Dendrimer-Based Drug Delivery Systems: From Theory to Practice

By Donald A. Tomalia

The opportunities and challenges of using dendrimers to improve drug delivery

Among pharmaceutical and biomedical researchers, the use of dendrimers in drug delivery systems has attracted increasing interest. In particular, researchers have noted that the volume of a dendrimer increases when it has a positive charge. If this property can be applied effectively, dendrimers have enormous potential in drug delivery systems, directly supplying medication to targeted human organs.

With contributions from an international team of pioneers and experts in dendrimer research, this book provides a comprehensive overview of the latest research efforts in designing and optimizing dendrimer-based drug delivery systems. The book analyzes key issues, demonstrating the critical connections that link fundamental concepts, design, synthesis, analytical methodology, and biological assessment to the practical use of dendrimers in drug delivery applications. Topics covered include:

• Dendrimer history
• Synthesis
• Physicochemical properties
• Principles of drug delivery
• Applications in diverse biomedical fields

Dendrimer-Based Drug Delivery Systems reflects the authors' thorough review and analysis of the current literature as well as their own firsthand experience in the lab. Readers will not only discover the current state of the science, but also gain valuable insights into fruitful directions for future research. References at the end of each chapter serve as a gateway to the growing body of literature in the field, enabling readers to explore each individual topic in greater depth.

Pharmaceutical and biomedical researchers will find this book a unique and essential guide to the opportunities, issues, and challenges involved in fully exploiting the potential of dendrimers to improve drug delivery.

• Language: English
• Publisher: Wiley
• Release date: May 23, 2012
• ISBN: 9781118275221

Book preview

Chapter 1

Dendrimer Chemistry: Supramolecular Perspectives and Applications

Charles N. Moorefield, Sujith Perera, and George R. Newkome

There are many beautiful molecular architectures, it is just that some are easier to access than others.

Roald Hoffman, Nobel Prize in Chemistry, 1981

1.1 Introduction

1.1.1 Historical Background

Dendritic chemistry, from its initial development to its application in the construction of utilitarian devices and materials, has provided a great amount of proverbial cement for interdisciplinary integration. Similar to polymer (or macromolecular) chemistry, conceptualized and postulated by luminaries such as Flory [1–3] (Nobel—1974) and Staudinger (Nobel—1953) who provided a new foundation for material sciences, dendrimer chemistry has generated another new level of scaffolding upon which a myriad of potential uses are being explored and exploited.

First introduced as cascade molecules due to their repeating motif by Vögtle and coworkers [4] in 1978, materials analogously termed arborols (derived from the Latin word arbor for tree) and dendrimers (derived from the Greek word dendro for tree) were reported by Newkome et al. [5] and Tomalia et al. [6] both in 1985, respectively. While these reports specifically addressed the potential to craft branching molecular architectures with multiple terminal functionality and repetitive branch junctures (Tomaila, 1 → 2 branching based on linear building blocks; Newkome, 1 → 3 branching based on modular building blocks with preconstructed branching centers) another notable report appeared by Aharoni and coworkers [7] in 1982 describing the "Size and Solution Properties of Globular tert-Butoxycarbonyl-poly(α,ε-l-lysine). Their study involved the characterization of 1 → 2, asymmetrically branched materials that were termed nondraining globular biopolymers" that were iteratively prepared and reported in 1981 (U.S. patent 4289872, Denkewalter et al. [8]). Other notable and interesting reports prior to the explosive advent of dendritic chemistry, include the iterative synthesis of ultralong, linear paraffins reported by Bidd and Whiting [9], the early observation by Ingold and Nickolls [10] of the entrapment of gas molecules by methanetetraacetic acid, and Lehn's elegant modular approach [11] to cryptate syntheses.

1.1.2 Architectural Concepts

Dendritic molecules can be envisioned by considering the repetitive layering of multifunctional building blocks based on a protection and deprotection scheme or the addition of increasing numbers of linear, complementary monomers. This generally results in a branched, tree- or fractal-like, molecular motif whereby each incorporated layer provides a foundation for the successive layer. Since the number of reactive sites and branching centers increases with each layer, a mushrooming framework is produced. The synthetic protocol can be visualized (Scheme 1.1) by considering the attachment of a generic 1 → 3 branched building block 1 that possesses three reactive sites differentiated from the 4th. Thus, treatment of monomer 1 with three equivalents of a like monomer produces a new monomer 2 with the same functional group characteristics as the starting materials, except that the periphery has now grown and expanded to a 1 → 9 branched construct.

Scheme 1.1 Divergent and convergent routes to branched architecture.

The iterative dendritic strategy has developed into two general modes of construction. The divergent route, initially introduced by Vögtle et al. [4], whereby molecular growth essentially proceeds from the inside outward and the convergent route, introduced in 1990 by Fréchet et al. [12], resulting in growth from the outside inward. Differences in the two methods arise from building block order of addition and can be affected by the control over functional group activation and deactivation. Thus, logical choices of protection–deprotection strategies derived from classical synthetic chemistry are a prime importance in dendritic chemistry. Addition of nine equivalents of a triprotected monomer 1 to the surface of a growing specie 2 will lead to the progressively greater branched construct 3. The same material (i.e., 3) can be derived convergently by inverting the process to add three equivalents of the 1 → 9 higher–order, branched monomer to the simple monomer. Both methods allow the construction of dendritic material and also have their individual strengths and weaknesses. For example, divergent syntheses requires an ever increasing number of monomer attachment reactions leading to a higher probability of incomplete reactions at the ever-expanding periphery leading to a greater number of imperfections; whereas, convergent methods instill a greater probability to generate perfect structures due to few required reactions for layer construction, albeit at lower molecular weights. The potential to locate and connect at a single site within a growing multifunctional monomer diminishes with size and the attendant steric hindrance. Predicated on these features and a comprehensive mass spectrometry analysis, divergent and convergent methods have been compared to polymer and organic syntheses, respectively, by Meijer et al. [13].

As with most other unique areas that attract much attention, descriptive terminology has been developed within the dendritic chemistry community. While much is intuitive, a brief discussion is warranted. The central point from which all branching emanates is described as a core; whereas, the outer surface, or peripheral region, is populated with terminal groups (4; Fig. 1.1). Branching centers define the branching multiplicity based on the number of functional groups or reactive sites that they possess (i.e., 2, 3, or greater) and layers are often referred to as generations to easily denote the number of iterations used in construction. Notably, dendritic void volume is a valuable and useful property and has been employed by many research groups for purposes such as micellar entrapment, host–guest interactions, and catalytic site construction, to mention but a few. This feature has given rise to a new area of study upon which this book is largely based—drug delivery and pharmacological agents using dendritic species.

