DTD 5

Nanotechnology: applications and potentials

for heterogeneous catalysis
Harold H. Kung
*
, Mayfair C. Kung
Chemical and Biological Engineering, Northwestern University, Evanston, IL 60208-3120, USA
Abstract
Nanotechnology offers the potential to design, synthesize, and control at nanometer length scale. While the catalysis community has many
techniques at their disposal to synthesize catalytic materials at such a scale, the ability to fully design and control is still lacking. Examples are
presented to illustrate what can nanotechnology do for heterogeneous catalysis to help achieve the goal of designing catalysts for perfect
selectivity in a chemical reaction. Some current state-of-the-art approaches and potential limitations are discussed. Some examples of what
can catalysis do for nanotechnology are also presented. However, this aspect is much less studied, although it offers rich opportunities for the
catalysis community.
# 2004 Elsevier B.V. All rights reserved.
Keywords: Nanotechnology; Catalysis; Catalyst; Design; Selectivity
1. Introduction
Nanotechnology has gained substantial popularity
recently due to the rapidly developing techniques both to
synthesize and characterize materials and devices at the
nano-scale, as well as the promises that such technology
offers to substantially expand the achievable limits in many
different elds including medicine, electronics, chemistry,
and engineering. In the literature, there are constantly
reports of new discoveries of unusual phenomena due to the
small scale and new applications. Nano-size noble metal
particles have occupied a central place in heterogeneous
catalysis for many years, long before recognition of nano-
technology. Thus, it is tting to critically evaluate the impact
of such development on heterogeneous catalysis [1,2].
In the discussion here, nanotechnology refers to
techniques that offer the ability to deign, synthesize (or
manufacture), and control at the length scale ranging from
<1 to >100 nm. The emphasis in this denition of scope is
design and control, and not only synthesis. Synthesis of
materials at nanometer scale has already become routine
practice for supported noble metal catalysts after decades of
research on the subject. However, there is much room for
development to design and control. Thus, our discussion
starts with dening what we would like to achieve in
heterogeneous catalysis. This is followed by selective
examples of how nanotechnology has helped researcher
advance toward these goals, that is, examples of what can
nanotechnology do for (heterogeneous) catalysis.
There is a complementary aspect in the relationship
between catalysis and nanotechnology, which is: what can
catalysis do for nanotechnology? Can catalysis help in the
development of nanotechnology and in overcoming critical
barriers? This is a much less explored aspect but it deserves
attention.
2. Nanotechnology for catalysis
A major goal in catalysis research is to design catalysts
that can achieve perfect selectivity and desirable activity.
Between activity and selectivity, it is commonly accepted
that the latter is much more difcult to achieve and control.
Thus, it is the focus of our discussion. A reaction of perfect
selectivity would generate no waste products, thereby reduce
energy and process requirements for separation and
purication.
www.elsevier.com/locate/cattod
Catalysis Today xxx (2004) xxxxxx
* Corresponding author. Tel.: +1 847 491 7492; fax: +1 847 467 1018.
E-mail address: hkung@northwestern.edu (H.H. Kung).
0920-5861/$ see front matter # 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2004.07.055
In nature, biological enzymes operate with very high
selectivity. An enzyme is made of protein. The specic order
of amino acids in the protein generates functionalities
critical for the operation of the enzyme. Like other catalysts,
an enzyme has an active site where chemical transformation
takes place, such as bond breaking and forming. What
distinguishes an enzyme from many other catalysts is the
environment of the active site. Often the active site is in a
cavity formed by the protein. Along the wall of the cavity are
functional groups that help bind the reactant and product.
One interesting and perhaps critical characteristic is the fact
that the protein is exible [3], such that the cavity, the active
site, and the binding sites move during the course of the
reaction to accommodate the adsorption and release of the
reacting species.
Therefore, in the synthesis of a designed heterogeneous
catalyst, it would be desirable to have complete control over
the formation of the active site, the environment around the
active site, the binding sites and their locations relative to the
active site, and the path to access these functionalities. In
other words, we would like to have complete control from
the atomic scale to tens or hundreds of nanometer scale.
2.1. Active sites for metals
It is now accepted that for metallic catalysts, control at
the atomic level is needed to design active sites because the
chemical and catalytic properties of atoms at terraces,
corners, and edges of a metal crystallite are different, and
can be different from atoms at the metal-support interface
[4]. In addition, the location of the metal cluster on the
support may be important also if its properties are affected
by its proximity to the support defects [5].
