Dendrimer/inorganic nanomaterial composites: Tailoring preparation ...

SCIENCE CHINA Chemistry • REVIEWS • · SPECIAL ISSUE · In Honor of the 80th Birthdays of Professors SHEN JiaCong, SHEN ZhiQuan and ZHUO RenXi

February 2011 Vol.54 No.2: 286–301 doi: 10.1007/s11426-010-4205-7

Dendrimer/inorganic nanomaterial composites: Tailoring preparation, properties, functions, and applications of inorganic nanomaterials with dendritic architectures ZHAO FuGang & LI WeiShi* Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032, China Received October 31, 2010; accepted November 20, 2010

Inorganic nanomaterials have a variety of fascinating properties and a wide range of promising applications. However, they often suffer from instability and poor processibility. To solve it, dendrimers, a special family of macromolecules having a unique three-dimensional architecture, provide one of the excellent solutions. In addition, the site-selective functionalization of the specific elements in the dendritic structure endows the nanohybrid system new functions and applications. Inspired by such ideas, a variety of dendrimer/inorganic nanomaterial composites have been designed and exploited. This review article selects a number of representative examples, and illustrates their preparation, characterization, properties, and applications. The influence and the unique features that originate from the introduced dendritic structures are particularly discussed. inorganic nanomaterials, nanoparticles, dendrimers, nanocomposites, carbon nanotubes

1

Introduction

Inorganic nanomaterials, including metal nanoclusters and inorganic semiconductor quantum dots, possess a variety of fascinating properties that their bulk materials do not have [1]. For example, all nanomaterials have an extremely high ratio of surface area to volume. Metal and semiconductor nanoparticles have special light absorption and emission properties that originate from surface plasmon resonance. These intriguing properties make them highly attractive for the next generation of (opto)electronics, sensors, catalysts, and biomedical applications. However, due to their high surface area to volume, naked nanomaterials are kinetically unstable and have a strong tendency to agglomerate. Therefore, how to make their size stable on the nanometer-scale, as well as how to finely tune their size, has become one of the central research topics in the field of nanotechnology. *Corresponding author (email: [email protected]) © Science China Press and Springer-Verlag Berlin Heidelberg 2011

To date, a variety of strategies, such as electrostatic stabilization using columbic repulsion, steric stabilization using a protecting shell or a viscous polymer media, anchoring to solid supports, have been demonstrated [2]. Among which, the approach to build a protecting shell has been considered as a practical and versatile method because of its many merits as compared with other ones. For example, one can control over the morphology and the size of nanomaterials, finely tune their solubility or dispersibility, and enable new functions and applications, just by simply modifying the protecting agents. Although the commonly used protecting agents are organic small compounds and polymers, this review article does not intend to cover all of them but focuses on a special family of macromolecules called dendrimers. Dendrimers are monodispersed macromolecules having a tree-like morphology with a regularly branched, three-dimensional architecture (Figure 1) [3–5]. A dendrimer is generally composed of a core, a number of repetitive branching units, and a plenty of surface groups. The number of the repeating layer of branched units from the core to the periphery is chem.scichina.com

www.springerlink.com

Zhao FG, et al. Sci China Chem

Figure 1 Schematic structural illustration of a dendrimer and a dendron.

called the generation number. The first compound having dendritic structure was reported by Vögtle and coworkers in 1978 [6]. Several years later, Tomalia developed famous poly(amidoamine) dendrimers (PAMAM) and first used the phrase “dendrimer” to describe this special family of macromolecules [7]. The convergent synthetic methodology inverted by Fréchet et al. in the early of the 1990s [8] has greatly promoted the development of dendrimer science since it allows the preparation of defect-free dendrimers. To date, a variety of dendritic structures including three famous types: PAMAMs, Fréchet-type poly(aryl ether)s (PAEs), and poly(propyleneimine)s (PPIs) can be often seen from literature. One of the important merits of dendrimers is their threedimensional architecture, which endows themselves a significantly important function: encapsulation [9]. Model simulation revealed that small dendrimers have open structure, while larger dendrimers (G > 3) have a compact structure with a spherical or cylinder shape depending on the elongation of the core [10]. When a functional unit is embedded inside a large dendrimer, the large different local environment from outside and its accessibility may have significant effect on its properties. Furthermore, its solubility is predetermined by the nature of the terminal groups of the dendritic structure. With the same functional group, one can realize different versions with completely different solubility just simply modification on the terminal groups of the dendrimer. The encapsulation function of a dendritic structure is very attractive for researchers in the field. Aida, Jiang and Li have applied this function to overcome the shortages of conjugated polymers and oligomers that are strong interchain interactions, low solubility, and deactivation of photoexcited state [11]. They attached two thirdgeneration dendritic wedges to a monomer having ethynylene units at both ends. Upon Glaser oxidative coupling oligomerization and separation by preparative size exclusion chromatography, a series of discrete conjugated oligomers up to 64-mer, which has an end-to-end contour length of 147 nm, were produced. The obtained organic nanowires have excellent solubility, a high fluorescent quantum yield, and a long migration distance for excitation energy. The encapsulation of the conjugated backbones with