Figure 1.1 2D and 3D representations of dendritic components.

Branched monomers, or building blocks, used in dendritic construction are now commonly referred to as dendrons, in analogy to synthons in classical organic chemistry. Many dendrimers have been reported [14] using nonbranched monomers; however, their monomers are usually not described as dendrons owing to their linear characteristic. Arising from the convergent protocol, the single reactive site on a multifunctional dendron is described as the focal site. The individual layers of building blocks that comprise dendritic structures are generally denoted as generations, which in turn allow for easy descriptive terminology and a ready understanding of the potential number of surface moieties provided the multiplicity of the core and dendron(s) are known. The concept of dense packing arises from the consideration of increasing numbers of surface groups and a proportionately decreasing amount of available surface area; hence, at some level of construction there will not be enough surface area to accommodate a stoichiometric number of building blocks. This aspect may or may not be problematic and will depend on the desired end characteristics of the material(s) in question.

Ultimately, consideration of dendritic generation leads to the question – structurally, what constitutes a dendrimer? Numerous reports in the literature describe new dendritic species comprising only a single generation. In many cases, a zeroth-generation construct is reported. The importance, elegance, and usefulness of these materials notwithstanding, they are not dendrimers in an historical or idealized sense. They do not possess repeating architectural details at different generations. Therefore, we will herein only describe those materials possessing the attributes of greater than two generations as belonging to a dendrimer family and they must be structurally characterized.

1.1.3 Initial Reduction to Practice

In 1978, a branched covalent molecular architecture was initially reported (Scheme 1.2) by Vögtle et al. [4]. Their scheme represented the first report of a repetitively branched, polyfunctional molecule whereby all to the intermediates were isolated, purified, and substantially characterized in contrast to the traditional synthesis of a polymer whereby only the starting materials and products are isolated and verified. The synthetic protocol utilized Michael-type, nucleophilic amine addition to an electron-poor cyanoalkene followed by reduction of the cyano groups to generate new amine moieties used for further reaction. Thus, for example, amine 5 was treated with acrylonitrile in the presence of glacial acetic acid to give bis-nitrile 6 that was then reduced with NaBH4 and CoCl2·6H2O to afford diamine 7. Repetition of the sequence generated polynitrile 8 and subsequently polyamine 9 possessing 3 tertiary and 4 primary amino moieties. The procedure was also undertaken with diamines such as 2,6-diaminomethylpyridine and diaminoethane to give the corresponding 16 amine constructs.

Scheme 1.2 Vögtle's original cascade preparation of a 1 → 2 N-branched polyamine.

The authors described these new materials as cascade- or nonskid-chain-like owing to the repeating pathway for bond formation and they devised the scheme for the construction of large molecular cavities capable of host–guest interactions. This general procedure was also applied to diaza-monocyclic rings for the construction of polycyclic medium- and large-ring materials.

Approximately 1 year later, in 1979, Denkewalter et al. [8] reported in a patent the construction of high molecular weight materials based on step-wise coupling (Scheme 1.3). This was the first example of dendritic materials construction using a protection–deprotection strategy and a preformed 1 → 2 C-branching center inherent in the building block. Their scheme employed the 4-nitrophenol-activated ester of N,N-bis(tert-butoxycarbonyl)-l-lysine (11), a chiral-protected amino acid, as the dendron. Treatment of an initial diamine 10 with the BOC-protected diamino-activated ester 11 followed by removal of the BOC groups (CF3CO2H) afforded the tetraamine trisamide 12. Repetition of the sequence generated the octaamine 13 and eventually led to dendrimers with theoretically 512 terminal lysine groups corresponding to nine generations. With no characterization reported in the patent, Aharoni et al. [7] subsequently aided in the characterization of these impressive materials by examining the viscosity, photo correlation spectroscopy (PCS), and size exclusion chromatography (SEC). It was concluded that each generation was monodisperse and that these materials behaved as nondraining spheres.

Scheme 1.3 Synthetic method for Denkewalter et al. polylysine dendrimers.

During 1985, Newkome and coworkers [5] published the first example of a 1 → 3 C-branched dendrimer, then termed an arborol for its likeness to tree architecture (specifically, the Leeuwenberg model [15] that branched 1 → 3 in a similar manner to that of tetrahedral, tetravalent carbon) and its terminal alcohol groups. In the same year, Tomalia and coworkers [6] reported their work with 1 → 2 N-branched materials, which they described as starburst dendrimers (derived from the Greek root dendro- for tree-like); these were the first series of polyamines to be prepared in high generation. These two dendritic examples are the first fractal families that were fully characterized.

Newkome's synthesis [5] (Scheme 1.4) began with a polyalcohol (14) that was extended by reaction with chloroacetic acid, under basic conditions, and subsequently esterified to give triester 15. Reduction with LiAlH4 and treatment with tosyl chloride afforded the activated triol 16 that was next reacted with the Na+ salt of methanetricarboxylic triethyl ester (17) to generate the nonaester 19 followed by amidation with tris(hydroxymethyl)aminomethane (18); the resulting 27-alcohol 20 was isolated as a white solid that was freely soluble in water. The requisite extension of the alcohol moieties was necessary due to substitution of the bulky triester nucleophile, which precluded repetition of the scheme, however, this was the first example of dendrons possessing preconstructed 1 → 3 branching centers.

Scheme 1.4 Newkome's preliminary 1 → 3 C-branched dendritic scheme.

Other arborols constructed using these building blocks included the bolaamphiphile, dumbbell-shaped [9]-(CH2)n-[9] and [6]-(CH2)n-[6] series [16, 17], where [9] or [6] denotes the number of hydroxyl groups connected by alkyl chains with n equal to carbons. These materials formed thermally reversible gels upon cooling of aqueous and alcoholic solutions at low concentrations. Gel formation was characterized by electron microscopy and predicated on maximizing lipophilic–lipophilic and hydrophilic–hydrophilic interactions [28]. Arborols [18] constructed with an aromatic benzene core also formed spherical aggregates in solution with diameters of approximately 20 nm and have recently been shown to assemble into large, hollow, spherical motifs [18]. Notably, the globular shape was postulated to be reminiscent of a unimolecular micelle [5].

