Supramolecular assembly of melamine and its derivatives: nanostructures to functional materials

Abstract

The last twenty years have witnessed increasing research activity in the area of supramolecular chemistry of 1 : 1 co-assembly of melamine (M)–cyanuric acid (CA), since the historic discovery of the M·CA aggregate in crystal form and its structural analysis by Wang and his coworkers in 1990. Its useful chemical structure and fascinating H-bonding interaction sites distinguish M and its analogous derivatives as scaffolding components in the field of supramolecular chemistry to develop desired nano-to-micro scaled architecture. To date, M-based supramolecular assemblies are known in diverse forms which include fascinating nano- to micromorphological structures, molecular guest boxes, small molecular gels, membrane, sensor and liquid crystal development, polymeric scaffolds etc. In this review, we have covered the development of M and its derivatives, encompassing both nano/micro-ordered structures and advanced functional materials.

---

Bappaditya Roy

Bappaditya Roy received his M.Sc. degree in Applied Chemistry from Bengal Engineering and Science University in 2008, and then he joined in Prof. Arun K. Nandi’s group in 2008 at The Polymer Science Unit, Indian Association for the Cultivation of Science as a CSIR fellow. He received his Ph.D. degree in Nov. 2012, thesis entitled “Supramolecular Assembly of Melamine and its derivatives”. Presently he is working as a postdoctoral fellow in Prof. Seiji Shinkai’s lab at Kyushu University, Japan. His current research interests focus on host–guest chemistry and molecular recognition.

Partha Bairi

Partha Bairi received his M.Sc. degree in Chemistry from Indian Institute of Technology (IIT) Madras, India in 2009. He joined Prof. Arun K. Nandi's group in July, 2009 at the Polymer Science Unit, Indian Association for the Cultivation of Science (IACS) as a CSIR fellow. His Ph.D. thesis “Studies on Photoluminescence Active Supramolecular Gels” is submitted for Ph.D. degree. His current research interests focus on energy transfer in supramolecular assembly and soft materials.

Arun K. Nandi

Arun K. Nandi obtained Ph.D. degree on “Studies on Polymer–Polymer and Polymer–Solvent mixing” and joined Chemistry Department, North Bengal University, Darjeeling. He did post doctoral work at Florida State University with Prof. L. Mandelkern in crystallization of polymers. In 1992 he was appointed at Polymer Science Unit at Indian Association for the Cultivation of Science and he is presently senior professor and head of the unit. His research interests focus on polymer blends, polymer crystallization, polymer and supramolecular gels, polymer nanocomposites, polymer grafting and biomolecular hybrids. He is author of more than 150 papers, and supervised 22 Ph.D. students.

---

Introduction

The term ‘supramolecule’ stresses the key role of non-covalent interactions over the covalent interactions in molecules focussed on by classical chemistry. The amazing alternation of covalent bond is successfully filled up by these non-covalent interactions, which play a major role in the self-assembly processes of supramolecular chemistry.^1–3 The term ‘non-covalent’ encompasses an enormous range of attractive and repulsive forces, amongst which π–π interaction, hydrogen bonding, metal–ligand interaction etc. are those most often considered in the self-assembly process.³ They have low bonding strength and are considered to be secondary interactions, but they operate in such a co-operative fashion that the overall strength of the self-assembly almost equals that of a covalent bond. π-Stacking is generally used to describe the aromatic ring interaction between neighboring molecules. In the early 90s, Hunter and Sanders proposed^4a a mathematical model describing the role of substituents in the π-stacking processes. According to their model, the electron-withdrawing groups reduce the negative quadrupole of the aromatic ring, and thereby favor parallel displacement, producing a sandwiched structure with another aromatic ring. Although, in general, π-stacking and π–π stacking are not differentiated much by users, recently Martinez and Iverson^4b briefly illustrated the differences between these two. In addition to this, hydrogen bonding⁵ is the attractive interaction between a hydrogen atom covalently bonded to an electronegative atom of a certain molecule and an electronegative atom of another molecule or chemical group. Supramolecular chemistry deals with areas beyond those of molecules, and focuses on systems comprised of discrete numbers of assembled molecular subunits or components.

1,3,5-Triazine, also called s-triazine, is a six membered heterocyclic aromatic organic compound. Melamine (M, 1; Fig. 1), in this respect, is a special class with a 1,3,5-triazine skeleton. At a glance, it contains 66% nitrogen by mass. It contains nine H-bonding sites, amongst which the sp² hybridized nitrogen atoms of the triazine ring of each monomer provide three unshared pairs of electrons, which can act as hydrogen bonding acceptors (A), and an additional six H-bonds can be shared (D) from the three exocyclic unsubstituted primary amine groups. This systematic donor–acceptor arrangement reveals M as three donor–acceptor–donor (D–A–D) groups with each side having a length of 4.8 Å.⁶ The aromatic planar ring structure also facilitates the π-stacking process during ordered structure formation in the self-assembly. The useful chemical structure and fascinating H-bonding interaction sites distinguish M and its corresponding derivatives as scaffolding components in the field of supramolecular chemistry that can be used to develop desired nano- or micro-scaled architecture.

Fig. 1 Chemical structure of melamine (1).

This review represents an overview of development of the field, from the commencement of supramolecular chemistry depending on the 1,3,5-triazine ring backbone of M and its related derivatives. To date, M-based supramolecular assemblies are known in a variety of forms, which include fascinating morphological structures, small molecular gels, molecular guest boxes, membrane and sensor development etc. The journey from the 1 : 1 co-assembly of melamine and cyanuric acid reported by Wang and his coworkers in 1990 to its modern applications in supramolecular chemistry, M is a long and far-reaching one.^7a

Nanostructures formation

Assembly of M with cyanuric/barbituric acid

M and cyanuric acid (CA), and also their derivatives, can form various types of aggregates characterized by different H-bonding patterns. A two-dimensional (2D) hydrogen bonded molecular network based on the complementary H-bonding interactions between M and CA was first reported by Wang and his coworkers in 1990 (ref. 7a) by resolving the crystal structure obtained from HCl solution. Whitesides and co-workers have recognized that these molecules can form three different sub-motifs in the solid state: (i) infinite linear tapes, (ii) infinite crinkled tapes, and (iii) finite cyclic rosette motifs (Scheme 1).^7b,c Construction of a well-defined H-bonded extended 1D network structure has been successfully achieved using the “infinite linear tapes” of the M·CA crystal structure. The extended D–A–D and A–D–A arrangements via H-bonding of M and CA make them an indivisible unit pair and the assembly and/or recognition of CA by M, and their derivatives, has been studied by a number of researchers both in aqueous and organic medium.⁸ By considering them as a useful complementary unit in the self-assembly process, researchers have been able to develop the desired architecture by modifying the outer part while keeping the basic core intact.

Scheme 1 Different supramolecular H-bonding architectures formed by M and CA (taken from ref. 7 and 8).

Subtle structural changes, either in the acids or in the amines, strongly affect the stability of linear or crinkled tapes in the solid state. Whitesides and co-workers have reported that derivatives of CA/barbituric acid (BA) and M assemble into an astonishing variety of structural motifs.^8f (Fig. 2) Specifically, minor variation in the structural subunits results in the direction of aggregation changing in different ways than that of simple 2D assembly. For example, M derivatives containing relatively small para substituents (X = F, C1, Br, I, CH₃), generate a linear tape structure (2).^8c In contrast, when the substituent is an ethyl ester, the product assumes a crinkled tape structure (3).^8f This may be due to the steric interactions between the ethyl ester moieties on adjacent M in the linear structure, interactions that apparently are alleviated in the crinkled motif. A cyclic hexamer (4) is formed in cases where the para substituent is a tert-butyl group i.e. X = tert-butyl. The relationship between functional group modulation in the monomeric units and the harmonizing effect in the multi-component aggregated states was thus an area of interest to the scientists.

Fig. 2 Set of H-bonded 2D supramolecular structures of substituted M and BA (taken from ref. 8f).

Molecular sheets

Zang et al. have reported the synthesis of 2D organic micro sheets⁹ using dialkylated M derivatives with halogen acid (HX). Pure M, with HX, produces a 3D structure through multiple hydrogen bonding interactions, as well as π–π and electrostatic interactions. Alkyl groups present in the dialkylated M derivatives block the interaction of hydrogen bonds and restrict growth in the corresponding dimension, which leads to the formation of 2D micro-sheet structure (Fig. 3). The assemblies of 2D microsheets have lamellar structure and the lamellar d-spacing varies with the length of the alkyl chain. The aspect ratio of the organic micro sheet can be varied with the addition of different amount of acids and the micro sheet can be reversibly tuned by alternating addition of acid and base.

Fig. 3 (a) The 3D network structure produced from protonated M through multiple H-bonds, counterion bridging and π–π stacking and (b) 2D microsheets produced from dialkylated melaminium derivatives through chloride ion bridged hydrogen bonds, as well as electrostatic interaction and π–π stacking, while alkyl chains are introduced to eliminate the interaction of multiple hydrogen bonds in the previous 3D network (taken from ref. 9).

Channel structure

Of many interesting H-bonding systems, M·CA adducts have been investigated rigorously and the results are fascinating, as M·CA forms H-bonding on both sides. CA can also form H-bonded adducts with other molecules, and bis(4-pyridyl)ethene and 4.4′-bipyridyl^10a,b are among them. The resulting network forms layered structures of different shapes^7b and the rosette aggregate can have a different structure, which was first established by Ranganathan et al. in 1999.^11a The 1 : 1 adduct of M·CA, grown under hydrothermal conditions, has an asymmetric unit consisting of superimposed M and CA. The crystal structures of CA and tri-thiocyanuric acid (TCA) adduct with melamine form a hexamer arrangement in 2D planar sheets, and these molecular 2D sheets can further stack in either direction perpendicular to the hexamer to produce molecular channels (Fig. 4a), comparable with the cavities in cryptands. The best stacking (50%) is reported to be alternating stacking of CA/TCA with M, while the least stable is reported to be regular stacking of CA with CA or M with M. The diameter of these channels is approximately 4 Å.

Fig. 4 Three-dimensional stacking arrangement of (a) M·CA and (b) M·U adducts forming channels along the crystallographic a-axis (taken from ref. 11a and b).