A much studied method attempting to generate metal
clusters with atomic control is to use well-dened
organometallic complexes as precursors. In recent years,
uniform clusters of identiable geometry up to ve or six
metal atoms can be synthesized [6]. Extension of this
method to larger clusters, however, is limited by the
availability of precursor complexes.
A different approach taken by Somorjais group is to
attempt control of particle size using lithography [7]. Metal
clusters are deposited onto a masked, patterned surface. This
technique, while offering control of location and separation
between metal clusters, does not yet have control at the
atomic level. Owing to the current limit in size resolution of
the technique, even with the size-reduction lithography, it
will be a challenge to produce atomically uniform (identical
number of atoms per cluster and its shape) clusters at the
nanometer size of interest in catalysis without major
breakthroughs in the technology.
Deposition of small clusters using a STM tip has been
demonstrated [8] that could also position metal clusters at
desired locations on a at substrate. The technique can be
scaled up using the millipede approach to deposit a large
number of clusters simultaneously [9]. Potentially, it can be
coupled with dip-pen technology for a continuous supply of
the metal precursor [10]. At present, however, these
techniques have not yet achieved atomic control. In other
words, there is no technique yet to generate supported metal
clusters of uniform composition, shape, and size (see note
added in proof).
When considering controlling metal clusters at the
atomic level, it is important to keep in mind that the
clusters may be dynamic, that is, the location of the atoms
may respond to the environment. For example, there is direct
observation using high resolution electron microscopy that
the surface layer of a Pd crystallite restructures [11] and the
shape of the crystallite changes [12] upon exposure to
oxygen. A thorough understanding of the dynamic behavior
of the atoms in a metal crystallite during catalysis would be
greatly benecial to establishing the limit and desirability of
atomic control in these systems.
2.2. Active sites for oxides
Among oxides, zeolite is the most studied and best
understood. For zeolite acids, the active sites are associated
with the trivalent substituents in the zeolite framework.
Other functionality, such as framework titanium for
epoxidation and framework Co for oxidation, are also
established. Extraframework active sites, including those
introduced by ion-exchange and those synthesized by the
ship-in-a-bottle techniques can be readily characterized.
For other oxides, however, the nature of the active site is
less well dened. It is common that coordination unsatura-
H.H. Kung, M.C. Kung / Catalysis Today xxx (2004) xxxxxx 2
DTD 5
Scheme 1. Illustrating the soft-chemical technique to synthesize alumina. Starting from the left, an aminoalkoxyaluminum monomer is formed, which is
hydrolyzed in a controlled manner to form an alumina the surface of which is covered with Al ions bound to amine adducts. These sites are Lewis acid sites,
active for catalysis without post-synthesis calcination.
tion of a cation in the surface is necessary for catalytic
activity [13]. The general method to generate coordination
unsaturation is by high temperature dehydroxylation of an
oxide, which generates a wide range of surface sites.
Recently, we explored a different method of preparation, in
which we protect the coordination unsaturation site of the
precursor during preparation of the oxide, so as to preserve it
until the catalyst is formed. We demonstrated this with
alumina [14]. Thus, we rst converted the aluminum
alkoxide precursor into aminoalkoxyaluminum monomers
(Scheme 1). The amine ligand serves to protect the
coordination unsaturation of the aluminum ion. Controlled
hydrolysis of this precursor eventually led to the formation
of an alumina surface that contains a much higher
concentration of surface coordination unsaturation sites
(Lewis acid sites) relative to surface hydroxyls than a
conventionally prepared alumina. The new preparation is
also much more active for Lewis acid catalyzed aminolysis
of epoxide [15]. While this method offers some control of
the active site at the atomic level, there is much to be
developed before we can achieve the goal of controlling the
active sites in general.
2.3. Environment around the active sites
Dening and controlling the active site is insufcient to
have complete control of a catalytic reaction. As
mentioned earlier, using an enzyme as example, the
environment plays an important role as well. A very
interesting example of how the environment affects
selectivity is in the selective terminal oxidation of hexane
by Co-AlPO-18 [16]. The high selectivity for 1,6-diacid
was interpreted as due to the interaction of hexane with the
wall of the cavity, such that the molecule is aligned in such
a manner that the terminal carbons are positioned near the
framework Co ions to favor their activation. If one desires
to exploit such interaction to align the reactant molecule
with respect to the active site in other systems, it would be
necessary to be able to position the active sites in a cavity
by design. Unfortunately, there is yet no known method to
control the spatial distribution of cations in a molecular
sieve framework.