Tomalia's protocol [6] was similar to that of Vögtle's [4] in that it relied on the reaction of linear monomers and generated branching centers by the Michael-type reaction of electron-poor alkenes with a nucleophilic amine during the construction of successive layers. Thus, in an early example (Scheme 1.5), three equivalents of methyl acrylate (21) were reacted with ammonia to give the triester 22 followed by generation of a new triamine core (24) by treatment with diaminoethane (23). Based on the minimally sterically demanding building blocks, repetition of the sequence to afford hexamine 25 and higher generations was smoothly facilitated. This initial report described the second instance of an iterative synthesis accessing materials up to seven generations.

Scheme 1.5 Tomalia's original dendrimer synthesis based on linear building blocks.

These manuscripts provided the foundation for the burgeoning field that dendritic chemistry is today, however, as it is with most scientific advances, chemistry before and after has played a major role. Thus, advances in the field of macromolecular science by pioneers such as Flory, who reported theoretical [1–3] and experimental [19] evidence for the existence of branched-chain, three-dimensional materials in 1941 and 1942, respectively, began to focus attention on the potential that macromolecules might eventually play in the chemical and materials science arenas. Stockmayer [20] added to the interest by developing equations for branched-chain size distribution and the extent of reaction where a gel, or network, should be formed. Flory [21] later considered the formation of 1 → 2 branched polymers and their scaling properties, notably describing what has now become the well-known area of hyperbranched dendrimers. Along with growing interest in macromolecules during the formative years of polymer chemistry, scientists such as Staudinger [22] postulated that materials like rubber were really high molecular weight polymers and not aggregates of smaller species. Studies by Carothers [23] on condensation polymerizations supported this idea. Lehn [11] subsequently introduced step-wise strategies for the construction of macrocyclic rings in 1973 and later received the Nobel Prize in Chemistry (1987) for work on the host–guest chemistry of designed molecular cavities [24] (e.g., cryptands).

Following these initial reports of Vögtle [4], Newkome [5], Denkwalter [8], and Tomalia [6], research into dendrimer properties and chemistry began to accelerate. Balzani and coworkers [25] introduced metallodendrimers; Hawker and Fréchet [26] developed the convergent protocol; Masamune et al. [27] reported the first preparation of silicon-based dendrimers; de Gennes (Nobel 1991) and Hervet [28] described the first theoretical study of dendrimers; Seebach et al. [29] delineated their work in the preparation of chiral dendrimers; Hudson and Damha [30] described the construction of DNA-based dendrimers; Moore and Xu [31] exploited phenylacetylene chemistry for dendrimer construction; Meijer and de Brabander-van den Berg [32, 33] along with Wörner and Mülhaupt [34] reported, in back-to-back manuscripts, improved procedures for the large-scale preparation of Vögtle-type, polypropylenimine (PPI) dendrimers; Majoral and coworkers [35] reported the first phosphorous-based dendrimers; Zimmerman et al. [36] described the self-assembly of a complex dendrimer based on hydrogen-bonding at the core; and Schlüter et al. [37] reported their work on dendrimerization of a classic polymer framework.

This abbreviated historical account, while not all-inclusive, is intended to give the reader a flavor of the beginnings, or roots, of the current dendritic arena. There are many scientist and researchers, who have contributed to the milestones of dendritic chemistry, which not only strives for new synthetic methods for theoretical and utilitarian applications, but also include an element of artistic style. It is the relative simplicity of design and construction of these complex polyfunctional architectures along with their ease of integration and synergy with other areas of chemistry that affords dendritic chemistry its unique position among materials building blocks. Numerous accounts of the history [38], theory [39], syntheses [40], and applications [38, 41] of dendrimers exist in the literature and it is assumed the reader will pursue their topic of choice; a selected survey is herein presented.

1.2 Supramolecular Perspectives

1.2.1 Unimolecular Micelles—The Advent of the Container

Early reports heralding the potential of dendritic architecture include Newkome and coworkers [42, 43] construction of the first example of a unimolecular micelle (defined in the seminal 1985 report [18]) possessing an all saturated hydrocarbon infrastructure and charged carboxylate surface groups. The unimolecular micelle concept (28) is illustrated in (Fig. 1.2) along with representations of a classical micelle (26) comprising a collection of associated long chain hydrocarbons with polar head groups that are bound together by noncovalent van der Waals- and ionic-based forces and a surface-networked, micellar aggregate 27 accessed from a classical micelle with polymerizable head groups. Surfactant-based, micellar aggregates have been known and used in numerous applications for many years, however, structural dependence on temperature, surfactant concentration, ionic strength, and hydrophilic–hydrophobic environment adds several degrees-of-freedom to their utilitarian considerations. Thus, dendrimer chemistry has provided a means to eliminate or control these aggregate phenomena.

Figure 1.2 Idealized representations of a micelle, a polymerized aggregate, and a unimolecular micelle.

Synthesis of the unimolecular micelle [42] was facilitated by the crafting of 1 → 3 C-branched dendrons (Scheme 1.6) possessing functional groups sufficiently removed (3 CH2 moieties) from the quaternary branching center to allow for smooth end group transformation [44]. Beginning with the Michael-type addition of acrylonitrile to nitromethane to generate a nitrotrinitrile, followed by hydrolysis of the nitrile groups to carboxylic acids and subsequent reduction to the corresponding alcohols, the nitrotriol 29 was obtained. Relying on electron-transfer and free-radical chemistry developed by Newkome et al. [45], Ono et al. [46], and Geise et al. [47], a novel route to the synthesis of quaternary carbon centers was developed. This new method allowed the preparation of dendrons with differentiated functionality in contrast to a 17 step synthesis reported by Rice et al. [48] leading to a similar framework with identical termini (i.e., tetrabromide 32).

Scheme 1.6 Synthesis of 1 → 3 C-branched, hydrocarbon-based dendrons.

Thus, benzyl protection of the alcohol groups in triol 29 allowed radical initiated substitution of the nitro group with acrylonitrile (AIBN, toluene, n-Bu4SnH) to give the mononitrile trisbenzyl ether 30. Reaction with (1) KOH, (2) BH3-THF, (3) SOCl2, and lithium acetylide–TMEDA complex then afforded the desired terminal alkyne 31; whereas, treatment of an intermediate monoalcohol trisbenzyl ether with HBr in H2SO4 gave the starting tetrabromide core 32.