In 2007, Thomas and Kulkarni showed that Uracil (U) can also form a rosette structure, resembling the M·CA adduct, and also gives rise to channel-like structures in the perpendicular direction.^11b In order to perceive the H-bonded superstructure of melamine–uracil, an equimolar mixture (1 : 1) in ethanol was made and evaporated slowly to develop the co-crystals. On close inspection, the crystal structure of melamine and uracil produced an asymmetric unit with a C2/c space group. The triply bonded melamine–uracil pairs are stacked in parallel and the chains run along both a and b axes. Such an arrangement of the chains produces an aperture structure, as shown in Fig. 4b. Four M and two uracil molecules thus enclose the aperture. The apertures are connected through N–H⋯O (H of –NH₂, M and O of >C=O, U) bonds giving rise to infinite supramolecular H-bonded channels along both a and b directions. These are shown from one perspective in Fig. 4b, where the channels run along the a direction. The shape and dimensions of M·U channels are different to those of M·CA rosettes, which demonstrates clearly the potential of developing molecular directed channel structure.

Nanoropes and nanotubes

A perylene bisimide (PBI) anchored with M (5) units has been used to develop supramolecular polymerization upon binding with N-dodecylcyanurate (dCA, 7) via H-bonding by Yagai and co-workers.^12a The resulting flexible supramolecular polymer self-organizes via π-stacking interaction between the perylene chromophores, affording ribbon-like aggregates in cyclic alkanes and rope-like aggregates in acyclic alkanes that form gels (Fig. 5). In the case of the former, the formation of the nanoobjects occurs quantitatively, detected by absorbance spectra and optical microscopy. These one-dimensionally elongated nanoobjects consist of electronically active chromophores, which can be applied as nanoscale conducting wires.

Fig. 5 Representation of the chemical structures of different M and CA derivatives (taken from ref. 12).

They have also constructed and tuned the optical properties using a M-linked oligo(p-phenylenevinylene), OPV (6) unit as a π-electronic supramolecular building block with three different cyanurates (7).^12b They have used ddCA (7c), dCA (7b) and CA (7a), which are monotopic, ditopic and tritopic triple H-bonding molecules, respectively. The OPV capped monotopic triple H-bonding site of M shows affinity with the complementary mono-to-triple H-bonding cyanurates and form assemblies in exact ratios of 1 : 1, 2 : 1 and 3 : 1, respectively. The assembly process has been studied successfully with the help of the UV-titration method. At higher concentration of the melamine-linked OPV in methyl cyclohexane solution, the dimeric unit is inert in the presence of cyanurates, which is detected by the λ_max shifting with the addition of an equivalent amount of different cyanurates. The complementary molecules follow stoichiometric/quantitative formation of the D–A–D to A–D–A H-bonding pairs in the assembly process.

The self-assembly of the two complementary molecules, 5-(4-N-methyl-N-dodecylaminobenzylidene)-2,4,6-(1H,3H)-pyrimidine-trione and 4-amino-2,6-didodecylamino-1,3,5-triazine, in chloroform is reported by Li and coworkers.¹³ The molecules form a hollow supramolecular nanotube with a diameter of 6 nm and the hollow nanotube shows a supercoil structure of 300 nm in diameter on staining (Fig. 6). They have explained the nanotube formation with the help of X-ray diffraction patterns of the self-assembled state. The diffraction pattern showed strong peaks at a 2θ value of 17.5°, which indicates a periodic gap between the two parallel cyclic hexamers of 4.9 Å. The nanotube fiber formation can be summarized as two phases. First, the cyclic hexamer is formed based on the network of triple H-bonds between 3 and 4. In the next step, during the nanotube formation, the cyclic hexamers stack together layer by layer. In this stacking procedure, the expression of chirality in the hexamers prevents production of a nanotube, as is evident from the TEM image. However, when staining is applied to the nanotubes, the interaction between the hexamers in the nanotubes became flexible and the chirality is expressed in the generation of a mesoscopic supercoil.

Fig. 6 (a) Schematic representation of (left) the supramolecular cyclic hexamer and (right) the stacking of the hexamers into the nanotube and TEM photographs of the self-assembly observed (b) without staining treatment and (c) after staining by a saturated solution of uranyl acetate for 15 min (taken from ref. 13).

Columnar structure

Amongst the various supramolecular synthons which can be used for bottom-up construction of multi-component nanostructures formed by the so-called M·CA cyclic hexamers, π-stacking units play an important role. Beyond the framework of the discrete M·CA rosette architecture, some rosettes are designed to assemble into extended columnar structures.¹⁴ Recently, Kitamura and his coworkers have reported^14a the construction of well-defined columnar assemblies decorated with functional π-systems, using M·CA rosette 2D hexamers as primary H-bonded motifs. Perylene-3,4:9,10-tetracarboxylic acid diimide is functionalized with ditopic cyanuric acid (CA·PBI, 8) to produce columnar structures (Fig. 7). The incorporation of additional non-covalent interactions, e.g. amide or urea H-bonding networks, in the parallel direction to the π–π stacking axis would result in the PBI stacks forming stable assemblies in solution. In contrast, CA can provide more directional multiple H-bonding interactions in the orthogonal direction to the stacking axis. Various spectroscopic and microscopic investigations of the complex of CA-functionalized PBI dye (CA·PBI) and the complementary melamine possessing aliphatic chains show that the complexes form well defined columnar structures in organic solvents, like methyl cyclohexane (MCH), under dilute conditions (Fig. 7). The nanostructures of the CAPBI (9b) complex were observed by transmission electron microscopy (TEM) and atomic force microscopy (AFM) in a dilute MCH solution at a concentration of 1 × 10⁻⁴ M. The images displayed hierarchical organization of the elementary fibrils into higher order bundled structures.

Fig. 7 Structures of CA-functionalized PBI (CAPBI, 8) and substituted melamines (9,10) (taken from ref. 14a).

On increasing the concentration of the CA·PBI complexes in the MCH solution (C > 5.0 × 10⁻² M), a lyotropic liquid crystal phase is obtained. An interesting thermal convertibility of the columnar structure to the lamellar structure in the bulk state has been reported and the transformation of the nanostructure is revealed by the XRD study. A large structural change from columnar to lamellar structure via the formation of an intermediate phase has also been confirmed by thermometric study. This work also indicates the effect of substituents on the self-assembled structure formation.

The 2D rosette structure is a requirement for the columnar structure formation, but the stacking process requires some additional elements to form a favorable arrangement. The residual entities attached to the primary M or CA units are responsible for ensuring stacking for structure formation (Fig. 8). Kitamura and co-workers have found some components which could serve as building blocks to form the rosette structure, which is essential for columnar structure formation.^14e,15 They have used azobenzene-appended M (11) and more sterically N-dodecylcyanurate (12) to synthesize rosette sheets which can further stack in a face-to-face arrangement to form a hexagonal columnar mesophase in the bulk state and hierarchically organize into elongated fibrous aggregates (Fig. 8). The use of N-dodecylcyanurate instead of bulky tridodecyloxyphenyl (13) substituent provides peripheral flexibility to stabilize the columns. They have also followed the conventional method by using photochromic modules (azo group) to control the molecular self-assembly, and the transformation from columnar to rosette sheets can be triggered externally by light illumination.

Fig. 8 Chemical structures of functionalized melamine, barbiturate and cyanurate molecules (taken from ref. 14e and 15).

Molecular boxes

Development of molecular guest boxes is a fascinating idea to build a cage-like structure using non-covalent bonding. Reinhoudt et al. have reported that three calix[4]arene dimelamine derivative molecules and six diethylbarbiturate (DEB) molecules produce molecular receptor (boxes) through self-assembly (Fig. 9).^16a,b This non-covalent molecular receptor selectively encapsulates different anthraquinone derivatives in a highly organised manner. The driving force of encapsulation is π–π stacking between the electron-deficient centre ring of the anthraquinone derivatives and the relatively electron-poor two melamine units of the receptor, which is explained by the Hunter–Sanders model for π–π stacking. After encapsulation of anthraquinone derivatives, the conformation of the receptor changes from staggered D₃ symmetry to eclipsed C_3h symmetry in the solid state, as well as in the solution state, which is evident from the single crystal X-ray and NMR studies. Complex formation and its binding energy largely depend on the position of substitution on the anthraquinone molecule.

Fig. 9 General molecular structure for guest box including the top and side view of the box (taken from ref. 16).

It has previously been found that these supramolecular assemblies form a racemic mixture of (P)- and (M)-enantiomers that could be separated as diastereomers by using chiral barbiturates (Fig. 10), and the chiral barbiturates can then be replaced with achiral cyanurates to give the enantiomerically pure species.^16b,d Shinkai and co-workers have subsequently found that appending each calix[4]arene moiety with two 2-pyridyl functionalities, attached to opposite phenyl rings, allows the pyridyl groups to interact with external chiral dicarboxylic acids.^16e Thus, when 3 equiv. of an enantiopure dicarboxylic acid, such as dibenzoyl tartaric acid, is added to the racemic rosette, there exists the possibility of formation of diastereomers. Each diacid binds via interaction with two of the six 2-pyridyl moieties, which, from Corey–Pauling–Koltun (CPK) model analysis, is situated across the two halves of the rosette in a clip-like manner.

Fig. 10 Enhancement of enantioselectivity by incorporating chiral dicarboxylic acid (taken from ref. 16e).

Supramolecular chirality

The assembly of achiral M with achiral complementary components can result in chirality in the self-assembled structures, often known as supramolecular chirality. That is to say, the chirality of achiral molecules could be formed by the intermolecular interactions in a chiral matrix or environment, no matter whether the chiral environment is composed of chiral molecules or not.^13b,17 Generation of chirality of achiral molecular assemblies induced by chiral molecules is described often, but, for the first time, in 2005 Huang et al.^17areported on the chirality of molecular assemblies produced by achiral molecular entities at the air–water interface. A Langmuir–Blodgett (LB) film of achiral melamine–barbituric acid (M·BA) is formed through in situ complementary H-bonding at the air–water interface. The complex LB film has shown supramolecular chirality, which has been verified by circular dichroism (CD) spectral measurements. It is well known that when achiral molecules interact with chiral molecules or are put into chiral environment, induced CD can be observed in the absorption band of the achiral molecules.¹⁸

When the M·BA film is subjected to CD spectroscopic studies, a cotton effect at 212 nm is found. The absorption band at this region is assigned to the localized absorption band of melamine. This CD band at 212 nm appears to be due to the excitonic coupling among M aligned in the linear complementary hydrogen-bond networks. The CD spectra may be attributed to the overcrowded and cooperative stacking of the triazine ring molecules which form H-bonding with BA components and adopts a non-coplanar arrangement (Fig. 11).

