Volume 77, Number 44
CENEAR 77 44 pp. 27-35
ISSN 0009-2347
C&EN London
The chemistry of highly branched macromolecules known as dendrimers is blossoming, judging from a conference held in Frankfurt last month. But can this relatively new field of research bear fruit in terms of applications?
The answer is a definite but qualified "yes," according to several speakers at the first-ever international symposium devoted to the topic. Suggested applications include drug delivery, energy harvesting, ion sensing, catalysis, and information storage.
The 1st International Dendrimer Symposium was organized by DECHEMA, a Frankfurt-based German nonprofit society concerned with chemical engineering, chemical technology, and biotechnology. Topics included not only potential applications of dendrimers and their close cousins--the hyperbranched polymers--but also their design, synthesis, structure, analysis, and properties.
The conference, which was attended by 183 participants from 21 countries, attracted many of the pioneers and leading lights in the field of dendrimer chemistry.
One of them, Fritz Vögtle, chemistry professor and a director at the University of Bonn, Germany, pointed out in his welcoming address that a large variety of dendrimers with different cores, branching units, and end groups has been synthesized in the past few years.
"It is astonishing how creative dendrimer chemists are," he said. "The area is booming at the moment."
Only a handful of research papers on dendrimers was published each year during the 1980s, according to Vögtle. The number of publications increased to almost 50 a year in the early 1990s, and by 1997 the figure had climbed exponentially to almost 500 a year. Vögtle's group in Bonn published a paper on the preparation of well-defined branched structures in 1978. The researchers used an iterative sequence of reaction steps, or "cascade synthesis" as they called it, for the synthesis.
The term "dendrimer," which derives from the Greek words dendron, meaning tree, and meros, meaning part, was introduced in 1984 by Donald A. Tomalia, scientific director of the Center for Biologic Nanotechnology at the University of Michigan, Ann Arbor, and coworkers.
Dendrimers are typically well-defined globular macromolecules constructed around a core unit. During synthesis, each successive reaction step leads to an additional "generation" of branching. Ideally, dendrimers exhibit monodispersity, which means that all the molecules are exactly the same in terms of their structure, composition, and molecular weight. However, because the syntheses of higher generations require numerous steps, the final products often contain defects.
Hyperbranched polymers, on the other hand, can be synthesized in a single step but lack the monodispersity and well-defined structures of perfect dendrimers (C&EN, Sept. 6, page 37).
"Both dendrimers and hyperbranched polymers are dendritic molecules characterized by branching of branches," Vögtle noted.
From left, dendrimer pioneers Tomalia, Vögtle, and Newkome. [Photos by Michael Freemantle]
As a dendrimer grows larger, the end groups on the surface of the globule become more densely packed and eventually, because of steric hindrance, the dendrimer reaches its upper generation limit. This is known as the "starburst effect" or "de Gennes dense packing" after Pierre-Gilles de Gennes, the French physicist and winner of the 1991 Nobel Prize in Physics who first reported it.
The growth of successive generations of a dendrimer radially outward from a central core is called "divergent" synthesis. Tomalia reported the use of this approach in 1985 to synthesize poly(amidoamine) dendrimers--known as PAMAM dendrimers, or "starburst polymers," as he called them--starting from an ammonia core.
"PAMAM dendrimers are now being used commercially as immunodiagnostic agents and gene transfection vectors," Tomalia told C&EN.
Dendrimers closely match the sizes and contours of many important proteins and bioassemblies, Tomalia observed. "Fifth- and sixth-generation PAMAM dendrimers have diameters that are about the same as those of the ubiquitous membranes encasing all biological cells," he noted.
Another dendrimer pioneer, George R. Newkome, chemistry professor and director of the Center for Molecular Design & Recognition at the University of South Florida, Tampa, also published a paper in 1985 on the use of the divergent approach to prepare dendrimers. Newkome and his team synthesized poly(amidoalcohol)s with micellar structures that he called "arborols" because they are based on an architectural model of trees. The outer surfaces of these molecules are covered with polar functional hydroxyl groups.
