of three such approaches, by which
(or London dispersion, or Hamaker) force. A sum of these colloidal spheres have been organized into closely packed,
two potential energies of interaction is plotted against the 2D hexagonal arrays.
center-to-center separation in Figure 5B. For highly
charged colloidal spheres suspended in a solution contain-
ing stray electrolytes, the effective pair potential is still un-
der debate because the experiments are usually compli-
cated by many body effects.[34,35] In general, a third term
has to be added to the potential energy: that is, the much
stronger, long-range Coulomb repulsion shielded by elec-
trolytes (the so-called Yukawa potential):
!2
Ze2 eka eÿkr
Ur (1)
e 1ka r
where r is the center-to-center distance between the two
spheres, Z is the number of charges per sphere, a is the
radius of the sphere, and e is the dielectric constant of the
dispersion medium. The magnitude of this electrostatic re-
pulsive force decreases strongly with increasing concentra-
tion of stray electrolytes due to the screening effect caused
by the counterions present in the double layers.[36] This re-
lationship is also characterized by the Debye±Hückel
screening length k±1 = [4p(nsZ + ni)lB)]±1/2, a parameter
that scales inversely with the square root of the ionic
strength of the suspension and measures the distance over
which the repulsive Coulomb potential is canceled by the
screening effect of the counterions. In this equation, ns and
ni are the concentration of spheres and counterions, and lB
= e2/ekBT is the characteristic separation (or the Bjerrum Fig. 6. Schematic illustrations of three methods that have been demonstrat-
length) between ions carrying a single electronic charge, e, ed for organizing monodispersed colloidal spheres into 2D hexagonal ar-
at temperature T in a fluid of dielectric constant e. The sum rays: A) at the air±liquid interface via long-range attractive interactions
[41]; B) in a thin liquid film spread on a solid substrate via attractive capil-
of these three potentials gives the well-known Derjaguin± lary forces [45]; and C) on the surface of a solid electrode via electrophoret-
Landau±Vervey±Overbeek (DLVO) potential,[35] which is ic deposition [52].
plotted against the center-to-center separation in Fig-
ure 5C. As illustrated in Figure 5D, high levels of stray In the first method (Fig. 6A) a 2D array of colloidal
electrolytes can effectively shield the repulsive interactions spheres is formed at the air±liquid interface; and this array
between colloidal particles, and thus the van der Waals can be subsequently transferred onto the surface of a solid
attraction prevails at all separations and the particles substrate. The surfaces of these colloidal spheres have to
agglomerate. Only at very low electrolyte concentration be modified such that they will only be partially immersed
(<10±5 M) is the intensity of electrostatic repulsion suffi- into the surface of a liquid after they have been spread
cient to stabilize an ordered configuration of the particles onto the air±liquid interface through a spreading agent
at separations greater than a particle diameter. The interac- (usually an alcohol).[41] It is the strong attractive interac-
tions between charged colloidal spheres could be strongly tions (e.g., those between dipoles induced by the asym-
influenced by the presence of other highly charged parti- metric interface) among the colloidal spheres that lead to
cles or external charged surfaces. In this case, long-range the spontaneous formation of a 2D aggregate at the inter-
attractive interactions have also been observed among face. The morphology of the aggregate usually exhibits
similarly charged colloidal particles, which could be ex- fractal characteristics; but it can also be changed by varying
plained by the redistribution of the electric double layers a number of parameters such as the size, the number con-
of ions and counterions around the particles.[37] centration, the surface hydrophobicity, or the charge den-
sity on the colloids, and the electrolytic properties of the
underlying liquid.[41] In a recent demonstration, for exam-
3. Formation of 2D Arrays of Colloidal Spheres ple, Kondo et al. were able to generate 2D arrays of silica
colloids (1 mm in diameter) with relatively large domain
Monodispersed colloidal spheres can be self-assembled sizes by controlling the degree to which the silica colloids
into ordered 2D arrays on solid supports or in thin films of were immersed into the liquid surface.[42] Deckman et al.,
Lenzmann et al., and Fulda and Tieke applied the Lang- The third method is usually referred to as electrophoretic
muir±Blodgett (LB) film technique to this first method and deposition.[51±53] In this approach (Fig. 6C), a liquid disper-
they were able to obtain polycrystalline 2D arrays of poly- sion of colloidal spheres is confined between two parallel
mer latexes over areas as large as several square centi- solid electrodes such as indium tin oxide (ITO)-coated mi-
meters.[43] More recently, Burmeister et al. also demon- croscope coverslips. In the presence of a sufficiently strong
strated a similar technique that was capable of forming electric field (50±100 V/cm), the colloidal spheres that have
ordered 2D arrays of colloidal spheres on various types of been randomly deposited on the anode will move toward
substrates.[44] each other to form a stable 2D hexagonal array. The entire
The second method was largely explored by Nagayama process can be modulated by changing the amplitude of the
and his co-workers.[40,45] It uses the attractive capillary applied electric field. It has been suggested that the long-
forces among colloidal spheres to organize them into a hex- range attraction between the colloidal spheres was caused
agonal 2D array in a thin film of liquid supported on a flat by electrodynamic flows which, in turn, were induced by
substrate. In a typical experiment, a liquid dispersion of distortions in the applied electric field and the passage of
colloid spheres is spread onto the surface of a solid sub- ionic current through the solution.[52,53] This method has
strate. When the solvent evaporates slowly under a con- been applied to a number of colloidal systems, such as mi-
trolled condition, these colloidal spheres are self-assembled crometer-sized silica colloids and polymer latexes,[52] as
into a closely packed, hexagonal array (Fig. 6B). Nagaya- well as nanometer-sized gold colloids.[54] Recently, López
ma and co-workers also followed this self-assembly process and co-workers have also extended this method to the fab-
experimentally by using an optical microscope.[40a,46] They rication of 3D opaline structures from silica colloids.[55]
found that a nucleusÐan ordered region that consists of a At the current stage of development, all of these meth-
number of colloidal spheresÐwas first formed when the ods are only capable of generating colloidal arrays built up
thickness of the liquid layer approached the diameter of by small domains, and the largest single domain usually
the colloids. More colloids were driven toward this nucleus contains fewer than 10 000 colloidal spheres.[47] These
by a convective transport, and eventually organized around methods can also only form 2D hexagonal arrays in which
the nucleus due to the attractive capillary forces. A flat, the colloidal spheres are in physical contact. As a result, it
clean, and chemically homogeneous surface has to be used is very hard to independently vary the lattice constant and
in order to generate a highly ordered array with relatively the particle size. The approach based on optical forces
large domain sizes. Solid substrates such as glass slides or seems to have the potential to overcome these difficul-
silicon wafers have been used directly in this technique. La- ties.[56] In this method, colloidal spheres are organized into
zarov et al. have also explored the use of liquids such as a highly ordered 2D structure in a liquid by creating an op-
perfluorinated oil (F-oil) or mercury as the substrates in tical standing wave pattern having a regular array of inten-
forming highly ordered 2D arrays of colloidal spheres.[47] sity antinodes. The colloidal spheres are then driven to the
More recently, Colvin and co-workers demonstrated the antinode maxima by the optical forces. Depending on how
capability and feasibility of this method in forming 3D opa- many laser beams are used to create the standing wave,
line lattices with well-controlled numbers of layers along patterns as complex as a quasicrystal have also been pro-
the [111] direction.[48] duced. As demonstrated by Misawa et al. and Mio and
The evaporation of solvent can also be accelerated by Marr, the method based on optical forces was also capable
carefully spin-coating a colloidal dispersion onto a solid of generating an arbitrary 2D or 3D pattern by adding indi-
substrate.[49] In order to obtain a uniform monolayer, it is vidual colloidal spheres to an array one at a time.[57]
critical that the colloidal dispersion is able to wet com-
pletely the surface of the solid substrate and there exists
an electrostatic repulsion between the colloidal spheres 4. Fabrication of 3D Arrays of Colloidal Spheres
and the solid substrate. In many cases, the wetting could
be greatly improved by adding a surfactant to the colloidal A rich variety of methods are available for organizing
dispersion or simply by precoating the substrate with a monodispersed colloidal spheres into highly ordered 3D ar-
thin layer of surfactant. Van Duyne and co-workers also rays. Figure 7 shows the schematic diagrams of three repre-
exploited the use of this technique to form single- and sentative approaches, by which colloidal spheres have been
double-layered structures from colloidal spheres.[50] As ob- successfully assembled into 3D crystalline lattices with rela-
served by Deckman et al.,[49] the surfactant might form re- tively large domain sizes.
siduals on the substrate, and the domain size of the or-
dered region often decreases with the dimension of
colloidal spheres (ordering did not occur when colloidal 4.1. Sedimentation in a Force Field
spheres were smaller than 50 nm). Multilayers could be
obtained at relatively high solid concentrations and low Sedimentation in a gravitational field (Fig. 7A) seems to
spin speeds, while many defects appeared under these con- be the simplest approach to the formation of 3D crystalline
ditions. arrays from colloidal spheres.[1] Although it looks simple,
Fig. 7. Schematic outlines of three experimental procedures that have been