Fig. 11 Upper, linear tape formation of M·BA complex by complementary hydrogen bonding at the air–water interface; below, the helical linear tape of M·BA: side view (left) and top view (right) (taken from ref. 17a).

In another interesting work, Kimizuka and his coworkers^17b have used a complementary pair of quaternary ammonium-derivatized cyanuric acid, as the hydrophilic subunit, and alkylated melamine, as the hydrophobic subunit. The azobenzene chromophore is introduced as aromatic moieties enhance the stability of the preformed H-bond network into water¹⁸ and, in addition, the chromophoric unit (azobenzene) displays unique spectral characteristics depending on its orientation, i.e. photo-isomerization, in the presence of external light source. While injecting ethanolic solution of 15 into the aqueous solution of 14 at an equimolar ratio, helical superstructures are formed. The helical superstructures are 14–28 nm in thickness and 30–50 nm in width, having pitches of 180–430 nm. When CD spectrometry is used to investigate the generation of chirality, the achiral component 15 failed to generate a CD signal in water. On the other hand, upon addition of subunits 10(L-) to the aqueous solution of 15, two intense couples of positive and negative cotton effects appear in the spectrum. A mirror symmetric CD spectrum is obtained when 14(D-) is employed in place of the L-form. The strong excitonic interaction among the π–π* transition dipoles of melamine and azobenzene chromophores and the blue shift of azobenzene absorption observed upon mixing the subunits (Fig. 12) are ascribed to the parallel orientation of the chromophores in the assemblies. Also, the appearance of the intense exciton-coupled induced circular dichroism (ICD) confirms the regularly fixed orientation of the azobenzene against the glutamate moiety in M. The observed absorption blue-shift and the ICD intensities are less eminent compared to those of the in situ formed assembly, i.e. the assembly formed by mixing the components in water. By mixing of suitably designed complementary units in water, amphiphilic networks of linear complementary H-bonds of 14–15 are spontaneously formed. The presence of bulk water directs the supramolecular organization, in which the more hydrophobic melamine subunits are organized in the interior of the assembly while the exterior part is decorated by the tertiary ammonium containing hydrophilic subunits. This amphiphilic supramolecular organization satisfies both the solubility and the complementary H-bonding network in aqueous solution. As the bilayer is formed by the injection of ethanolic solution of M into the aqueous solution of cyanuric acid derivative, Kimizuka and coworkers checked the solvophobic effect in the organization process and the highest molecular orientation has been found at an ethanol content of 40 vol%. This improved molecular alignment in the presence of ethanol may be a consequence of the enhanced hydrophobicity and crystallinity.¹⁹ Strong π-stacking of the complementary H-bonding networks and of azobenzene chromophoric units in the complex might be relaxed with moderate ethanol concentration, which lowers the dielectric constant of the solvent and allows their ordering in the assemblies.^17b

Fig. 12 Chemical structures of the components (14 and 15) and schematic illustration of the hybridization of amphiphilic complementary H-bonded network in water, and their hierarchical self-assembly into the supramolecular membranes (taken from ref. 17b).

The origin of a helical columnar architecture resides in the structure of the molecule itself, independent of whether it is chiral or not, however, the generation of supramolecular chirality often lies behind the formation of a twisted molecular assembly. Design of the interaction between the M derivative and the acid functionalized banana like molecule with unique molecular packing can result in helicity along the stacking direction, allowing other supramolecular interactions to occur.²⁰ The construction of a column like structure has been achieved by Sierra and his co-workers. They selected 4-diamino-6-dodecylamino-1,3,5-triazine (T) as a disk at the centre of a radial design, so that stacking within the mesophase would be possible during aggregation of the complexes. The M derivative offers H-bond interactions in a 3-to-1 ratio (acid-to-melamine), thus allowing construction of a complex with the V-shaped acids (n = 1–6). The smart introduction of the dodecyl alkyl chain into the M units not only increases the solubility in different solvents, but also creates a negligible amount of steric crowding without affecting the complementary association. The V-shaped acids rotate through a certain angle with respect to the triazine plane. In this way, the acid arms can lie up and down with respect to the triazine core (Fig. 13a). Such an arrangement could give rise to a columnar structure consistent with propeller-like supramolecules interpenetrated along the column, in which the tilted V-shaped acids of neighboring complexes can interact in a way parallel to each other. Compact columnar packing should then occur by the formation of a helical arrangement within the column (Fig. 13b).

Fig. 13 (Above) Tetrameric complex and (below) A–C, the helical columnar organization within the mesophase. Both enantiomeric stacking possibilities, right-handed (P) and left-handed (M), are shown (taken from ref. 20).

Supramolecular chirality is also generated by Ag(I) coordinated M molecules at a molar ratio of 1 : 2 in aqueous medium. The detail of this work is briefly described in the “Metal selective gelation” section.

Assembly with mono/diimides

The A–D–A arrangement, similar to cyanuric/barbituric acid, is also seen in the aromatic imides,which may be mono or diimides. Imides have thus been chosen as alternate and useful H-bonding moieties to M. Kimizuka et al. have reported^21a sonication induced supramolecular assembly of M derivatives and the aromatic imides in organic solvent. After collecting the powder from the evaporation of equimolar mixture of diimides (1 mM) in dimethylsulphoxide and M derivative (1 mM) in ethanol, this is transferred into MCH solution, and the mixture is sonicated. They planned to produce the rosette like H-bonding assembly, as mentioned earlier for the cyanurate/barbiturates, with a 1,8:4,5-napthalenebis(dicarboximide) and ditopic M modified with long aliphatic chains (Fig. 14), and from TEM microscopy bundles of 100 Å wide strands are found to be abundantly present along with twisted helical bundles recognized by AFM study. The 1 : 1 complexes of melamine and the diimide in MCH solution have produced a broad absorption peak centered at 370 nm with significant decrease in absorption intensity with respect to the UV-vis spectra of diimide in DMSO solution. This hypo-chromism and the broadening of the absorption peak are associated with the aromatic chromophoric stacking that was reported earlier by some research groups.^21b,c The H-bonded networks between M and diimide were extensively characterized by FTIR study and they have proposed the aggregate-geometry as three possible architectures, e.g. (i) cyclic structure, (ii) helical growth structure and (iii) linearly extended structure. The tubular structure generated by the columnar stacking of a cyclic heterododecamer of melamine derivatives and diimide contains a circular combination of the two components via complementary H-bonds. In contrast, aromatic stacking of linear tapes in the extended linear arrangement exhibit a sheet like structure with a 50 Å width. The essential extended H-bonded network is absent when monoimide is used instead of diimide. Similarly, the 1 : 2 complex of M derivative and monoimide, prepared from ethanol with the same procedure, does not form a stable dispersion in MCH. The general strategy described by them has thus been extended to design and synthesis of a wide range of mesoscopic superstructure analogous to the tobacco mosaic virus.

Fig. 14 Chemical structures of the M and naphthalene derivatives (taken from ref. 21a).

Würthner's group, being inspired by this type of M-imide interaction, have developed a mesoscopic superstructure via π–π stacking and H-bond formation. They have also studied the binding constants of the functional perylene chromophores with mono and/or ditopic melamine.^22a The process of superstructure formation has been shown to involve multiple orthogonal intermolecular interactions, appropriate solubilizing substituents, and a solvent of low polarity (Fig. 15).

Fig. 15 Concept for hierarchically organized functional superstructure based on orthogonal intermolecular forces (taken from ref. 22a).

The complementary supramolecular assemblies of imides with N²,N⁴-didodecyl-1,3,5-triazine-2,4,6-triamine in a ratio of 1 : 1, 1 : 2 and 1 : 3 have been investigated with the help of crystal engineering by Meijer and co-workers.^22b They have demonstrated that the availability for H-bonding of the carbonyl groups in the imides is strongly influenced by subtle differences in its molecular structure i.e. the structural variation of the molecule complimentary to N²,N⁴-didodecyl-1,3,5-triazine-2,4,6-triamine has large consequences. Three different imides, (i) succinimide, (ii) glutarimide and (iii) 1-N-propylthymine are used to develop supramolecular structures with M (Fig. 16). A single crystal grown in water or DMSO shows that the 1 : 1 complex of M and succinimide has formed a sheet-like structure, glutarimide has produced a 1 : 2 complex with herringbone structure and the third one, i.e. 1-N-propylthymine, formed a 1 : 3 C₃-symmetrical structure. The availability of the carbonyl groups in the imide components allows tuning of the supramolecular structure with melamine. These results have inspired researchers to design and arrange complementary molecules with the aim of forming giant supramolecular architecture.

Fig. 16 X-ray crystal structures of the 1 : 1, 1 : 2 and 1 : 3 complexes of melamine with (a) succinimide, (b) glutarimide and (c) 1-N-propylthymine, respectively (taken from ref. 22b).

2D assembly on surface

Recently, the concepts of supramolecular organization have been applied to 2D assemblies on surfaces. Supramolecular organization using selective and directional non-covalent interactions has mostly been applied in solution and in crystalline phases.²³ In two dimensions and at surfaces, these organizations had not been fully exploited until the development of the scanning tunneling microscopy (STM) technique motivated researchers to understand the factors controlling the supramolecular ordering, in order to control the outcome of the self-assembly process. STM is an instrument for imaging surfaces at the atomic level and it is based on the concept of quantum tunneling. A metallic tip is brought very close to a conductive substrate for scanning. The surface height and electron density causes changes in current which are usually displayed as an image. The initial applications of STM have dealt with imaging of semiconductor, inorganic and metal surfaces and, later on, experiments on molecules adsorbed on surfaces began. The method requires conductive substrates for immobilization of the molecules and, due to the distance dependence of the tunneling process, only very thin layers—typically monomolecular—can be probed. The molecules are deposited on the surface under sublimation conditions. De Feyter and De Schryver have briefly described in their review²⁴ that STM is a useful tool for probing the two dimensional assemblies used for construction of supramolecular architecture.