February (2011) Vol.54 No.2

287

three-dimensional huge dendritic wedges is vital to such achievements, since it not only makes the separation between oligomers very easy, but also suppresses unfavorable interchain interactions, thus enhancing photoluminescence properties. By changing the terminal functionalities to carboxylate (CO2 K+), the dendronized conjugated polymers become water-soluble [12]. Of interest, these water-soluble conjugated polymers can be used as sensitizers for photoreduction of H2O to H2. The overall quantum yield is 13%, much larger than small organic dyes. Another important merit of a dendrimer is site-specific functionalization. That is, one can freely introduce specific functionalities into the desired position of a dendritic scaffold. This unique feature originates from a step-by-step fashion of dendrimer synthesis, which either follows a divergent or a convergent approach. Therefore, one can rationally design and control the structure, properties, and functions of a dendrimer. By utilizing this unique feature, Aida, Jiang, Li et al. have developed a series of multiporphyrin dendrimers, which possess a number of porphyrin units, all residing in the same layer of a dendritic scaffold [13, 14]. These compounds served as models to study the energy migration and transfer processes in natural photosynthetic systems [15, 16]. When these multiporphyrin dendrimers entrapped a layer of fullerene units via metal-ligand interactions, a concentric electron donating (D) and accepting (A) double layer structure formed [17]. Time-resolved fluorescence and transient absorption spectroscopy revealed that such a D-A layered structure has remarkable photoinduce electron transfer properties. Owing to the outstanding encapsulation and the specific functionalization of dendritic structures, scientists have recognized that dendrimers could serve as a very nice material for the protecting shell of nanomaterials. The first example of such a hybrid system was reported in 1998 by Crooks and coworkers [18] and shortly afterwards by Tomalia et al. [19], in which Cu nanoparticles (NPs) were prepared by entrapping Cu2+ into the interior of a poly(amidoamine) dendrimer (PAMAM) framework (Figure 2) and subsequently followed by a chemical reduction. In the same year, Esumi et al. reported the extremely small Au nanoparticles could be achieved by using PAMAM as a stabilizer under UV irradiation [20]. One year later, Pd and Pt NPs with dendrimer templates have been prepared and exhibit excellent catalytic properties in hydrogenation [21]. After these pioneering works, a variety of dendrimer/inorganic nanocomposites, including metal-, semiconductor quantum dot-, and carbon-based systems, have been exploited and showed fascinating functions and specific applications in catalysis, optoelectronics, sensors, and biomedical treatments. In this field, researchers are often asked with questions: what is the advantage of dendrimers over small organic agents and polymers; and what are special properties, functions, and applications that dendritic structures bring?

288



Figure 2 Molecular structures of 4th-generation poly(amidoamine) dendrimers with hydroxyl (PAMAM-G4-OH), amine (PAMAM-G4-NH2), alkyl (PAMAM-G4-C12), acetamide (PAMAM-G4-NHAc), and glycidol hydroxyl (PAMAM-G4-GlyOH) terminal groups.

With an attempt to well answer such questions and to shine light on basic principles of structure-property-function of dendrimer/inorganic nanomaterial hybrid systems, this review article selects a number of representative examples, and highlights the novel features endowed by dendritic structures. 2

Dendrimer/non-noble metal nanoparticle composites

As mentioned in INTRODUCTION part, dendrimer/inorganic nanomaterials composites were firstly demonstrated with dendrimer/Cu NP composites using poly(amidoamine) (PAMAM) dendrimers (Figure 2) following a coordinationreduction method (Scheme 1) [18]. The PAMAM bears a lot of tertiary amine units through out its structure which can coordinate Cu2+. Thus, when a solution of PAMAM-G4-OH dendrimer (Figure 2) was mixed with a Cu2+ aqueous solution, Cu2+ was rapidly entrapped into dendritic box. In the presence of NaBH4 that can reduce Cu2+ to Cu(0), a golden

brown dendrimer/Cu nanoparticle composite solution resulted. Transmission electron microscopy (TEM) confirmed the formation of Cu cluster with an average size less than 1.8 nm. Despite the extremely small size of the resultant Cu NPs, the obtained dendrimer/Cu NP composites are stable for at least one week in an oxygen-free solution. Based on the observed phenomena, the authors concluded that Cu clusters reside in the dendrimer interior. Later together with a plenty of other works [22, 23], the authors named this kind

Scheme 1 General procedure for synthesis of dendrimer-encapsulated metal nanoparticles.



289

of composites with dendrimer-encapsulated nanoparticles (DENs). Just after the work of Crooks, Tomalia et al. reported a similar work but with hydrazine as a reductant [19]. Here, it should be noted that stable dendrimer-encapsulated metal nanoparticles are strongly dependent on the chemical composition of the dendrimer. For example, when PAMAM-G4-NH2 (Figure 2) was used in place of PAMAM-G4-OH, Cu nanoparticles with a larger size (> 5 nm in diameter) formed [18]. The larger size is a consequence of the agglomeration of Cu particles absorbed to the unprotected dendrimer exterior. In such a case, both intraand inter-dendrimer encapsulation of Cu nanoparticles are possible. Later in 2008, Yang and coworkers studied the effect of dendrimer size and preparation conditions, especially the ratio of Cu2+ to dendrimer, on the resultant Cu nanoparticle in a detailed way with PAMAM dendrimers having a trimesyl core and amine terminal groups [24]. They found that the nanoparticle size decreases progressively when the dendrimer generation of PAMAM goes from 3 to 6. In contrast, Cu nanoparticles become larger upon increasing the feeding ratio of Cu2+/dendrimer. In addition to copper-based, other non-noble transition metal-based DENs have been reported. For example, Crooks et al. prepared iron nanoparticles of extremely monodispersed small size using a 6-generation PAMAM dendrimer (PAMAM-G6-C12, refer to PAMAM-G4-C12 in Figure 2) having 256 dodecyl terminal groups as a template [25]. The synthesis followed the similar protocol that used for Cu DENs. Firstly, a 2.00 M solution of PAMAM-G6C12 was mixed with sufficient 30.0 mM FeCl3 in THF to yield Fe3+/dendrimer composites of iron-to-dendrimer ratio of either 55 or 147. Next, a 10-fold excess of NaEt3BH (1.00 M in toluene) was added, which lead to nearly colorless solutions of DEN containing 55 or 147 atoms of Fe. The size of resulting Fe nanoparticles is extremely small and monodispersed: 0.9 ± 0.2 nm for Fe55 DEN, 1.1 ± 0.2 nm for Fe147 DEN. Furthermore, the obtained Fe DENs are indefinitely stable in N2, but decompose quickly after exposure to O2. Such small Fe nanoparticles with nearly sizemonodispersity and high stability have been seldom achieved with other common surfactants or polymers as template and stabilizers, indicating the superior of dendrimers. Magnetic analysis indicates that 55-atoms nanoparticles are superparamagnetic, while that of 147 atoms undergo a transition to ferromagnetic at 6 K. Both materials exhibit suppression of the magnetic saturation as compared to bulk Fe. Following the same approach, Ni DENs were synthesized with PAMAM-G6-C12 as template and NaEt3BH as a reductant [26]. The Ni DENs are ferromagnetic at 5 K with magnetic saturation.