Five years after Tomalia's and Newkome's seminal papers, Cornell University chemists Jean M. J. Fréchet, now a chemistry professor at the University of California, Berkeley, and Craig J. Hawker, now a staff scientist at IBM Almaden Research Center, San Jose, Calif., introduced the "convergent" approach to dendrimer synthesis. With this approach, dendrimer segments, or dendrons, are constructed and then assembled around the central core.
"The field of dendrimers continues to yield wonderful findings," Fréchet told C&EN. "It is amazing how much can be done with such lovely molecules that owe much of their properties to the unique architectural and functional control achieved during their synthesis."
Fréchet, who delivered the opening lecture at the Frankfurt symposium, pointed out that, while numerous types of dendrimers have been developed over the past 15 years, only a few structures have been used widely, mostly in the research environment.
"Our group at UC Berkeley is deeply involved not only in mission-oriented syntheses but also in learning more about the properties of dendrimers and developing applications that take full advantage of their precise nanometer size, high functionality, and regular structural features," he noted.
Fréchet's lecture in Frankfurt largely focused on the design, synthesis, and applications of polyether and polyester dendritic polymers. "We have recently developed two totally new families of aliphatic dendrimers built with ether or ester linkages," he said.
The polyether dendrimer is a branched analog of poly(ethylene glycol) with a 2:1 carbon-to-oxygen ratio and multiple hydroxyl or protected hydroxyl chain ends. It is soluble in water. The aliphatic polyester dendrimer, when terminated by alcohol groups, is also water soluble and, according to Fréchet, appeared to be harmless in recent animal studies.
"We believe that these two families of polymers are ideally suited for applications as varied as targeted drug delivery, light harvesting, and energy conversion," he said.
In collaboration with Francis C. Szoka, professor of pharmacy and pharmaceutical chemistry at the University of California, San Francisco, Fréchet and his team are working on the use of dendrimers for targeted delivery of toxic drugs used in chemotherapy.
"The drug is attached to the dendrimer via a cleavable linkage," Fréchet explained. "Solubilizing groups are added, and a chemical moiety capable of targeting the dendrimer to the target organ is attached.
"Our studies confirmed that the aliphatic dendrimers we use are nontoxic," he continued. "Biodistribution studies carried out using a dendrimer with radioactive iodine-125 attached to a phenolic core show that the dendrimers can be completely eliminated through the kidneys as urine or through the feces. Our next round of animal studies will start soon with a more advanced model containing the anticancer drug doxorubicin."
Another dendrimer that is attracting a lot of attention is a highly branched macromolecule complexed with gadolinium(III) ions. The dendrimer, called Gadomer-17, is a promising candidate for magnetic resonance angiography and is about to enter Phase I clinical trials, reported Heribert Schmitt-Willich, senior scientist at Schering, Berlin.
Small gadolinium(III) chelates have been used clinically since 1988, he pointed out. Gadomer-17, however, is the first polymeric gadolinium contrast agent designed for angiography and tumor differentiation that has been shown in animal tests to fulfill the essential requirement of complete elimination of the heavy metal from the body.
"Gadomer-17 has a molecular weight of 17 kilodaltons," Schmitt-Willich told C&EN. "It is built up of a central trimesoyl [benzene-1,3,5-tricarbonyl] core, contains two generations of l-lysine residues, and has 24 macrocyclic gadolinium(III) chelates at its surface."
The macromolecular complex was characterized by matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry, capillary electrophoresis, gel permeation chromatography, and other analytical techniques. "It took more than one year to obtain the monodisperse compound," Schmitt-Willich said. "High-purity Gadomer-17 can now be synthesized on the kilogram scale for the planned clinical trials."
Fréchet remarked that "this is a real accomplishment considering the scale and the very high purity of the product."