M, in this respect, is used as a constructive component, with its tritopic H-bonding nature, in building two dimensional supramolecular assemblies. There are few reports in support of the arrangement of M molecules on the conductive surface.²⁵ Silly and co-workers^25a have shown in high resolution that the M arrangement is stabilized by a double H-bond, although their result is in conflict with some other reports.^25b,c They have used a fast Fourier transformation of the STM image obtained from a monomolecular layer of M on a gold surface (Fig. 17), and the transformed image (inset: Fig. 17a) supports that the M 2D hexagon centers are aligned in the [32̄1̄] direction with a center–center separation measured in real space of 9.6 ± 0.5 Å.

Fig. 17 STM image of a M domain on a Au(111) surface: (a) 24 × 32 nm², with the dotted line highlighting the M hexagon center alignment (inset: fast Fourier transform (FFT) of the image); (b) 15 × 10 nm², with the network unit cell indicated (inset: high resolution STM image (3 × 3 nm²)). V_s = −1.0 V, I_t = 0.5 nA. (c) Model of a M molecule (gray balls are carbon atoms, white balls are hydrogen atoms and blue balls are nitrogen atoms). (d) Model of M ordering [taken from ref. 25a].

As expected, M can also form a bimolecular layer with CA^25,26 on a metallic surface under high vacuum, and the assembly successfully develops a bimolecular layer forming a honeycomb-like network. The molecular recognition on the surface between M and CA also forms a supramolecular 2D sheet-like structure, in the same way as in solution or in bulk state. The complementary triple anti-parallel H-bond gives an idea about the initial growth procedure. Besenbacher and coworkers^25c have studied the variation in the complex composition during deposition of the molecules on the metallic surface. Simultaneous deposition of M and CA leads to the formation of M₁·CA₁ structure, whereas M₃·CA₁ structure is formed through sequential addition of CA after M. They have used a theoretical model from the SCC-DFTB method to explain the stability of the different networks observed and to provide insights into the intermolecular interactions leading to the layer formation (Fig. 18).

Fig. 18 (a) STM image of a self-assembled network resulting from simultaneous deposition of M and CA on Au(111). (b) Optimized model for CA₁M₁ network. The hexagon marked “2” in (a) and (b) shows corresponding areas as a guide to the eye (taken from ref. 25c).

Again, perylene tetra-carboxylic di-imide (PTCDI) with its monotopic straight edges can also form networks with M. Beton and co-workers²⁷ have demonstrated that the PTCDI–M forms a H-bonded network on a silver terminated silicon surface. M and PTCDI are chosen for this application because they are expected to exhibit much stronger hetero-molecular hydrogen bonding compared to homo-molecular hydrogen bonding. In Fig. 19, the 2D growth of the PTCDI–M complex shows open honeycomb-like hexagonal molecular networks resulting from three hydrogen bonds per PTCDI–M pair, as compared with just two for each PTCDI or M pair. The open areas in the networks are much smaller and comparable in size when M itself is used as a single molecule to develop a supramolecular array (cf. description earlier). The incorporation of PTCDI into the bimolecular supramolecular building blocks results in much larger pores that can serve as traps, or vessels, for the co-location of several large molecules. The molecular network design can be tuned using carefully chosen component ratios and the temperature during sampling of the components on the metal surfaces at high vacuum. This is particularly efficient when various molecular interactions are possible. The epitaxial relationship between the molecule and the substrate surface can also play an important role. All these parameters can considerably regulate the supramolecular arrangements. Prediction of the supramolecular structure is thus complicated. The previous example of PTCDI–M pair exhibits an open non-chiral honeycomb-like structure at a molar ratio of 3 : 2. In contrast, after a few years, Silly et al.^25a observed a chiral “pinwheel” structure (Fig. 20) of the PTCDI–M assembly in a 3 : 4 ratio on Au(III) surfaces. They have also investigated the variation of the epitaxial relationship between the molecules and the substrate for single molecular networks and supramolecular networks. The domains of hexagons are composed of six M molecules, but the incorporation of PTCDI produces a chiral superstructure. These results open the door for materials researchers to investigate supramolecular assemblies by tuning the ratios of the participating components.

Fig. 19 Self-assembly of a PTCDI–M supramolecular network. (a) Chemical structure of PTCDI (b) and M and (c), schematic diagram of a PTCDI–M junction (taken from ref. 27).

Fig. 20 (a) STM image of mixed PTCDI and M “pinwheel” domain on a Au(111) surface. (b) The proposed supramolecular structure model (taken from ref. 25a).

M in crystal engineering

Apart from the most intriguing assemblies involving M with CA forming the rosette architecture, there is a large field left for researchers interested in crystal engineering. Over the last few years, attention has been directed towards M forming co-crystals with aromatic acids,^28a,b like benzoic acids, di/trimesic acids, hydroxy benzoic acids etc., and also with tartaric acid,^28c,d and this has been studied extensively. In order to produce self-assembled supramolecular architectures, an important and simple method is the use of acid-functionalized small molecules with N–H⋯O, O–H⋯N H-bonds and other weak intermolecular interactions that can create 1D, 2D and 3D networks in crystalline solids. M, in the solid state, can form complementary arrays of N–H/N hydrogen bonds resulting in a 2D network and, depending on the extent of protonation in the triazine ring, it can also add different dimensionalities. Mono-protonation of the triazine ring can restrict the growth to1D only, while di-protonation can result in a 0D hydrogen bonded network (Fig. 21).²⁹

Fig. 21 Different types of H-bonding pattern presented by M molecules in different dimensions depending on the extent of protonation (taken from ref. 29).

In an exploratory work, Pedireddi and co-workers³⁰ compiled an exhaustive and systematic study of the host–guest structures of 2,4-di-amino-6-methyl-1,3,5-triazine (Me–M, 23) with various dicarboxylic acids (24a–i) to produce supramolecular assemblies with a variety of exotic architectures. The crystal structure of Me–M shows fascinating packing of the molecules in three dimensions, and it forms a host–guest type assembly. The arrangement shows a channel-like structure, where the channels are occupied by the solvent molecules. However, the molecular complexes formed by co-crystallizing with different dicarboxylic acids provide two types of host network and all the complexes serve as representative examples of host–guest assemblies. The differences are attributed to variations in the pK_a values of the acid molecules under consideration. The existing host structure is composed only of molecules of Me–M when the pK_a of the acid (24c,d,f,g,i) is >3.0, whereas a host structure comprised of both Me–M and the carboxylate is observed (24a,b,e,h) when the pK_a is <3.0 (Fig. 22). These observations may be significant for the development of specific types of H-bonding patterns, such as ionic and neutral, by using different types of species.

Fig. 22 Chemical structures of Me–M (23) and different types of aliphatic dicarboxylic acids (24a–i) (taken from ref. 30).

Champness and co-workers³¹ have worked extensively with molecules having high molecular weight and low solubility and they have established the structural chemistry of these with M, the so-called best supramolecular synthon. They have targeted hydrogen-bonding motifs consisting of multiple hydrogen bonds between aromatic heterocyclic compounds (tecton, 25a–c), which offer the possibility of directionality and tunable strength. However, they have been able to produce two co-crystals M·25a and M·25b, (25c did not form co-crystals with M), typically with DMSO and benzene. The complementary supramolecular H-bonds between uracil/thymine moieties and M; both in M·25a and M·25b, are certainly responsible for driving the assembly of those co-crystals. In M·25a, there is a 1 : 1 stoichiometry of the two main components, and they combine to produce a 2D layer structure, despite the presence of DMSO molecules. In the crystal structure of M·25b, there is also a 1 : 1 stoichiometry of the main components and, again, the overall result of the complementary H-bonds between 25b and M is a 2D layer (Fig. 23). The formation of 2D H-bonded sheet-like structures by these combinations of molecules raises the possibility of using such tectons for surface-based self-assembly studies.

Fig. 23 (A) Chemical structures of three different tectons (25a–c) having complementary moieties of M, and presentation of the H-bonded sheet like structure formation of (B) M·25a and (C) M·25b (the oxygen atoms of DMSO molecules are represented as yellow spheres, with the rest of the DMSO molecule omitted to aid clarity) (taken ref. 31).

Crystal engineering, in its true sense, has been employed stoichiometrically with different imides, e.g. (i) succinimide, (ii) glutarimide and (iii) 1-N-propylthymine, in order to develop supramolecular structures with M by Meijer and co-workers.^22b The design of the crystallization has been driven by restricted groups of the related complexes and the preferred stoichiometric ratio (cf. “Assembly with mono/diimides” section).

Melamine-based supramolecular gels

Small molecular gelators (SMGs)³² have gained the attention of investigators, due to their interesting structural divergence and resultant properties in the bulk state. Recently, there has been significant research interest in the various types of complementary molecules that can act as SMGs and, in part, attempts to clarify the structural features responsible for gelation are also in progress. Usually this sol–gel transformation requires some external strategy e.g. heating–cooling,^32a,33 sonication,³⁴ irradiation,³⁵ external binding agent (metal, salts, acid–base or ions)³⁶etc. The gel properties depend on various factors, including solvent, temperature, composition, structure of gelators, and even the use of different isomers.³⁷ Understanding of the underpinning mechanisms and various influencing factors at the basic level is, therefore, of particular significance for fabricating smart small molecular gels. These small molecular gels are generally produced by supramolecular i.e. physical interactions, and that is why these are usually known as physical gels. The self-assembly of small molecules or macromolecules in a solution entraps a large amount of solvent, immobilizing it and producing the gel. The high surface area of the nanostructures developed in the self-assembly process is the main cause of physical gelation.