logical importance because of their promising applications as catalysts in a variety of reactions. In general, the catalytic activities depend strongly on the particle size and the unoccupied surface area. As compared with common stabilizers, dendrimers featured with a three-dimensional architecture can provide an extremely efficient protect against agglomeration, but less anchoring to the nanoparticle surface. Therefore, dendrimers are ideal capping agents and templates to prepare small, but stable noble metal nanoparticles with a high catalytic activity. Such works have firstly been conducted by Crooks and coworkers [21]. They followed the same method of Cu DENs [18] to prepare Pt and Pd DENs. That is: a predetermined quantity of metal ions (PtCl42 or PdCl42) was firstly extracted into the interior of PAMAM-G4-OH, followed by chemical reduction with BH4 to yield zerovalent metal particles. Of interest, the Pd and Pt DENs so prepared have a spherical shape and extremely small size with nearly monodispersity, as revealed by high-resolution transmission electron microscopy (HRTEM) images (Figure 3). For examples, t h e s i z e i n d i a m e t er o f P t 4 0 @PAMA M- G4- OH, Pt60@PAMAM-G4-OH, and Pd40@PAMAM-G4-OH are 1.4 ± 0.2, 1.6 ± 0.2 and 1.3 ± 0.3 nm, respectively. The catalytic performance of the obtained DENs was tested for hydrogenation of a simple alkene (allyl alcohol) and an electron-deficient alkene (N-isopropyl acrylamide) in water

3 Dendrimer/noble metal nanoparticle composites

Figure 3 HRTEM images of (a) Pt40@PAMAM-G4-OH, (b) Pt60@PAMAM-G4-OH, and (c) Pd40@PAMAM-G4-OH, which show the monodisperse size and shape distribution. Adapted with permission from ref. [21]. Copyright 1999 Wiley-VCH Verlag GmbH & Co. KGaA.

Noble metal nanoparticles are of fundamental and techno-

290


(Table 1). As compared with Pt DENs, Pd DENs exhibit a much higher catalytic activity for both hydrogenations. Most importantly, the size of the dendrimer strongly affects the catalytic activity of DENs, which shows a decrease when dendrimer becomes larger. The turnover frequencies (TOFs, mol of H2 per mol of metal atoms per hour) in the hydrogenation of N-isopropyl acrylamide for Pd40@PAMAMG6-OH and Pd40@PAMAM-G8-OH are only 11% and 5%, respectively, of that for Pd40@PAMAM-G4-OH. The authors explained this result with different porosity in the dendrimer with a different generation number. Higher generation dendrimers are less porous and thus less likely to admit substrates to interior metal nanoparticles. That is, the dendrimer acts as a selective nanoscopic filter that controls the catalytic activity of the composite (Figure 4). Thus, one can control reaction rates and do selective catalysis by adjusting the “mesh” of the dendrimer “nanofilter”. Although the activity of Pd40@PAMAM-G4-OH, the highest level among the family examined, is comparable to PVP-stabilized, colloidal Pd dispersions in water [27], the stability of DENs are superior. They did not aggregate for up to four months in water. They can be isolated as a black powder and redissolved in water to yield a stable, dark-brown solution that is Table 1 Hydrogenation activity of Pt and Pd DENs in water a) Catalyst

TOF b) N-isopropyl acrylamide

Allyl alcohol

Pd40@PAMAM-G4-OH

372

218

Pd40@PAMAM-G6-OH

42

201

Pd40@PAMAM-G8-OH

17

134

Pt40@PAMAM-G4-OH

57

25

Pt40@PAMAM-G6-OH

30

–

a) Hydrogenation reactions were carried out at 20 ± 2 °C with 2 × 104 M Pd(0) or Pt(0) composite catalysts [21]. The turnover frequency (TOF) was calculated on the basis of H2 uptake. b) Measured as mol of H2 per mol of metal atoms per hour.

Figure 4 The hydrogenation rate can be controlled by using dendrimers of different generation: Pd clusters present within dendrimers of lower generation exhibit the highest catalytic activity. The sterically crowded terminal groups of higher generation dendrimers hinder substrate penetration. Adapted with permission from ref. [21]. Copyright 1999 Wiley-VCH Verlag GmbH & Co. KGaA.


identical to that present before drying. After hydrogenation reaction were run for up to 3 h, the solution of Pd DENs remained clear. No evidence was presented for agglomeration, and also no changes were observed in the UV-vis or IR spectra. Although Pt DENs are less active in hydrogenation reactions, they have been demonstrated to act as effective O2 reduction catalyst when anchoring to Au electrode [22]. In addition to PAMAM, dendrimers based on propyleneimine [28], ethyleneimine [29], diaminobutane [30] phosphorus-containing [31], and phenylene-pyridyl units [32] have been used to prepare dendrimer/noble metal nanoparticle composites. More recently, Astruc and coworkers developed an interesting family of dendrimers (Figure 5) synthesized by click reaction and applied them as templates and hosting agents for Pd NPs [33, 34]. They found that 1,2,3-triazole units formed by click reaction are capable of binding Pd(OAc)2 in one-to-one stoichiometry. As evidenced by the change of redox wave of ferrocenyl units in cyclic voltammetry, the complexation occurs in order, from inner to outer, if the dendrimer contains more than one layer of triazole units. This means the inner triazole units bind Pd(II) first, and the ferrocenyltriazoles residing at the periphery bind Pd(II) only when the inner triazole ligands are already saturated with Pd(II). Of interest, the presence of methanol [34] or water [35] is required for the complexation. The authors explained this observation by that methanol can break the trimetric species of Pd(OAc)2 upon dechelating one or two acetate ligands, making Pd(II) ready for binding. The more interesting is that methanol not only promotes the complexation but also acts as a reductant to reduce Pd(II) to Pd(0). Thus, Pd NPs can be prepared by just simply adding Pd(OAc)2 to the dendrimer in methanol. The size of the dendrimer and the type of the reductant were found to have significant influence on the formation of Pd NPs. Owing to its small size and open structure, PTA-G0Fc could not encapsulate a Pd NP within a dendrimer. Alternatively, interdendrimer-stabilized Pd NPs (Pd DSNs) was produced in this case. Their size depends strongly on the nature of the reducing agent that was used (Table 2). When methanol was used as the reducing agent, the size of Pd@PTA-G0-Fc DSNs are 2.8 ± 0.3 nm in diameter, which was estimated to be 766 Pd atoms stabilized by 85 dendrimers. Whereas, reduction with NaBH4 afforded Pd DSNs with a diameter of 1.2 ± 0.2 nm (60 Pd atoms surrounded by 7 dendrimers), much smaller than that produced with methanol. In contrast to the G0 dendrimer, the G1 and G2 dendrimers are large enough for hosting a Pd NP. Thus, in both cases, intradendrimer-encapsulated Pd NPs (Pd DENs) formed (Figure 6). Their sizes are fairly consistent with the preloaded amount of Pd(II) that bind to one dendrimer. The catalytic properties of prepared Pd DSNs and DENs were tested for hydrogenation of styrene at 0.1% mol Pd. The results (Table 2) have demonstrated that the prepared dendrimer/Pd nanoparticle composites are highly efficient catalysts for the hydrogenation. They are stable under