Fréchet and coworkers at UC Berkeley are also investigating the use of dendrimers to harvest broadband light and convert the energy into monochromatic light with amplification, into electricity through charge separation, or into chemical energy. The systems consist of light-harvesting dendrons with numerous laser dye chromophores, such as coumarins, at the periphery. A single chromophore--an oligothiophene, for example--lies at the core of the dendrimer. The light is harvested over a wide area on the surface of the dendrimer and funneled by dipole-dipole interactions to the single active site at the core where energy conversion takes place.
"Energy transfer is through space, not through bonds," Fréchet said. "It is remarkably efficient, almost quantitative. The system more closely mimics the behavior found in natural light-harvesting systems."
In work yet to be published, the group shows that the amount of light harvested by a system containing a pentathiophene acceptor core can be doubled by increasing the generation or size of the dendrimer. Furthermore, the light emitted by the pentathiophene doubles with each successive generation of the dendrimer.
"The system is an optical amplifier," Fréchet said.
The group has also used mixed self-assembled monolayers of light-harvesting and light-emitting chromophores on the surface of a silicon wafer. Energy harvested by the light-harvesting donor molecules is transferred through space to the acceptor molecules, which emit fluorescence with amplification. The donor is a blue coumarin chromophore built into a first- or second-generation dendrimer. The acceptor molecule, also a coumarin, is a red chromophore. Both coumarins are very stable laser dyes, according to Fréchet.
"All the light harvested by the blue chromophore is transferred through space to the red chromophore that emits light," Fréchet explained. "Optical amplification is achieved as the red chromophore emits much more light than would be the case by its direct excitation."
Another speaker in Frankfurt, Vincenzo Balzani, chemistry professor at the University of Bologna, Italy, described recent and ongoing work on dendrimers containing photo- and redox-active units. "Dendrimers are likely to play an important role in artificial photosynthesis and information processing at the molecular level," he said. "Our group is particularly interested in dendrimer-photon and dendrimer-electron interactions.
"Since the beginning of our studies we have realized that transition-metal complexes, because of their outstanding photochemical, photophysical, and electrochemical properties, may be more useful than organic molecules as building blocks for the construction of dendrimers capable of elaborating photon and electron inputs," he told C&EN.
Balzani's group has developed dendrimers that can not only be used for harvesting sunlight but are also "outstanding" electron reservoirs. The work is a collaborative project with chemistry professor Sebastiano Campagna and research worker Scolastica Serroni at the University of Messina, Italy; Gianfranco Denti, chemistry professor at the University of Pisa, Italy; and chemistry professor Sergio Roffia's group at the University of Bologna.
"We have shown that a small dendrimer containing six ruthenium(II) polypyridine-type units undergoes as many as 26 reversible ligand-centered reduction processes," Balzani said. "Since the same compound also exhibits six reversible metal-centered oxidation processes, it is capable of exchanging as many as 32 electrons altogether. This is the most extensive redox series reported so far."
Balzani's group also has been collaborating with Vögtle and his team in Bonn on a number of dendrimer projects. They include an investigation of the photophysical properties of first- and second-generation dendrimers built around a core of the ruthenium(II) bipyridine complex [Ru(bpy)3]2+ and bearing 12 or 24 naphthyl units on the periphery.
Balzani: dendrimers for harvesting sunlight
"We have shown that the fluorescence of the peripheral naphthyl units is completely quenched as a consequence of a very efficient energy-transfer process to the metal-based dendritic core," Balzani told C&EN. "Furthermore, since the dendrimer wedges protect the [Ru(bpy)3]2+core from oxygen quenching, in aerated solutions the quantum yield of ruthenium-based phosphorescence of the second-generation dendrimer is three times that of [Ru(bpy)3]2+."
He pointed out that dendrimers containing photoactive units can also be exploited for sensing metal ions. Balzani, Vögtle, and their teams have studied up to fifth-generation poly(propylene amine) dendrimers--also known as POPAM dendrimers--functionalized at the periphery with strongly fluorescent dansyl [1-dimethylamino-1-naphthalene-5-sulfonyl] units. The fifth-generation dendrimer contains 126 tertiary amine units: 64 peripheral dansyl units and 62 amine units in the branches. The dendrimer therefore has 126 protonation or metal-coordination sites.
