‘Bricks and mortar’ nanoparticle self-assembly using polymers

Polym Int 56:461–466 (2007) Polymer International Mini Review ‘Bricks and mortar’ nanoparticle self-assembly using polymers Palaniappan Arumugam, Hao Xu, Sudhanshu Srivastava and Vincent M Rotello∗ Department of Chemistry, University of Massachusetts Amherst, 710 North Pleasant Street, Amherst, MA 01003, USA Abstract: Developments in self-assembly methods allow access to hierarchical materials featuring a wide range of functionality and applications. Polymer-based self-assembly of nanoparticles opens up new avenues for the fabrication of highly structured nanocomposites that can serve as bridges between ‘bottom-up’ and ‘top-down’ methods. Of various interactions leading to self-assembly of nanocomposites, hydrogen bonding and electrostatic interactions are commonly utilized. In this review, we illustrate the design and subsequent property tuning of various self-assembled nanocomposite materials that were developed based on these interactions.  2007 Society of Chemical Industry Keywords: nanoparticle assembly; self-assembly; nanocomposites; polymer-based self-assembly INTRODUCTION The constant demand for technologically advanced devices has led to the development of a wide range of functional materials, including nanoparticle-based nanocomposite materials.1 – 3 Nanoscale structures are traditionally achieved using various lithographic techniques (‘top-down’ approach), and are currently reaching their inherent limitations with respect to their smallest readily achievable features (<20 nm). Smaller structures can be easily synthesized, but assembling them on substrates (for device applications) in an ordered fashion still remains as a challenge.4,5 The recent advancements in synthesis and self-assembly of polymer-based nanocomposite materials provides a useful approach to this sort of fabrication.6 – 8 In addition, the self-assembly process provides versatile ways to tailor physical and chemical properties of nanocomposite materials by varying the polymer chain length, nanoparticle (NP) size, number of assembly layers, and interparticle spacing between neighboring NPs.9 – 11 The non-covalent interactions that dictate selfassembly are hydrogen bonding, dipole–dipole, electrostatic, van der Waals, and hydrophobic interactions, between the polymers (supramolecules) and NPs. Selfassembly of NPs and polymers can also be achieved through covalent bonding but is beyond the scope of this review.12,13 Of the above non-covalent interactions, electrostatic and hydrogen bonding interactions play crucial roles in supramolecule–NP self-assembly. The binary nature of electrostatic interactions provides a straightforward approach to assembly, whereas the interest in hydrogen bonding-based self-assembly arose initially from biological system where such interactions led to the formation of well-ordered system, e.g. the formation of α-helix and β-sheet peptide structures. In this article, the use of hydrogen bonding for the self-assembly of nanocomposite materials is reviewed. NANOPARTICLE ‘BRICKS’ NPs provide the ‘bricks’ for polymer-mediated bricksand-mortar self-assembly. Synthetic routes are well established for the preparation of different types of NPs, including metal, metal oxides, chalcogenides, pnictides, and semiconductors, with controlled size and shape.14 – 17 The tunable size-dependent properties exhibited by these materials further enhance their utility as building blocks for self-assembly. The first step towards self-assembly is the functionalization of NP surfaces to provide controlled interactions with polymers and/or substrates, and is generally achieved in a two-stage process. Initially the NPs are synthesized with a passive organic monolayer which prevents particle aggregation. The next step involves complete or partial replacement of the passive organic layer with appropriate functional group(s) in what is known as a place-exchange process. This process has been demonstrated for variety of NPs including metal, metal oxide, and semiconductors.18 – 22 A typical displacement reaction is represented schematically in Fig. 1(a). POLYMER ‘MORTAR’ The next stage of the self-assembly involves introduction of functional/terminal groups in the polymer moiety (Fig. 1(b)) to provide an affinity towards ∗ Correspondence to: Vincent M Rotello, Department of Chemistry, University of Massachusetts, Amherst, MA 01002, USA E-mail: rotello@chem.umass.edu Contract/grant sponsor: NSF; contract/grant number: CHE-0518487 Contract/grant sponsor: University of Massachusetts Center for Hierarchical Manufacturing; contract/grant number: NSEC, DMI-0531171 (Received 4 August 2006; revised version received 6 September 2006; accepted 24 October 2006) Published online 6 February 2007; DOI: 10.1002/pi.2210  2007 Society of Chemical Industry. Polym Int 0959–8103/2007/$30.00 P Arumugam et al. Figure 1. General schematic of NP self-assembly. (a) Place exchange of NP surface with appropriate functional groups. (b) Addition of the complementary functional groups to polymers. (c) Electrostatic and/or hydrogen bonding interaction between NP surface and functional polymers leads to self-assembly of NPs. surface-functionalized NPs and to direct self-assembly (Fig. 1(c)). As with NPs, there are numerous strategies for polymer synthesis, including living radical polymerization (LRP) and ring-opening metathesis polymerization (ROMP).23 – 25 The ‘living’ nature of both of these processes allows the ready synthesis of block copolymers26,27 that can be further functionalized either before or after the polymerization process. The introduction of functional monomers during the polymerization process allows growth of polymers with versatile functionality. Nonetheless, there are challenges in terms of processability as the polymerization conditions and rates may differ for different functional monomers. As an alternative, functionality in polymers can be introduced after the polymerization process. Thus, using one polymer scaffold, a series of polymers with similar structures and varying functionality can be synthesized.28 – 30 In addition to linear polymers, highly branched macromolecules (such as dendrimers) can also be used as nanometric building blocks, as these molecules can offer wide range of functionality and structures.31,32 The chemistry in this area is well developed for accurate placement of functionality along the branching chains.33,34 DENDRIMER-MEDIATED SELF-ASSEMBLY OF NANOPARTICLES Precise control of the optical and magnetic properties of the NPs is essential for further progress in fields such as optical storage, optical computing, magnetic storage, magnetic refrigeration, and biosensors.35 – 37 Traditionally, these properties of NPs are affected by the particle size and/or shape. Through dendrimer-mediated self-assembly, Frankamp et al. demonstrated a novel way to regulate the interparticle spacing of gold38 and magnetic properties of iron oxide (γ -Fe2 O3 ) NPs.39 They fabricated thin films of gold nanocomposites using carboxylic acid-functionalized gold NPs (‘brick’) and amine-terminated polyamidoamine (PAMAM) 462 dendrimers (‘mortar’). The precise control of spacing between NPs by different dendrimer generations, generation G0 to G4, enabled them to blue-shift the surface plasmon resonance (SPR) of as-formed thin films. Thus, the modulation of dipolar coupling between NPs provides a new approach to tune the optical properties of nanocomposite materials.40 The same group has influenced the spacing between magnetic NPs (e.g. γ -Fe2 O3 ) again by the use of a PAMAM dendrimer39 and observed sequential changes in the blocking temperature (the temperature below which a superparamagnetic material shows ferromagnetic behavior) with an increase in dendrimer generations. The general sense of increase in interparticle spacing (d) between NPs upon dendrimer assembly is represented in Fig. 2. In another approach, Boal et al. developed a complementary technique, using a triazine-functionalized linear polymer (mortar) and γ -Fe2 O3 NPs (brick), to control the spacing between magnetic NPs in a polymer–NP assembly.41 First, γ -Fe2 O3 NPs of ∼6.5 nm diameter were synthesized by a literature method42 followed by a place-exchange reaction with thymine-functionalized diol ligand. The ligand favors hydrogen bonding interactions with triazine polymer and leads to self-assembly of γ -Fe2 O3 nanocomposites. Upon assembly, a decrease in blocking temperature was observed, compared to blocking temperature from