Reaction of the monoalkyne with the tetrabromide (LDA, TMEDA, and HMPA) followed by concomitant Pd-C-mediated benzyl ether hydrogenation and alkyne reduction afforded the first-generation 12 alcohol construct 33. Subsequent bromination (HBr, H2SO4) and treatment with more of the alkyne dendron (LDA, TMEDA) gave the second-generation, benzyl-protected alcohol dendrimer that was reduced (Pd-C, H2), oxidized (RuO4), and treated with tetramethyl ammonium hydroxide to give the 36 tetramethyl ammonium carboxylate 34 (Scheme 1.7). This dendrimer was described as a [8²·3] micellanoate, where 8²·3 represents two generations of an eight carbon spacer with 1 → 3 branching.

Scheme 1.7 Newkome's synthesis of a unimolecular micelle with a saturated hydrocarbon infrastructure.

Dendrimer aggregation promoted in solution by carboxylic acid H-bonding was inhibited by ion exchange to the tetraalkylammonium carboxylate as evidenced in the observed 30 Å diameters in electron micrographs of 34 that compared favorably to the calculated values [43]. Fluorescence lifetime and anisotropy decay values obtained by phase resolved anisotropy experiments with diphenylhexatriene (DPH) as a molecular probe were similar to that observed with DPH in phosphatidylcholine vesicles demonstrating the micellar host–guest relationship in an aqueous environment [43]. Other molecular probes used to explore the micellar properties of these dendrimers include chlortetracycline (fluorescence), phenol blue, naphthalene (UV absorbance), and pinacyanoyl chloride (color change).

Newkome and coworkers [49] also reported the construction of the unique dendron tetraacid core [50] 35 (prepared by reaction of pentaerythritol and acrylonitrile followed by hydrolysis) and aminotriester [51] 36 (accessed by Michael-type addition of tert-butyl acrylate to nitromethane; commonly referred to as Behera's amine [51] in honor of Prof. Rajani K. Behera who, while working in Prof. Newkome's laboratories, first prepared and used this material) that were employed for the construction of a series of amide-based dendrimers [52, 53]. The stable isocyanate of Behera's amine has been used for facile combinatorial surface modification [54, 55]. Coupling of the amine under standard peptide conditions (DCC, 1-HOBT, DMF) afforded the first-generation, 12 tert-butyl ester dendrimer 37 (Scheme 1.8). Liberation of the carboxylic acid surface groups (HCO2H) generated the new acid-terminated periphery 38 that could be treated with more aminotriester. Dendrimers in this family were prepared and isolated through generations 1 to 5 corresponding to 12, 36 (i.e., 39), 108, 324, and 972 theoretical terminal groups.

Scheme 1.8 1 → 3 C-branched dendrimers constructed using amide-connectivity.

Initial studies [56] of these amide-based dendrimers involved the systematic evaluation of the pH effect on the hydrodynamic radii using two-dimensional, diffusion-ordered NMR spectroscopy (DOSY NMR). Termination of the carboxylic acid series with dendrons crafted to incorporate amine and hydroxylated surfaces [56] (Fig. 1.3; 40 and 41, respectively) generated the complementary basic and neutral surfaces, respectively. Accordingly, the acid-terminated dendrimers were found to be largest or in an expanded state in neutral and basic pH; whereas, the amine-terminated species exhibited contraction in basic media; the hydroxyl-terminated constructs showed a constant hydrodynamic radius over the range from basic to acidic pH. A study of dendrimer expansion and contraction based on ionic strength has also been reported [57].

Figure 1.3 1 → 3 C-Branched dendrons for the incorporation of amine and hydroxylated surfaces.

Kuzdzal and coworkers [58, 59] have used these acid-terminated dendrimers based on Behera's amine as a micellar substitute for the pseudostationary phase in electrokinetic capillary chromatography for the separation of a series of parabens. Tanaka et al. [60] were the first to report the use of dendrimers in electrokinetic chromatography, and Muijselaar et al. [61], have also investigated this dendritic property. Newkome et al. [62] further reported the incorporation of H-bonding sites on the arms of the interior dendritic framework; the encapsulation of AZT based on complementary H-bonding was achieved. The micellar properties were also employed by Miller et al. [63] to construct an electronic nose, whereby selective dendritic encapsulation of organic solvents provided a means of detection.

Meijer and coworkers [64] studied extensively the polypropylenimine (PPI) dendrimers and reported the dendritic box, whereby the amine termini were capped with the activated ester of a Boc-protected chiral amino acid (42) to generate sterically demanding surface that traps molecular guests (Scheme 1.9; 42). Molecular probes used to investigate entrapment include 3-carboxypropyl radical 43, tetracyanoquinodimethane (TCNQ) 44, and Rose bengal 45 along with the corresponding analytical techniques of EPR, UV, and fluorescence spectra, respectively. Notably, the diffusion of guest molecules after being locked in was unmeasurable.

Scheme 1.9 Topological trapping of guest molecules in a Molecular Box.

Zimmerman et al. [65, 66] have provided the first example of dendrimer construction employing self-assembly based on H-bonding of isophthalic acid moieties attached at the focal positions (Scheme 1.10) of Fréchet-type dendrons [26]. Synthesis of the requisite dendrons (generations 1 to 4) began with conversion of pyridine dibromide 46 to the bis-boronic acid that was then transformed to the bis(dimethyl isophthalate) using aryl iodide Pd(0) coupling. Attachment of the dendritic wedge to the phenolic position of 47 was accomplished by KOH-promoted substitution at the focal benzylic bromide 48 to give the poly(isophthalic acid) substituted dendron 49. Self-assembly into ordered hexameric aggregates (i.e., 50) was studied by SEC, VPO, and LLS. Molecular weights determined by SEC retention times using polystyrene for calibration were in agreement with NMR data that showed monomeric species in THF and hexameric structure in noncompeting CH2Cl2 solution. However, observed SEC traces for the lower generation dendrons suggested a greater percentage existed in linear and dimeric forms due to less steric pressure to form the hexameric species.

Scheme 1.10 Self-assembly of dendrimer architecture based on H-bonding.

Zimmerman and coworkers [67] have explored the potential to use chromogenically modified, peripherally cross-linked dendrons as amine chemosensors. Their strategy involved coupling Fréchet-type dendrons modified at the termini with alkene groups (either homoallyl or allyl ether moieties) attached to the phenolic positions of a trifluoroacetylazo dye that has been shown to be an amine chemosensor, based on its ability to trap amines by reaction with the trifluoroacetyl unit, thereby exhibiting a 50 nm shift in λmax in the visible region from red-orange to yellow (λ = 475–425 nm for uncomplexed to complexed, respectively). The requisite sensor-modified dendrons (Scheme 1.11) were prepared by standard Fréchet-based attachment (KF promoted coupling with 18-crown-6) of two dendrons to the dye. Treatment of two equivalents of the dendron-substituted dye 51 with butane 1,4-diazide in the presence of triphenylphosphine produced the bis-imino didendron 52. Metathesis with Grubbs catalyst then effected the surface cross-linking (Scheme 1.12) to give the encapsulated amine active site 53. Treatment with aqueous HCl transformed the trifluoroimine moieties to the starting trifluoroacetyl groups 54. Extensive systematic host–guest studies using these unique materials with a library of amines and alcohols revealed the selective signaling of certain diamines, although it was determined not to arise due to template-mediated imprinting.