"Upon addition of acid, the absorption and emission bands of the dansyl units are gradually replaced by those of the protonated dansyl units," Balzani explained. "The protonation first involves the tertiary amine groups of the interior and then the dansyl units in the periphery. Dendrimers of different generations exhibit different titration curves showing that energy-transfer processes take place in the partially protonated species.
"On addition of cobalt(II) or copper(II) ions to solutions of such dansyl-functionalized POPAM dendrimers, the dansyl fluorescence is strongly quenched," he continued. "This is a very interesting result since it shows that the extraction of transition-metal ions from solution by suitable dendrimers can easily be monitored with sensory signal amplification by exploiting the quenching action of the metal ions on fluorescent labels appended to the dendrimers.
"Furthermore, fluorescence recovery is observed on standing," he added. "We can therefore follow the time dependence of the penetration of metal ions in the interior of the dendrimers."
Chemistry professor David N. Reinhoudt and coworkers at the University of Twente, the Netherlands, have been investigating single molecules of metal-containing dendrimers.
"We had a dream about 10 years ago that we could communicate with single molecules," he told the Frankfurt audience. "Single molecules might be the ultimate way to store information, but they need to be large. And dendrimers are large molecules.
"We are using a noncovalent supramolecular approach to synthesize molecules with dimensions between 5 and 20 nm," he noted. "We are exploring two different ways of connecting building blocks--that is, either through hydrogen bonded networks or through coordination chemistry. For the synthesis of metallodendrimers we have used divergent and convergent approaches, and also self-assembly."
In a paper to be published in the Journal of Physical Chemistry A, Reinhoudt's group reports that single dendritic molecules containing a single fluorescent rhodamine dye in the core have been observed with near-field scanning optical and confocal microscopy. The dendrimers, which have dimensions ranging from a few to tens of nanometers depending on the number of generations, are organopalladium spheres constructed using coordination chemistry.
The full three-dimensional orientation of each individual fluorescent core can be resolved, the authors report. "This is the first simultaneous measurement of the topography and fluorescence of single dendritic molecules adsorbed on a glass surface," they state.
The group also is attempting to use single molecules of the fluorescent dendrimer to make a reversible on/off switch. "We can address single molecules," Reinhoudt explained. "We can switch the fluorescence off by passing ammonia through a cell containing the molecules. When the cell is flushed with nitrogen, the fluorescence is switched back on."
Functional dendrimers can mimic globular proteins. François Diederich, professor of organic chemistry at the Swiss Federal Institute of Technology in Zürich, presented a lecture on their development.
"In our research, we introduce defined binding and catalytic sites into the cores of dendrimers and explore how the dendritic shells create a distinct micropolarity around the core, thereby influencing its molecular recognition and catalytic properties," he said. "The distinct microenvironment inside dendrimers promises interesting applications in catalysis. Not only can the potentials of electrocatalysts be tuned to order, but the microenvironment can also be used for catalytic processes inside dendrimers that are accelerated in a specific apolar or polar environment."
The bulk of his talk was devoted to research to be published in Angewandte Chemie International Edition on dendritic iron porphyrins that are functional mimics of globular electron-transfer heme proteins, particularly the cytochromes. The dendrimers have triethylene glycol monomethyl ether surface groups and two imidazoles covalently tethered as axial ligands to their porphyrin cores.
"We have provided the first quantitative evaluation of microenvironmental effects inside dendrimers," Diederich told C&EN. "We showed that the redox potentials of an iron heme encapsulated inside dendrimers are strongly dependent on the dendritic generation, but not the solvent."
He pointed out that the redox potential of the Fe(III)/Fe(II) couple becomes more positive as the dendrimer generation increases, even in solvents of widely differing polarity.
"The Fe(III)/Fe(II) redox potential of the second-generation dendrimer is exactly the same in solvents as different as dichloromethane and water," he said. "Our results clearly demonstrate that the dendritic branching creates a unique local microenvironment around the isolated electroactive core. The porphyrin in the second-generation dendrimer no longer recognizes the nature of the outside solvent.