used to assemble colloidal spheres into 3D crystalline lattices: A) sedimen- Fig. 8. Phase diagrams for colloidal systems whose constituent units are
tation in the gravitational field [58]; B) ordering via repulsive electrostatic treated as A) hard, and B) soft spheres [9c]. The vertical axis of (A) rep-
interactions [69]; and C) crystallization through physical confinement and resents the reduced pressure, which has been normalized by the thermal
hydrodynamic flow [76]. energy kT and the number density (nccp) of the ccp structure. As indicated
in (A), hard spheres begin to order when they occupy approximately 50 %
of the volume. This structural transition from a topologically disordered liq-
this method actually involves a strong coupling of several uid into a long-range, highly ordered packing (the so-called Kirkwood±Al-
complex processes such as gravitational settling, transla- der transition) involves no change in heat or energy. It is solely driven by
the entropic effect. The sedimentation of non-charged colloidal spheres can
tional diffusion (or Brownian motion), and crystallization
be treated according to the hard-sphere model, if the dispersion medium is
(nucleation and growth). The success of this method relies non-polar and has a refractive index that nearly matches that of the colloidal
on tight control over several parameters such as the size spheres. As shown in (B), charged colloidal spheres can order into either
bcc or fcc lattices as the volume fraction or the screening length increases.
and density of the colloidal spheres, as well as the rate of In either case, the disorder-to-order (or Kirkwood±Alder) transition occurs
sedimentation. The colloidal spheres can always settle com- at much lower volume fractions than the hard-sphere system.
pletely to the bottom of a container as long as the size and
density of these spheres are sufficiently high. Only when opal.[59,62] The preference of a ccp structure over a hex-
the sedimentation process is slow enough, the colloidal agonal-close-packed (hcp) one has been suggested to be a
spheres concentrated at the bottom of the container will result of the difference in entropy between these two struc-
undergo a hard-sphere disorder-to-order phase transition tures.[63] Recently, van Blaaderen et al. demonstrated the
(see the phase diagram in Fig. 8A) to form a three-dimen- use of lithographically defined surfaces as templates to
sionally ordered lattice.[58] If the colloidal spheres are small grow 3D crystalline arrays with desired spatial struc-
in size (less than 0.5 mm) and/or their density is close to tures.[64] This process is the mesoscopic equivalent of epi-
that of the dispersion medium, they will exist as a dis- taxial growth: highly ordered and well-controlled arrays as
persed, equilibrium state in which the number of colloidal large as ~1 cm3 could be generated. Colvin and co-workers
spheres per unit volume varies with height according to the also developed a layer-by-layer sedimentation method for
Boltzmann distribution function. fabricating ccp arrays of silica colloids; these arrays have a
Monodispersed silica colloids are most commonly em- tightly controlled number of layers along the [111] direc-
ployed in sedimentation due to the high density of amor- tion.[48]
phous silica. Opalescent structures (usually referred to as The major disadvantage of the sedimentation method is
synthetic or artificial opals[11]) have been obtained from that it has very little control over the morphology of the
these colloidal materials under carefully controlled condi- top surface and the number of layers of the 3D crystalline
tions.[59±61] It is generally accepted that the 3D crystalline arrays.[65] Layered sedimentation may occur, which usually
arrays produced by this method have a cubic-close-packed leads to the formation of a number of layers of different
(ccp) structure (or a face-centered-cubic (fcc) lattice with a densities and orders along the direction of the gravitational
packing density of ~74 %) similar to that of a natural field. It also takes relatively long periods of time (weeks to
months) to completely settle submicrometer-sized particles. sphere, number density of spheres, and concentrations of
The 3D arrays produced by the sedimentation method are counterions in the dispersion medium. In most cases, a
often polycrystalline in nature. Kumacheva and co-workers shear flow is necessary in order to form translational order-
recently showed that sedimentation under an oscillatory ing over a long range.[74] The particles in the crystalline ar-
shear could greatly enhance the crystallinity and ordering rays formed by this method are always separated by a cer-
in the resulting 3D arrays.[66] In addition to its function as a tain distance due to the repulsive interactions among these
means for increasing the concentration of colloidal spheres, particles. Optimal fabrication of useful devices often re-
the gravitational field may also have a profound effect on quires an understanding how to control the crystalline
the crystalline structure. In a recent study, Chaikin and co- structure by changing the interactions among colloidal
workers were able to eliminate this effect by performing spheres, as well as the kinetics of the colloidal spheres.
the experiment on a space shuttle.[67] They found that the
colloidal spheres were organized into a purely random hex-
agonal close packed (rhcp) structure at volume fractions up 4.3. Self-Assembly under Physical Confinement
to 61.9 %.
Monodispersed colloidal spheres often organize them-
selves into a highly ordered 3D structure when they are
4.2. Crystallization via Repulsive Electrostatic Interactions subjected to a physical confinement.[75] We recently dem-
onstrated such a method (Fig. 7C) by which 3D opaline ar-
Under appropriate conditions, highly charged colloidal rays of colloidal spheres could be produced with domain
spheres suspended in a dispersion medium can sponta- sizes as large as several square centimeters.[76] In this meth-
neously organize themselves into a variety of crystalline od, colloidal spheres (regardless of their surface and bulk
structures as driven by minimization of electrostatic repul- properties) with a diameter ranging from 50 nm to 1 mm
sive interactions. These ordered 3D arrays are usually re- were assembled into a highly ordered structure in a spe-
ferred to as colloidal crystals (Fig. 7B).[9,11] In 1988, Rob- cially designed packing cell. A continuous sonication was
bins and Grest noted that the phase diagram of such a the key to the success of this method. Only under this con-
colloidal suspension could be described using two thermo- dition was each colloidal sphere placed at the lattice site
dynamic variables: that is, the volume fraction of colloidal represented as a thermodynamic minimum. The 3D crystal-
spheres and the electrostatic screening length (or the De- line array was found to be in the ccp structure with a pack-
bye±Hückel length, k±1) of the dispersion system.[68] The ing density very close to ~74 %, and the (111) face was par-
crystal structures that have been observed in this system in- allel to the glass substrates. Figure 9 shows the SEM
clude body-centered-cubic (bcc), fcc, rhcp, and AB2.[9,69±71] images of two typical examples. They were fabricated from
Figure 8B shows a typical phase diagram for this system. electrically neutral polystyrene beads that were ~220 and
When the screening length is shorter than the center-to- ~480 nm in diameter. This method is relatively fast, and it
center distance between two spheres, the colloidal spheres also provides a tight control over the surface morphology
act like ªhard spheresº and they will not influence each and the number of layers of the crystalline assemblies. Be-
other until they are in physical contact. In this case, a fcc cause the opaline arrays formed by this method are three-
crystal structure is formed, and no heat or energy change is dimensionally periodic structures, they can be directly used
involved upon crystallization. If the screening length is as tunable optical notch filters that are capable of selec-
longer than the center-to-center distance, the colloidal tively rejecting a very narrow wavelength interval of light
spheres behave like ªsoftº spheres and will only crystallize (as determined by the Bragg equation) in the spectral re-
into a bcc lattice, similar to that of a one-component plas- gion extending from ultraviolet to near infrared.[76b]
ma system.[71] In both crystalline structures, the colloidal
spheres are separated by a distance comparable to their
size. From the phase diagram, it is clear that the disorder- 5. Template-Directed Synthesis of 3D
to-order transition can be provoked either by increasing Macroporous Materials
the volume fraction of colloidal spheres or by extending
the range of the screening length. A more recent synchro- Template-directed synthesis is a convenient and versatile
tron diffraction study indicated that charge-stabilized method for generating porous materials. It is also a cost-ef-
colloidal spheres could form fcc structures at all volume fective and high-throughput procedure that allows the com-
fractions up to ~60 %.[72,73] plex topology present on the surface of a template to be
The method based on repulsive electrostatic interactions duplicated in a single step. In this technique, the template
seems to be the most powerful and successful route to simply serves as a scaffold around which other kinds of ma-
large-scale 3D crystalline arrays of colloidal spheres. This terials are synthesized. By templating against supermolecu-
method, however, has a very strict requirement on the ex- lar assemblies self-organized from small molecules, surfac-
perimental conditions: such as the temperature, monodis- tants, and block co-polymers, it has been possible to
persity in size, density of charges on the surface of each prepare various types of porous materials with pore sizes in
Fig. 10. Schematic outline of the experimental procedure that generates 3D

porous materials by templating against crystalline arrays of colloidal
spheres, followed by selective etching. Polystyrene beads and silica spheres
are the two most commonly used templates. The polystyrene template can
be removed either by calcination in air or by dissolution with toluene. The
silica colloids are usually removed by etching with an aqueous HF solution.
such as an ultraviolet (UV) or thermally curable organic

prepolymer,[80] an ordinary organic monomer (plus an ini-
tiator),[82] a sol±gel precursor to a ceramic material,[79±83] a
solution containing an inorganic salt,[84] or a dispersion of
nanoparticles with sizes in the range of 1±50 nm.[85] Subse-
quent solidification of the precursor and removal of the
colloidal spheres gives a 3D porous structure that has a
highly ordered architecture of uniform air balls (intercon-
nected to each other by small circular ªwindowsº). The
void spaces among colloidal spheres have also been filled
with a variety of materials through electrochemical deposi-
tion.[86] The fidelity of this procedure is mainly determined
Fig. 9. SEM images of two 3D crystalline arrays that were assembled in by van der Waals interactions, the wetting of the template
12 mm thick packing cells from: A,B) ~220 nm PS beads; and C) ~480 nm
PS beads. By using such specially designed cells, it was possible to obtained
surface, kinetic factors such as the filling of the void spaces
highly ordered arrays over several square centimeters in area, with well-con- in the template, and the volume shrinkage of precursors
trolled numbers of layers along the [111] direction from one to a few hun- during solidification. The porous materials obtained by this
dred. These SEM images suggest that the ordering extends over all three di-
mensions, and could be assigned as a ccp structure [76]. approach have also been referred to as ªinverse opalsº or
ªinverted opalsº because they have an open, periodic 3D
the range of 0.3±10 nm.[77] With the use of mesoscale ob- framework complementary to that of an opaline structure.
jects as templates, the dimension of these pores can be sig- Templating against opaline arrays of colloidal spheres
nificantly extended to cover a wide range that spans from has been successfully applied to the fabrication of 3D
~10 nm to 10 mm.[78] In particular, templating against opa- macroporous structures from a wide variety of materials,
line arrays of colloidal spheres offers a generic route to including organic polymers, ceramic materials, inorganic
macroporous materials that exhibit precisely controlled semiconductors, and metals.[79±88] Fabrication based on this
pore sizes and highly ordered 3D porous structures.[79±81] approach is remarkable for its simplicity, and for its fidelity
Figure 10 illustrates the schematic procedure for this ap- in transferring the structure from the template to the repli-
proach. After the opaline array of colloidal spheres has ca. The size of the pores and the periodicity of the porous
been dried, the void spaces (~26 % in volume) among the structures can be precisely controlled and readily tuned by
colloidal spheres are fully infiltrated with a liquid precursor changing the size of the colloidal spheres. Table 2 gives a
Table 2. Macroporous materials fabricated by templating against 3D colloidal arrays [a].
[a] Information contained here is in accordance with our knowledge of related publications at this time. [b] Ordered refers to
long-range order within the sample. Partially ordered refers to smaller domain sizes.
partial list of such 3D porous materials that have been fab-

ricated and characterized. There is no doubt that a similar
approach is extendible to other materials. The only require-
ment seems to be the availability of a precursor that can in-
filtrate into the void spaces among colloidal spheres with-
out significantly swelling or dissolving the template
(usually made of polystyrene beads or silica spheres). Some
gaseous precursors have also been employed in this process
albeit the initial product deposited on the surfaces of sam-
ples might block the flow in of gaseous precursors.[87] At
present, the smallest colloidal spheres that have been suc-
cessfully used in this method are ~35 nm in diameter;[82]
the lower limit to the particle size that can be incorporated
into this technique has not been completely established.
Recently, Stucky and co-workers demonstrated a simple
and convenient method for fabricating hierarchically por-
ous materials by combining colloidal arrays and block co-
polymers into a single system.[88]
Figure 11 shows SEM images of several typical examples
of 3D porous materials that were fabricated by our re-
search group: macroporous thin films of polyurethane,
SiO2, SnO2, and TiO2.[80] As illustrated by these images,
these membrane-type films have a periodic structure com-
plementary to that of an opalÐeach of them consists of a
highly ordered, 3D array of uniform air balls that are inter-
Fig. 11. SEM images of 3D porous materials that were made of: A±C) poly-
connected to each other by a small ªwindowº. For organic urethane; D) SiO2; E) SnO2; and F) TiO2 [80]. They were fabricated by tem-
polymers, it has been possible to fabricate such 3D porous plating a UV-curable prepolymer (A±C) and sol±gel precursors (D±F)
against crystalline arrays of polystyrene beads, followed by etching in tolu-
structures as free-standing thin films that are several square ene. The images of (A±C) were taken from the same film, and C) shows a
centimeters in area.[80c] These porous membranes can be cross-sectional view of the sample. It can be clearly seen that each spherical
subsequently incorporated into a flow system to measure pore is connected to 12 adjacent spherical pores by 12 small circular win-
dows, as indicated by an arrow. The dimensions of these windows is largely
and study the permeabilities of various kinds of gases and determined by the degree of physical contact between adjacent colloidal
liquids. For metals and ceramic materials, the porous struc- spheres, and the amount of liquid precursor infiltrated into the void spaces.
tures still remain too fragile to be released from their solid