Scheme 1.11 Coupling of dendrimerized chromogenic sites as chemosensors.

Scheme 1.12 Idealized representation of matrix encapsulated, chemosensor site for diamine recognition.

Zimmerman and coworkers [68] have used dendrimer surface cross-linking based on Grubbs-promoted alkene metathesis for the modification of nanoparticles. They showed that the degree of dendrimer cross-linking can be controlled, thereby leading to nanoparticles with predictable rigidity. Control over cross-linking has also been examined by internal placement of the alkene moieties [65]. The distribution of alkene cross-linking placement between subunits on 1 → 2 branched, aryl ether dendrons has also been studied [69] along with the reversibility of dendrimer metathesis [70] and cross-linked arylether dendrimers with arylester cores have been hydrolytically decored without significant degradation [71].

Zimmerman and coworkers [66] initially reported the use of dendrimer-based surface metathesis chemistry in concert with the potential to co-facially connect and linearly arrange porphyrin moieties with the goal of creating new organic nanotubes (Scheme 1.13). Treatment of dendrimerized, Sn-metallated porphyrin 55, with succinic acid and excess Ag2O generated the oligomerized dendrimer 56; it was noted that successful oligomerization hinged on reaction mixture concentration by solvent evaporation during the transformation. Notably, the oligomer was treated with Grubbs catalyst without delay due to an observed increase in molecular weight over time. Following metathesis, the newly formed rod 57 was reacted with NaOMe to liberate the porphyrin core and generate the decored organic nanotube 58. SEC comparison of the hollow constructs to polystyrene and dendrimer standards revealed tetramer and dodecamer formation with corresponding molecular weights of 23,000 and 72,600 Da, respectively.

Scheme 1.13 A novel adaptation of self-assembly for the construction of organic nanotubes.

Percec et al. [40, 72] investigated dendrimer supramolecular self-assembly by constructing a library of conical dendrons, whereby the focal groups embed into the central core of spherical motifs that can be envisioned as the dendritic equivalent of a surfactant based micelle. Porous columns were also obtained. The dendritic library was constructed using 1 → 2 arylether, Fréchet-type synthesis with C4 to C12 chiral or achiral carbon chains attached to the periphery with ester, alcohol, or dipeptides as focal units. The self-assembly process is exemplified in Scheme 1.14 where the benzyl alcohol dendron 59 with C6 alkyl chain termini assembles into a hollow sphere 60 with an 83.5 Å diameter and core diameter of 26.4 ± 4 Å. Further assembly based on spherical packing into a Pm3n cubic lattice 61 was determined. Whereas, the starting dendrons were fully characterized by NMR, HPLC, and MALDI-TOF; the supramolecular assemblies were analyzed by small-angle X-ray diffraction and DSC. Reconstruction of the electron density maps afforded three-dimensional mapping of the hollow cubic phases showing the electron density profiles; aliphatic and aromatic regions were clearly discernible. Low temperature TEM imaging combined with electron diffraction also revealed circular objects arranged in square lattices. A dendritic crown derived from dendron-modified cyclotriveratrylenes [73], semifluorinated, Janus-type, dendritic benzamides that form bilayer pyramidal columns [74], dendritic crown ethers [75], dendronized poly(carbazoles) [76], dendritic dipeptides [77], dendronized polyphenylacetylenes [78], π-stacked, semifluorinated dendrons [79], and thixotropic dendritic organogelators [80] have also been studied; a comprehensive review [40] is available.

Scheme 1.14 Using dendron shape for the assembly of complex architectures.

Hirsch and coworkers [81] have reported the switchable supramolecular assembly (Fig. 1.4) based on an amphiphilic fullerene possessing Newkome-type, carboxylic acid-terminated dendrons. In contrast to an amphiphilic fullerene 62 with ester-based dendron connectivity and in comparison to the calixarene-cored analog [82] 63, the dendronized fullerene 64 has been shown to exhibit globular micellar character at basic pH and predominately rod-shaped character at near neutral pH. Using electron cryogenic microscopic data (cryo-TEM), a three-dimensional reconstruction of the pseudospherical shape was obtained. The globular motif (diameter = 85 ± 5Å) was modeled as a packed aggregate of eight molecules of 64 in a C2-symmetrical arrangement consisting of two interlocked, U-shaped species, each consisting of four molecules, with 180° opposing domes, essentially capping one another, and perpendicular planes. At a lower pH of 7.2, rod-shaped, double-layered aggregation was observed in electron micrographs (diameter = 65 ± 5Å with variable lengths). The spherical aggregation possesses the expected attributes of a micellar structure in that as arranged all of the hydrophobic alkyl chains are shielded from the aqueous environment and aid in the structural stability by solubilizing each other on the interior of the superstructure. A later report discussed the control of self-assembly of the structurally persistent micelles by specific-ion effects and hydrophobic guests, while a dendronized fullerene and porphyrin hybrid have also been studied with respect to their electrostatic attraction and resulting 1.1 μs charge separated state [83] and their efficient light harvesting and charge-transfer character [84].

Figure 1.4 Dendronized fullerenes that exhibit pH switchable, globular micellar aggregation and linear, rod-shaped assembly.

Hirsch and coworkers [85] have also reported the dendronization with Newkome-type dendrons of perylene bis-imides that exhibit electronic communication with graphene in solution or following surface deposition. Noncovalent π-system binding provided the association and facilitated the interaction. Perylene dendronization and its utility as a rigid spacer for terpyridine-based metal connectivity used in metallosupramolecular self-assembly have also been reported by Würthner and coworkers [86–89].