Hydrogel

Recently, riboflavin–melamine (R·M),³⁸ melamine–uric acid (M·U),³⁹ gallic acid–melamine (G·M),⁴⁰ and some other bicomponent thermoreversible hydrogels⁴¹ have been reported. The hydrogels are produced by using M as a synthon, because of its symmetrical, planar and rigid structure, which helps to produce the 3-D fibrillar networks in solution through supramolecular interactions. The R·M bicomponent system, in this context, was first reported by our research group in 2006.^38a In this report, we recognized the 3 : 1 composition as the most stable compared to the other compositions, showing maximum absorbance in the Job's plot. The result is quite predictable, as R has monotopic H-bonding arrangement with which M forms 3 : 1 complex. The 3 : 1 composition produces a physically rigid gel, but in the other compositions, e.g. 1 : 1, 1 : 2, 1 : 3, 4 : 1 and 8 : 1, physical softness gradually appears. At 1 : 3 and 8 : 1 molar compositions of R·M, the system is discontinuous. However, the interesting feature is its morphology tuning with changes to its composition, and this 1D morphological tuning (Scheme 2) (from helical fiber to rod-like to tubular morphology) produces an interesting photoluminescence (PL) property by changing the composition of the R·M components.

Scheme 2 Representation of energy minimized molecular structure of R, M, and their aggregated form in 1 : 1 and 1 : 3 molar ratios. The construction of self-assembled structures through H-bonding and π-stacking is also shown (taken from ref. 38a and b).

In a recent extended work in 2011, we established that the fiber morphology can be tuned in different architectures by in situ formation of silver nanoparticles (AgNPs) in the R·M11 (molar ratio of R : M = 1 : 1) hydrogel matrix.⁴² The nanotube superstructure of R·M11 gel fibers can be tuned into nanorod, superhelices or spheroidal morphologies by adjusting the concentration of the AgNO₃ in the bulk gel material (Scheme 3).

Scheme 3 Schematic presentation of the formation of different morphologies in the R·M11 system from mechanical effect of in situ formation and stabilization of Ag NPs (red balls) by R (taken from ref. 42).

The tubular morphology of the complex originates from the planar sheet structure of the complex, due to the π-stacking causing interdigitation of ribityl chains of R. The inherent chirality of the ribityl chain causes the sheet to bend into a tube structure.^38b A slight slippage of the R molecular plane from that of the complementary plane takes place due to the mechanical force exerted by the growing AgNP (formed and stabilized by R). The size of the AgNPs controls the amount of slippage of the molecular planes causing the change in the morphology, as the mechanical force depends on the size of the AgNPs. In the R·MAg0.02 system (where 0.02 indicates that 0.02 mg fine AgNO₃ powder is used in 2 ml gel sample), the size of AgNPs on the R side is low, but sufficient to disturb the orienting effect of the chiral ribityl chain to bend the sheet, preventing tube formation and forming a rod morphology from stacking of the sheets. For R·MAg0.2 (0.2 mg AgNO₃ powder is used), due to the larger size of the AgNPs, slipping of the R plane from the M plane is relatively large. AgNPs thus inhibit the growth of the sheets laterally, but promote its longitudinal growth due to a strong π-stacking force among the twisted supramolecular complexes. This gives rise to the helical fibrillar morphology of the R·M11 (molar ratio of R : M = 1 : 1) complex, which further self-assembles to produce super helical fiber networks. However, in the case of R·MAg2 there are formations of bigger AgNPs destabilizing the supramolecular complex formation due to the large mechanical force. This causes a spheroidal morphology, with a Ag core and a R shell, and M molecules may remain absorbed on its outer shell.

Sonication is more commonly used to increase the dissolution rate of insoluble compounds, particularly those with strong hydrogen bonding functionality. Materials scientists are now using this sonication wave to make an ordered arrangement: they have thought to cleave the intramolecular π-stacking interaction of the components inducing rapid and spontaneous aggregation through interpenetrating stacking interactions.^34a Steed and co-workers have studied detailed structural calculations using DFT (PBE1PBE/6-31+G*) of the dihydrate crystal of M·U (1 : 1) complex,³⁹ and also matched the simulated powder X-ray diffraction pattern (PXRD) with the xerogel PXRDs. Although co-crystal formation in the same solvent as that used for gelling is much more difficult, the energy minimized structure correlation has become very important in the field of bicomponent systems. They have found distinct evidence of M·U·2H₂O phase in the ¹³C NMR spectroscopy of the xerogel. They have generated computerized crystal structures of both M·U and M·U·2H₂O for the possible patterns of supramolecular interactions and have generated an energy minimized structure (Fig. 24). The crystal structure prediction of M·U·2H₂O has been a challenge, as it contains four independent molecules in the asymmetric unit (Z′′ = 4). The simulated PXRD patterns from the computer-generated fourth lowest energy minimized crystal structures are matched to those observed from the dried xerogel, both visually and using de Gelder's normalized weighted cross-correlation similarity function.⁴³ The unit cell is highly anisotropic and exhibits sheets of strong H-bonding interaction, which have resulted in tape-like morphology (SEM images) of the xerogels.

Fig. 24 Calculated fourth-lowest-energy crystal structure of M·U·2H₂O; monoclinic, P2₁/c, a = 3.818, b = 25.810, c = 14.088 I, b = 71.178. (a) View down the axis, and (b) view side-on to the hydrogen bonding (taken from ref. 39).

In the search for new complementary molecules, gallic acid (G) appears to be a suitable candidate which can form a stable hydrogel with M.⁴⁰ Its structural advantages make G a good hydrogelator. The variations in G·M molar ratio in aqueous solution produce different kind of gel networks (Fig. 25). The physical properties of the different compositions also change due to the structural differences. The normalized UV-vis spectra of GM13 has the highest absorption intensity with respect to others having almost comparable intensity. The constancy of π–π* peak positions and the abrupt rise in intensity of GM13 compared to those of other complexes is attributed to the symmetry deterioration of G in the GM13 complex compared to that of other complexes. In the GM13 complex, G is at the center and the three M molecules are at the periphery, increasing the dissymmetry of G significantly. This increased dissymmetry causes an increase in the transition probability and hence an enhancement in the intensity of absorption maximum. The different types of physical appearances in morphology in GM11, GM13 and GM31 hydrogel systems may thus arise from different molecular arrangement during complex formation. The participation of phenolic –OH groups in the supramolecular network formation during hydrogelation is confirmed by a number of techniques (cf. FTIR). Further, we have also used individual –OH groups of o, m, p-isomers of hydroxybenzoic acid to observe the changes in physical properties^41e (Fig. 26). The basic structural differences influence the supramolecular structure formation which is reflected in the physical appearances, as well as the physical properties, of the hydrogels. The hydrogels are prepared by heating–cooling methods in a 1 : 1 molar ratio, and the hydrogels exhibit different texture, e.g. the o- and p-hydrogels are opaque in appearance, while the m-hydrogel is translucent. Their structure differences also affect the critical gelation concentration (CGC), and the values are 0.5, 1 and 0.1% for o, m and p-isomer, respectively.

Fig. 25 FESEM images of the GM hydrogel systems (i) GM11 has twisted fiber (ii) GM13 has intertwined fiber and (iii) GM31 has rod like fiber morphology [taken from ref. 40].

Fig. 26 Chemical structures of the monohydroxybenzoic acid isomers (27a–c).

Molecular complexes of M with hydroxybenzoic acids have been analyzed by Guru Row and co-workers using a crystal engineering approach.^28a In their work, they investigated the collective role of hydroxyl (OH) and carboxylic (COOH) functionalities in the recognition processes (cf. M in crystal engineering).

Malik and co-workers⁴⁴ have recently introduced a unit in the bicomponent hydrogel by using acid functionalized perylene diimide (PDI)m which has a pronounced hydrophobicity owing to the presence of an aromatic perylene core (Fig. 27). The supramolecular network formation with M brings the PDI moieties into H-type arrangements which result in patterned fluorescencent nanofibers.

Fig. 27 Chemical structures of the perylene-based hydrogel system reported by Malik and co-workers (taken from ref. 44).

Organogel

The potent organogel-forming ability of triazines functionalized with chiral α-amino acid substituents and a study on the influence of structural and stereochemical factors on the formation of the organogels was reported by Zanda and co-worker in 2008.⁴⁵ In order to show structural effects on gelation, as well as on the sol–gel transition temperature, they have synthesized chiral α-amino derivatives of M. A selection of the synthesized molecules are summarized in Fig. 28. They have also tested gelation ability in various organic solvents and the obtained result is summarized in Table 1. They observed that M derivatives with amide or hydroxamate functional groups can form gel in toluene, with the exception of tert-butyl ester derivative 29d and bis-glycine derivative 31. The chiral bis-phenylalanine triazines methyl or ethyl esters, which bear a glycine functionalized as hydroxamate (29a) or amide (29b,c), have the strongest gelation capacity (checked by falling drop method) in chloroalkanes, in aromatic solvents and even in ethyl acetate. In contrast, the corresponding achiral meso-structures 30a,b have lower gelating capacity in chloroalkanes, but they give very strong gels in aromatic solvents, like toluene. They have observed the very interesting fact that the racemic mixture of 29a and 29b forms a gel in toluene, but surprisingly the enantiomerically pure compounds form weaker gels with lower sol–gel temperatures. The presence of hydroxamic acid or amide functions of 29a and 29b probably gives rise to different the supramolecular structures of these M derivatives, depending on their enantiomeric purities. They also raise the issue that the C₂-symmetry of organogel-forming triazine 29, compared to lower symmetry molecules like 31 and 32, has lower gelating power. The C_s-symmetrical mesostructures of 30a,b have also shown much lower gel-forming capacity in organic solvents.

Fig. 28 Set of 1,3,5-triazine-based M molecules investigated for their organogel-forming ability (taken from ref. 45).