supports. The high percentage of volume shrinkage in-
volved in the solidification of these inorganic materials (in
particular for sol±gel precursors) usually results in random
cracks in these thin films.
Although this templating procedure may lack the charac-
teristics required for high-volume production, it does pro-
vide a simple and effective route to 3D porous materials
having tightly controlled pore sizes and well-defined peri-
odic structures. These porous structures can serve as unique
model systems to study many interesting subjects such as
adsorption, condensation, transport (diffusion and flow),
and separation of molecules in macroporous materials, as
well as mechanical, thermal, and optical properties of 3D
porous materials. With appropriate surface modifications,
these 3D open structures (porosity ( 74 %) should also be
useful for fabricating prototype sensors with enhanced sen-
sitivities, or as templates to generate complex 3D structures
of various functional materials that can not be produced
using conventional lithographic techniques. Because of
their long-range ordering, these 3D periodic structures are
particularly promising as PBG crystals that may exhibit
complete bandgaps in the spectral regime that covers the
UV, through visible, to near-infrared (see Sec. 6.3).[89±91]
Other types of applications that have been proposed for
these 3D macroporous materials include their potential
usage as collectors of solar energies, as model systems for
studying quantum confinement, as electrodes for fuel cells
and other types of electrochemical processes, as supports
for catalysts, and as low-dielectric materials for capacitors. Fig. 12. A) Illustration of a 1D PBG crystal made of alternating thin layers
of two dielectric materials. B) The photonic band structure (normal inci-
dence) calculated for a typical 1D PBG crystal where e1 and e2 are equal to
13 and 12, respectively. The dashed region indicates the photonic bandgap.
6. Fabrication of PBG Crystals from Colloidal Any electromagnetic wave whose frequency falls anywhere in this gap can-
Spheres not propagate through this PBG crystal. C) The photonic band structure of
a PBG crystal can be probed experimentally by measuring its transmission
spectra using electromagnetic waves with different wavelengths. The band-
A PBG crystal (or photonic crystal) is a spatially periodic gap is usually characterized by three parameters: the position of the midgap
(lmin), the width of the gap (Dl) or the gap/midgap ratio (Dl/lmin), and the
structure fabricated from materials having different dielec- maximum attenuation of the gap (10 log(Imax/Imin) in the unit of dB) [92].
tric constants.[92] It can influence the propagation of elec-
tromagnetic waves in a similar way as a semiconductor principle, to produce light-emitting diodes (LEDs) that dis-
does for electronsÐthat is, there exists a bandgap that ex- play coherence properties, to fabricate thresholdless semi-
cludes the passage of photons of a chosen range of frequen- conductor diode lasers, and to significantly enhance the
cies (Fig. 12). The concept of this new class of material was performance of many other types of optical, electro-opti-
first proposed independently by Yablonovich[93] and cal, and quantum electronic devices.
John[94] in 1987, and since then a wide variety of applica- Research on PBG crystals has extended to cover all three
tions has been envisioned or demonstrated for this new dimensions, and the spectral region ranging from ultravio-
class of material. A photonic crystal, for example, provides let to radio frequencies. The real push in this area, how-
a convenient and powerful tool to confine, control, and ma- ever, has been the strong desire to obtain a 3D bandgap
nipulate photons in all three dimensions of space: for ex- around 1.55 mmÐthe wavelength now used in optical fiber
ample, to block the propagation of photons irrespective of communications. Through an interplay between computa-
their polarization or direction; to localize photons to a spe- tional simulation and experimentation, the goal has now
cific area at restricted frequencies; to inhibit the sponta- been successfully accomplished to produce one-, two-, and
neous emission of an excited chromophore; to modulate or three-dimensional bandgaps for microwave and millimeter-
control stimulated emission; and to serve as a lossless wave radiations.[95±97] For wavelengths shorter than near-in-
waveguide to direct the propagation of photons along a frared, true photonic gaps have only been achieved in one-
specific direction.[95,96] All of these photonic properties are and two-dimensional systems, although computational
technologically important because they can be exploited, in studies have pointed out the conditions (e.g., the crystal
structure, the shape of the building block, the minimum The plane wave expansion method (PWEM) seems to be
contrast between the high and low dielectric regions, and the most commonly used tool for 3D systems, albeit it can-
the lattice parameter) under which a 3D photonic crystal not be applied to those systems whose dielectric constants
will display a complete gap in the optical regime.[98] The exhibit large imaginary parts.[103] A viable alternative to
fundamental challenge at present is one of materials this method is the photonic analogue of the Korringa±
science: that is, the fabrication of such 3D structures in a Kohn±Rostoker (KKR) approach, which has been able to
controllable way and at a reasonable cost. At the current produce results similar to those obtained by the
stage of development, it is still difficult to apply conven- PWEM.[104] The traditional scalar-wave approximation, on
tional microlithographic techniques to the fabrication of the other hand, has been found to be inadequate in fully
3D periodic structures that meet all the proposed criteria describing many important aspects of 3D photonic crys-
when the feature size becomes comparable to the wave- tals.[105]
length of near-infrared or visible light. Nevertheless, a Theoretical studies that involve 3D systems have been
number of 3D structures amenable to layer-by-layer fabri- largely focused on the fcc structure because the Brillouin
cation or patterning have already been generated in proto- zone of this lattice has the least derivation from a sphere
type forms and are expected to display complete 3D band- and thus appears to be favored for the formation of a PBG.
gaps in the optical regime.[99] An alternative route to 3D A detailed description of the computational results was re-
periodic structures is based on self-assembly in which cently summarized by Busch and John[98b] and Haus.[106]
building blocks are spontaneously organized into stable, For a fcc lattice consisting of colloidal spheres, there only
well-defined structures by non-covalent forces.[100] The key exists a pseudo gap in the photonic band structure, no mat-
idea in self-assembly is that the final structure is close to or ter how high the dielectric contrast.[106,107] The existence of
at thermodynamic equilibrium; it thus tends to reject de- a complete gap in this simple system is inhibited by a sym-
fects. A wide variety of self-assembling strategies have been metry-induced degeneracy at the W-point. When the sym-
successfully demonstrated for fabricating three-dimension- metry of such a fcc lattice is reduced to a diamond struc-
ally ordered structures with feature size ranging from mo- ture, a true gap evolves between the second and third
lecular, through mesoscopic, to macroscopic scales. Those bands in the photonic band structure, with a maximum gap/
that have been explored for making 3D photonic crystals midgap ratio of ~15.7 % located at a volume fraction of
include phase separation of block co-polymers,[101] crystalli- ~37 % and a dielectric contrast of ~13 (or ~3.6, in terms of
zation of monodispersed colloidal spheres,[102] and tem- refractive index).[98] The symmetry-induced degeneracy at
plate-directed synthesis.[89±91] Next we will concentrate on the W-point can also be lifted by choosing non-spherical
the last two approaches; both of them involve the use of objects as the building block to form the fcc lattice.[107] Gu
monodispersed colloidal spheres as the building blocks. and co-workers have theoretically shown that a full gap
(between the second and third bands) can develop in the
photonic band structure when the building block is a di-
6.1. Search for 3D PBGs meric object consisting of two dielectric spheres intercon-
nected to each other.[108] This full bandgap has a width of
A complete (or full, true) gap is defined as one that ex- ~11.2 % for a volume fraction of ~30 % and a dielectric
tends throughout the entire Brillouin zone in the photonic contrast of ~13. Recent calculations also suggested that the
band structure.[92] An incomplete one is often referred to stop band of a fcc lattice could be significantly widened or
as a pseudo gap (or the so-called stop bandgap), because it fully opened by using colloidal spheres made of materials
only shows up in the transmission spectrum along a certain with large magnetic susceptibilities or intensity-dependent
propagation direction. A 3D complete gap can thus be con- refractive indices.[109]
sidered as a stop band that has a frequency overlap in all Cubic lattices consisting of interconnected air balls em-
three dimensions of space. For a 1D system (Fig. 12),[92] bedded in a continuous dielectric matrix (i.e., inverse
there always exists an infinite number of full gaps in the opals) represent another promising structure type that
band structure as long as there exists a dielectric contrast. might exhibit a complete 3D bandgap.[98,110] Theoretical
For a 2D or 3D system, however, the evolution of a full studies indicate that the minimum contrast between the re-
bandgap strongly depends on the crystal structure, as well fractive indices at which a complete gap (between the
as the dielectric contrast (i.e., the strength of the potential eighth and ninth bands) is formed depends on the structur-
well).[95±99] In principle, the band structure of a PBG crystal al type and varies from 1.9 for a layered structure, through
can be obtained by solving the Maxwell equations that con- ~2.1 for a diamond structure, to ~2.8±2.9 for a fcc inverse
tain a spatially periodic function for the dielectric constant. opal structure. In the fcc case, a gap/midgap ratio as large
Because the Maxwell equations enjoy scale invariance, one as ~14 % can be achieved. More recently, Busch and John
can shift a gap theoretically to any frequency range by scal- showed that an inverse opal can also display a fully tunable
ing all the sizes of a given periodic structure. A number of bandgap if the surface of the 3D porous structure is coated
techniques have already been demonstrated for calculating with an optically birefringent nematic liquid crystal.[110]
photonic band structures of various types of PBG crystals. This 3D bandgap can be opened or closed by applying an
CMYB
electric field that rotates the axis of the liquid crystal mole-
cules with respect to the backbone of the inverse opal. In-
verse opals with a hcp structure of air balls also exhibit
complete bandgaps if the dielectric contrast is sufficiently
high.[98b]
6.2. Crystalline Arrays of Colloidal Spheres as 3D Photonic