1.2.2 Framework Conformational Control

Supramolecular self-assembly predicated on complementary H-bonding interactions has also been investigated by Hirsch and coworkers [90]. Employing a 2,2′-bipyridinyl-4,4′-dicarboxylic acid as the starting point for metal-centered core construction (Scheme 1.15), a 1 → 2 aryl branched, bis(2,6-diamidopyridine) receptor 65 was attached by standard coupling procedures to diacid 66 to generate the bipyridine dendron 67. Reaction with RuCl3 afforded the trisbipyridine Ru(II) core that was subsequently coordinated at the binding sites with cyanuric acid-modified dendrons that introduce potential repetition for continued dendritic growth, chirality, or electronic possibilities (68). Generation of the final, poly(H-bonded), Ru complex was accomplished both by sequential addition or in a single-step reaction. The complementary H-bonding was monitored by NMR titration and determination of the stepwise formation constants K1–K6.

Scheme 1.15 An innovative use of H-bonding for dendritic construction.

Parquette and Huang [91] demonstrated conformational restriction of the dendritic framework based on the incorporation of internal H-bonding. Their sequence for convergent dendron construction focused on 4-chloropyridine-2,6-dicarbonyl chloride (Scheme 1.16; 69, obtained from the treatment of chelidamic acid with POCl3), which was initially treated with dodecyl anthranilate to generate the terminal species 70. Subsequent reaction with NaN3, followed by hydrogenation (H2, Pd-C), and treatment with more diacid dichloride 69 afforded the second-generation dendron 71; repetition gave generation 4 72. X-ray structure determination of the second-generation dendron revealed a monoclinic asymmetric unit exhibiting a P21/c space group and a propeller-like secondary motif resulting from symmetrical assembly of entwined dimers in the solid state. Along with the X-ray data, NMR and IR spectra further evidenced the pyridine N–H amide structure; distances ranged from 2.13 to 2.33 Å.

Scheme 1.16 H-bonding-based conformational framework restriction.

Parquette and coworkers [92] have also investigated H-bond-based dendrons using a 2-methoxyisophthalamide moiety that were designated as 2-OMe-IPA (i.e., 76) and compared it to the corresponding 2,6-dicarboxamidopyridine-based dendrons that were designated as 2,6-Pydic (i.e., 77). The 2-OMe-IPA building blocks were prepared (Scheme 1.17) starting with 2,6-dimethylanisole, which was sequentially oxidized (KMnO4), nitrated (HNO3, H2SO4), and carbonylated [(COCl)2] to give the nitrobis(carbonyl chloride) 73. Introduction of the capping species, provided by the tetraethyleneglycol ester of anthranilic acid 74, afforded the first-generation dendron 75. Two iterative treatments with SnCl2 for reduction of the aryl nitro group followed by reaction with the bis(acid chloride) 73 gave the third-generation dendron 76. The 2,6-pydic-based materials were accessed using 4-chloropyridine-2,6-dicarbonyl chloride. Both materials were evaluated computationally with respect to the energy requirements for syn–syn, syn–anti, and anti–anti conformations. Although subtle differences were found, both systems exhibited organization, based on the syn–syn conformations; thus, the carboxamide protons were oriented inward. This preference in 2-OMe-IPA was attributed entirely to H-bonding; whereas, dipole moment minimization effects also played a part in the 2,6-pydic dendrons. Although it was noted that solvophobic compression was deemed to be a more important effect on hydrodynamic properties than solvent-based shifts in repeat unit conformational equilibria for both series.

Scheme 1.17 Both H-bonding systems were determined to exist in syn–syn conformations.

Using the 2,6-dicarboxamidopyridine-based protocol for dendron construction and oxazoline capping groups, Preston et al. [93] described the synthesis and properties of folded metallodendrons exhibiting shell-selectivity toward metal coordination. Circular dichroism and X-ray studies verified the selectivity and confirmed the helical properties of these dendrons. This offers added flexibility to design and construct dendrimers with redox potential gradients and the ability to fine tune dendritic electronic materials and components. Conformational properties of the folded metallodendrons have also been reported [94].

2,6-Dicarboxamidopyridine-type dendrons integrated with alanine-based oligomers (denoted as peptide–dendron hybrids, PDHs) have been reported [95] to undergo a reversible conversion from an amyloid fibrillar structure to a nanotube assembly in water effected by a change in pH or ionic strength; the phenomena was exploited for the encapsulation and release of a dye (Nile Red) upon a pH change from high to low. Dendrons crafted with these unique building blocks have been used as catalysts for Aldol reactions [95] where the dendritic effects on stereoselectivity were analyzed.

1.2.3 Harnessing Electronic Properties

Nierengarten and coworkers [96] have described the self-assembly of fullerodendrimers using the four-fold H-bonding specie 2-ureido-4-[1H]pyrimidinone. Synthesis of the requisite building block (Scheme 1.18) was accomplished [97, 98] beginning with the alkylation of 3,5-dihydroxybenzyl alcohol using 1-bromohexadecane to give the dialkylated benzyl alcohol 78 that was subsequently treated with Meldrum's acid affording the malonic acid 79. Coupling with the di(benzyl alcohol) 80 (accessed from dimethyl 5-hydroxyisophthalate in two steps [98] using DIBAL reduction, followed by reaction with tert-butyl bromoacetate) gave the bis-malonate 81, which when reacted (I2, DBU) with C60, generated the tert-butyl-protected dendron 82; deprotection with CF3CO2H afforded the carboxylic acid-modified, fullerodendron 83. Reaction with the tetraalcohol dendron 84 (prepared [99] in an analogous procedure to that of 82, whereby 2-bromo-6-benzyloxyhexane is used in place of 1-bromohexadecane for alkylation of 3,5-dihydroxybenzyl alcohol, followed by debenzylation) with the focal acid moieties of 83 gave the pentafullerodendron 85. Transformation of the focal tert-butyl ester to the corresponding acid [CF3CO2H (TFA)], coupling (DCC, DMAP) to BOC-protected, 2-amino-6-(4-hydroxybutyl)-[1H]pyrimidine-4-one (86), followed by amine liberation (TFA) and alkylation with octylisocyanate (H17C8NCO) produced the desired 2-ureido-pyrimidone-modified, dodecaC60, fullerodendron dimer (87). Proof-of-dimerization was verified by mass spectrometry, albeit in low abundance (5%), and NMR, which clearly revealed the large downfield shifts corresponding to the relevant protons positioned between the H-bonding units (i.e., ureaNHs, 11.81 and 10.06 ppm; pyrimidoneNH, 13.23 ppm). The corresponding smaller diC60dimer was also prepared. These materials demonstrated the potential to craft novel supramolecular architectures exhibiting fullerene-based photoinitiated properties. Mass spectrometry, along with electrochemical analysis (CV) of fullerodendrons possessing 3, 5, and 7 C60 moieties, revealed independent redox behavior of the methanofullerene groups [99]. A review on the use of [61] fullerenes as photoactive cores for dendrimers is available [100].