291


Figure 5 Molecular structures of 1st-generation poly(triazole) dendrimers with ferrocenyl (PTA-G1-Fc) and triethyleneglycol tethers (PTA-G1-TEG) terminal groups.

Table 2 Size and catalytic activity of Pd DSNs and DENs based on poly(triazole) dendrimers for hydrogenation of styrene at 0.1% mol Pd [34] Number of Pd atoms per dendrimer

Calculated diameter (nm)

Pd@PTA-G0-Fc DSN

9

Pd@PTA-G1-Fc DEN

Catalyst

Measured diameter (nm)

TOF (mol H2 (mol Pd)1 h1)

TON

method 1

method 2

method 1

method 2

method 1

method 2



1.2  0.2

2.8  0.3

200

1200

30000

31500

36

1.0

1.1  0.2

1.3  0.2

310

1620

10000

9300

Pd@PTA-G2-Fc DEN

117

1.5

1.6  0.3

1.6  0.3

200

1280

20000

16650

Pd36@PTA-G2-Fc DEN

36

1.0

1.1  0.2

1.3  0.3

280

1380

7400

10000

Pd@PAMAM DEN

40

1.0

1.1  0.2



56



7500



292


catalytic conditions and can be reused for several times (10 circles for Pd@PTA-G1-Fc and 20 cycles for Pd@PTAG2-Fc). The turn over number (TON) is extremely large: ca. 10000 for Pd@PTA-G1-Fc, ca. 20000 for Pd@PTAG1-Fc, whereas ca. 30000 for Pd@PTA-G0-Fc. Of interest, the TOF of the DSNs and DENs prepared with methanol as the reductant is much higher than that of DENs obtained using NaBH4. The authors argued that certain residue (B(OCH3)3 and Na+) leaved by NaBH4 would inhibit the hydrogenation reaction. Later, a water soluble version of Pd

Figure 6 Schematic representations of DSN and DEN formed with PTA-G0-Fc and PTA-G1-Fc, respectively.


DSNs and DENs were prepared with poly(triazole) dendrimers having sulphonated terminal groups, and exhibited a high stability against air and moisture and a high catalytic efficiency for olefin hydrogenation and Suzuki coupling in water [35]. More recently, an interesting work done by Christensen et al. [36] has demonstrated that the chirality of a dendrimer can be transferred to its encapsulated metal nanoparticles. In this work, a series of specific PAMAM dendrimers having chiral carbons through out the dendritic structure (Figure 7(a)) were prepared from ()-1,2-diaminopropane. Then, Pd2+ and Rh3+ were loaded into the prepared chiral PAMAM dendrimers and subsequently reduced to yield DSNs upon addition of NaBH4. TEM confirmed the formation of nanoparticles with a nearly monodispersed size. A broad absorption band originating from the electronic structure of nanoparticles appeared in the visible region up to 500 nm in UV-vis spectra (Figure 7(b)). Of interest, in the same visible region, a strong cotton effect was clearly observed in circular dichroism (CD) spectra. As a control, when Pd nanoparticles were prepared using a racemic PAMAM dendrimer as the template, no CD signal was observed. These observations illustrate a strong interaction between metal nanoparticles and the chiral dendrimers, which can not be observed in UV-vis spectroscopy. More importantly, the observations also indicate the nanoparticles themselves have certain chirality, which must be transferred from the chirality of their hosting dendrimer. As the authors pointed out, this approach opens a new avenue to endow the chirality to metal nanoparticles, which might have a promising application as chiral catalysts for asymmetric synthesis. In relation to this work, Haag and coworkers have reported that Pt DSNs allows the modification of Pt NP surface with a small chiral agent, and then can be used to catalyze hydrogenation with a relative high enantiometric excess (ee) value [37]. Au is another important noble metal. The dendrimer/Au nanoparticle composites were first reported by Esumi and coworkers in 1998 [20]. They prepared Au colloids from HAuCl4 using PAMAM-Gn-NH2 as stabilizer upon UV irradiation. The molar ratio of surface amino groups of PAMAM to HAuCl4 was found to significantly affect the size of nanoparticle. When this ratio is 1:1, the particle size of gold colloids was in 2–18 nm range with a broad distribution. The average particle size was not dependent on the dendrimer generation number. However, when the ratio is 4:1, ultrafine nanoparticles with a diameter less than 1 nm and relative small dispersity were formed with dendrimers larger than G3. Furthermore, the dendrimer generation number has no effect on the average size of particles. However, the authors did not point out what kind of dendrimer/ nanoparticle composite is formed under different molar ratio of surface amino groups to AuCl4. The results imply that in the case of 1:1 ratio, DSNs were formed while in the case of 4:1, DENs generated. Two years later, Amis et al. reported



293

Figure 7 (a) Schematic structural representation and (b) UV-vis absorption and CD spectra of Pd@PAMAM-G3-Bn and Rh@PAMAM-G3-Bn prepared with chiral dendrimer PAMAM-G3-Bn. Adapted with permission from ref. [36]. Copyright 2008 The Royal Society of Chemistry.