Crystals
Although crystalline arrays of colloidal spheres are not

expected to exhibit full bandgaps due to the relatively low
dielectric contrast that can be achieved for these materials,
they offer a simple and easily prepared model system to ex-
perimentally probe the photonic band diagrams of certain
types of 3D periodic structures.
Colloidal crystals[11] assembled from highly charged poly-
styrene beads or silica spheres have been known for a long
time to produce Bragg diffraction of light in the optical re-
gion.[111] Spry and Kosan and Asher and co-workers no-
ticed that the position, width, and attenuation of the Bragg
diffraction peak could be described by the dynamic scatter-
ing theory that was originally put forward by Zachariasen
CMYB
for X-ray diffraction.[112] These highly ordered systems

were recently, studied in more detail as photonic crystals
by Vos et al.,[113] Watson and co-workers,[114] and several
other groups. Vos et al. also concluded that the dynamic
scattering theory had to be modified to take into account
the excluded volume effect.[115]
Lopez and co-workers,[116] Vlasov and co-workers,[117]
Zhang and co-workers,[118] and Colvin and co-workers[119]
have extensively investigated the photonic properties of ar-
tificial opals[11] fabricated from monodispersed silica col-
loids. In some cases, the void spaces among the colloidal
spheres could be infiltrated with a variety of other materi-
als to change the dielectric contrast.[120] Colvin and co-
workers also measured the dependence of stop band at-
tenuation on the number of layers along the [111] direc-
tion.[119] Our group studied the photonic properties of opa-
line structures assembled from polystyrene beads.[76,121]
Fig. 13. A typical 3D PBG crystal fabricated from monodispersed PS beads
Figure 13A shows the photograph of such a PBG crystal [121]. A) A photograph of the cell that was used to assemble PS beads into
that contains a set of bands with different colors. Each col- a crystalline structure consisting of a set of colored bands. The red-, green-,
and blue-colored regions are all ccp lattices made of 270, 220, and 206 nm
or band corresponds to a 3D crystalline array formed from
PS beads, respectively. B) The UV-vis transmission spectra obtained from
PS beads of a specific size: 270 nm for the red color, each colored region. C) The linear relationship between the midgap position
220 nm for the green color, and 206 nm for the blue col- and the size of the PS beads.
or.[121b] All of these PS beads were essentially uncharged
on their surfaces and the resulting crystals had a ccp struc- position of the stop band can be changed to cover the
ture. Figure 13B gives the UV-vis transmission spectra that whole spectral region from UV to near-IR by choosing PS
were obtained from the red-, green-, and blue-colored re- beads with different diameters. All of these studies are con-
gions of the sample. In principle, it should also be straight- sistent with the computational results: that is, there only ex-
forward to incorporate a large number of different PS ists a pseudo bandgap for any fcc lattice self-assembled
beads into the same photonic crystal. By simply scanning from monodispersed colloidal spheres. More recently, Lo-
across the surface of this sample, one should have access to pez and co-workers and Xia and co-workers demonstrated
photonic crystals that display a wide range of different stop the capability to fine tune the stop band of the 3D crystal-
bands. Figure 13C illustrates the dependence of the posi- line array fabricated from PS beads or silica colloids by sin-
tion of stop band on the size of PS beads. Obviously, the tering the samples to elevated temperatures.[122,123] In this
case, the position (and intensity) of the stop band could be of the existence of a complete photonic bandgap is still
changed in a controllable way to cover a narrow spectral missing for these 3D porous materials. Part of the reason
region. lies in the fact that the filling of the void spaces was not
complete in most cases and the resulting materials might
not be dense enough to acquire a refractive index close to
6.3. Inverse Opals as 3D Photonic Crystals that of the bulk material. For example, Vos and co-workers
have estimated the refractive index of their anatase porous
Computational studies have suggested that a porous ma- structure to be 1.18±1.29, a value that is much lower than
terial consisting of a opaline lattice of interconnected air that of single crystalline bulk anatase (~2.6).[89]
balls (embedded in an interconnected matrix with a higher
refractive index) should give rise to a complete gap in the
3D photonic band structure.[98b] Optimum photonic effects 7. Other Applications of Crystalline Arrays of
require the volume fraction of the matrix material to fall Colloidal Spheres
anywhere in the range of 20±30 %. Although such a 3D
structure can be built up layer by layer through conven- Crystalline arrays of colloidal spheres have also attracted
tional microlithographic techniques, it has been very diffi- considerable attention for their unique applications in
cult to achieve that goal when the feature size becomes many other areas. For instance, they have been extensively
comparable to the wavelength of visible light.[124] Process- exploited as a model system with which to study the mech-
ing difficulties have also limited the formation of such 3D anisms of structural phase transitions.[126] In some cases,
structures with more than a few layers, or from materials these transitions can mimic the complex, many-body prop-
other than those currently employed in microelectronics. erties of condensed matter on the atomic scale. The long
An alternative approach is based on template-directed syn- time scale of the motion of colloids allows one to make real
thesis against opaline arrays of colloidal spheres (see time observation and analysis of various phenomena, which
Sec. 5). This method is attractive because the periodicity of are difficult or impossible to achieve for atomic and molec-
this system can be conveniently tuned and a wide variety of ular systems. Crystalline arrays of colloidal spheres also
materials with relatively high refractive indices can be easi- provide a potentially useful route to high-strength ceram-
ly incorporated into the procedure (see Table 2). ics.[127] In this approach, ceramic materials are produced as
The most promising candidates for the matrix seem to be uniform colloidal spheres and subsequently crystallized
some wide bandgap semiconductors such as diamond, II± into closely packed lattices so that all grains and pores are
VI semiconductors (e.g., CdS and CdSe), titania, and tin di- the same size. As we are limited by space, next we will only
oxide because they have a refractive index higher than 2.5 describe two more applications that are based on crystal-
and are optically transparent in the visible and near-IR re- line arrays of monodispersed colloidal spheres.
gion.[125] Other semiconductors with strong absorption in
the visible region (such as Si and Ge) can be applied to the
near-IR regime. Since the first demonstration by Velev et 7.1. 2D Crystalline Arrays of Colloidal Spheres as Masks in
al. in 1997, many advances have recently been made in this Lithography
area.[79a] For instance, Vos and co-workers have demon-
strated the fabrication with polycrystalline titania (anatase) When colloidal spheres form a two-dimensional, closely
by using a sol±gel process and also measured the reflec- packed, hexagonal array on the surface of a solid substrate,
tance spectrum of this crystal.[89] Baughman and co-work- there exists a triangular void space among the three spheres
ers have incorporated chemical vapor deposition (CVD) that are in physical contact to each other. This ordered array
into this procedure and generated inverse opals containing of void spaces can serve as a physical mask through which
different forms of carbon.[87] Norris and co-workers and other materials such as metals or dielectric materials can be
Braun and Wiltzius were able to obtain 3D periodic struc- selectively deposited onto the surface of the underlying sub-
tures from II±VI semiconductors such as CdS and CdSe, al- strate using ordinary lift-off procedures that involve vacuum
beit no optical measurement was presented in their publi- evaporation or sputtering. This method for generating pat-
cations.[85a,86b] Stein and co-workers, Velev et al., and terned arrays of microfeatures (usually as discrete dots) was
Colvin and co-workers also fabricated highly ordered 3D initially referred to as ªnatural lithographyº,[128] and more
porous materials from metals that might display interesting recently as ªnanosphere lithographyº.[50] Figure 14 shows
photonic properties.[84,85c,86a] Pine and co-workers and Sub- the schematic diagram of such an approach. The deposited
ramania et al. have fabricated inverse opals of titania by material forms a hexagonal array of dots on the solid sup-
filling the void spaces among colloidal spheres with slurries port. Each dot has the geometry of the largest inscribed
of nanometer-sized titania particles.[85b,91] They also ob- equilateral triangle. One of the most attractive features of
served stop bands (between the second and third bands) this technique is that the lateral dimension of the resulting
for these 3D porous materials made of rutile- and anatase- structures is only about one seventh of the diameter of the
phase titania. Despite these advances, a definitive signature colloidal spheres. As a result, it is relatively easy and
fabricated high-Tc Josephson effect devices in YBCO thin

films.[135] Lenzmann et al. fabricated 2D arrays of ZnS
hemispheres by lift-off, followed by thermal growth.[43b]
Burmeister and co-workers patterned the surfaces of sub-
strates such as glass, ITO, and WSe2.[136] Wiesendanger et
al. formed 2D arrays of magnetic nanoparticles using this
method and characterized these arrays by scanning probe
microscopy.[137] Van Duyne and co-workers systematically
studied this lithographic method. They extended it to mi-
cropatterning of molecular materials, and also introduced
the use of double-layered arrays of polystyrene beads in
masked evaporation of silver.[50] Douglas et al. extended
the resolution of this method to the scale of ~10 nm by
making 2D arrays of spheres from protein molecules (the
so called S-layer).[138]
ªNanosphere lithographyº is an intrinsically parallel pro-
cess. It also provides a simple, convenient, low-cost, and
materials-general method for generating 2D arrays of nano-
structures from a broad range of materials on the surfaces
of a wide variety of substrates. The patterned nanostruc-
tures fabricated using this method are useful in a number of
areasÐfor example, to serve as antireflection coatings, reg-
ular arrays of microelectrodes, and selective solar absorb-
ers; to function as active layers to enhance Raman scatter-
ing or to improve photovoltaics; to work as supports in
biosensing or heterogeneous catalysis, and to perform as ar-
rays of quantum dots or single-electron transistors. The ma-
Fig. 14. Schematic illustration of ªnanophase lithographyº in which a 2D jor problem that hinders the use of this method in practice
hexagonal array of colloidal spheres is used as the physical mask for selec-
tive deposition. After the colloidal spheres have been removed, a regular is the lack of a tight control over the density of defects
2D array of nanofeatures is left behind on the surface of the substrate [50]. (point defects and dislocations), and the size of domains.
Also, in the original setup of this approach, it is difficult to
straightforward to produce regular arrays of isolated nano- change the lateral dimensions of the features and the sepa-
structures from micrometer-sized colloidal spheres using ration between these features independently.
this method. If the colloidal spheres are robust enough to re-
sist energetic species, the triangular voids in the 2D array of
colloidal spheres can be transferred directly into the surface 7.2. 3D Crystalline Arrays of Colloidal Spheres as Optical
of a substrate by using dry processes such as reactive ion Sensors
etching (RIE) or argon ion milling.[129] If the colloidal
spheres are chemically linked to the surface of a substrate, it A 3D crystalline array of colloidal spheres can strongly
might also be possible to achieve pattern transfer by using diffract light of a specific wavelength as determined by the
solution-phase deposition or wet etching. Bragg condition:[111]
Since its first demonstration by Fisher and Zingsheim in
the early 1980s,[130] this lithographic method has been ex- mlmin = 2ndhklsiny (2)
plored by a number of groups for generating 2D arrays of
nanostructures from a variety of materials in (or supported where m is the order of the diffraction, lmin is the wave-
on) the surfaces of solid substrates: Fisher and Zingsheim length of the diffracted light (or the so-called stop band), n
demonstrated the fabrication of 2D arrays of Pt dots on is the mean refractive index of the 3D array, dhkl is the in-
glass substrates by lift-off.[130] Deckman and Moustakas terplanar spacing along the [hkl] direction, and y is the an-
fabricated 2D arrays of Si and GaAs posts by reactive ion gle between the incident light and the normal to the (hkl)
milling and 2D arrays of Ag dots by lift-off.[131] Roxlo et al. planes. This equation indicates that the position of the stop
employed this method to texture the surface of MoS2, a band is directly proportional to the interplanar spacing.
layered compound having catalytic activities.[129,132] Bun- Any variation in the spacing should result in a observable
cick et al. fabricated 2D arrays of Ag ellipsoidal particles shift in the diffraction peak. As a result, a 3D array of col-
using this method and studied their optical properties.[133] loidal spheres can serve as an optical sensor that is capable
Fang et al. and Green et al. fabricated quantum pillar struc- of displaying and measuring environmental changes
tures in the surfaces of GaAs substrates.[134] Dozier et al. (Fig. 15). Optical sensors can thus be fabricated by combin-
ample, a hydrogel responsive to Pb2+, Ba2+, and K+ was fab-