"The dendritic shell therefore fully mimics the protect-ing peptide shell, which modulates the redox potential of the Fe(III)/Fe(II) couple in cytochromes," he continued. "This model study confirms the large contributions of the densely packed protein shell to the strong positive shifts of the Fe(III)/Fe(II) potential in cytochromes."
Fréchet, Hawker, and their coworkers at UC Berkeley and IBM have developed dendrimers as nanoscale reactors for catalysis [J. Am. Chem. Soc.,121, 9471 (1999)]. They introduced polar groups in the "inner" building blocks of a convergent dendrimer to build a dendritic macromolecule that contains a high concentration of the polar groups within the cavity of the dendrimer. The researchers demonstrated that the internal polarity of the dendrimer enabled simple nucleophilic substitution or elimination reactions to occur in catalytic fashion.
"This is the first demonstration of the use of the special high-polarity environment of the inner cavity of a dendrimer to catalyze a reaction," Fréchet told C&EN. "Turnover is achieved as the product of the reaction is ejected from the reactive site. There is no reaction without the dendrimer."
Steven C. Zimmerman, chemistry professor at the University of Illinois, Urbana-Champaign, focused on the synthesis and potential applications of dendrimers that have had their cores chemically removed. He reminded participants at the Frankfurt symposium that three structural components are common to all dendrimers: a core unit, peripheral groups, and multiple branching units that span the other two components. The core unit covalently links the dendritic "wedges," or dendrons, as they are called. Hollow dendrimers, he suggested, could greatly facilitate the development of new dendritic catalysts, delivery vehicles, and recognition systems.
Earlier this year, Zimmerman and University of Illinois graduate student Michael S. Wendland reported the first example of what they called a "cored dendrimer" synthesized by cross-linking the dendrimer's peripheral groups and then removing the core by hydrolysis [J. Am. Chem. Soc., 121, 1389 (1999)].
"Chemists have dreamed about being able to create a mold around any molecule, which would then bind that molecule tightly and specifically," Zimmerman told C&EN. "We have embarked on a major new initiative to develop synthetic antibodies and novel nanostructured capsules by a molding process in which a single dendritic structure is cross-linked around a removable template. This process bears conceptual similarity to polymer imprinting. Molecular-imprinted polymers are well-established materials and widely acknowledged for their extraordinary potential to impact biotechnology and biomedicine."
Molecular-imprinted polymers are synthesized by carrying out a polymerization reaction in the presence of a template. However, they have yet to achieve commercial applicability because of a number of limitations, including insolubility and difficulty in quantitatively removing the template.
From left, Reinhoudt, Meijer, and Zimmerman discussed chemical applications using functionalized dendrimers.
"Two strategies are possible for hollowing out the core of a dendrimer," Zimmerman said. "One is self-assembly. The other, which we adopted, is direct synthesis. We chose this approach because we can work at the high dilutions required to avoid oligomerization. With the noncovalent approach, much higher concentrations are needed."
Zimmerman's group synthesizes the hollow dendrimers by covalently attaching dendrons to a molecular core that acts as a template.
"The covalent attachments of dendrons to the template are designed to be cleaved without destroying the dendrons," he said. "Once we do the cross-linking and then cleave the core or template, the whole macromolecular structure holds together. We have removed quite complicated cores such as chiral nonracemic cores and porphyrin cores."
E. W. (Bert) Meijer, professor of macromolecular and organic chemistry at Eindhoven University of Technology, the Netherlands, reviewed developments in the chemistry and potential applications of poly(propylene imine) dendrimers. In 1994, Meijer's group synthesized and characterized dendritic boxes that are able to encapsulate a large variety of guest dye molecules. The boxes are constructed from chiral shells of protec-ted amino acids and fifth-generation poly(propylene imine) dendrimers.