Table 1 Gel forming capacity of triazines 29–34 in different solvents at 3 wt% and room temperature (taken from ref. 45)^a

Triazine	CH2Cl2	CHCl3	(ClCH2)2	Toluene	EtOAc
a Key: Gel = gel; Sol = solution; Wgel = weak gel that is unable to sustain the steel ball on the surface. A precipitate (Prec) tends to form at a concentration of 2 wt%.
29a	Gel	Gel	Gel	Gel	Sol
29b	Gel	Sol	Gel	Gel	Gel
29c	Gel	Gel	Gel	Gel	Gel
29d	Sol	Sol	Sol	Sol	Sol
29e	Sol	Sol	Sol	Gel	Sol
30a	Sol	Sol	Sol	Gel	Sol
30b	Sol	Sol	Sol	Gel	Sol
31	Sol	Sol	Sol	Wgel	Sol
32	Gel	Prec	Prec	Gel	Prec
33	Sol	Sol	Sol	Sol	Sol
34	Sol	Sol	Sol	Sol	Sol

---

In 1993, Hanabusa and co-workers first developed^46a a two component organogel-based on a M like molecule (35). From their previous work on a gelator of long chain alkylamide of N-benzyloxycarbonyl-L-valyl-L-valine in different organic solvent at a very low concentration,^46b they have assumed that “the driving factors for the physical gelation are intermolecular hydrogen bonding interactions which build up the macromolecular-like aggregates and van der Waals interactions which juxtapose and interlock the large aggregates”. Since then, this bicomponent approach has been used by several research groups and the self-aggregation between the two dissimilar molecules has become a well-known driving force for the physical gelation (Fig. 29).

Fig. 29 Chemical structures of the molecules used in formation of organogel by Hanabusa and coworkers (taken from ref. 46a).

In 2005, a similar type of binary organogels has been prepared from the coaggregation of a flexible azobenzene-tethered M dimer and cyanurate/barbiturates⁴⁷ (Fig. 30). In their work, Yagai and Kitamura used this D–A–D/A–D–A H-bonding arrangements between M and cyanurate or barbiturate as a primary intermolecular glue. They have introduced a rigid azobenzene tether between two M binders, which supplies some rigidity that is required for regular molecular packing, into the gelator molecule and additionally provides π–π interactions, enhancing the interchain ordering of the quasi-1D supramolecular polymer. During gelation studies, they found that the gelation phenomena strongly depend on the complementary cyanurate/barbiturate. These complementary parts bring the differences in the secondary structure of the flexible quasi-1D supramolecular polymers. Also, the differences in the minimum gelation concentration (MGC) and gel-to-sol phase transitions come from the structural variations in the M dimers.

Fig. 30 Possible non-covalent interactions are shown for the aggregated structure formation of azobenzene-tethered M dimer and cyanurate/barbiturates (taken from ref. 47).

In place of azobenzene, they have further attempted to create a structure with a new linker tri(phenyleneethylene) (TPE), in collaboration with Ajayaghosh's group.⁴⁸ Ditopic cyanurate (dCA) with a long alkyl chain acts as a complementary molecule. They have checked the self-assembly behavior, as well as the gelation properties, in the presence and absence of the complementary dCA module. TPE, by itself, can produce a self-assembled structure with the help of the amide linkage and, at higher concentration, it produces blue light emitting gel in aliphatic solvents like hexane and MCH (Scheme 4). The physical appearance, critical gelation concentration (CGC), and glass transition temperature (T_g) are different when TPE forms a bicomponent aggregate with dCA. The higher number of H-bonds, as well as the π-stacking of the TPE moiety, in the 1 : 1 complex generates the higher stability and the transparency, which in turn produce lower aggregates. The complex can also self-assemble in 2 : 2 fashion. The 2 : 2 dimer is invoked during the initial stage of the self-assembly. In the presence of a larger number of H-bonds; the dimers further self-assemble to form a hierarchical structure in a J-type fashion through amide H-bonding. Temperature dependent fluorescence measurements indicate the selective breakage of the amide linkage of the inter-aggregate up to 45 °C and in this way a H-type dimer with favorable π-stacking is formed. The H-type dimers are subsequently dissociated at higher temperature. The role of cyanurate during complex formation with melamine-linked TPE is hence responsible for the physical properties and also the morphological features of the resultant supramolecular organogel.

Scheme 4 Schematic representation of different types of self-assembly processes of TPE linked M in the presence and absence of complementary molecules (taken from ref. 48).

Ion and metal selective gelation

Zhang and co-workers^49a have found that M, in its protonated form, can be triggered by the oxo-anions such as NO₃⁻, PO₄⁻³, ATP and SO₄⁻² to form a superstructure and, in addition, this superstructure is able to gelate large amount of water molecules (Scheme 5). This gelation can further be reversibly switched on/off by tuning the pH or temperature of the systems. Initially they tried with PO₄³⁻ ion at the molar ratio of 6 : 1 to M and at high concentration (700 mM), but this did not lead to any observable change. However, it was unexpectedly observed that a stable and opaque gel is obtained at room temperature when the solution pH was adjustaed with dilute HCl solution to 2–5. On increasing the system pH to 7 or above, a clear solution is regenerated and this process can be reversibly repeated many times by alternatively increasing and decreasing the pH. The gel shows a thermoreversible nature, i.e. on increasing the temperature, it shows gel to sol physical transformation and the earlier gel state can also be achieved by decreasing the experimental temperature to normal room temperature.

Scheme 5 Schematic illustration of protonated melamine self-assembly and hydrogel triggered by oxoanions and XO_yⁿ⁻ represents PO₄³⁻, NO₃⁻, SO₄²⁻, or ATP [taken from ref. 49a].

Other oxo-anions such as NO₃⁻, SO₄²⁻, and the bulky anion adenosine 5′-triphosphate (ATP), are also able to produce protonated M to form supramolecular hydrogels. However, other anions, like AcO⁻, Cl⁻, F⁻, are not able to produce the same gelation seen with oxo-anions or ATP.

In recent times, Luo and co-workers^49b have carried out excellent work by trying to elucidate the mechanism of M–phosphate ion superstructure formation with the help of several spectroscopic and crystallographic techniques and they have shown that the gelation process is highly pH dependent. Changes in pH dramatically influence the molar ratio of M to melaminium (HM⁺) ion and this subsequently affects the synergistic nature of hydrogen bonding patterns, electrostatic and π–π stacking interactions, resulting in the reversibility of the aggregation and dissolution process. The mechanism of the pH-induced aggregation reaction between M and phosphate is proposed as follows:

M + H⁺ ↔ HM⁺

HM⁺ + M + H₂PO₄⁻ ↔ HM⁺·M·H₂PO₄⁻

and

HM⁺·M·H₂PO₄⁻ ↔ 2HM⁺·HPO₄²⁻

They have established the composition of M–phosphate aggregate as 2[HM⁺]·[HPO₄²⁻]·3H₂O and the aggregation reaction is very much pH dependent.

In an interesting work, Song and co-workers have assessed the interaction of ionic groups (using phosphate) with M by demonstrating the Flory–Huggins interaction parameter as a model to predict the gelation behavior of a known gelator in a range of untested solvents.^49c There is one report, in which Fan and co-workers have followed the previously reported Flory–Huggins interaction theory, and they have applied it for the same gelator system in water (insoluble) methanol (soluble) system.^49d

So far, M acts as a good supramolecular synthon for gel preparation, and, for the first time, our research group has reported a silver-coordinated M metallogel.^50a In aqueous medium, addition of a quantitative amount of AgNO₃ to M produces a thermoreversible and responsive Ag(I)-coordinated metallogel. The N-atoms in the triazine ring are excellent coordinate sites for metal atoms. Ag(I) coordinates with M molecules through N-atoms forming either a linear or a bent geometry around the Ag(I) ion. A Ag(I)–M composition at 1 : 2 molar ratio produces a stable AgM12 gel which is composed of a helical fibrillar network. However, a 1 : 1 composition (AgM11) produces a metastable gel which, on aging, gradually converts into gel crystals in the form of precipitate at the bottom of the gel tube. We studied the stability contrast of AgM11 and AgM12 gels, guided by their respective crystal structures.^50b,c The AgM12 complex consists of alternating bilayer structures with extensive H-bonding.^50c Ag(I) metal remains perfectly linearly coordinated with the nitrogen atoms of two triazine rings and NO₃⁻ remains sandwiched between two planes through N–H–O hydrogen bond formation. This bilayer structure can grow longitudinally by H-bonding and laterally it grows through Ag coordination and π–π stacking (Fig. 31). The fibers grow in the direction of H-bond and the AgM12 complex is both kinetically and thermodynamically stable. The helicity of the fibers, as evident from the SEM and AFM images, is attributed to the chiral symmetry breaking, as has been observed in the coordination polymer gel of Ag(I) and imidazole derivatives.^50d,e In the AgM11 system, Ag(I) metal centre coordinates two nitrogen atoms of two different triazine rings, not in a linear fashion as in AgM12, but bent with a N–Ag–N bond angle of 127°.^50b This bent N–Ag–N bonding is involved in the lateral growth, and the fiber growth is assisted through H-bonding with other M molecules. The bending nature causes lower stability in the system by disturbing the H-bonds. During gelation, rapid cooling produces kinetically controlled fibers, but, on aging at room temperature (30 °C), more lateral growth of the crystal occurs by mere Ag(I) coordination and π–π stacking, causing breakage of the longitudinally strained fibers.

Fig. 31 ChemDraw model of AgM11 and AgM12 complexes (drawn according to ref. 50b (CCDC no. 141627) and ref. 50c (IUCr electronic archive series DN 3109), (NO₃⁻ ions are not shown for clarity)) (taken from ref. 50a).

Based on supramolecular assembly of M and Ag(I), Li and co-workers^50f have used this assembly as a template to synthesize silver nanoparticles by photo-reduction technique. With this simple and efficient strategy, they produced Ag nanochains of high potential for incorporation into functional electronics, optoelectronics, and sensing devices etc., which leads to extreme miniaturization enhancing the performance.

Sensors

During the last couple of decades, considerable efforts have been devoted to the development of fluorescent and colorimetric sensors, which can selectively detect targeted molecules or metal ions. This section deals with this molecular recognition property as a branch of the supramolecular family, in the sense of real applications of the supramolecular interactions to the quantitative analysis of small molecules. Due to the high nitrogen content (66%), M is unethically used by milk manufacturers in adulterating milk to make it appear more protein-rich. This results in severe renal disease for both humans and animals, due to the formation of insoluble M·CA crystals in the kidneys.⁵¹ A certain maximum level of M ingestion is fixed in infant foods, e.g. 2.5 ppm in the USA and 1 ppm in China. There is thus an urgent need to develop some simple, reliable and easily accessible tool to detect and allow quantitative measurement of M contamination in milk with a high sensitivity. Some analytical methods have been developed on the basis of gas chromatography–mass spectrometry (GC-MS),^52a gas chromatography,^52b liquid chromatography–mass spectrometry (LC-MS),^52c capillary electrophoresis,^52d matrix-assisted laser desorption ionization (MALDI) MS^52e and enzyme-linked immunosorbent assay (ELISA)^52f methods. A M sensor could also be developed by recognition method by using anti-parallel H-bonding interaction of M·CA aggregate. Bearing this in mind, some suitable molecules have been reported in scientific journals (Fig. 32).⁵³

Fig. 32 Chemical structures of the reported molecules suitable for M recognition and used as sensing components (taken from ref. 53).