ricated by incorporating crown-ether (4-acryloyl
aminobenzo-18-crown-6) groups into the gel matrix.[140]
The crown ethers selectively trapped these cations, drawing
in counterions, increasing the Donnan osmotic pressure of
the hydrogel, and swelling the matrix. The magnitude of
this reversible swelling depended on the number of charged
groups covalently attached to the gel. An ~150 nm red shift
was observed when the concentration of Pb2+ was increased
from ~20 ppb to ~2000 ppm. In another demonstration, an
enzymatic-based glucose sensor was also demonstrated
with reversible swelling in response to the presence of
10±12 M (in the absence of O2) to 0.5 mM glucose. The avi-
dine-functionalized enzyme glucose oxidase was attached to
a biotinylated composite of crystalline colloidal spheres in a
polyacrylamide gel.[140] Conversion of glucose to gluconic
acid by the enzyme swells the gel via the formation of an-
ions, increasing the lattice spacing of the crystalline frame-
work. Sensing devices have also been fabricated by placing
a stabilized crystalline array of colloidal spheres at the end
of an optical fiber. The optical sensors fabricated using this
approach are potentially useful in a number of areas such as
clinical diagnostics and environmental monitoring.
8. Summary and Outlook
Monodispersed colloidal spheres have emerged as the ma-

terial of choice for a wide variety of niche applications that
Fig. 15. Schematic illustration of sensing protocol that involves 3D crystal-
line arrays of colloidal spheres [140]. When the array is exposed to an ana-
range from nanopatterning to fabrication of photonic de-
lyte (or change in temperature), the spacing between colloidal spheres may vices. These materials may represent the simplest class of
change accordingly. Depending on the position of the Bragg diffraction building blocks that can be readily assembled into three-di-
peak, this small change in lattice constant can be easily visualized by the
naked eye or observed using a spectrometer. mensionally ordered structuresÐcolloidal crystals and opa-
line arrays.[11] Furthermore, new types of 3D periodic struc-
ing the responsiveness of polymer gels (or other kinds of tures (e.g., inverse opals) can also be easily fabricated by
functional materials) to environmental stimuli with the dif- templating various kinds of precursors against crystalline ar-
fractive power of 3D colloidal arrays. rays of colloidal spheres. These two simple approaches pro-
A number of smart sensors have been demonstrated vide a flexible and cost-effective route to high-quality, 3D
through the work of Asher and co-workers.[139,140] In their structures with remarkably little of the investment required
work, highly charged colloidal spheres were assembled into by the more familiar clean-room methods commonly used in
3D colloidal crystals with lattice constants in the range of microfabrication. The feature size of these highly ordered
50±500 nm. Embedding the crystalline array in a polymer structures can also be easily varied in a well-controlled way
matrix helps to retain its ordered structure when exposed to to cover an extremely broad range that spans from a few
varying environments. Two diffractive devices sensitive to nanometers to a few hundred micrometers. The ability to
temperature changes have been fabricated from poly(N- generate such periodic 3D structures with varying feature
isopropylacrylamide) (PNIPAM), which reversibly dehy- sizes allows one to obtain useful functionalities not only from
drates and thus changes its conformation with a moderate the constituent materials but also from the long-range order-
change in temperature.[141] When the colloidal spheres are ing that characterizes these structures. For example, both
made of PNIPAM, a variation in temperature changes the colloidal arrays and inverse opals can be exploited as a plat-
diameter of the colloidal spheres and leads to a modifica- form to fabricate novel types of optical or electro-optical de-
tion in the intensity of the Bragg diffraction peak.[139] When vices such as photonic crystals and smart diffractive sensors.
the matrix around polymer beads is made of PNIPAM, the Monodispersed colloidal spheres and most of the applica-
reversible shrinkage of this material between 10±35 C tions derived from these materials are still in an early stage
changes the lattice spacing and subsequently results in a of technical development. There are a number of issues that
shift of the Bragg diffraction peak.[139] Chemically respon- remain to be addressed before these materials can reach
sive colloidal crystals have also been demonstrated. For ex- their potential in core industrial applications. First, the di-
versity of materials must be greatly expanded to include