nanoparticle precipitate, and attributed to the increased spacing between the particles. Thus, the controlled manipulation of interparticle spacing provides new dimensionality to influence the optical and magnetic properties (other than shape and size) of NPs. MULTIPOINT INTERACTION-INDUCED SELF-ASSEMBLY The reversibility, controlled affinity, and high specificity of hydrogen bonding-based self-assembly allow one to design ordered arrays of NPs. By using Polym Int 56:461–466 (2007) DOI: 10.1002/pi ‘Bricks and mortar’ nanoparticle self-assembly Figure 2. Schematic representation of sequential increase in interparticle spacing d upon assembly with different generations of PAMAM dendrimer. Figure 3. Multipoint hydrogen bonding-based self-assembly. (a) Six-point hydrogen bond formation between barbituric acid-functionalized gold NPs and Hamilton receptor-functionalized block copolymer (BCP). (b) Self-assembly of NPs on a thin film of microphase separated BCP. multipoint hydrogen interactions, two- and threedimensional self-assembled nanocomposites are generated and the NPs of such assembly are found to exhibit different physical properties compared to the isolated particles.43 Based on six-point hydrogen bonding interactions, Binder et al. controlled the binding of barbituric acid-functionalized gold NPs onto specific sites of microphase-separated block copolymers (BCPs)44 (Fig. 3). One block of the poly(oxynorbornenes) BCP was functionalized with a receptor to favor six-point hydrogen bonding with the NPs, while the other block was functionalized with a fluorinated side chain to favor microphase separation. The immobilization of NPs by the use of supramolecular interactions leads to generation of a new class of hybrid nanoscale materials with potential applications in multifunctional biosensors and novel electronic, mechanical, and photonic devices.30 As in the case with block copolymers, different functionality can be introduced onto predefined regions of surfaces to make the surface a suitable substrate Polym Int 56:461–466 (2007) DOI: 10.1002/pi for orthogonal self-assembly. Xu et al. demonstrated the formation of orthogonal self-assembly by taking advantage of electrostatic, hydrogen bonding, and hydrophobic interactions.45 First, they made a silicon wafer surface positive by spin casting poly(4-vinyl-Nmethylpyridinium iodide) (PVMP) on the wafer. Then spin-casting of a thymine-functionalized polystyrene (Thy-PS) overlayer was performed followed by exposing the surface to UV light under a photoresist mask. The overall process resulted in the generation of Thy-PS squares and positively charged PVMP lines (Fig. 4(a)).45 The generation of orthogonal patterns was then demonstrated by the use of diaminopyridinefunctionalized polystyrene (DAP-PS) and carboxylatederivatized CdSe@ZnS core–shell NPs (COONP). The three-point hydrogen bonding between diamidopyridine-thymine and electrostatic interactions between pyridinium-carboxylate led to the fabrication of self-assembled orthogonal patterns (Fig. 4(b). The selective depositions of NPs on specific 463 P Arumugam et al. Figure 4. Schematic representation of an orthogonal self-assembly process. (a) Functionalization of silicon wafer surface with PVMP and Thy-PS polymer. (b) Orthogonal pattern generation through PS-Thy:PS-DAP recognition and PVMP:COO-NP electrostatic interactions. sites of the polymer templated surfaces gives the possibility of generating complex self-assembled patterns and also provides new horizons for the hierarchical organization of nanostructure materials on surfaces. THREE-DIMENSIONAL SELF-ASSEMBLY: CONSTRUCTION AND INCLUSION OF TAILORED PROPERTIES The reversible nature of hydrogen bonding and electrostatic interactions allows one to tailor the morphology and functional properties of supramolecular nanostructures at the same time. Using diblock copolymers [P4VP(MSA)1 (PDP)1 and PS] composed of a coil-like block and a polymer–amphiphile complex block, Ruokolainen et al. controlled a polymeric microstructure at two length scales.46 They took advantage of the effects of hydrogen bonding at different temperatures to transform the microstructure morphology. As shown in Fig. 5(a), below 100 ◦ C, the P4VP(MSA)1 (PDP)1 and PS blocks form a lamellar structure (alternating layers) with a long period Lb ≈ 35 nm and further microphase separates into another lamellar structure with a period Lc ≈ 4.8 nm. In the 100–150 ◦ C temperature region, the second lamellar structure disappears along long period Lc , while retaining the first lamellar structure along long period Lb (Fig. 5(b)). This leads to the formation of an order–disorder assembly which can be further manipulated at higher temperatures (>150 ◦ C) to form an order–order assembly. The second lamellar structure converts into a cylindrical shape above 150 ◦ C. The system shows tunability of the protonic conductivity upon change in morphology. Shenhar et al. regulated the spatial distribution of guest molecules within block copolymer films using molecular recognition methods.47 Block copolymers, made of polystyrene in one block (apolar domain) 464 and polystyrene bearing 2,6-diamidopyridine (DAP) in another block (polar domain) (represented by PS-b-PS/DAP), were used as a polymer scaffold. Thymine (Thy), and N(3)methyl-thymine (MeThy) both bearing polyether dendron backbones were used as guests. Three-point hydrogen bonding between Thy and DAP regulated the distribution of the guest while the increase in dendrimer generation regulated the morphology of the block copolymer to lamellar, cylindrical, and spherical shapes (Fig. 6). Thus, hierarchical materials showing a wide range of applications were accessed by the integration of block copolymer self-assembly with molecular recognition processes. The possibility of using self-assembled nanocomposites for potential drug delivery systems was demonstrated by Thibault et al.48 where they utilized the complementary interaction between the DAP- and Thy-functionalized polymers to form recognitioninduced polymersome (RIP) vesicles. By making use of competition between monovalent (flavin) and multivalent guest molecules (Thy-Au) to bind to RIP, they controlled the stability of RIP. Tailored stability of RIP provided a platform for controlled release of guest molecules attached to it. CONCLUSIONS AND FUTURE DIRECTIONS Self-assembly of nanocomposites based on recognition between polymers and NPs provides new dimensionality in nanofabrication processes where the generation of high-performance and low-cost devices is of paramount importance. The ability to direct reversible hydrogen bonding and/or electrostatic interactions in one, two, and three dimensions, at predefined positions, provides immense flexibility to self-assembly processes. The tunable properties exhibited by these nanocomposite materials can be exploited in fields ranging from electronics to molecular biology. Polym Int 56:461–466 (2007) DOI: 10.1002/pi ‘Bricks and mortar’ nanoparticle self-assembly Figure 5. Graphical representation of the self-organized structures of PS-block-P4VP(MSA)1 (PDP)1 . (a) Alternating two-dimensional PS layers and one-dimensional P4VP(MSA)1 and PDP layers at temperature below 100 ◦ C. (b) Alternating PS and disordered P4VP(MSA)1 (PDP)1 lamellae in the temperature region 100–150 ◦ C. (c) One-dimensional disordered P4VP(MSA)1 (PDP)x (with x ≪ 1) cylinders within the three-dimensional PS-PDP above 150 ◦ C. (Reprinted with permission from Ruokolainen J, Makinen R, Torkkeli M, Makela T, Serimaa R, Brinke Gt et al., Science 280:557–560 (1998)). Figure 6. Strategy for molecular recognition-based self-assembly. (a) Three-point hydrogen bonding between 2,6-diaminopyridine (DAP)-functionalized block copolymers (BCPs) and thymine-functionalized dendron macromolecules. (b) Transmission electron micrographs show change in morphology of BCPs with change in dendron generations. Polym Int 56:461–466 (2007) DOI: 10.1002/pi 465 P Arumugam et al. While researchers have made considerable progress in understanding and harnessing self-assembly, the extent of complex structures that can be obtained through the self-assembly is limited. Simple entropy gain, cooperative non-covalent interactions, and nonspecific packing would be insufficient to obtain multicomponent complex structures. As such, enhanced understanding of the fundamental nature of selfassembly, coupled with new developments in polymer and particle synthesis should open up new avenues for the creation of functional materials. ACKNOWLEDGEMENTS The support of the NSF (CHE-0518487) and the University of Massachusetts Center for Hierarchical Manufacturing (NSEC, DMI-0531171) is gratefully acknowledged. REFERENCES 1 Bakueva L, Musikhin S, Sargent EH, Ruda HE and Shik A, in Handbook of Organic–Inorganic Hybrid Materials and Nanocomposites, vol. 2. American Scientific Publishers, Stevenson Ranch, California, pp. 181–215 (2003). 