Scheme 1.18 Supramolecular self-assembly of a fullerodendrimer.

Connection of the pentaC60 dendrons (85) acid focal group followed by coupling to third-generation, PEG-terminated, Fréchet-type dendron generated amphiphilic diblock dendrimers that were examined [97, 101] for their ability to form Langmuir and Langmuir–Blodgett films. The potential to form ordered films arises largely from the hydrophilic and hydrophobic peripheral chains on the opposing hemispheres of the dendrimers and also provided structural attributes to facilitate efficient transfer to a multilayer Langmuir–Blodgett array. Langmuir films have also been prepared by attachment of these methanofullerenes to bipyridine, followed by generation of a tris(2,2′-bipyridine)Ru(II) complex to act as a polar head group [102]. Ruthenium(II) complexes attached to C60 through polyethylene glycol units, based on bipyridine and terpyridine, have been reported as well [103]. A bis-phenanthrolene–Cu(I) complex used as a tetradirectional core with grafted G1 through G3, C60-based, dendrons has been reported [104]; the encapsulation of fullerene-modified dendritic frameworks was shown to isolate the central complex from electrode oxidation and electrochemical oxidation was not observed for the G2 and G3 constructs, leading to a dendritic black box description. Additional reviews are available regarding the supramolecular and photophysical aspects of fullerene-rich dendrimers [105, 106].

Ceroni and coworkers have investigated the electronic and excited state attributes of fullerene-modified phenyleneethynylene dendrons characterized by 1 → 2 aryl branching employing either a 1,2,4- or 1,3,5-substitution pattern (88–91; Fig. 1.5). Synthesis of these phenylacetylenes is exemplified by the preparation of the 1,3,4-motif 89, which began with Corey–Fuchs dibromoolefination (CBr4, PPh3, Zn) of 3,4-dibromobenzaldehyde followed by transformation to the alkyne (LDA) and trapping with triethylsilyl chloride (TESCl) to give the protected dibromoalkyne 92 (Scheme 1.19). Sonogashira coupling [Pd(PPh3)2Cl2, Cu(I)] of the capping agent 93 then gave the silated alkyne 94 which was next deprotected (TBAF) to afford the terminal alkyne 95. Repetitive coupling with the Sonogashira reagents and the starting dibromoaldehyde generated the G2 dendron 96. Treatment of the aldehyde moiety with N-methylglycine to give an intermediate azomethine ylide facilitated reaction with fullerene to afford the desired C60-modified dendron 89. Whereas all these dendron hybrids were shown to facilitate ultrafast energy transfer, the 1,2,4- versus 1,3,5-branching motifs showed dramatic differences in their absorption and emission spectra with the former pattern exhibiting a lower absorption commencement and a broadened profile relative to the latter pattern; thus, the 1,2,4-architecture was revealed to possess enhanced light-harvesting potential.

Figure 1.5 Different branching patterns have explored for optimum energy transfer.

Scheme 1.19 Synthesis of a phenylacetylene-based dendron.

Fréchet et al. [107] synthesized a multichromophoric light-harvesting dendrimer possessing two complementary donor dyes. The donors absorb a broad spectrum of light energy in the UV and visible regions as well as facilitate red region emission by a porphyrin core through a florescence resonance energy transfer mechanism. Dyes chosen as chromophores include the carboxylic acid-modified, naphthopyranone [Scheme 1.20; prepared by Peckmann condensation of 2,7-dihydroxynaphthalene and ethyl trifluoroacetoacetate, followed by reaction with tert-butyl bromoacetate (K2CO3) with subsequent acid-mediated, tert-butyl group removal] and commercially available coumarin 3-carboxylic acid. The dye-functionalized, 1 → 3 C-branched building blocks were derived by carbodiimide-based coupling (EDC with 1-HOBT being used in the case of the naphthopyranone) of each dye to 5-amino-5-hydroxymethyl-2,2-dimethyl-1,3-dioxane [108] (97) [prepared by treatment of tris(hydroxymethyl)aminomethane (TRIS) with commercially available 2,2-dimethoxypropane] to generate the intermediate alcohols 98 and 99, followed treatment with succinic anhydride, in the presence of DMAP to give the desired functional dendrons 100 and 101.

Scheme 1.20 Dye incorporation into dendritic building blocks.

Dendritic preparation proceeded (Scheme 1.21) by polystyrene-carbodiimide coupling of the hydroxyl-terminated porphyrin 102 with the naphthopyranone building block 100 in the presence of pyridine to generate the first-generation polyol 103 after removal of the acetonide protecting moieties with (NH4)2Ce(NO3)2. Attachment of the coumarin-based, dye 101 layer, thereby generating the 24 dye-modified dendrimer 104, was conducted similarly to the octa(naphthopyranone) dye dendrimer. Fluorescence spectra recorded after excitation at 335 or 358 nm (coumarin and naphthopyranone absorbance values, respectively) reveal predominant emission at 651 and 717 nm corresponding only to the porphyrin core. This modular approach to dendritic construction is a notable example of framework modification employing preconstructed, application-oriented dendrons, thereby instilling a gradient property characteristic, in this case, light absorption.

Scheme 1.21 Construction of a dye-based gradient within a dendritic framework.

1.2.4 Drug Delivery Systems

Fréchet and coworkers [109] have employed dendrimer–polymer hybrids as micellar capsules for event-triggered drug release. Dendrimer assembly (Scheme 1.22) relied on the use of benzylidene-protected, 2,2-bis(hydroxymethyl)propionic anhydride 105 (prepared [110] by treatment of 2,2-bis(hydroxymethyl)propionic acid with benzaldehyde dimethyl acetal, followed by coupling and dehydration with DCC), as the dendritic building block, followed by Pd-C hydrogenolysis of the resulting polybenzylidene surface to give a new polyhydroxy periphery that could be further elaborated. Using this protocol and starting with the polyethyleneoxide (PEO) core 106, the PEO–dendrimer composite 107 was generated, which was next capped with diazonaphthoquinone to afford a dendritic framework (108) that was able to undergo Wolff rearrangement to an indene carboxylic acid, upon irradiation with UV light; hence, the transformation would change the micellar character and disrupt the aggregation. Since Nile Red excimer fluorescence could be used to follow micelle formation in the noncapped PEO-dendrimer hybrids, it was reasoned that the capped hybrid upon irradiation and rearrangement would provide a release mechanism useful for drug delivery. Upon irradiation at 355 nm and termini rearrangement, the resulting indene carboxylic moieties changed the dendritic character from hydrophobic to hydrophilic, disrupted micellar formation, and released the encapsulated dye as revealed using fluorescence emission studies and dynamic light scattering experiments.