every major class of functional material. Currently, only sil-
ica and a few polymers (e.g., PS and PMMA) have been
prepared as truly monodispersed colloidal spheres. These
materials, however, do not exhibit any particularly interest-
ing optical, nonlinear optical, or electro-optical functional-
ity. Second, formation of complex crystal structures other
than ccp, bcp, and rhcp lattices has to be demonstrated. As
a major limitation to the self-assembly approaches de-
scribed in this review, all of them are not capable of gener-
ating arbitrary 3D structures. Recent demonstrations based
on optical trapping may provide a potential solution to this Fig. 16. The SEM image of several peanut-shaped colloidal particles that
were synthesized by swelling monodispersed PS beads (480 nm in diameter)
problem, albeit this approach might be too slow to be use- with an MMA (methylmethacrylic acid) monomer, followed by radical poly-
ful in practice.[56,57] Another potentially useful approach is merization of the MMA into a polymer, PMMA. The PMMA phase was
based on the self-assembly of polydispersed colloidal separated from the PS sphere during the polymerization process to form a
peanut-shaped dimer consisting of two spheres: one is PS and the other is
spheres having a well-defined distribution in size.[142] Third, PMMA [146].
the density of defects in the crystalline arrays of colloidal
spheres must be well-characterized and kept below the lev-
[2] Microscopy & Histology Catalog, Polysciences, Warrington, PA 1993±
el tolerated by a certain type of photonic application. 1994.
Fourth, the compatibility of the crystallization techniques [3] Reviews on metal colloids: a) A. Henglein, Chem. Rev. 1989, 89, 1861.
b) G. Schmid, Chem. Rev. 1992, 92, 1709. c) G. Schon, U. Simon, Col-
with the processes used in the production of other types of
loid Polym. Sci. 1995, 273, 101. d) Z. L. Wang, Adv. Mater. 1998, 10,
integrated electro-optical devices (such as diode lasers) still 13. e) G. Schmid, L. F. Chi, Adv. Mater. 1998, 10, 515.
needs to be defined and improved. [4] Reviews on semiconductor nanocrystallites: a) M. L. Steigerwald,
L. E. Brus, Acc. Chem. Res. 1990, 23, 183. b) M. G. Bawendi, M. L.
Monodispersed colloidal spheres will, of course, continue Steigerwald, L. E. Brus, Annu. Rev. Phys. Chem. 1990, 41, 477. c) Y.
as the dominant subject of research in colloid science. They Wang, Acc. Chem. Res. 1991, 24, 133. d) H. Weller, Angew. Chem. Int.
are, however, not necessarily the best and/or the only op- Ed. Engl. 1993, 32, 41. e) A. P. Alivisatos, Science 1996, 271, 933.
[5] General reviews on nanophase materials: a) J. H. Fendler, Chem. Rev.
tion for all fundamental studies and applications that in- 1987, 87, 877. b) G. A. Ozin, Adv. Mater. 1992, 4, 612. c) Nanophase
volve the use of colloidal particles. They are not, for exam- Materials: Synthesis, Properties, Applications (Eds: G. C. Hadjipanayis,
R. W. Siegel), Kluwer Academic, Norwell, MA 1994. d) D. A. Toma-
ple, suitable building blocks for generating 3D PBG
lia, Adv. Mater. 1994, 6, 529.
crystals with complete bandgaps.[98] Non-spherical colloidal [6] R. K. Iler, The Chemistry of Silica, Wiley, New York 1979.
particles may offer some immediate advantages over their [7] a) Emulsion Polymerization (Ed: I. Piirma), Academic, New York
1982. b) Science and Technology of Polymer Colloids Vol. II (Eds:
spherical counterparts in applications that require periodic G. W. Poehlein, R. H. Ottewill, J. W. Goodwin), Martinus Nijhoff,
structures with lower symmetries. Colloidal cubes, ellip- Boston, MA 1983.
soids, and rods have been directly synthesized using a rich [8] General reviews: a) E. Matijevic, Acc. Chem. Res. 1981, 14, 22. b) Fine
Particles (Ed: E. Matijestic), a special issue in MRS Bull. 1989, 14(12),
variety of chemical methods, but none of these samples was 18. c) E. Matijestic, Chem. Mater. 1993, 5, 412. d) E. Matijevic, Lang-
truly monodispersed in size and/or shape.[8,143] Two indirect muir 1994, 10, 8.
[9] Reviews on colloidal arrays: a) P. Pieranski, Contemp. Phys. 1983, 24,
approaches seem to be more promising in this regard, both
25. b) W. van Megan, I. Snook, Adv. Colloid Interface Sci. 1984, 21,
of them are based on monodispersed colloidal spheres. In 119. c) A. P. Gast, W. B. Russel, Phys. Today 1998, December, 24.
the first method, monodispersed PS beads were swollen d) From Dynamics to Devices: Directed Self-Assembly of Colloidal Ma-
terials (Ed: D. G. Grier), a special issue in MRS Bull. 1998, 23(10), 21.
with an immiscible monomer with the help of an aprotic [10] J. V. Sanders, Acta Crystallogr. 1968, A24, 427.
solvent. Subsequent polymerization of the new monomer [11] Two types of crystalline arrays of colloidal spheres (or colloidal arrays)
led to the formation of peanut-shaped colloidal particles have been extensively studied: the first type includes fcc lattices
formed from highly charged colloidal spheres and their volume frac-
due to the phase separation effect.[144,145] Figure 16 shows tions of colloids are often less than 10 %; the second type is a ccp
the SEM image of such particles that were fabricated from structure (also a fcc lattice) and its volume fraction of colloids is al-
ways close to 74 %. In our discussion, the first type will be referred to
PS and PMMA.[146] In the second method, monodispersed
as colloidal crystals, and the second type as the so-called synthetic
PS or PMMA spheres were embedded in a polymer matrix opals, opaline structures, or opaline arrays.
and subsequently converted into ellipsoids during a ther- [12] S. Hayashi, Y. Kumamoto, T. Suzuki, T. Hirai, J. Colloid Interface Sci.
1991, 144, 538.
mal-stretching process.[147,148] These non-spherical particles [13] Y. Xia, J. Tien, D. Qin, G. M. Whitesides, Langmuir 1996, 12, 4033.
will certainly provide complexity and new types of func- [14] C. Murray, MRS Bull. 1998, 23(10), 33.
tionalities that can not be offered by spherical colloids. [15] G. Mie, Ann. Phys. (Leipzig) 1908, 25(4), 377.
[16] a) E. M. Zaiser, V. K. LaMer, J. Colloid Interface Sci. 1948, 3, 571.
Received: February 25, 2000 b) V. K. LaMer, R. H. Dinegar, J. Am. Chem. Soc. 1950, 72, 4847.
± c) V. K. LaMer, Ind. Eng. Chem. 1952, 44, 1270.
[1] a) D. H. Everett, Basic Principles of Colloid Science, Royal Society of [17] W. Stöber, A. Fink, J. Colloid Interface Sci. 1968, 26, 62.
Chemistry, London 1988. b) W. B. Russel, D. A. Saville, W. R. Schowal- [18] P. Calvert, Nature 1994, 367, 119.
ter, Colloidal Dispersions, Cambridge University Press, Cambridge [19] M. Ocana, Rodriguez-Clemente, C. J. Serna, Adv. Mater. 1995, 7, 212.
1989. c) R. J. Hunter, Introduction to Modern Colloid Science, Oxford [20] G. H. Bogush, M. A. Tracy, C. F. Zukoski IV, J. Non-Cryst. Solids
University Press, Oxford 1993. d) Ordering and Phase Transitions in Col- 1988, 104, 95.
loidal Systems (Eds: A. K. Arora, B. V. R. Tata), VCH, Weinheim 1996. [21] M. Ueda, H.-B. Kim, K. Ichimura, J. Mater. Chem. 1994, 4, 883.
[22] See, for example: a) N. A. M. Verhaegh, A. V. Blaaderen, Langmuir [49] H. W. Deckman, J. H. Dunsmuir, S. Garoff, J. A. McHenry, D. G. Peif-
1994, 10, 1427. b) S. Y. Chang, L. Liu, S. A. Asher, J. Am. Chem. Soc. fer, J. Vac. Sci. Technol. 1988, B6, 333.
1994, 116, 6745. c) L. M. Liz-Marzan, M. Giersig, P. Mulvaney, Lang- [50] a) J. C. Hulteen, R. P. V. Duyue, J. Vac. Sci. Technol. 1995, A13, 1553.
muir 1996, 12, 4329. b) J. C. Hulteen, D. A. Treichel, M. T. Smith, M. L. Duval, T. R. Jen-
[23] A. J. I. Ward, S. E. Friberg, MRS Bull. 1989, 14(12), 41. sen, R. P. V. Duyne, J. Phys. Chem. B 1999, 103, 3854.
[24] J. Ugestad, A. Berge, T. Ellingsen, R. Schmid, T. N. Nilsen, P. C. Mork, [51] P. Richetti, J. Prost, P. Barois, J. Phys. Chem. 1984, 45, L1137.
P. Stenstad, E. Hornes, O. Olsvik, Prog. Polym. Sci. 1982, 17, 87. [52] a) M. Trau, D. A. Saville, I. A. Aksay, Science 1996, 272, 706. b) M.
[25] A. L. Aden, J. Appl. Phys. 1951, 22, 1242. Trau, D. A. Saville, I. A. Aksay, Langmuir 1997, 13, 6375.
[26] Previous work: a) T. Sugimoto, MRS Bull. 1989, 14(12), 23. b) M. Oh- [53] a) S.-R. Yeh, M. Seul, B. I. Shraiman, Nature 1997, 386, 57. b) Y. Solo-
mori, E. Matijevic, J. Colloid Interface Sci. 1992, 150, 594. c) W. P. mentsev, M. Bohmer, J. L. Anderson, Langmuir 1997, 13, 6058.
Hsu, R. Yu, E. Matijevic, J. Colloid Interface Sci. 1993, 156, 56. [54] M. Giersig, P. Mulvaney, Langmuir 1993, 9, 3408.
[27] Recent studies: a) S. Y. Chang, L. Liu, S. A. Asher, J. Am. Chem. Soc. [55] M. Holgado, F. Garcia-Santamaria, A. Blanco, M. Ibsate, A. Cintas, H.
1994, 116, 6739. b) A. P. Philipse, M. P. B. Van Bruggen, C. Pathmama- Miguez, C. J. Serna, C. Molpeceres, J. Requena, A. Mifsud, F. Mese-
noharan, Langmuir 1994, 10, 92. c) D. Walsh, S. Mann, Nature 1995, 377, guer, C. Lopez, Langmuir 1999, 15, 4701.
320. d) R. D. Averitt, D. Sarkar, N. J. Halas, Phys. Rev. Lett. 1997, 78, [56] See, for example: a) M. M. Burns, J. M. Fournier, J. A. Golovchenko,
4217. e) M. Giersig, L. M. Liz-Marzan, T. Ung, D. Su, P. Mulvaney, Bun- Science 1990, 249, 749. b) W. Hu, H. Li, B. Chang, J. Yang, Z. Li, J. Xu,
senges. Phys. Chem. 1997, 101, 1617. f) H. Bamnolker, B. Nitzan, S. Gura, D. Zhang, Opt. Lett. 1995, 20, 964.
S. Margelo, J. Mater. Sci. Lett. 1997, 16, 1412. g) S. J. Olderburg, R. D. [57] a) H. Misawa, K. Sasaki, M. Koshioka, N. Kitamura, H. Masuhara, Appl.
Averitt, S. L. Westcott, N. J. Halas, Chem. Phys. Lett. 1998, 288, 243. Phys. Lett. 1992, 60, 310. b) C. Mio, M. D. W. Marr, Langmuir 1999, 15, 8565.
[28] Proc. Mater. Res. Soc., Vol. 372 (Eds: D. L. Wilcox, Sr., M. Berg, T. [58] a) K. E. Davis, W. B. Russel, W. J. Glantschnig, Science 1989, 245, 507.
Bernat, D. Kellerman, J. K. Cochran, Jr.), Materials Research Society, b) P. N. Pusey, W. van Megen, Nature 1986, 320, 340.
Pittsburgh, PA 1994. [59] J. V. Sanders, Nature 1964, 204, 1151.
[29] See, for example: H. Huang, E. E. Remsen, T. Kowalewski, K. L. [60] a) P. J. Darragh, A. J. Gaskin, J. V. Sanders, Aust. Gemmol. 1977, No-
Wooley, J. Am. Chem. Soc. 1999, 121, 3805. vember, 109. b) T. C. Simonton, R. Roy, S. Komarneni, E. Breval, J.
[30] See, for example: a) A. Garg, E. Matijevic, J. Colloid Interface Sci. Mater. Res. 1986, 1, 667.
1988, 126, 243. b) N. Kawahashi, E. Matijevic, J. Colloid Interface Sci. [61] a) A. P. Philipse, J. Mater. Sci. Lett. 1989, 8, 1371. b) V. N. Bogomolov,
1990, 138, 534. c) X. C. Guo, P. Dong, Langmuir 1999, 15, 5535. d) E. S. V. Gaponenko, I. N. Germanenko, A. M. Kapitonov, E. P. Petrov,
Kumacheva, O. Kalinina, L. Lilge, Adv. Mater. 1999, 11, 231. e) S. M. N. V. Gaponenko, A. V. Prokofiev, A. N. Ponyavina, N. I. Silvanovich,
Marinakos, J. P. Novak, L. C. Brousseau III, A. B. House, E. M. Ediki, S. M. Samoilovich, Phys. Rev. E 1997, 55, 7619.
J. C. Feldhaus, D. L. Feldheim, J. Am. Chem. Soc. 1999, 121, 8518. [62] H. Miguez, F. Meseguer, C. Lopez, A. Mifsud, J. S. Moya, L. Vazquez,
[31] a) M. A. Correa-Duarte, M. Giersig, L. M. Liz-Marzan, Chem. Phys. Langmuir 1997, 13, 6009.
Lett. 1998, 286, 497. b) T. Ung, L. M. Liz-Marzan, P. Mulvaney, Lang- [63] L. V. Woodcock, Nature 1997, 388, 235.
muir 1998, 14, 3740. [64] A. van Blaaderen, R. Ruel, P. Wiltzius, Nature 1997, 385, 321.
[32] a) F. Caruso, H. Lichtenfeld, M. Giersig, H. Mohwald, J. Am. Chem. [65] R. C. Salvarezza, L. Vazquez, H. Miguez, R. Mayoral, C. Lopez, F.
Soc. 1998, 120, 8523. b) F. Caruso, R. A. Caruso, H. Mohwald, Science Meseguer, Phys. Rev. Lett. 1996, 77, 4572.
1998, 282, 1111. [66] O. Vickreva, O. Kalinina, E. Kumacheva, Adv. Mater. 2000, 12, 110.
[33] Z. Zhong, Y. Yin, B. Gates, Y. Xia, Adv. Mater. 2000, 12, 206. [67] J. Zhu, M. Li, R. Rogers, W. Meyer, R. H. Ottewill, STS-73 Space
[34] See, for example: a) D. G. Gier, Nature 1998, 393, 621. b) C. A. Mur- Shuttle Crew, W. Z. B. Russel, P. M. Chaikin, Nature 1997, 387,
ray, Nature 1997, 385, 203. c) A. E. Larsen, D. G. Grier, Nature 1997, 883.
385, 230. [68] M. O. Robbins, G. S. Grest, J. Chem. Phys. 1988, 88, 3286.
[35] a) B. V. Derjaguin, L. Landau, Acta Physicochim. URSS 1941, 14, 633. [69] Early work: a) N. Ise, Angew. Chem. Int. Ed. Engl. 1986, 25, 323. b) S.
b) E. J. W. Verwey, J. T. G. Overbeek, Theory of the Stability of Lyopho- Dosho, N. Ise, K. Ito, S. Iwai, H. Kitano, H. Matsuoka, H. Nakamura,
bic Colloids, Elsevier, Amsterdam 1948. c) J. Visser, in Surface and Col- H. Okumura, T. Ono, I. S. Sogami, Y. Ueno, H. Yoshida, T. Yoshiya-
loid Science, Vol. 8 (Ed: E. Matijevic), Wiley, New York 1976, pp. 3±84. ma, Langmuir 1993, 9, 394.
[36] See, for example: a) G. M. Kepler, S. Fraden, Phys. Rev. Lett. 1994, 73, [70] Improved methodology: a) N. A. Clark, A. J. Hurd, B. J. Ackerson,
356. b) J. C. Crocker, D. G. Grier, Phys. Rev. Lett. 1994, 73, 352. Nature 1979, 281, 57. b) P. Pieranski, L. Strzelecki, B. Pansu, Phys. Rev.