Two years later, the group showed that inverted unimolecular dendritic micelles could be synthesized by attaching hydrophobic alkyl chains to the end groups of poly(propylene imine) dendrimers.
"We have shown that these dendritic micelles have a number of potential applications in the fields of drug delivery, optical data storage, and dyeing of polypropylene fibers," Meijer said.
In his lecture, which he titled "Self-Assembly and Guest-Host Properties of Dendrimers," Meijer revealed recent work, yet to be published, on fifth-generation dendritic macromolecules modified by alkyl or adamantyl end groups via urea linkages. Pairs of these end groups trap guest molecules by what Meijer calls a "clicking mechanism."
"We have designed guest molecules that specifically bind to the peripheral scaffold of the dendrimer by both acid-base interactions and hydrogen bonding," Meijer said. "NMR analysis shows that 32 molecules are clicked in between 64 end groups of the fifth-generation dendrimer. The synthetic methodology is straightforward and allows easy modification of the dendrimers by clicking of a variety of designed molecules. We foresee several uses for these dendrimers, for example, in catalysis."
Newkome and coworkers at the Center for Molecular Design & Recognition have used a combinatorial approach for the rapid preparation and screening of dendritic families for desirable bulk properties such as solubility, viscosity, and reactivity.
Dendrimers have numerous variable structural parameters, he pointed out in Frankfurt, such as degree of core and monomer branching, choice of branch point, distance between successive branch junctions, and the number and locations of desired internal functional groups. The most important structural unit that addresses all of these parameters is the branched monomer or molecular building block.
Newkome's group has used families of branched isocyanate and functional amines for combinatorial dendrimer construction. "We can put functional groups on the outside of dendrimers and play with them," Newkome said.
He outlined the concept of "turnstile" bond rotation by which a cone segment of the dendrimer containing a triad of surface functional components rotates about a bond in the same way that three substituents on a saturated tetrahedral carbon atom rotate around the fourth carbon bond. This combinatorial-based macromolecular construction technique allows the surface functional groups to be varied and could lead to the creation of dynamic heterogeneous surfaces, he suggested.
This concept, Newkome said, is reminiscent of a "Rubik's Cube." Dendritic macromolecules constructed in this way may be considered to be molecular "Rubik's Spheres," he added.
Percec: designing dendritic building blocks.
The South Florida group has also shown, in work to be published in Angewandte Chemie International Edition, that dendrimers can be used as counterions for positively charged polymetallic arrays to improve the aqueous solubility of the arrays. One example is a third-generation carboxylate-terminated dendrimer used as a counterion for a hexaruthenium macrocycle that is poorly soluble in water. A 1:1 mixture of the hexamer and the dendrimer yielded a deep red solution at room temperature that could be analyzed by nuclear magnetic resonance spectroscopy.
"We anticipate that the use of dendrimers in this manner will lead to new avenues for construction of polymetallic arrays via self-assembly," stated Charles N. Moorefield, assistant director of the center.
Virgil Percec, chemistry professor at the University of Pennsylvania, described how libraries of self-assembling dendrons could be used as building blocks for the rational design of complex nanoarchitectures.
"We have designed dendritic building blocks which self-assemble into cylindrical, spherical, and three-cylindrical bundle supramolecular dendrimers," he said. "These supramolecular dendritic objects have the perfection to self-organize into two- and three-dimensional liquid-crystalline and crystalline lattices and superlattices."
The group has recently synthesized and structurally analyzed a polymer containing twin-dendritic benzamide side groups [Chem. Eur. J., 5, 1070 (1999)]. The disklike side groups of this polymer self-assemble into supramolecular cylindrical dendrimers through hydrogen bonding along the long column axis. The authors point out that this creates a novel supramolecular dendritic architecture consisting of a polymer chain coated by a bundle of three cylinders. This polymer then self-organizes into a thermotropic nematic liquid-crystalline phase.