One approach to attempt to control the relative positions
of sites (active sites and binding sites) is to synthesize the
area of interest by an unit-by-unit approach. That is, starting
with an active site, spacer atoms or units are added one by
one until a desired separation is obtained before the second
functional site is added to the ensemble. We have started
research along this direction. Fig. 1 shows a siloxane chain
synthesized by such unit-by-unit approach. Starting from a
single unit siloxane, a disiloxane is synthesized by adding
another single unit to it. By repeating the procedure,
trisiloxane, tetrasiloxane, and pentasiloxane were synthe-
sized. [17].
Other approaches have been explored. One is molecular
templating technique. Starting with a templating molecule
that contains functional groups that can be linked to
siloxane, a silica is formed by reacting the siloxane with
other silica precursors. Consequently the template is
enclosed in a cavity of silica. Cleaving the template from
the silica gel with generation of catalytically active
functional groups then generates active sites anchored to
the silica wall at positions directed by the template [18]. This
is schematically shown in Fig. 2 At present, such method
offers some, but not precise control over positioning of the
active sites.
H.H. Kung, M.C. Kung / Catalysis Today xxx (2004) xxxxxx 3
DTD 5
Fig. 1. Schematic drawing showing the unit-by-unit synthesis of a siloxane chain. Starting with a single siloxane, additional units are added stepwise to form
disiloxane, trisiloxane, tetrasiloxane, and pentasiloxane. The
29
Si NMR spectra show the presence of different Si atoms in each of the chains [17].
Molecular templating is also used to shape the cavity
around an active site. In one example [19], a template that
resembles the reaction transition state bound to the active
site was used. After forming the cavity around this template,
the portion nonessential for the active site was removed. The
resulting cavity can accommodate reactants that t the shape
better than molecules that do not, thus favoring reaction of
these reactants. That is, the technique offers some control of
selectivity. This technique was used to generate a more
selective catalyst for hydrogenation of aryl ketones.
Templating using ordered surfactant molecules is a
heavily studied technique rst applied to prepare MCM41
type of materials [20]. Using this technique, regular
channels of diameter ranging from about 2 nm up to
10 nm can be prepared. The original application was to
prepare siliceous materials, but the technique has been
successfully extended to many other oxides [21]. Using the
same principle and with spherical styrofoam beads, siliceous
materials containing regular spherical cavities can be
prepared (Fig. 3) [22]. In fact, cavities of other shapes
can be prepared using the appropriate templates. Such
techniques, while versatile in preparing cavities and
channels, are limited by the templates available. Further-
more, it will take further development to design cavities
containing functional groups at specic positions.
2.4. Reaction engineering
An area of great potential for newdiscovery is application
of nanotechnology to reaction engineering. Here, the
discussion emphasizes on control at the macromolecular
level to inuence chemical reaction, more specically
controlling when an active site is active and selectivity.
Researchers have successfully modied enzymes to
effect on-off switching of catalytic activity. For example, by
incorporating a polymer whose conformation responds to
small changes in temperature into an enzyme near the active
site, it was possible to switch on and off the enzyme
reversibly by a small temperature swing [23]. Extending this
concept to catalysts in general, it should be possible to
design catalyst systems where there is a movable segment,
controlled by a hinge, that brings a blocking group to or
retrieve it from an active site, thereby turning the catalyst off
and on (Fig. 4). One could even think of constructing a
molecular-sized box big enough just to contain a stoichio-
metric number of the reactant molecules (Fig. 5). Then, only
H.H. Kung, M.C. Kung / Catalysis Today xxx (2004) xxxxxx 4
DTD 5
Fig. 2. Schematic drawing to showcontrol of positioning of active sites on a
cavity wall by molecular templating technique. A matrix is formed around a
template molecule T that possesses three functional groups F as linking
groups to the cavity wall. After removal of T and, at the same time,
converting the linkage into catalytic active sites F, a cavity containing
active sites F at locations directed by T is formed.
Fig. 3. SEM picture of a siliceous material containing uniform cavities of
about 400 nm diameter, formed using regularly packed, 400 nm diameter
styrofoam beads as template.
Fig. 4. Schematic drawing illustrating the concept of on-off switchable active sites. The circle in a square planar ligand represents the active site where reactant
R reacts to form product P. When the active site is blocked by a blocking group (shaded sphere), the active site is at rest.
the desired product can be formed, since a wrong mixture in
the box would not react. The necessary ingredients needed to
bring this concept to reality are already known. For example,
wall materials of different properties, from carbon as in
carbon nanotubes to organic-inorganic hybrids [24], to
inorganic oxide (e.g. MCM41) are known. Molecular
hinges, the most important component, are also known.