Scheme 1.22 Surface modification able to undergo light-induced degradation.

Gillies and Fréchet [111] have also used a micelle disruption mechanism for the controlled release of doxorubicin. As an example, the micelle-forming, diblock copolymer 109 (Scheme 1.23) comprised a polyethylene chain (~ 10,000 MW) and a third-generation dendron terminated with 2,4,6-trimethoxybenzaldehyde acetal moieties (prepared [110, 112–116] by iterative reaction of hydroxyl terminal groups with benzylidene-protected, 2,2-bis(hydroxymethyl)propionic anhydride, treatment with an aminodiol and subsequent acetal formation) was subjected to a pH decrease. Following the pH change, the surface acetals begin to hydrolyze and generate a polar protic dendron surface 110 that changes the polar character of the micelle thereby releasing the doxorubicin. Several variations of these polyester dendrimers have been reported [117–120]. A marvelous example of dendritic utility is realized in the report [121] of one dose of a doxorubicin-modified dendrimer acting to cure mice inflicted with C-26 colon tumors.

Scheme 1.23 Dendrimer-based micelles modified for acid-promoted drug release.

1.2.5 Dendritic Self-Assembly, Sensors, and Devices

Chow and coworkers [122] have crafted 1 → 2 C-branched, all hydrocarbon dendrons, focally attached to bis(2-ureido-4-pyridinone) (UPy)2, which can self-assemble to form novel dendronized polymers [123]. Synthesis of the aliphatic dendrons (Scheme 1.24) began by diallylation of diethyl malonate with dimethyl allyl chloride 111 followed by Pd(OH)2-mediated reduction of the alkene moieties to give the dialkylated malonic ester 112. Subsequent saponification (KOH), decarboxylation (pyridine, H2O, heat), and reduction (LAH), followed by PCC oxidation, bis(trimethoxy)phosphonoacetate homologation, and finally DIBAL reduction afforded the first-generation allylic alcohol 113. Reaction of 113 with Meldrum's acid using Shin's modification [124] of the Mitsunobu reaction [PPh3, diisopropylazodicarboxylate, (DIAD)] gave predominantly the C-alkylated product, which was then subjected to the same sequence beginning with the Pd(OH)2-mediated reduction of the alkene moieties and ending with the DIBAL reduction to generate the second tier dendron 114; the third-generation construct 115 was obtain analogously stopping at the focal carboxylic acid, prior to the homologation steps.

Scheme 1.24 Synthetic protocol for 1 → 2 C-branched, aliphatic dendrons.

Construction of the H-bonding, dimeric (2-ureido-4-pyridinone) units with all aliphatic dendrons [122], exemplified using the third-generation dendron, (Scheme 1.25) was achieved by coupling [(PPh3)2PdCl2, Cu(I), NEt3] the alkyne-modified dendron 116, prepared by transforming the corresponding focal aldehyde obtained from 115 (Scheme 1.24) with vinyl dibromide (CBr4, PPh3, K2CO3), followed by n-BuLi elimination using the Corey–Fuchs method to ditriflate 117 (accessed by treatment of the corresponding diol with trifluoromethylacetic anhydride) affording the dendronized aryl diester 118. Reduction of the alkyne moieties (H2, Pd-C) gave the alkyl-modified diester 119 that was subsequently saponified to the diacid 120 and treated with diphenyl phosphorazidate (DPPA) and NEt3 to produce, via a Curtius rearrangement, the diisocyanate 121 in situ.

Scheme 1.25 Construction of a dendritic monomer for use in polymeric, H-bonding, self-assembly.

Attempts to obtain the desired di(2-ureido-4-pyridinone) by reaction with 6-methylisocytosine proved difficult due to solubility, reduced nucleophilic character of the aryl amine, and decomposition when forcing conditions were employed. Use of an O-benzyl protected, 2-amino-6-methylpryimidine, a method for the synthesis of 2-ureido-4-pyridinones reported by Meijer [125], allowed the desired transformations. Thus, treatment of commercially available 2-amino-4-chloro-6-methylpyrimidine with benzyl alcohol in the presence of base (NaH) gave the O-benzyl-protected aminopyrimidine 122 (Scheme 1.26), which was subsequently reacted with the dendronized diisocyanate 121 to afford the dibenzyl ether 123. Hydrogenolysis [Pd(OH)2, H2] to remove the benzyl groups generate the H-bonded motif, and the desired supramolecular polymer 124 was effected under dilute conditions to enable the removal of the catalyst and avoid product precipitation. Notably, upon isolation of the dendronized polymers, the first- and second-generation materials were difficult to resolubilize with the second-generation construct being described as devoid of solubility in any solvents. However, the third-generation polymer exhibited excellent characteristics and was highly soluble in nonpolar solvents. For this third-generation, supramolecular polymer 124, the specific viscosity showed a nonlinear increase with increasing concentration; above 26 mM, a very strong monomer association was observed—a behavior consistent with other 2-ureido-4-pyridinone-based polymers [125]. The UV-vis absorption maximum of the polymer exhibited a significant bathochromic shift with increasing polymer concentration; thus, the fibers formed by spin coating and imaged by SEM were described as J-type aggregates (i.e., end-to-end arrangement and narrow, red-shifted absorption peak). A model proposed for fiber assembly suggests adjacent, linear H-bonded arrays with alternating dendritic wedge interdigitation.

Scheme 1.26 Generation of a H-bonding-based, supramolecular polymer with a dendritic, all-hydrocarbon coating.

An interesting use of branched architecture has been reported by Tang and coworkers [126], based on the construction of materials that possess high molecular compressibility and exhibit aggregation-induced emission (AIE). These polymer nanoaggregates were shown to detect explosive material with a superamplification effect. Theoretical and experimental evidence has shown the main cause of AIE to arise from restriction of intermolecular rotation of the phenyl moieties in such propeller-shaped molecules as tetraphenylethylene and hexaphenylsilole; thus, they are nonemissive when in the dissolved state and emit efficiently in the aggregate state.

Construction of the hyperbranched polymers was envisioned to incorporate spring-like flexibility between the triazine-based sites of connectivity, thereby introducing an element of compressibility, which was