The team also produced a novel hexagonal columnar liquid-crystalline superlattice by coassembling the three-cylinder polymeric dendrimer with a single-cylinder supramolecular dendrimer. The latter was prepared by self-assembly of a nonpolymeric bisdendritic benzamide. The self-assembly of the cylinders in both the three-cylinder and single-cylinder components relies on the ability of successive twin-dendritic benzamide side groups or bisdendritic benzamide molecules to rotate through 90 °.
Frey: hyperbranched modular approach.
"X-ray analysis of these thermotropic liquid-crystalline lattices allowed for the first time the direct determination of the shape and size of supramolecular dendrimers and of their self-assembling dendrons in the bulk state," he said.
Percec suggested that the new architecture provides a synthetic concept that could become as powerful as that of the three- and four-helix bundle protein motifs used by nature to create binding sites, even though the detailed self-assembly mechanisms of these two systems are quite different.
Tomalia and coworkers at the University of Michigan Center for Biologic Nanotechnology have been investigating the use of dendrimers as fundamental buildings blocks for the synthesis of macromolecular clusters with a higher order of complexity. The group synthesizes the clusters from two poly(amidoamine) dendrimer modules either by direct regiospecific bond formation or by self-assembly through charge neutralization followed by in situ covalent bond formation.
"We call these covalently bonded clusters 'core-shell tecto-(dendrimers),' " Tomalia told the Frankfurt symposium. "They are a new class of regiospecifically cross-linked polymers. We have made dozens of these materials by the gram."
The core-shell tecto-(dendrimer) products consist of a fifth-, sixth-, or seventh-generation core dendrimer surrounded by a shell of third- or fifth-generation dendrimers [Langmuir, 15,7347 (1999)]. The group has used atomic force microscopy, MALDI-TOF mass spectrometry, and other analytical techniques to provide evidence of these spheroidal nanoscale clusters and has shown that their diameters are in the range 1 to 10 nm.
"These core-shell constructs may be utilized as core-type templates to add additional shells," Tomalia pointed out. "These strategies will now allow the systematic construction of relatively precise macromolecular structures with diameters ranging from 1 to 100 nm."
He pointed out that such nanoscale structures result in ultrahigh surface areas and quantumlike confinement effects that produce properties uniquely different from those observed for comparable bulk or conventional materials.
The clusters also have another attractive feature--the shells can either be saturated or partially filled. "Our present synthetic results demonstrate that our strategies may be utilized to produce partially filled core-shell structures by direct regiospecific bond formation," Tomalia explained. "Preorganization followed by in situ regiospecific bond formation produces a totally saturated core-shell structure."
He suggested that the dendrimers are reactive molecular core-shell analogs of atoms. "We have used dendrimers to mimic at the molecular level what Dalton proposed for atoms 150 years ago," he said.
British scientist John Dalton (1766-1844) proposed that "simple" atoms consist of spherical particles and that their combination to form "compound atoms" takes place in such a way that the latter are as symmetrical as possible.
"We are very excited about dendrimer-dendrimer reactions and core-shell tecto-(dendrimers)," Tomalia concluded. "It is new chemistry."
Holger Frey, a senior researcher at Albert Ludwigs University, Freiburg, Germany, opened his talk on recent progress on the tailoring of hyperbranched polymer architectures by quoting a question posed in a recent article on the topic in C&EN (Sept. 6, page 37).
"Can hyperbranched polymers rival dendrimers?" he asked. "This question may sound like blasphemy at a conference of dendrimer chemists."
He revealed that the Freiburg group, which also includes professor of macromolecular chemistry Rolf Mülhaupt and Ph.D. student Alexander Sunder, has recently developed a modular approach that uses dendronlike hyperbranched structures to synthesize hyperbranched polymers. The approach relies on anionic ring-opening multibranching polymerization of glycidol, a highly reactive hydroxy epoxide, using an anionic alkoxide initiator.
"Using this modular concept, we can vary the structure of the initiator and the nature of the monomer," he said. "And by using copolymerization of other epoxides with glycidol, we can introduce a well-defined number of different end groups in the molecules. In just two steps we can build a large variety of hyperbranched polymers with this approach. And in only three steps, we can prepare core-shell molecules that possess an interior clearly distinguished from the end groups."