For example, the orientation of ligands around a metal ion
can be reversibly changed through change in the oxidation
state of the ion [25]. Other switching mechanisms employ-
ing light excitation, pH changes, and others are possible
[26]. With ingenuity, these could be adapted for reaction
engineering applications.
2.5. Catalysis for nanotechnology
Thus far, applications of catalysis to nanotechnology are
limited. Two examples will be cited here for illustration.
Catalysts are used in forming carbon nanotubes. Metal
catalyzed decomposition of carbon monoxide or hydrocarbon
to formcarbon nanotubes has been observed over twodecades
ago by T. Baker, but it was until recently that their unique
properties were recognized. For practical applications, it is
important to be able to control the dispersion in properties of
the nanotubes such as wall thickness (single wall or
multiwall), tube diameter, helicity, and length. Research
has shown that catalytic synthesis offers much better potential
to control the formation of the tubes than less-well controlled
thermal decomposition. Various catalysts have been inves-
tigated, and some have been found to be effective in
generatingmultiwall tubes, whereas others effective for single
wall tubes [27]. Recently, synthesis of tubes of a rather narrow
dispersion of tube diameter and helicity was reported,
catalyzed by a Co-Mo catalyst [28].
The second example involves using catalytic reaction to
supply energy for motion of small particles. Mallouk et al. at
Penn State University coated one end of small particles with
Pt, and the remaining part with Au. Upon placing the
particles in a hydrogen peroxide solution, Pt-catalyzed
decomposition of hydrogen peroxide occurs and the action
propels the particles into a forward motion [29]. The exact
mechanism that causes the motion is still under investiga-
tion. One possibility is the change of surface tension at the
reaction end.
These examples demonstrate that catalysis can play an
important role in the development of nanomaterials and
nanomachines. However, at present, there are very few
examples reported although there are many opportunities
available.
3. Conclusion
There is little doubt that the catalysis community is
following developments in nanotechnology closely and
attempts to capitalize on these developments to further the
goal of custom-designing catalysts to achieve perfect
selectivity in a catalytic reaction. Much progress has been
made, but many challenges remain. These challenges
include the need to control the relative locations of the
active sites and binding sites, controlling the atomic details
of the active sites, and incorporate exible openings and
cavities. In reaction engineering, nanotechnology offers the
opportunity to control reactor operation at the molecular
level. One example is incorporation of molecular switches
for active sites. It is expected that as the achievable limits are
expanded, new applications will arise. In all these
developments, in addition to inventing methods to achieve
them in a laboratory, a signicant hurdle is the need to
develop reliable methods for mass production before
signicant commercialization is possible.
On the other hand, there is much less attention paid to use
catalysis to enable nanotechnology. While catalytic synth-
esis of carbon nanotubes is being pursued, there is generally
a lack of examples in which catalysis plays a pivotal,
enabling role. It is likely that such opportunities are present
and waiting to be discovered. It should be highly benecial
to pursue them.
Note added in Proof
Preparation of metal nanoparticles of a narrower size
distribution than conventional methods has been reported
using the dendrimer-assisted method [30], which is
consequent to starting with uniform-size dendrimer mole-
cules and close to stoichiometric binding of metal precursors
to them. Bimetallic particles can be made also [31], although
a uniform composition has yet to be achieved.
Acknowledgement
Support by the Basic Energy Sciences, Ofce of Science,
Department of Energy is greatly appreciated.
H.H. Kung, M.C. Kung / Catalysis Today xxx (2004) xxxxxx 5
DTD 5
Fig. 5. Nanoreactor of the size of the mean-free path of the reactants,
containing molecular hinge to control the reaction by closing the reactor.
The two circles represent two reactant molecules, which can react to form
the product represented by the oval. Since the box can only contain two
molecules, only the desired reaction can proceed in the nanoreactor.
References
[1] G.A. Somorjai, Y.G. Borodko, Catal. Lett. 76 (2001) 1.
[2] H.H. Kung, M.C. Kung, Appl. Catal. A: Gen. 246 (2003) 193.
[3] E.Z. Eisenmeser, D.A. Bosco, M. Akke, D. Kern, Science 295 (2002)
1520.