2 Shenhar R, Norsten TB and Rotello VM, Adv Mater 17:657 (2005). 3 Sanchez C, Soler-Illia G, Ribot F, Lalot T, Mayer CR and Cabuil V, Chem Mater 13:3061 (2001). 4 Cui Y and Lieber CM, Science 291:851 (2001). 5 Arumugam P, Shinozaki SS, Wang R, Mao G and Brock SL, Chem Commun 1121 (2006). 6 Leunissen ME, Christova CG, Hynninen AP, Royall CP, Campbell AI, Imhof A, et al, Nature 437:235 (2005). 7 Zeng H, Li J, Liu JP, Wang ZL and Sun S, Nature 420:395 (2002). 8 Boal AK, Ilhan F, DeRouchey JE, Thurn-Albrecht T, Russell TP and Rotello VM, Nature 404:746 (2000). 9 Gittins DI, Bethell D, Schiffrin DJ and Nichols RJ, Nature 408:67 (2000). 10 Taton TA, Mirkin CA and Letsinger RL, Science 289:1757 (2000). 11 Sandrock ML and Foss CA, J Phys Chem B 103:11398 (1999). 12 Peng H, Tang J, Pang J, Chen D, Yang L, Ashbaugh HS, et al, J Am Chem Soc 127:12782 (2005). 13 Zin MT, Yip H-L, Wong N-Y, Ma H and Jen AKY, Langmuir 22:6346 (2006). 14 El-Sayed MA, Acc Chem Res 34:257 (2001). 15 Masala O and Seshadri R, Annu Rev Mater Res 34:41 (2004). 16 Murray CB, Kagan CR and Bawendi MG, Annu Rev Mater Sci 30:545 (2000). 17 Brock SL, Perera SC and Stamm KL, Chem Eur J 10:3364 (2004). 18 Templeton AC, Wuelfing WP and Murray RW, Acc Chem Res 33:27 (2000). 466 19 Templeton AC, Hostetler MJ, Warmoth EK, Chen S, Hartshorn CM, Krishnamurthy VM, et al, J Am Chem Soc 120:4845 (1998). 20 Boal AK, Das K, Gray M and Rotello VM, Chem Mater 14:2628 (2002). 21 Hong R, Fischer NO, Emrick T and Rotello VM, Chem Mater 17:4617 (2005). 22 Peng XG, Wilson TE, Alivisatos AP and Schultz PG, Angew Chem Int Ed 36:145 (1997). 23 Chiefari J, Chong YK, Ercole F, Krstina J, Jeffery J, Le TPT, et al, Macromolecules 31:5559 (1998). 24 Hawker CJ and Wooley KL, Science 309:1200 (2005). 25 Bielawski CW and Grubbs RH, Angew Chem Int Ed 39:2903 (2000). 26 Hawker CJ, Bosman AW and Harth E, Chem Rev 101:3661 (2001). 27 Coessens V, Pintauer T and Matyjaszewski K, Prog Polym Sci 26:337 (2001). 28 Shenhar R, Sanyal A, Uzun O, Nakade H and Rotello VM, Macromolecules 37:4931 (2004). 29 Carroll JB, Jordan BJ, Xu H, Erdogan B, Lee L, Cheng L, et al, Org Lett 7:2551 (2005). 30 Malkoch M, Thibault RJ, Drockenmuller E, Messerschmidt M, Voit B, Russell TP, et al, J Am Chem Soc 127:14942 (2005). 31 Jikei M and Kakimoto M, Prog Polym Sci 26:1233 (2001). 32 Zimmerman SC, Zeng FW, Reichert DEC and Kolotuchin SV, Science 271:1095 (1996). 33 KukowskaLatallo JF, Bielinska AU, Johnson J, Spindler R, Tomalia DA and Baker JR, Proc Natl Acad Sci USA 93:4897 (1996). 34 Vogtle F, Gestermann S, Hesse R, Schwierz H and Windisch B, Prog Polym Sci 25:987 (2000). 35 Murray CB, Kagan CR and Bawendi MG, Annu Rev Mater Sci 30:545 (2000). 36 Spaldin N, Magnetic Materials: Fundamentals and Device Applications. Cambridge University Press, Cambridge, UK, p. 250 (2003). 37 Zhu M-Q, Zhu L, Han JJ, Wu W, Hurst JK and Li ADQ, J Am Chem Soc 128:4303 (2006). 38 Frankamp BL, Boal AK and Rotello VM, J Am Chem Soc 124:15146 (2002). 39 Frankamp BL, Boal AK, Tuominen MT and Rotello VM, J Am Chem Soc 127:9731 (2005). 40 Srivastava S, Frankamp BL and Rotello VM, Chem Mater 17:487 (2005). 41 Boal AK, Frankamp BL, Uzun O, Tuominen MT and Rotello VM, Chem Mater 16:3252 (2004). 42 Rockenberger J, Scher EC and Alivisatos AP, J Am Chem Soc 121:11595 (1999). 43 Huang C-H, McClenaghan ND, Kuhn A, Bravic G and Bassani DM, Tetrahedron 62:2050 (2006). 44 Binder WH, Kluger C, Straif CJ and Friedbacher G, Macromolecules 38:9405 (2005). 45 Xu H, Hong R, Lu T, Uzun O and Rotello VM, J Am Chem Soc 128:3162 (2006). 46 Ruokolainen J, Makinen R, Torkkeli M, Makela T, Serimaa R, Brinke Gt, et al, Science 280:557 (1998). 47 Shenhar R, Xu H, Frankamp BL, Mates TE, Sanyal A, Uzun O, et al, J Am Chem Soc 127:16318 (2005). 48 Thibault RJ, Galow TH, Turnberg EJ, Gray M, Hotchkiss PJ and Rotello VM, J Am Chem Soc 124:15249 (2002). Polym Int 56:461–466 (2007) DOI: 10.1002/pi