The Freiburg group has shown, in work to be published in Angewandte Chemie International Edition, that it is possible to prepare unimolecular micelles by simple esterification of hyperbranched polyglycerols with fatty acids. The group has also shown that such polymers can be used for dendrimer-like encapsulation of dyes and for polymer-supported synthesis of organic compounds.
"Such hyperbranched polymers, prepared in a controlled manner in two or three steps, might be able to perform just as well as dendrimers prepared in 10 to 12 steps," Frey concluded. "However, these hyperbranched materials will inevitably possess a certain polydispersity. The question of whether or not this polydispersity is tolerable in potential applications will certainly be an issue in the future."
Dirk Muscat, a research chemist at DSM Research, Geleen, the Netherlands, described the synthesis of a family of multifunctional hyperbranched polymers that DSM launched under the trade name Hybrane earlier this year.
The materials are hyperbranched polyesteramides synthesized by polycondensation of a monomer prepared by the addition reaction of diisopropanolamine and a cyclic anhydride. Hybrane polymers with aliphatic and aromatic carboxylic ester, unsaturated fatty acid ester, and tertiary amine end groups have been prepared.
"Properties of the hyperbranched polyesteramides, such as melt viscosity and solubility, can be controlled and adjusted by varying the molecular weight, by modifying the end groups, and by choice of cyclic anhydride," Muscat noted.
The polymers may find potential applications as cross-linkers in coatings, as toner resin, and as surfactants, according to Muscat. "Especially impressive is the potential use for disperse dyeing of polypropylene," he noted. "This has been a problem for several decades. In order to keep the dyes in the polypropylene fibers, the dye booster has to fix the dye and to be compatible with the polypropylene."
He points out that hyperbranched polyesteramides based on phthalic anhydride and diisopropanolamine and half-functionalized with stearic acid are able to fix the dyes via their polar core, and at the same time their long alkyl chains make them compatible with the polypropylene matrix.
Other hyperbranched polymers have been commercially available for many years, according to Frey. "Hyperbranched aliphatic polyesters have been on the market for three or four years and have been tested as a component for novel coatings by various companies," he told C&EN. "And if you polymerize aziridine, you obtain hyperbranched polyethyleneimine, which has been sold by BASF for more than 30 years. This hyperbranched product possesses a broad polydispersity and, in comparison to the perfectly branched dendrimers, is not well defined. Even so, it has a large market."
Ulrich S. Schubert, a senior research chemist at the Technical University of Munich, Germany, and a participant at the Frankfurt symposium, asked if the special properties and architectures of dendrimers were needed for applications.
"In my opinion, I only see possibilities for applications for certain medical purposes," he said. "On the other hand, the less perfect systems--the hyperbranched polymers--have great potential.
"They combine some of the features of dendrimers, such as highly branched architectures and a high content of functional groups, with cheap and easy preparation procedures," he continued. "However, they lack the monodispersity and well-defined structures of perfect dendrimers that offer unique possibilities for drug delivery and imaging applications and for the understanding of such architectures. The present ongoing research in dendrimers therefore prepares the ground for potential applications in this and related fields."
Fréchet is optimistic about the future of highly branched materials. "We are only beginning to exploit the unique properties of dendrimers and, despite claims to the contrary, not all useful dendritic structures have been made," he remarked. "There is still room for some imaginative syntheses since the few families of dendrimers and hyperbranched polymers in widespread research use today cannot accommodate the huge variety of applications that could benefit from their unique properties and unusual behavior. The future of dendritic macromolecules still lies in large part with clever syntheses that can deliver the structures best suited for interesting applications."
Meijer is also excited by advances in the science and engineering of dendritic materials in recent years. "Our understanding has progressed enormously, while a number of applications are close to being used," he said. "As a result, the field is highly competitive and this is great to see because friendly competition is what science needs for rapid progress."
Chemical & Engineering News
Copyright © 1999 American Chemical Society