that the type and the size of gold nanoparticles using PAMAM as template and NaBH4 as reductant are highly dependent on the reaction conditions and the dendrimer generation [38]. Au nanoparticles with homogeneous size can be only obtained under a certain range of concentration and the slow reduction rate. For generation 2–4, PAMAM dendrimers are not large enough. Therefore, Au DSNs, in which several dendrimers are surrounding one nanoparticle, were resulted using PAMAM dendrimers of generation of 2–4. For G6–G10, nanoparticles can be completely entrapped inside individual dendrimers. TEM clearly indicate that one dendrimer of generation 6–9 can accommodate only one nanoparticle inside (Figure 8). The nanoparticle is located offset from the center of the dendrimer. For G10 PAMAM, multiple smaller gold particles per dendrimer were observed. More recently, certain (co)polymers containing tertiary amine units have been demonstrated to act simultaneously both a reducing agent and a stabilizer [39]. Inspired by such

finding, Baker, Jr. et al. investigated the spontaneous formation of dendrimer/Au NPs composites using PAMAM [40]. They found that PAMAM dendrimers modified with glycidol hydroxyl units at the periphery (For an example, see PAMAM-G4-GlyOH in Figure 2) can serve as reducing agent and stabilizers for Au NPs. Thus, just by simple mixing of HAuCl4 with PAMAM-G5-GlyOH, Au DSNs form spontaneously. However, by mixing acetamide-terminated PAMAM dendrimers (For an example, see PAMAM-G4NHAc in Figure 2) with HAuCl4, no stable Au nanoparticles but black precipitation occurred, implying that the acetamideterminated PAMAM can reduce AuCl4 to Au(0), but cannot stabilize the nanoparticles. Of surprising, when primary amino-terminated PAMAM dendrimers were used first to bind AuCl4, then subjected to acetylation with acetic anhydride in the presence of triethylamine, Au DSNs generated. The result implies that the acetylation reaction not only converts amine units at the periphery of the dendrimer to acetamide groups, but also promotes the formation of Au

Figure 8 TEM of PAMAM/Au nanoparticle composites with G8 (a), G9 (b), and G10 (c) dendrimer obtained from 1:1 loading and slow reduction. The dendrimers have been stained with phosphotungstic acid. Adapted from ref. [38].

294


DSNs. Although the mechanism is not very clear, this method has been successfully used to prepare dye-functionalized PAMAM/Au nanoparticle DSNs. In a followed report, glycodendrimers prepared from modification of poly(propyleneimine) (PPI) dendrimers with oligosaccharide at periphery can also serve as effective reducing and stabilizing agents for the formation of Au and Ag nanoparticles [41]. More recently, water-soluble PEGylated click dendrimers terminated with Percec-type triethyleneglycol (TEG) tethers have been applied as templates for synthesis of Au nanoparticles (Figure 5) [42]. The research has disclosed an interesting phenomenon: the presence of both triazole units and PEGlated Perce-type dendrons is required for the formation of Au DENs using NaBH4 as the reductant in methanol. Related dendrimers lacking one of these structural features cannot stabilize Au nanoparticles upon NaBH4 reduction in this solvent. The authors interpreted the observation by the different roles of these two units in the formation of Au NPs. The triazole ligands serve to trap the Au(III) ions inside the “click” dendrimers, then stabilization of the reduced Au atoms in Au NPs is provided by the nearby semicavitand formed by the PEGylated Percec-type dendrons. However, in the absence of external reductant in water, the semicavitand effect of Percec-type PEGylated dendritic structure is crucial to the formation of Au NPs. Thus, all the dendrimers having Perce-type PEGylated periphery can reduce HAuCl4 to produce Au NPs as a form of DSNs. Apart from the above works, Zheng and coworkers have developed a series of Au nanoparticle-cored dendrimers (NCDs) [43]. NCDs is another type of dendrimer/nanoparticle composites in addition to DENs and DSNs. The structural feature of NCDs is that a nanoparticle is the core of a dendrimer. In the work from Zheng et al., a series of Percec dendrons (PD-Gn) having a 4-pyridone functionality at their focal points but different generation numbers were used as capping agents for Au nanoparticles (Figure 9). The preparation of such a dendrimer/gold NP composite followed a liquid-liquid phase-transfer method: Au3+ was firstly trapped into organic phase (toluene/CH2Cl2 2/1) from water by (n-C8H17)4NBr as a phase transfer surfactant, followed by adding a CH2Cl2 solution of dendritic capping agents, and subsequent freshly prepared aqueous solution of NaBH4. In all cases, wine-red solutions having a characteristic surface

Figure 9 Schematic representations of nanoparticle-cored dendrimers.


plasmon resonance at about  = 525 nm for gold nanoparticle were obtained. The phase transfer surfactant was used to avoid the formation of a significant black precipitate. The prepared dendrimer/gold nanoparticle composites are stable enough to be isolated and purified by recrystallization. The resulted dark crystalline solids can be redissolved in CH2Cl2 to form a clear wine-red solution again without noticeable visual change. Elemental analysis of Au-G1 showed that it is composed of 5% G1 and 95% Au, corresponding roughly an Au:PD-G1 ratio of 49:1. Thus, a 2-nm particle has a formula of Au247@(PD-G1)5 based on the calculation with gold bulk packing density. The small number of dendrons required for a particle is not unreasonable considering the coverage offered by the umbrella-shaped dendrons. Interestingly, the size of dendritic agents has a strong impact on the size of Au nanoparticles, as revealed by TEM. Going from PD-G1 to PD-G3, the average size of nanoparticles measured over 200 particles in TEM image become larger and larger: 2.0 ± 1.0 for PD-G1, 3.3 ± 1.1 for PD-G2, and 5.1 ± 1.7 nm for PD-G3. The authors interpreted this observation by higher steric requirement of with a larger generation number. Soon after the above work, Kim et al. reported Au nanoparticles with a much narrow size distribution could be prepared with a serious Fréchet-type poly(benzyl ether) dendrons possessing a single thiol group at their focal point and 3,5-dimthoxy phenyl units at their periphery (PBE-GnOMe, Figure 9) [44]. As revealed by high solution TEM, Au@PBE-G1-OMe has an average core size of 2.8 ± 0.3 nm. Most remarkably, the size of Au@PBE-G1-OMe (2.4 ± 0.2 nm) is nearly monodispersed and not significant affected by different ratio of PBE-G2-OMe:HAuCl4, such as 0.5 and 6, in the preparation. The authors suspected that PBE-G2-OMe dendron may have a “magic” size to protect Au NP with particular size leading to impressing monodispersity. When PBE-G3-OMe was used as a capping agent, Au nanoparticle with average size 3.1 ± 0.6 nm were resulted. As compared with the cases of PBE-G1-OMe and PBE-G2-OMe, Au particle becomes larger with less narrow dispersity, which is similarly observed by Zheng and coworkers [43]. However, Fox and coworkers reported a completely different result, in which Au NPs capped with similar Fréchet-type poly(benzyl ether) dendritic sulfides but with phenyl at their


