<- click here for height profile plot of PAHs in
a 70mm CH4 flamecluster research group, university of hawaii at manoa, honolulu
In a collaboration with Asea Brown Bovery (ABB) Corporation and the Swiss Federal Institute of Technology (ETH) in Zuerich we studied Nanoparticle Aerosols. In particular we are interested in the early stages of soot and fullerene formation in laminar hydrocarbon diffusion flames, with participation of polycyclic aromatic hydrocarbons (PAH).
<- click here for height profile plot of PAHs in
a 70mm CH4 flame
Using various experimental techniques we studied electrical and electronic properties of free and assembled fullerenes.
Photoluminescence spectrum of ZnS-particles
 |
 |
UV and IR parts of the luminescence spectrum |
 |
Deconvoluted UV luminescence spectrum and assigned optical transitions into and out of deep energy levels in the band gap of ZnS particles . |
Soot formation in combustion is
a serious problem leading to air pollution and a reduction of the performance of
coal power plants. The high-temperature growth of carbon particles (soot) in
flames however is little understood. A major problem is that there is much more
soot produced in a flame than expected from the 'standard growth model' , which
assumes a successive nucleation process based on acetylene decomposition. In a
set of experiments we showed that coalescence of pylycyclic aromatic
hydrocarbons (PAH) the soot precursors, is an important process leading to a
much higher rate of soot production. With these results we could find ways to
suppress the formation of primary carbon particles by using ferrocene and
nanoparticles of palladium.
Evaluation of an in-situ Sampling Probe for its Accuracy in Determining Particle Size Distributions from Flames, M. Kasper, K. Siegmann, and K. Sattler, J. Aerosol Sci. 28, 1569 (1997)
Formation
of Carbon in Combustion: The Influence of Fuel Additives
The Effect of Ferrocene Addition on Particle Formation and Burnout in
Combustion, M. Kasper, K. Sattler, K. Siegmann, and U. Matter, J. Aerosol
Sci. 29, Suppl. 1, 617 (1998)
Increasing lung cancer mortality is found in many countries all over the world. It had been reported that in China more women than men die of lung cancer even though men are much heavier smokers. In fact, most Chinese women with lung cancer have adenocarcinoma, which is believed to have little relationship with smoking. Since women are cooking with very hot oil, we suspected that the fumes from the woks could contain carcinogenous PAH molecules. We built an experimental setup to produce oil fumes in a controlled way and to analyze the size distribution of the oil particles. Cooking oils from food stores in Switzerland were heated and the fumes were studied. Using a scanning mobility particle sizer (SMPS) we found that the oil fume particles are small enough to pass the filters of the human nose so that the oil droplets are deposited in the lung. Using gas chromatographic mass spectroscopy (GC/MS) and high-performance liquid chromatography (HPLC) we determined the distribution of the PAH's. We showed that due to the carcinonenic action of some the the PAHs, oil fumes therefore may be health hazards for women or kitchen professionals preparing food using very hot vegetable oils.
Aerosol from Hot Cooking Oil, a Possible Health Hazard, K. Siegmann and K. Sattler, J. Aerosol Sci. 27 , Supl. 1, 493 (1996)
Fabrication of structures with nano-scale dimensions could lead to a substatial decrease in the size of electronic devices. The techniques for this yet have to be develloped. One of the tools can be the atomically sharp tip of a scanning tunneling microscope. The tip can be placed with atomic precision at a selected spot on a surface and can influence single atoms or layers. Using the tip of a STM we could manipulate single atomic sheets of carbon. The graphene sheets can be folded and unfolded repeatedly. One can imagine that the developed technique could be used to wrap single atoms with carbon sheets.
There is currently a strong interest in the prospect of producing new materials consisting of small atomic clusters. Such cluster-assembled materials may vary significantly from their crystalline counterparts. Mechanical, electronic, optical and other properties are expected to be different for such assemblies which should make them good potential candidates as new building materials for electronic devices. Also, the quantum effects which occur in such materials of finite size and dimension, lead to their special properties.
It has been about 15 years since atomic clusters were produced and investigated in laboratories. Since then, knowledge about clusters has enjoyed a rapid and sustained growth, and cluster research became a new branch of science. Most of the work focused on the properties of free clusters. Only few results have been obtained for deposited or assembled clusters. Aggregates, consisting of less than a few hundred atoms cover the transition region between single atom and solid. Over the years, stunning results have been obtained for this size range. From such results it became clear that the same chemical element could be used to make a series of new materials with entirely different and new properties. The most prominent example is carbon, which exists in two crystalline forms, graphite and diamond, but also in a cluster-assembled structure, the fullerite solid. C60 clusters are the units, assembled in fullerite, arranged in a face-centered cubic lattice at normal conditions. Other carbon cluster units, like C70, C84, C96, have also been assembled to form solids.