c) J. C. Crocker, D. G. Grier, Phys. Rev. Lett. 1996, 77, 1897. Lett. 1983, 50, 900.
[37] W. R. Bowen, A. O. Sharif, Nature 1998, 393, 663. [71] Recent studies: a) T. Okubo, Langmuir 1994, 10, 1695; Langmuir 1994,
[38] C. A. Murray, D. H. V. Winkle, Phys. Rev. Lett. 1987, 58, 1200. 10, 3529. b) E. A. Kamenetzky, L. G. Magliocco, H. P. Panzer, Science
[39] A. T. Skjeltorp, P. Meakin, Nature 1988, 335, 424. 1994, 263, 207. c) H. B. Sunkara, J. M. Jethmalani, W. T. Ford, Chem.
[40] a) N. D. Denkov, O. D. Velev, P. A. Kralchevsky, I. B. Ivanov, H. Mater. 1994, 6, 362. d) M. Weissman, H. B. Sunkara, A. S. Tse, S. A.
Yoshimura, K. Nagayama, Nature 1993, 361, 26. b) G. Picard, Lang- Asher, Science 1996, 274, 959.
muir 1998, 14, 3710. [72] See, for example: a) T. B. Mitchell, J. J. Bollinger, D. H. E. Dubin,
[41] a) A. J. Hurd, D. W. Schaefer, Phys. Rev. Lett. 1985, 54, 1043. b) Z. Hor- X. P. Huang, W. M. Itano, R. H. Baughman, Science 1998, 282, 1290.
volgyi, M. Mate, M. Zrinyi, Colloids Surf. A: Physicochem. Eng. Asp. b) T. M. O'Neil, Phys. Today 1999, February, 24.
1994, 84, 207. c) H. H. Wickman, J. N. Korley, Nature 1998, 393, 445. [73] W. L. Vos, M. Megens, C. M. van Kats, P. Bösecke, Langmuir 1997, 13,
[42] M. Kondo, K. Shinozaki, L. Bergstrom, N. Mizutani, Langmuir 1995, 6004.
11, 394. [74] T. Okubo, Langmuir 1994, 10, 1695.
[43] a) H. W. Deckman, J. H. Dunsmuir, S. M. Gruner, J. Vac. Sci. Technol. [75] a) D. H. Van Winkle, C. A. Murray, Phys. Rev. 1986, 34, 562. b) P. Pier-
1989, B7, 1832. b) F. Lenzmann, K. Li, A. H. Kitai, H. D. H. Stover, anski, L. Strzelecki, B. Pansu, Phys. Rev. Lett. 1993, 50, 900. c) P. Lei-
Chem. Mater. 1994, 6, 156. c) K. U. Fulda, B. Tieke, Adv. Mater. 1994, derer, T. Palberg, Phys. Rev. Lett. 1997, 79, 2348.
6, 288. [76] a) S. H. Park, D. Qin, Y. X. Xia, Adv. Mater. 1998, 10, 1028. b) S. Park,
[44] F. Burmeister, C. Schafle, T. Matthes, M. Bohmisch, J. Boneberg, P. Y. Xia, Langmuir 1999, 15, 266. c) B. Gates, D. Qin, Y. Xia, Adv. Ma-
Leiderer, Langmuir 1997, 13, 2983. ter. 1999, 11, 466.
[45] See, for example: a) N. D. Denkov, O. D. Velev, P. A. Kralchevsky, [77] See, for example: a) C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C.
I. B. Ivanov, H. Yoshimura, K. Nagayama, Langmuir 1992, 8, 3183. Vartuli, J. S. Beck, Nature 1992, 359, 710. b) D. D. Archibald, S. Mann,
b) A. S. Dimitrov, K. Nagayama, Langmuir 1996, 12, 1303. c) O. D. Nature 1993, 364, 430. c) P. Behrens, G. D. Stucky, Angew. Chem. Int.
Velev, N. D. Denkov, V. N. Paunov, P. A. Kralchevsky, K. Nagayama, Ed. Engl. 1993, 32, 696. d) H. Yang, N. Coombs, G. A. Ozin, Nature
Langmuir 1993, 9, 3702. d) S. Rakers, L. F. Chi, H. Fuchs, Langmuir 1997, 386, 692. e) M. Trau, N. Yao, E. Kim, Y. Xia, G. M. Whitesides,
1997, 13, 7121. e) A. S. Dimitrov, T. Miwa, K. Nagayama, Langmuir I. A. Aksay, Nature 1997, 390, 674.
1999, 15, 5257. [78] a) S. A. Davis, S. L. Burkett, N. H. Mendelson, S. Mann, Nature 1997,
[46] A. S. Dimitrov, C. D. Dushkin, H. Yoshimura, K. Nagayama, Lang- 385, 420. b) A. Imhof, D. J. Pine, Nature 1997, 389, 948. c) M. Anto-
muir 1994, 10, 432. nietti, B. Berton, C. Goltner, H.-P. Hentze, Adv. Mater. 1998, 10, 154.
[47] G. S. Lazarov, N. D. Denkov, O. D. Velev, P. A. Kralchevsky, K. d) A. Imhof, D. J. Pine, Adv. Mater. 1998, 10, 697.
Nagayama, J. Chem. Soc. Faraday Trans. 1994, 90, 2077. [79] a) O. D. Velev, T. A. Jede, R. F. Lobo, A. M. Lenhoff, Nature 1997,
[48] P. Jiang, J. F. Bertone, K. S. Hwang, V. L. Colvin, Chem. Mater. 1999, 389, 447. b) O. D. Velev, T. A. Jede, R. F. Lobo, A. M. Lenhoff, Chem.
11, 2132. Mater. 1998, 10, 3597.
[80] a) S. H. Park, Y. Xia, Chem. Mater. 1998, 10, 1745. b) S. H. Park, Y. [114] a) I. I. Tarhan, G. H. Watson, Phys. Rev. Lett. 1996, 76, 315. b) I. I. Tar-
Xia, Adv. Mater. 1998, 10, 1045. c) B. Gates, Y. Yin, Y. Xia, Chem. Ma- han, M. P. Zinkin, G. H. Watson, Opt. Lett. 1995, 20, 1571.
ter. 1999, 11, 2827. [115] W. L. Vos, M. Megens, C. M. van Kats, P. Bosecke, J. Phys.: Condens.
[81] a) B. T. Holland, C. F. Blanford, A. Stein, Science 1998, 281, 538. b) B. T. Matter 1996, 8, 9503.
Holland, C. F. Blanford, T. Do, A. Stein, Chem. Mater. 1999, 11, 795. [116] a) R. Mayoral, J. Requena, J. S. Moya, C. Lopez, A. Cintas, H. Miguez,
[82] S. A. Johnson, P. J. Ollivier, T. E. Mallouk, Science 1999, 283, 963. F. Meseguer, L. Vazquez, M. Holgado, A. Blanco, Adv. Mater. 1997, 9,
[83] J. S. Yin, Z. L. Wang, Adv. Mater. 1999, 11, 469. 257. b) H. Miguez, C. Lopez, F. Mesequer, A. Blanco, L. Vazquez, R.
[84] H. Yan, C. F. Blanford, B. T. Holland, M. Parent, W. H. Smyrl, A. Mayoral, M. Ocana, V. Fornes, A. Mifsud, Appl. Phys. Lett. 1997, 71,
Stein, Adv. Mater. 1999, 11, 1003. 1148. c) C. Lopez, L. Vazquez, F. Meseguer, R. Mayoral, M. Ocana, H.
[85] a) Y. A. Vlasov, N. Yao, D. J. Norris, Adv. Mater. 1999, 11, 165. b) G. Miguez, Superlattices Microstruct. 1997, 22, 399.
Subramanian, V. N. Manoharan, J. D. Thorne, D. J. Pine, Adv. Mater. [117] a) V. N. Astratov, Y. A. Vlasov, O. Z. Karimov, A. A. Kaplyanskii,
1999, 11, 1261. c) O. D. Velev, P. M. Tessier, A. M. Lenhoff, E. W. Ka- Y. G. Musikhin, N. A. Bert, V. N. Bogomolov, A. V. Prokofiev, Phys.
ler, Nature 1999, 401, 548. Lett. A 1996, 222, 349. b) Y. A. Vlasov, V. N. Astratov, O. Z. Karimov,
[86] a) P. Jiang, J. Cizeron, J. F. Bertone, V. L. Colvin, J. Am. Chem. Soc. A. A. Kaplyanskii, V. N. Bogomolov, A. V. Prokofiev, Phys. Rev. B
1999, 121, 7957. b) P. V. Braun, P. Wiltzius, Nature 1999, 402, 603. 1997, 55, R 13 357.
[87] A. A. Zakhidov, R. H. Baughman, Z. Iqbal, C. Cui, I. Khayrullin, [118] D. Mei, H. Liu, B. Cheng, Z. Li, D. Zhang, Phys. Rev. B 1998, 58, 35.
S. O. Dantas, J. Marti, V. G. Ralchenko, Science 1998, 282, 897. [119] J. F. Bertone, P. Jiang, K. S. Hwang, D. M. Mittleman, V. L. Colvin,
[88] P. Yang, T. Deng, D. Zhao, P. Feng, D. Pine, B. F. Chmelka, G. M. Phys. Rev. Lett. 1999, 83, 300.
Whitesides, G. D. Stucky, Science 1998, 282, 2244. [120] Y. A. Vlasov, K. Luterova, I. Pelant, B. Honerlage, V. N. Astratov,
[89] a) J. E. G. J. Wijnhoven, W. L. Vos, Science 1998, 281, 802. b) M. S. Appl. Phys. Lett. 1997, 71, 1616.
Thijssen, R. Sprik, J. E. G. J. Wijinhoven, M. Megens, T. Narayanan, [121] a) B. Gates, S. H. Park, Y. Xia, J. Lightwave Technol. 1999, 17, 1956.
A. Lagendijk, W. L. Vos, Phys. Rev. Lett. 1999, 83, 2730. b) B. Gates, Y. Xia, unpublished.
[90] K. Yoshino, S. B. Lee, S. Tatsuhara, Y. Kawagishi, M. Ozaki, A. A. Za- [122] H. Miguez, F. Meseguer, C. Lopez, A. Blanco, J. S. Moya, J. Requena,
khidov, Appl. Phys. Lett. 1998, 73, 3506. A. Mifsud, V. Fornes, Adv. Mater. 1998, 10, 480.
[91] G. Subramania, K. Constant, R. Biswas, M. M. Sigalas, K.-M. Ho, [123] B. Bates, S. H. Park, Y. Xia, Adv. Mater. 2000, 12, 653.
Appl. Phys. Lett. 1999, 74, 3933. [124] See, for example: a) S. Fan, P. R. Villeneuve, R. D. Meade, J. D. Joan-
[92] J. D. Joannopoulos, R. D. Meade, J. N. Winn, Photonic Crystals, nopoulos, Appl. Phys. Lett. 1994, 65, 1466. b) C. C. Cheng, A. Scherer,
Princeton University Press, Princeton, NJ 1995. J. Vac. Sci. Technol. B 1995, 13, 2696. c) S. Noda, N. Yamamoto, A. Sasa-
[93] E. Yablonovitch, Phys. Rev. Lett. 1987, 58, 2059. ki, Jpn. J. Appl. Phys. 1996, 35, L909. d) J. S. Foresi, P. R. Villeneuve, J.
[94] S. John, Phys. Rev. Lett. 1987, 58, 2486. Ferrera, E. R. Thoen, G. Steinmeyer, S. Fan, J. D. Joannopoulos, L. C.
[95] Recent reviews: a) S. John, Phys. Today 1991, May, 32. b) E. Yablono- Kimerling, H. I. Smith, E. P. Ippen, Nature 1997, 390, 143. e) S. Y. Lin,
vitch, J. Opt. Soc. Am. B 1993, 10, 283. c) J. D. Joannopoulos, P. R. Vil- J. G. Fleming, D. L. Hetherington, B. K. Smith, R. Biswas, K. M. Ho,
leneuve, S. Fan, Nature 1997, 386, 143. M. M. Sigalas, W. Zubrzycki, S. R. Kurtz, J. Bur, Nature 1998, 394, 251.
[96] Recent special issues: a) Photonic Band Gap Materials (Ed: C. M. [125] L. I. Berger, Semiconductor Materials, CRC Press, Boca Raton, FL
Soukoulis), Kluwer, Boston, MA 1996. b) A. Scherer, T. Doll, E. Ya- 1997.
blonovitch, H. O. Everitt, J. A. Higgins, a special issue in J. Lightwave [126] See, for example: C. A. Murray, D. G. Grier, Am. Sci. 1995, 83, 238.
Technol. 1999, 17, 1928. [127] See, for example: P. Calvert, Nature 1985, 317, 201.
[97] E. Yablonovitch, T. J. Gmitter, Phys. Rev. Lett. 1989, 63, 1950. [128] H. W. Deckman, J. H. Dunsmuir, Appl. Phys. Lett. 1982, 41, 377.
[98] See, for example: a) K. M. Ho, C. T. Chan, C. M. Soukoulis, Phys. [129] C. B. Roxlo, H. W. Deckman, B. Abeles, Phys. Rev. Lett. 1986, 57,
Rev. Lett. 1990, 65, 3152. b) K. Busch, S. John, Phys. Rev. E 1998, 58, 2462.
3896. [130] Y. C. Fisher, H. P. Zingsheim, J. Vac. Sci. Technol. 1981, 19, 881.
[99] a) C. T. Chan, K. M. Ho, C. M. Soukoulis, Europhys. Lett. 1991, 16, [131] H. W. Deckman, T. D. Moustakas, J. Vac. Sci. Technol. B 1988, 6, 316.
563. b) K. M. Ho, C. T. Chan, C. M. Soukoulis, R. Biswas, M. Sigalas, [132] C. B. Roxlo, H. W. Deckman, J. Gland, S. D. Cameron, R. R. Chianel-
Solid State Commun. 1994, 89, 413. li, Science 1987, 235, 1629.
[100] L. Isaacs, D. N. Chin, N. Bowden, Y. Xia, G. M. Whitesides, in Super- [133] M. C. Buncick, R. J. Warmack, T. L. Ferrell, J. Opt. Soc. Am. B 1987,
molecular Materials and Technologies (Ed: D. N. Reinhoudt), Wiley, 4, 927.
New York 1999. [134] a) H. Fang, R. Zeller, P. J. Stiles, Appl. Phys. Lett. 1989, 55, 1433. b) M.
[101] a) Y. Fink, A. M. Urbas, M. G. Bawendi, J. D. Joannopoulos, E. L. Green, M. Garcia-Parajo, F. Khaleque, Appl. Phys. Lett. 1993, 62, 264.
Thomas, J. Lightwave Technol. 1999, 17, 1963. b) S. A. Jenekhe, X. L. [135] W. D. Dozier, K. P. Daly, R. Hu, C. E. Platt, M. S. Wire, IEEE Trans.
Chen, Science 1999, 283, 372. c) J. T. Chen, E. L. Thomas, C. K. Ober, Magn. 1991, 27, 3223.
G.-P. Mao, Science 1996, 273, 343. d) F. S. Bates, G. H. Fredrickson, [136] a) J. Boneberg, F. Burmeister, C. Schafle, P. Leiderer, D. Reim, A.
Annu. Rev. Mater. Sci. 1996, 26, 501. Fery, S. Herminghaus, Langmuir 1997, 13, 7080. b) F. Burmeister, C.
[102] A. D. Dinsmore, J. C. Crocker, A. G. Yodh, Curr. Opin. Colloid Inter- Schafle, B. Keilhofer, C. Bechinger, J. Boneberg, P. Leiderer, Adv. Ma-
face Sci. 1998, 3, 5. ter. 1998, 10, 495.
[103] a) K. M. Leung, Y. F. Liu, Phys. Rev. Lett. 1990, 65, 2646. b) Z. Zhang, [137] R. Wiesendanger, M. Bode, M. Kleiber, M. Lohndorf, R. Pascal, A.
S. Satpathy, Phys. Rev. Lett. 1990, 65, 2650. c) H. S. Sozuer, J. W. Haus, Wadas, D. Weiss, J. Vac. Sci. Technol. B 1997, 15, 1330.
R. Inguva, Phys. Rev. B 1992, 45, 13 962. [138] a) K. Douglas, N. A. Clark, Appl. Phys. Lett. 1990, 56, 692. b) K.
[104] A. Moroz, C. Sommers, J. Phys.: Condens. Matter 1999, 11, 997. Douglas, G. Devaud, N. A. Clark, Science 1992, 257, 642.
[105] I. I. Tarhan, G. H. Watson, Phys. Rev. B 1996, 54, 7593. [139] a) J. M. Weissman, H. B. Sunkara, A. S. Tse, S. A. Asher, Science
[106] J. W. Haus, J. Mod. Opt. 1994, 41, 195. 1996, 274, 959. b) G. Pan, R. Kesavamoorthy, S. A. Asher, Phys. Rev.
[107] R. Biswas, M. M. Sigalas, G. Subramania, K.-M. Ho, Phys. Rev. B Lett. 1997, 78, 3860. c) G. Pan, R. Kesavamoorthy, S. A. Asher, J. Am.
1998, 57, 3701. Chem. Soc. 1998, 120, 6525.
[108] a) Z.-Y. Li, J. Wang, B.-Y. Gu, Phys. Rev. B 1998, 58, 3721. b) Z.-Y. Li, [140] a) J. H. Holtz, S. A. Asher, Nature 1997, 389, 829. b) J. H. Holtz, J. W.
J. Wang, B.-Y. Gu, J. Phys. Soc. Jpn. 1998, 67, 3288. Holtz, C. H. Munro, S. A. Asher, Anal. Chem. 1998, 70, 780.
[109] M. M. Sigalas, C. M. Soukoulis, R. Biswas, K. M. Ho, Phys. Rev. B [141] See, for example: R. Dagani, Chem. Eng. News 1997, June 9, 26.
1997, 56, 959. [142] a) S. Hachisu, S. Yoshimura, Nature 1980, 283, 188. b) P. Bartlett,
[110] K. Busch, S. John, Phys. Rev. Lett. 1999, 83, 967. R. H. Ottewill, P. N. Pusey, Phys. Rev. Lett. 1992, 68, 3801.
[111] See, for example: a) P. L. Flaugh, S. E. O'Donnell, S. A. Asher, Appl. [143] T. Sugimoto, M. M. Khan, A. Muramatsu, Colloids Surf. A 1993, 70,
Spectrosc. 1984, 38, 847. b) I. M. Krieger, F. M. O'Neill, J. Am. Chem. 167.
Soc. 1968, 90, 3114. c) P. A. Hiltner, I. M. Krieger, J. Phys. Chem. [144] A. T. Skjeltorp, J. Ugelstad, T. Ellingsen, J. Colloid Interface Sci. 1986,
1969, 73, 2386. d) J. W. Goodwin, R. H. Ottewill, A. Parentich, J. Phys. 113, 577.
Chem. 1980, 84, 1580. [145] H. R. Sheu, M. S. El-Aasser, J. W. Vanderhoff, Polym. Mater. Sci. Eng.
[112] a) R. J. Spry, D. J. Kosan, Appl. Spectrosc. 1986, 40, 782. b) P. A. 1987, 57, 911.
Rundquist, P. Photinos, S. Jagannathan, S. A. Asher, J. Chem. Phys. [146] Y. Lu, Y. Yin, Y. Xia, unpublished result.
1989, 91, 4932. [147] M. Nagy, A. Keller, Polym. Commun. 1989, 30, 130.
[113] W. L. Vos, R. Sprik, A. van Blaaderen, A. Imhof, A. Lagendijk, G. H. [148] K. M. Keville, E. I. Franses, J. M. Caruthers, J. Colloid Interface Sci.
Wegdam, Phys. Rev. B 1996, 53, 16 231. 1991, 144, 103.