periphery (PBE-Gn-H, Figure 9) have a wide size distribution [45]. The contrast results may be coming from the different preparation protocol: they used dendritic disulfides in place of dendritic thiols as precursors for the capping agents although the final products should be same. Later, Fox and coworkers changed the terminal groups of poly(benzyl ether) dendrimers to ester functional units (PBE-Gn-CO2Me, Figure 9) and prepared the Au NCPs thereof [46]. Upon hydrolysis to convert terminal ester groups to sodium salts, the Au NCPs became water soluble and exhibited micelle properties in aqueous solutions. Different from the above complexation and reduction strategy for preparation of NCPs, a strategy based on post modification on the ligands of the preformed nanoparticles has been proposed by Shon, Choi et al. [47]. In this approach, Au nanoparticles were first prepared from AuCl4 by Schiffrin protocol with hexanethiol as a protecting agent and NaBH4 as a reductant. After ligand exchange reaction, part of hexathiolated ligands were replaced with 11-mercaptoundecanoic acid, resulting Au nanoparticles having reactive functionalities (COOH groups) at their surface. Thus, in the presence of ester coupling reagents (1,3-dicyclohexylcarbodiimide and 4-dimethylamino- pyridine, dendrons having a hydroxy group at their focal point were anchored to the nanoparticle surface, giving NCDs (Scheme 2). Since the reaction is carried out only at the surface of nanoparticles, the core of nanoparticles remains intact. Their size and size distribution are preserved. Therefore, even with dendrons have different generation number, the resulting NCDs have similar coral size and size distribution. Another advantage of this approach is that the number of dendrons around nanoparticle core can be controlled by tuning the number of reactive functional groups on particles and the feeding ratio of dendritic reagents to nanoparticles.

4 Dendrimer-modified semiconductor quantum dots Semiconductor quantum dots (QDs) have attracted great attention in the past decade due to their unique optical properties [48–50]. In contrast to traditional organic dyes, QDs have many advantages in terms of optical properties, such as tunable and symmetric emission and photochemical stability. Generally, the key to employ QDs as a tool in biological systems is to achieve water solubility, biocompatibility, and photostability. As QDs themselves are hydrophobic, many kinds of methods have been exploited to prepare water-soluble QDs, for example, using capping ligands to modify the surface of QDs via a ligand-exchange method using thiol ligands [61]. Murphy and coworkers were the first to report using dendrimers to control the size and the solubility of QDs [52–54]. These materials have shown to be agglomerates of spatially segregated QDs stabilized by multiple dendrimers. That is, in these interdendrimer com-


295

posites, many dendrimers surround multiple QDs to yield large, aggregated structures. Wu et al. [55] reported the synthesis of CdS quantum dots with G8 amino- terminated PAMAM dendrimers as stabilizers. Zhang et al. reported the surface modification of CdSe nanoparticles with amphiphilic and flexible PAMAM dendrimers carrying different numbers of hydrophobic aliphatic chains [56]. However, most dendrimer-modified QDs were obtained in an organic solvent, with the result that the QDs are not able to be used as bioprobes. Liu et al. reported on the preparation of watersoluble CdSe/ZnS (core/shell) quantum dots through a solvent-evaporation method [57] (Scheme 3). This approach avoids complicated multiple step ligand- exchange procedures and allows to add to the versatility of QD-based probes for investigating intracellular transport and other cellular-signaling pathways in living cells. They synthesized 0.5, 1.5, 2.5, 3.5, and 4.5 generation (0.5G, 1.5G, etc.) PAMAM dendrimers via a divergent method from an ethylenediamine core. Monodisperse CdSe/ZnS core/shell QDs were prepared according to the above approach. As shown in Figure 10, it was observed that the fluorescence emission intensity increased, when the generation of PAMAM was changed from G0.5 to G4.5. There was no obvious emission-wavelength shift in comparison with the original QDs, which demonstrates that the water-soluble QD nanoparticles retain the fluorescent properties of the original QDs in water. The quantum yield of water-soluble QD nanoparticles increased as the generation of ester-terminated PAMAM increased. The TEM images in Figure 11 revealed that (G4.5 PAMAM/poloxamer 188)-modified QD nanoparticles have good dispersibility and size uniformity in water. As can be seen by the fluorescence microscopy images in Figure 12, the water-soluble QDs have strong fluorescence and dispersibility. The ability to use QDs as a bioprobe to track cellular processes was applied in cellular images. When incubated with live HeLa cells, (G4.5 PAMAM/poloxamer 188)-modified CdSe/ZnS nanoparticles could be found in the interior of the HeLa cells (Figure 13). The observation that water-soluble QD nanoparticles retained their fluorescence property shows that QD nanoparticles could accumulate without aggregation and that there is a general compatibility of the ester-terminated PAMAM/poloxamer 188 coating in a cellular environment.

5

Dendrimer-functionalized carbon nanotubes

Since their discovery by Iijima in 1991 [58], carbon nanotubes (CNTs) as a kind of carbon-based one-dimensional nanomaterials, have attracted a great scientific attention due to their outstanding mechanical, electronic, and optical properties and their promising applications in solar energy conversion [59–62], sensing [63], and biomedical treatments [64–66]. However, processing CNTs and integrating

296



Scheme 2 Preparation of Au NCD by post esterification of Au nanoparticles having –COOH functional groups at their surface and a G2 dendron with a hydroxyl unit at its focal point.