Self-assembly based colorimetric and spectrometric sensing of M became an important topic in 2009, when Lu et al.^53a, in the first approach of this kind, developed a thiol-capped gold nanoparticle (AuNPs) based M detection technique with a significant change in spectrographs. For this reason they have synthesized 1-(2-mercaptoethyl)-1,3,5-triazinane-2,4,6-trione (MTT) (37), a kind of thiol-functionalized CA derivative, and further used this MTT to stabilize the AuNPs via ligand exchange method with citrate stabilized AuNPs. The wine red colloidal AuNP solution exhibits a visual color change to blue (Fig. 33), induced by the triple hydrogen bonding recognition between M and the CA derivative grafted on the surface of the AuNPs. This color change allows self-assembly induced tuning in the optical properties of the AuNPs.

Fig. 33 H-bonding recognition between M and CA derivative: colorimetric detection of M using the MTT (37) stabilized AuNPs and visual color change of the optimized sensor (taken from ref. 53a).

Inspired by this work and our previously reported self-assembly of riboflavin (vitamin B2; R, 38) with Mvia supramolecular complex formation through H-bonding,^38a we have developed a M sensor based on supramolecular interaction with R molecule.^53b For this, the R stabilized gold nanoparticles (R–AuNPs) are prepared in aqueous medium, which allows us to develop an easier method to detect M in foods using the surface plasmon property of AuNPs. UV-vis spectroscopic technique was used to determine the optimal detection condition and the selectivity of the optimized sensor for M detection. A peak at the higher wavelength (700 nm) appears due to the longitudinal mode of the surface plasmon band,⁵⁴ which is a result of the coalescence of the AuNPs after adding M solution.

A surface-enhanced Raman spectroscopy (SERS) based method has also been developed by Tamer and his co-workers for the detection of M.⁵⁵ The spherical magnetic core in the gold-shell and rod-shaped AuNPs labeled with a Raman-active compound is used to form a complex with the M molecules. 5,5′-Dithiobis(2-nitrobenzoic acid) is used as the Raman active compound, because it is readily adsorbed by a gold nanoparticle surface forming a self-assembled monolayer (SAM) and has strong Raman scattering at 1330 cm⁻¹, because of the symmetric NO₂ stretching. The method is validated for linearity, sensitivity, and for good repeatability (intra-day), reproducibility (inter-day), and accuracy for detection.

In another interesting work, Kim and coworkers^53c have used a convenient M detection system based on polydiacetylene (PDA) (39–41) liposomes, which have rapid, selective, and sensitive detection, and dual signal capabilities e.g. colorimetric and the fluorescence detection (Fig. 34). The PDA sensory system shows colorimetric as well as fluorescence detection capability. A photo-polymerization technique has been used for the conversion of diacetylene monomers into PDA, which shows a blue color with a λ_max value of 640 nm. The PDA shows a transformation of the color from blue phase to red phase (550 nm) upon exposure to external stimuli, including pH, temperature,⁵⁶ ions⁵⁷ and mechanical pressure,⁵⁸ and the colorimetric change is believed to result from the conformational changes of the conjugated backbone of PDA. Interestingly, the stimuli-induced red phase of PDA has shown a weak fluorescent nature, therefore the system is able to provide dual signaling capability. An optimal sensory system has been synthesized using PCDA-EG-CA/PCDA (Fig. 34a) in 9 : 1 mole ratio and this system is able to detect 1 ppm level of M with a noticeable color change within 30 s of the addition of M. They have also developed analogous solid-state sensors and have investigated the detection limit using a microarray technique. They have micro-arrayed the PCDA-EG-CA/PCDA liposome solution to an amine modified glass substrate and photo-polymerized it before exposure to M solution. As the concentration of M is increased, more M·CA complex is formed and this induces a stronger perturbation of the conjugated PDA backbone, resulting in a more intense red fluorescence. They have found that the detection limit is 0.5 ppm lower than that of the colorimetric detection solution system.

Fig. 34 (A) Chemical structure of the investigated diacetylene monomers, PCDA, PCDA-CA and PCDA-EG-CA. (B) Schematic illustration of melamine and CA derived PDA liposome by intra/inter molecular hydrogen bond and resulting steric aggregation and repulsion (taken from ref. 53c).

Recently, in 2012, Gong and coworkers⁵⁹ developed photo-responsive molecularly imprinted hydrogel (MIH), through a facile methodology which provides desired structures and specific binding sites, for the detection of M in aqueous solution by combining the specificity of a molecularly imprinted technique, the solubility of the hydrogel and the photo-isomerization property of azobenzene chromophores. The rate of the trans–cis photo-isomerization of azobenzene chromophores within the imprinted receptor sites has been found to be affected by the concentration of M in the solution media. This has inspired them to use the trans–cis photo-isomerization rate of the azobenzene moieties within the hydrogel to reflect the concentration level of M in aqueous solution. This is an excellent approach for the chemosensing of trace organic pollutants in biomaterials.

In spite of forming selective supramolecular H-bonds with the complementary molecules, the triazine ring of M can be used for chemosensing. The triazine core derivatives are also able to enhance the energy transfer process in various supramolecular architectures, and the greater affinity of the triazine units demonstrates that they may act as useful spacers as well as selective receptors for sensors and supramolecules.⁶⁰

In an interesting work, Lin and his co-workers⁶¹ have used novel triarylamine-based dendrimers TPAD1 (42) and TPAD2 (43) with N⁴,N⁶-dibutyl-1,3,5-triazine-4,6-diamine probe for efficient electron/energy transfers in perylene bisimide (PBI)-based donor–acceptor–donor/parallel–antiparallel type H-bonded nano-sized supramolecular triads and, for the first time, they applied their sensing capabilities to various metal ions. Both the dendrimers TPAD1 and TPAD2 (Fig. 35) show high selectivity to Cu²⁺ ions via the turn-off fluorescence mechanism in the presence of other transition and non-transition metal ions in semi-aqueous media (THF : H₂O = 2 : 1 in vol) at pH 7.0. They have performed NMR titration to identify the binding sites of the dendrimers with the Cu²⁺ and they have found that both the triazine ring nitrogen and the diamine probe attached to the triazine core are responsible for sensing Cu²⁺ in solution (Fig. 36). The sensing competence of both the dendrimers towards Cu²⁺ were also verified and TPAD2 shows 3 times higher sensitivity than TPAD1, because of its higher generation structure, and hence greater energy transfer, in comparison to TPAD1.

Fig. 35 Chemical structures of triarylamine-based dendrimers (taken from ref. 61).

Fig. 36 Schematic representation of Cu²⁺ sensor responses provided by TPAD1 and TPAD2 dendrimers (taken from ref. 61).

The concept of aggregation induced emission (AIE) is employed by Sanji and co-workers^62a in a fluorescence “turn-on” effort to detect M. AIE-active molecules show no emission in solution, but emit intense light when aggregates are formed in the solid state, because of restriction of intramolecular rotations.^62b They have designed AIE-active tetraphenylethene (TPE) tethered with cyanuric acid moieties which can easily bind with tritopic M and, in turn, addition of M into TPE solution turns on the fluorescence through recognition and aggregate formation (Fig. 37).

Fig. 37 Chemical structure of the CA-tethered TPE fluorescence probe, and schematic representation of the turn-on detection process of M through AIE process (taken from ref. 62a).

In this sensing study, they have also developed a membrane filter by adsorption of TPE molecules on a hydrophobic poly(vinylidene difluoride) (PVDF) membrane. After spotting of M and solvent evaporation, the process is repeated several times, one drop of ethanol being spotted.

Gong and coworkers⁶³ have used the combination of the molecularly imprinted technique, water solubility of hydrogels and photoisomerization property of azobenzene chromophores to develop a quick detection method for traces of M in aqueous media, based on photo-responsive molecularly imprinted hydrogels (MIHs). MIHs with binding cavities can selectively adsorb substrates whose shape, size and functionality are exactly the same as the template, and are stable in extreme pH conditions and organic solvents. The MIH has been fabricated by using a water-soluble 4-[(4-methacryloyloxy)phenylazo]benzenesulfonic acid as the functional monomer, 1,3,5-benzenetriol as a mimic template, and tetramethacryloyl triethylene tetramine as the cross-linker. The azobenzene sulfonate groups offer multiple hydrogen-binding sites to adsorb M, and they can also photo-regulate the uptake and release of M in aqueous media. The M concentration can also affect the photo-isomerization process of azobenzene. This cooperative process makes the MIHs a reliable technique to detect M in dairy products with a limit of 0.1 ppm.