Since the discovery of C60, it has been speculated that similar-type hollow clusters could also be prepared with silicon. In fact there are silicon clathrate compounds which are composed of silicon polyhedra analogous to fullerenes. The polyhedra are connected three-dimensionally by face sharing, alkali metal atoms being placed in the center of the polyhedra. Such cluster-based materials with shared unit faces will be discussed for silicon and carbon.
Very few reports have been given for assembled metal clusters. Presumably, metal clusters after being assembled, tend to deform and grow together losing their identity, while covalent clusters, when 'magic' in their number of atoms, may much better keep their free cluster properties.
This book was written with the goal of introducing scientists and materials developers to clusters and cluster-assembled materials. The first two articles give general theoretical (Jena et al.) and experimental (Broyer et al.) aspects of cluster-based solids. It is pointed out that for building such materials the cluster unit should be very stable and its binding energy should be larger than the cohesive energy of the corresponding bulk. Therefore the building unit should be a 'magic number cluster' having closed atomic or electron shells and being less reactive than 'nonmagic number clusters'. A magic number cluster is chemically inert and can be assembled without coalescing. For example, 13, 55, and 147 are magic numbers for close-packed clusters, and C60 in its icosahedral cage-structure is another prominent example.
Experimentally it is difficult to produce in large quantities monodispersed clusters other than fullerenes. Such efforts are described for metal clusters in the articles by Broyer et al. and Schneider et al. . In Broyer's article it is shown that covalent clusters exhibit a memory effect after synthesis as amorphous material. For metal clusters the memory effect does not well exist. This suggests that it will be much easier to produce cluster-assembled materials from covalent clusters than from metal clusters. In Broyer's work, Si and C clusters were deposited and analyzed with electron and optical spectroscopies. A spectroscopic study of size-selected metal clusters is reported by Schneider et al.. Mass selected Ptn and Pdn (N=1-15) in submonolayer quantities were deposited on a Ag single crystal surface and characterized by X-ray and UV photoelectron spectroscopy. Discrete electronic structure features were found for the clusters. The structural stability of the Pd and Pt clusters was investigated and information about cluster-cluster and cluster-substrate interactions was obtained.
Two articles on silicon clusters follow, predicting atomic and electronic properties by use of quantum calculations (Kaxiras et al. and Chelikowsky et al.). In Kaxiras' article, structural models for Si clusters of intermediate size are reviewed. Surface reconstruction induced geometries are used for these models. They were developed to explain the occurrence of magic number clusters in beam experiments. Electronic density of states for Si33 and Si45, two of such magic number clusters, is given. Assembling these clusters into a solid will depend on how their structure can be accommodated in a solid form.
Chelikowsky et al. point out the importance of finite temperature simulations for clusters which may result in structures different from their ground state. Many problems with cluster calculations arise from the existence of multiple local minima in the potential energy surface. Also, kinetic effects and non-equilibrium morphologies may dominate clusters created under laboratory conditions. In beam experiments the clusters usually are ionized and interpretation of data requires to consider atomic relaxation within the charged cluster. Another important aspect is that covalent clusters may have atomic coordination numbers different from the bulk. Silicon atoms for example are four-fold coordinated in crystals, but may exhibit two-, three-, or possibly five-fold coordinated states in clusters. Therefore, it is useful to concentrate on ab initio methods for cluster calculations.
Fullerene-type Si-structures are experimentally realized in new silicon clathrate compounds containing barium with alkali metal, prepared from the ternary Zintl phases. The silicon clathrate compound is composed of a Si sp3 open network having two types of cages, Si20 pentagonal dodecahedra and Si24 tetrakaidecahedra. The two types of polyhedra are linked by shared faces. Yamanaka et al. found that the barium containing clathrate compounds became type II superconductors with TC of about 4K. This is the first superconductor found for a Si-sp3 covalent network.
Several articles about fullerene materials follow. These are currently the best characterized cluster-assembled materials. Studies by Sakurai et al. using scanning tunneling microscopy yield direct images of adsorbed fullerenes, their intramolecular structures and their interactions with semiconductor and metal surfaces. They investigated the fullerene thin film growth, on Si, Cu, and Ag single crystal substrates. Results are given for the fullerenes C60, C70, C84 and the metallofullerenes Sc@C74, Sc2@C74, Sc2@C84, Y@C82, and Gd@C82.