Scheme 3 Schematic illustration of the solvent-evaporation method to make (ester-terminated PAMAM/poloxamer 188)-modified CdSe/ZnS QD nanoparticles. Reprinted with permission from ref. [57]. Copyright 2006 Wiley-VCH Verlag GmbH & Co. KGaA.

them into real world applications is severely hampered by their inherent low solubility and the bundle aggregate. With the purpose to increase the solution processibility of CNTs, covalent and non-covalent approaches have been proposed [67]. The common feature of these two approaches is to attach soluble organic pendant groups to CNTs via depositing points either covalently or non-covalently. As compared with small organic molecules and traditional polymers, dendritic structures have an advantage: a huge soluble group with a three-dimensional architecture can be attached via a


Figure 10 Fluorescence spectra of CdSe/ZnS QDs in chloroform and CdSe/ZnS QDs modified with different generations of ester-terminated PAMAM/poloxamer 188 in water. Reprinted with permission from ref. [57]. Copyright 2006 Wiley-VCH Verlag GmbH & Co. KGaA.


single depositing point. In this regard, dendrimers represent a kind of particularly promising candidate for the functionalization of CNTs without significant destroy in their -conjugation nature. In 2001, Sun et al. reported the first example of functionalization of CNTs with dendritic wedges [68]. They utilized CNTs possessing carboxylic acid units at their surface as precursor. Upon esterification with dendritic alcohols or amidation with dendritic amine, dendritic wedges were attached to the wall of CNTs, yielding dendrimer-functionalized CNTs (for an example, see CNT@PBE-G1-C12 in Figure 14). Later in 2003, Hirsh and coworkers adopted another chemical reaction by the addition of oxycarbonyl nitrenes generated from dendritic azidocarbonates (for an example, see CNT@PBE-G1-TEG in Figure 14) [69]. In both cases, the functionalization with dendritic wedges greatly improves the solubility of CNTs in common solvents. The nature

Figure 11 TEM images of original QDs (a) and G4.5 PAMAM/poloxamer 188)-modified QD nanoparticles (b). Adapted with permission from ref. [57]. Copyright 2006 Wiley-VCH Verlag GmbH & Co. KGaA.

Figure 12 Fluorescence image of (G4.5 PAMAM/poloxamer 188)modified QD nanoparticles in water. Reprinted with permission from ref. [57]. Copyright 2006 Wiley-VCH Verlag GmbH & Co. KGaA.

Figure 13 Confocal images of HeLa cells after incubation with (4.5G PAMAM/poloxamer 188)-modified QD nanoparticles in (a) bright field and (b) fluorescence. Adapted with permission from ref. [57]. Copyright 2006 Wiley-VCH Verlag GmbH & Co. KGaA.

297

Figure 14 CNTs functionalized with dendritic wedges.

298


of their solubility is determined by the peripheral groups of the dendritic wedges. For example, CNT@PBE-G1-C12 that possesses lipophilic long alkyl chains at the surface is highly dissoluble in a variety of nonpolar and weakly polar organic solvents, such as hexane and chloroform, but does not essentially dissolve in highly polar solvents such as

Figure 15

Schematic representation of SWNT@PAMAM-G2-H2P.


ethanol and DMSO. While, CNT@PBE-G1-TEG having oligo(ethylene glycol) chains is soluble in both organic solvents, such as chloroform and DMSO, and DMF. A more interesting work was reported by Prato et al. [70]. They functionalized single-wall CNTs (SWNTs) with a large number of porphyrin-appended PAMAMs (Figure 15).


The composite SWNT@PAMAM-G2-H2P was prepared by allowing the dendritic branches to grow directly from the SWNT surface using a divergent approach, followed by anchoring the porphyrin units onto the periphery of the dendritic substituents. Such an approach guarantees an increase of the functional groups on the nanotubes without causing significant damage to the electronic properties of the nanotubes. TEM confirmed the presence of long highaspect-ratio objects that are several micrometers in length and a few to several tens of nanometers in diameter (Figure 16). These micrographs indicate that a large number of dendritic wedges prevent the heavy aggregation of pristine HiPCO SWNTs and provide dispersed SWNTs or very thin bundles. Namely, this methodology allows to envelope SWNTs with an electron donating layer to form a one-dimensional coaxial heterojunction, desirable for a variety of electronic or optoelectronic applications. Excitation of the porphyrin units results in a transfer of electrons to the SWNT core, as evidenced by quenching of the porphyrin fluorescence with an efficiency of ~85% (Figure 17). Transient absorption spectroscopy demonstrated the occurrence of a rapid charge separation (kCS = (1.5 ± 0.5) × 1010 s1) and a slow charge recombination (kCR = (2.9 ± 0.5) × 106 s1). Thus, a large number of porphyrin units in dendritic wedges play conceptually the role of an antenna: harvesting light to give more efficient electron transfers. This work provides an excellent example to illustrate that new properties and func-

Figure 16 TEM images of SWNT@PAMAM-G2-H2P from a DMF solution. Reprinted with permission from ref. [70]. Copyright 2006 American Chemical Society.


tions could be rewarded by specifically modification on the desired position of a dendritic structure with specific functionality.

6

Conclusions and outlook

The combination of inorganic nanomaterials and dendrimers produces a kind of completely new materials called dendrimer/inorganic hybrid nanocomposites, which merge the advantages from both sides. As exemplified by the selected works in this review, the use of a dendritic host with a three-dimensional architecture not only greatly improves the stability and processibility of inorganic nanomaterials, but also finely tunes their size and size distribution, thus strong affects and modifies the properties of inorganic nanomaterials. Furthermore, new properties and functions are often generated as the consequence of the interactions between inorganic nanomaterials and functionalities that are specifically deposited in the dendritic scaffold. Although dendrimer/inorganic nanocomposites have discovered for over 20 years, there are still ample room for the exploitation of new hybrid systems that have unusual properties and functions. For example, one may design a special gate that can be controlled by certain external stimuli, in a specific layer of a dendritic structure. When such dendrimers are used to encapsulate inorganic nanomaterials, one may control over the functions of inorganic nanomaterials, for example, to start or to shut down catalytic reactions, the release of nanomedicines, sensing, and so on, by external methods. In addition to development of novel hybrid systems, the whole field of nanotechnology including dendrimer/inorganic nanocomposites has shifted from nanoarchitecture synthesis to a greater focus on application. Catalysts, sensors and biomedical applications might be three important areas for dendrimer/inorganic nanomaterial composites. This work was financially supported by the National Natural Science Foundation of China (20974119, 90922019, and 21074147), Chinese Academy of Sciences, and the Science and Technology Commission of Shanghai Municipality (09PJ1411700). 1

2

3 4 Figure 17 Room-temperature fluorescence spectra of 5,10,15-tris(3,5-di(tert-butyl)phenyl)-20-[4-(carbonyloxy)phenyl]porphyrin (solid line) and SWNT@PAMAM-G2-H2P (dashed line) in THF with matching absorption at the excitation wavelength of 419 nm. Reprinted with permission from ref. [70]. Copyright 2006 American Chemical Society.