[4] R.I. Masel, Principles of Adsorption and Reaction on Solid Surfaces,
Wiley, New York, NY, 1996;
G.A. Somorjai, Introduction to Surface Chemistry and Catalysis,
Wiley, New York, NY, 1994.
[5] U. Heiz, A. Sanchez, S. Abbet, W.-D. Schneider, Chem. Phys. 262
(2000) 189.
[6] B.C. Gates, J. Molec, A. Catal, Chem. 163 (2000) 55.
[7] A.S. Eppler, G. Rupprechter, L. Guczi, G.A. Somorjai, J. Phys. Chem.
B 101 (1997) 9973.
[8] D.M. Kolb, Surf. Sci. 500 (2000) 722.
[9] P. Vettinger, G. Cross, M. Despont, U. Drechsler, U. Durig, B.
Gotsmann, W. Haberle, M.A. Lantz, H.E. Rothuizen, R. Stutz,
G.K. Binnig, IEEE Trans. Nanotechnol. 1 (2002) 39.
[10] R.D. Piner, J. Zhu, F. Xu, S. Hong, C.A. Mirkin, Science 283 (1999)
661.
[11] K. Hjrup Hansen, Z

. S

ljivancanin, E. Lgsgaard, F. Besenbacher, I.

Stensgaard, Surf. Sci. 505 (2002) 25.
[12] H. Graoui, S. Giorgio, C.R. Henry, Surf. Sci. 417 (1998) 350.
[13] H.H. Kung, Transition Metal Oxides: Surface Chemistry and
Catalysis, Elsevier Science, The Netherlands, 1989.
[14] A.I. Kozlov, M.C. Kung, W.M. Xue, H.H. Kung, Angew. Chem. Int.
Ed. 42 (2003) 2415.
[15] W.M. Xue, M.C. Kung, A.I. Kozlov, K.E. Popp, H.H. Kung, Catal.
Today 85 (2003) 219.
[16] (a) J.M. Thomas, R. Raja, G. Sankar, R.G. Bell, Nature 398 (1999)
227;
(b) R. Raja, J.M. Thomas, J. Mol. Catal. A Chem. 181 (2002) 3;
(c) R. Raja, G. Sankar, J.M. Thomas, Angew. Chem. Int. Ed. 39
(2000) 2313.
[17] Z. Chang, M.C. Kung, H.H. Kung, Chem. Comm. (2004) 206.
[18] A. Katz, M.E. Davis, Nature 403 (2000) 286;
V. Dufaud, M.E. Davis, J. Am. Chem. Soc. 125 (2003) 9403.
[19] K. Polborn, K. Severin, Chem. Commun. (1999) 2481.
[20] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D.
Schmitt, C.T.W. Chu, D.H. Olson, E.W. Sheppard, S.B. Mccullen, J.B.
Higgins, J.L. Schlenker, J. Am. Chem. Soc. 114 (1992) 10834.
[21] C.P. Mehnert, J.Y. Ying, Chem. Commun. (1997) 2215.
[22] R.C. Schroden, C.F. Blanford, B.J. Melde, B.J.S. Johnson, A. Stein,
Chem. Mater. 13 (2001) 1074.
[23] T. Shimoboji, E. Larenas, T. Fowler, A.S. Hoffman, P.S. Stayton,
Bioconjugate Chem. 14 (2003) 517.
[24] S. Inagaki, S. Guan, T. Ohsuna, O. Terasaki, Nature 416 (2002) 304.
[25] S. Zahn, J.W. Canary, Angew. Chem. Int. Ed. 37 (1998) 305.
[26] V. Balzani, A. Credi, F.M. Raymo, J.F. Stoddart, Angew. Chem. Int.
Ed. 39 (2000) 3348.
[27] B. Chen, G. Parker II., J. Han, M. Meyyappan, A.M. Cassell, Chem.
Mater. 14 (2002) 1891.
[28] S.M. Bachilo, L. Balzano, J.E. Herrera, F. Pompeo, D.E. Resasco, R.B.
Weisman, J. Am. Chem. Soc. 125 (2003) 11186.
[29] T. Mallouk, A. Sen, V. Crespi, private communication.
[30] R.M. Crooks, M. Zhao, L. Sun, V. Chechik, L.K. Yeung, Acc. Chem.
Res. 34 (2001) 181.
[31] R.W.J. Scott, A.K. Dayte, R.M. Crooks, J. Amer. Chem. Soc. 125
(2003) 3709.
H.H. Kung, M.C. Kung / Catalysis Today xxx (2004) xxxxxx 6
DTD 5