In the paper of Alonso et al. the experimental and theoretical work on clusters of C60, (C60)n, is reviewed. The magic numbers found experimentally for assemblies of C60 are very similar to those found for inert gas clusters. The assemblies show icosahedral structure, different from the face-centered cubic bulk structure of the solid fullerite crystal. A very large N is expected for the transition to the crystal structure to occur. In addition, the coating of C60 and assemblies of C60 by alkali metal and alkali earth metal atoms is considered. Using molecular dynamics simulations, also the melting behavior of (C60)n is calculated.
Experimentally it has been shown that the inter-fullerene coupling can significantly be altered with the application of light, pressure or temperature, leading to cross-linked structures, the C60 polymers (Rao et al.). C60 fullerenes crystallize at low pressures into a face-centered-cubic Van-der-Waals solid in which the molecules are spinning rapidly about their equilibrium lattice positions at room temperature. Cross-linking of the C60 units leads to significant changes of these properties.
Fullerene-related materials such as endohedral metallofullerenes have also been produced. It is a challenge to obtain such materials in large enough quantities for scientific characterization. In the article of Shinohara et al., they first discuss the production and sample purification procedures and then report various spectroscopic methods used. The success of the purification has been a big breakthrough for a further progress. STM studies show that the endohedral metallofullerenes are stacked as close-packed arrays on single crystal surfaces. Synchrotron X-ray diffraction studies of Y@C82 show that Y is encapsulated in the C82 cage. It also reveals that the encaged yttrium atom is not in the center of the cage but very close to the carbon network.
Silicon and carbon, two elements in the same column in the Periodic Table, are remarkably different in their properties as solids and as chemical compounds. However, as small clusters, there are similarities, and fullerene-type structures may be realized for silicon as well (Saito et al., and Yamanaka et al.). In Saito's theoretical work the electronic structure of C60 and Si20 fullerenes and their fullerides are obtained by density functional theory with emphasis on metal doped materials and their superconducting properties. The electronic structure of solid C60, K3C60 and Rb3C60 superconductors, alkaline-earth fullerides Sr6C60 and Ba6C60 and (Ba3Si3 Na@Si20)2 is presented. Similarities and differences between C and Si fullerene-based materials are addressed. Future materials such as C20-assembled solids are discussed. It is further pointed out that the fullerenes extracted so far do not have any fused pentagons (isolated pentagon rule). Fullerenes having fused pentagons in their network have dangling bond states where they can form chemical bonds with C atoms in the soot and therefore they are not extractable. It follows that the structural units in covalent cluster-assembled materials need to be free of dangling bonds, i.e., magic number clusters.
Another new exotic class of solids which are theoretically postulated, are so-called hollow diamonds (Benedek et al.). The proposed diamond-like carbon crystals combine a large nanoscale porosity and high stability with comparatively large stiffness. Their structures are similar to other four-fold coordinated covalent materials such as silicon clathrates. The authors describe the theory of hollow sp3 lattices which are obtained from the coalescence or assembling of small fullerenes. They give a theoretical description of the topology, structure, stability, elastic properties and electronic states of such hollow diamonds.
An article about inorganic fullerenes describes materials which have fullerene-type structures but are made of non-carbon substances like MoS2 (Tenne). This shows that inorganic compounds with layered structures can occur in morphologies similar to carbon. The analogy with carbon-based fullerenes can be extended to similar nanostructures, like nanotubes, nested fullerenes, fullerenes with negative curvature, etc. Various synthetic routes are used to obtain isolated phases of inorganic fullerene materials.
Cluster-assembled materials can be viewed as quasi-crystalline solids, intermediate between crystalline and noncrystalline. A crystalline solid is one in which the constituent atoms are arranged in an ordered array such as in a metal or semiconductor, as opposed to a disordered array as in a glass. In the ideal crystal, every atom is at its place and there is long-range ordering. In glass (which is essentially a frozen liquid) there is some short-range ordering, but no long-range order. In cluster-assembled solids there is the short-range order of the cluster unit, but this repeats at each lattice point, leading to the long-range order of the crystal. This coexistence of two different ordering properties makes these materials very difficult to describe theoretically. On the other hand it is a challenging new task for theorists to handle this interesting issue. Also, for the experimentalist, science of cluster-assembled solids is a great challenge. Production of such materials is very difficult. The cluster units need to be identical in size and structure and produced in macroscopic quantities, which requires far developed, effective generation and separation methods. Considering the many possibilities for new structures we can say that the field of cluster-assembled solids is just starting to develop. As such metastable materials may be realized for most of the elements of the periodic table, there is a great future potential for the science and technology of this new class of solids.
August 1996