299

5

6

Rao CNR, Müller A, Cheetham AK. The Chemistry of Nanomaterials: Synthesis, Properties and Applications. Weinheim: Wiley-VCH, 2004 Pachón LD, Rothenberg G. Transition-metal nanoparticles: Synthesis, stability and the leaching issue. Appl Organometal Chem, 2008, 22: 288–299 Newkome GR, Moorefield CN, Vögtle F. Dendrons and Dendrimers; Concepts, Synthesis and Applications. Weinheim: Wiley-VCH, 2001 Fréchet JMJ, Tomalia DA. Dendrimers and Other Dendritic Polymers. Chichester: John Wiley & Sons Ltd, 2001 Li WS, Jang WD, Aida T. Molecular design and self-assembly of functional dendrimers. Macromolecular Engineering. Precise Synthesis, Materials Properties, Applications. Eds. Matyjaszewski K, Gnanou Y. Leibler L. Weinheim: Wiley-VCH, 2007, V2: 1057–1102 Buhleier E, Wehner W, Vögtle F. “Cascade”- and “nonskid-chainlike” syntheses of molecular cavity topologies. Synthesis, 1978(2):

300

7

8

9

10

11

12

13

14 15

16

17

18 19

20

21

22

23

24

25 26

27


155–158 Tomalia DA, Baker H, Dewald J, Hall M, Kallos G, Martin S, Roeck J, Ryder J, Smith P. A new class of polymers: Starburst-dendritic macromolecules. Polym J, 1985, 17(1): 117–132 Hawker CJ, Fréchet JMJ. Preparation of polymers with controlled molecular architecture. A new convergent approach to dendritic macromolecules. J Am Chem Soc, 1990, 112: 7638–7647 Hecht S, Fréchet JMJ. Dendritic encapsulation of function: Applying nature’s site isolation principle from biomimetics to materials science. Angew Chem Int Ed, 2001, 40(1): 74–91 Tomoyose Y, Jiang DL, Jin RH, Aida T, Yamashita T, Horie K, Yashima E, Okamoto Y. Aryl ether dendrimers with an interior metalloporphyrin functionality as a spectroscopic probe: Interpenetrating interaction with dendritic imidazoles. Macromolecules, 1996, 29(15): 5236–5238 Li WS, Jiang DL, Aida T. Photoluminescence properties of discrete conjugated wires wrapped within dendrimeric envelopes: “Dendrimer effects” on -electronic conjugation. Angew Chem Int Ed, 2004, 43: 2943–2947 Jiang DL, Choi CK, Honda K, Li WS, Yuzawa T, Aida, T. Photosensitized hydrogen evolution from water using conjugated polymers wrapped in dendrimeric electrolytes. J Am Chem Soc, 2004, 126(38): 12084–12089 Li WS, Jiang DL, Suna Y, Aida T. Cooperativity in chiroptical sensing with dendritic zinc porphyrins, J Am Chem Soc, 2005, 127(21): 7700–7702 Li WS, Aida T. Dendrimer porphyrins and phthalocyanines. Chem Rev, 2009, 109(11), 6047–6076 Cho S, Li WS, Yoon MC, Ahn TA, Jiang DL, Kim J, Aida T, Kim D. Relationship between incoherent excitation energy migration processes and molecular structures in zinc(II) porphyrin dendrimers. Chem Eur J, 2006, 12: 7576–7584 Yang J, Cho S, Yoo H, Park J, Li WS, Aida T, Kim D. Control of molecular structures and photophysical properties of zinc(II) porphyrin dendrimers using bidentate guests: Utilization of flexible dendrimer structures as a controllable mold. J Phys Chem A, 2008, 112(30), 6869–6876 Li WS, Kim KS, Jiang DL, Tanaka H, Kawai T, Kwon JH, Kim D, Aida T. Construction of segregated arrays of multiple donor and acceptor units using a dendritic scaffold: Remarkable dendrimer effects on photoinduced charge separation. J Am Chem Soc, 2006, 128(32): 10527–10532 Zhao M, Sun L, Crooks RM. Preparation of Cu nanoclusters within dendrimer templates. J Am Chem Soc, 1998, 120(19): 4877–4878 Balogh L, Tomalia DA. Poly(amidoamine) dendrimer-templated nanocomposites. 1. Synthesis of zerovalent copper nanoclusters. J Am Chem Soc, 1998, 120: 7355–7356 Esumi K, Suzuki A, Aihara N, Usui K, Torigoe K. Preparation of gold colloids with UV irradiation using dendrimers as stabilizer. Langmuir, 1998, 14, 3157–3159 Zhao M, Crooks R. M. Homogeneous hydrogenation catalysis with monodisperse, dendrimer-encapsulated Pd and Pt nanoparticles. Angew Chem Int Ed, 1999, 38(3): 364–366 Zhao M, Crooks RM. Dendrimer-encapsulated Pt nanoparticles: Synthesis, characterization, and applications to catalysis. Adv Mater, 1999, 11(3): 217–220 Crooks RM, Zhao M, Sun L, Chechik V, Yeung LK. Dendrimerencapsulated metal nanoparticles: Synthesis, characterization, and applications to catalysis. Acc Chem Res, 2001, 34(3): 181–190 Jin L, Yang SP, Tian QW, Wu HX, Cai YJ. Preparation and characterization of copper metal nanoparticles using dendrimers as protectively colloids. Mater Chem Phys, 2008, 112, 977–983 Knecht MR, Crooks RM. Magnetic properties of dendrimerencapsulated iron nanoparticles containing an average of 55 and 147 atoms. New J Chem, 2007, 31(7): 1349–1353 Knecht MR, Garcia-Martinez JC, Crooks RM. Synthesis, characterization, and magnetic properties of dendrimer-encapsulated nickel nanoparticles containing