Xia Adv Mater 2000

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Xia Adv Mater 2000

Uploaded by

Copyright:

Available Formats

Monodispersed Colloidal Spheres:

Old Materials with New Applications**

1. Introduction semiconductor nanocrystallites, silica colloids, or polymer

Younan Xia (second from the left) was born in Jiangsu,

case in the microelectronic usage of a semiconductor. As

2. Monodispersed Colloidal Spheres

There are a number of reasons for choosing monodis-

2.1. Chemical Synthesis of Monodispersed Colloidal

A rich variety of chemical approaches are available for

2.3. Monodispersed Colloidal Spheres with Core±Shell

Table 1. Some commercial sources of monodispersed colloidal spheres [a].

(Fig. 2C).[26,28] In many cases, these hollow spheres may

sired material. Most of them involves the use of a con-

Fig. 7. Schematic outlines of three experimental procedures that have been

Fig. 10. Schematic outline of the experimental procedure that generates 3D

such as an ultraviolet (UV) or thermally curable organic

Table 2. Macroporous materials fabricated by templating against 3D colloidal arrays [a].

partial list of such 3D porous materials that have been fab-

tures still remain too fragile to be released from their solid

Y. Xia et al./Monodispersed Colloidal Spheres

6.2. Crystalline Arrays of Colloidal Spheres as 3D Photonic

Although crystalline arrays of colloidal spheres are not

for X-ray diffraction.[112] These highly ordered systems

fabricated high-Tc Josephson effect devices in YBCO thin

ample, a hydrogel responsive to Pb2+, Ba2+, and K+ was fab-

8. Summary and Outlook

Monodispersed colloidal spheres have emerged as the ma-

versity of materials must be greatly expanded to include

You might also like