Mesophase and liquid crystal formation

Self-assembly of small functional organic molecules is sometimes employed to achieve different states of matter, and liquid crystal^60b,64 is one such state that has properties between those of conventional liquid and solid crystal. The properties of these materials depend on the spatial arrangement of the molecules present in the system, i.e. the crystalline arrangement of the molecules. Supramolecular interactions such as van der Waals, quadrupole, charge transfer, H-bonding, etc. play an important role in the stability and properties of liquid crystals. The molecular alignment can be externally controlled by stimuli, such as heat, electric field, light etc., which in turn allow applications of the material, such as optical display and storage devices. The disk-like 1,3,5-triaxine molecule can provide an arrangement that could facilitate arrangement of molecular aggregates in a mesophase organization.⁶⁵

A reversible bistable nematic liquid-crystalline phase of small molecules has been successfully established by Xu et al. in 2008.⁶⁶ They have used nitrile azobenzene containing benzoic acid (A) (45) which acts as a proton donor during self-assembly with alkylated M derivative and T(46) acts as a proton acceptor. Both A and the A₃T complexes show liquid crystalline phases in a certain temperature range (Fig. 38). The presence of light sensitive groups like nitrile and azobenzene have not resulted in a distinct change between the A and A₃T complex during irradiation-induced photoisomerization. Instead of collapsing, the mesophase textures of the liquid crystalline phases are inclined to adopt another stable texture to fit the change in cis-isomerization. Stimuli responsiveness is also exhibited when A and A₃T are exposed to an electric field of 30 V under UV light. However, the effect of an electric field in the absence of UV light is not so productive in the case of the A aggregate. The A₃T assembly has shown a worm-like texture at 90 °C when cooled from isotropic phase and, when it is charged with an electric field, the direction of the texture changes into a planar-like texture, which is observed by polarized optical microscopy. A change also become pronounced in the crooked texture, which shows a birefringence nature when the complex is irradiated with UV light for 10 minutes at 90 °C. The A₃T self-assembled liquid crystal is able to hold the change of molecular orientation induced by both light and electric stimuli even after the removal of these stimuli. Study reveals that the multiresponsive characteristic is a consequence of hydrogen bonding and assembly-induced change of molecular dipole (Fig. 39).

Fig. 38 (a) The chemical structures of the compounds reported by Xu et al. and (b) formation of the A₃T complex (taken from ref. 66).

Fig. 39 A schematic overview of the dual switching behavior in the A₃T complex (taken from ref. 66).

Hydrogen bonded supramolecular complexes of 2,4,6-triarylamino-1,3,5-triazines (Tn) with semiperfluorinated benzoic acids are able to show columnar mesophase.⁶⁷ The mesophase formation has been analyzed as H-bonding between the Tn and benzoic acid in equimolar mixtures, which leads to discrete dimeric supramoleculear structure. For further conformation of the role of semi-perfluorinated alkyl chains, mesophase formation has been investigated with an equimolar mixture (1 : 1) of triazines with non-fluorinated three fold alkoxy-modified carboxylic acid and also with higher number of fluorinated alkyl chains. The H-bonding between Tn and carboxylic acids predominantly leads to formation of discrete dimeric structure and the H-bonds enhance the polarity of the core region. Essentially, an increase in the intramolecular polarity contrast upon replacing alkyl chains by semi-perfluorinated chains (A to C) favors a microsegregation, that leads to the columnar structure formation (Fig. 40).

Fig. 40 The self-assembled dimeric structure and corresponding polarized optical microscopic images showing formation of columnar phase (B and C) by introducing semi-perfluorinated alkyl chain into molecular periphery (taken from ref. 67).

Sierra and co-workers^68a have synthesised and characterized a novel series of asymmetric V-shaped acids derived from 1,3,4-oxadiazole, which is well known for exhibiting high photoluminescence and electron-accepting properties, as well as for its high thermal and hydrolytic stability and resistance to oxidative degradation. A mixture of the 2,4-diamino-6-dodecylamino-1,3,5-triazine and the oxadiazole acids in a 1 : 3 ratio gives supramolecular complexes through H-bonds. These oxadiazole acids are used to form supramolecular complexes through hydrogen bonding with three acid molecules around the 2,4-diamino-6-dodecylamino-1,3,5-triazine. These complexes have an H-bonded core designed to generate mesomorphism, fluorescence and helical superstructures, which have a tendency to form rectangular columnar mesomorphism.

In another work, they attempted to control and induce a supramolecular chiral response of the supramolecular organization by using light as an external stimulus.^68b The isomerization of azobenzene (Fig. 41) upon irradiation with light is a well-documented phenomenon and the possibilities that this reaction offers for the control of self-assembly in liquid crystals has been recognized. However, attempts to induce supramolecular chirality in columnar mesophases formed by azobenzene-containing mesogens has been demonstrated by controlling the supramolecular chirality of a helical stack of the propeller-like complexes using an external chiral stimulus, such as circularly polarized light (CPL). Furthermore, the chiral information of CPL is transferred to the columnar organization through photochromic azobenzene groups. This process enables reversible switching of the supramolecular chirality of the columnar arrangement upon illumination with CPL of the opposite handedness.

Fig. 41 Chemical structures of the V-shaped acid derivatives used by Sierra and co-workers (taken from ref. 68).

In an extension of work on “molecular boxes”,¹⁶ Reinhoudt has reported a liquid-crystalline material based on the self-organization of self-assembled 3D molecular boxes.^69a Each box consists of two rosette motifs connected through three calix[4]arene dimelamine molecules, and the circular rosette motifs^8e are formed by 36 hydrogen bonds^69b between the M moieties and added barbituric or cyanuric acid. The calix[4]arene dimelamine has been functionalized with octadecyl chains to promote self-organization of the double-rosette assemblies into a liquid-crystalline phase. The long alkyl chains are connected to the terminal benzamide moieties of calix[4]arene dimelamine to prevent their interference with the network of hydrogen bonds that holds the double rosette together. The double rosettes stack in an ordered columnar fashion and the driving force for the self-assembly of the columns is the nanoscale segregation of the double rosette cores and the lipophilic alkyl chains.

Towards the development of polymeric scaffolds

This would be an incomplete story without including development of fascinating polymers where M is used as an active site for the purpose of recognition. This is not an actual M molecule, but the 1,3,5-triazine part with two amine groups i.e. 2-vinyl-4,6-diamino-1,3,5-triazine monomer (VDAT) (Fig. 42) that functions like M, whether in recognition,^70a–c affecting the polymeric properties like glass transition point with multiple H-bonds,^70d as a modular scaffold for covalent inhibitor,^70e or in cell transfections.^70f

Fig. 42 Chemical structure of the 2-vinyl-4,6-diamino-1,3,5-triazine monomer, VDAT (47).

A major disadvantage with many current side-chain functionalized polymer systems based on hydrogen bonding is the lack of recognition motif density, or the lack of stability in the hydrogen bond, due to the use of only a single hydrogen bond per side-chain. The recognition target of molecules with high affinity is a fundamental step in biological systems. Natural receptors show exceedingly high selectivity and affinity by forming complementary hydrogen bonds with the target molecules. H-bonded sites like M can induce selectivity of the polymer chain, mimicking the molecular recognition by biological systems to the guest molecules. Kunitake and co-workers have synthesized a Langmuir–Blodgett film of an amphiphile, which bears a 4,6-diaminotriazine residue and binds thymidine at the air/water interface through the formation of a stable H-bonded monolayer.⁷¹ This H-bonding recognition has been employed to sense nucleic acids and bases in aqueous solution with poly(2-vinyl-4,6-diamino-1,3,5-triazine) (PVDAT). They have also demonstrated the effect of chemical modification of PVDAT on the guest binding activity. The adsorptions of pyrimidine and purine derivatives in water solution on PVDAT depend on the number of complementary hydrogen bonds. Neither a monomeric receptor VDAT nor a dimer model of PVDAT (BDAT) is active, only the polymeric receptors efficiently form hydrogen bonds with guests in water. The chemical modification of PVDAT affects the guest binding activity and the polymeric microenvironment is found to be essential for molecular recognition of nucleic acid bases in water solution.^70a–c

Komiyama and his co-workers have reported the adsorbing activities of PVDAT towards various nucleic acids, bases and nucleosides and this occurs in the following order: thymine, uridine, uracil, thymidine ≫ cytidine, cytosine > pyrimidine = 0. This order coincides fairly well with that of the number of H-bonding sites in guest molecules, which are complementary with those of the 4,6-diaminotriazine (DAT) residue. Uracil and thymine have three H-bonding sites, cytosine has two, and pyrimidine has only one. Apparently the complementary H-bonding between the guests and the DAT residues in the polymer is responsible for the guest adsorption.

A schematic polymeric scaffold to recognize a variety of guest molecules by designing and synthesizing polymer backbones with multiple recognition sites in their side chains can be functionalized via non-covalent interactions based on a “universal polymeric scaffold” (Scheme 6) concept. For rapid and reversible functionalization with suitable receptors, Weck and co-workers^72a have investigated the self-assembly properties of diaminotriazine- and diaminopyridine-functionalized norbornenes. They have presented a methodology of the synthesis of polymers containing H-bonding units and their functionalization is based on self-assembly, which has a number of advantages, including (i) high recognition motif density, (ii) controllable polymerization behavior, and (iii) low dimerization and high association constants of the recognition units towards hydrogen bonding based on a donor–acceptor–donor motif. In 2008, they reported thermally reversible polymeric networks with tunable physical properties by controlling the molecular structure and the amount of the cross-linking agent added.^72b The study has revealed that the microstructure plays an important role in the macroscopic mechanical properties of these hydrogen bonded networks in solution.

Scheme 6 Schematic representation of the design of the “Universal Polymeric Scaffolds” that can be used in the molecular recognition processes. The reddish orange and sky blue parts of the monomers/polymers are sites capable of recognizing complementary guest molecules through non-covalent interactions (taken from ref. 72a).

Conclusions

Melamine-based supramolecular chemistry is among one of the most interesting systems that utilizes a “bottom-up” approach to fabricate fascinating nano-to-micro scaled morphological architectures, and also in developing functional materials, such as gels, sensors, liquid crystals etc. Since the discovery of a melamine–cyanuric acid crystal aggregate, the donor–acceptor–donor hydrogen bonded pattern has been diversely used in synthetic supramolecular chemistry. Fascinating architectures like, molecular sheets, channel and columnar structures, nanotubes and nanoropes, and/or molecular boxes have been generated. These intriguing molecular assemblies have been further extended to explore the concept of crystal engineering. The achiral melamines have sometimes been used in generating chirality in the molecular assemblies by intermolecular interactions in chiral matrixes or environments. The concepts of supramolecular organization in donor–acceptor–donor fashion, specially with imides, have been applied to 2D assemblies on the surfaces of conducting materials. Despite several investigations regarding molecular architecture, some sensing systems have been developed based on melamines. Interplay of the non-covalent interactions provided by melamines could readily modulate the properties of assemblies by transforming them into bulk materials, like gels, liquid crystals etc.

Although many developments have been made in the field of supramolecular chemistry, progress still needs to be made and we believe that melamines may be able to fulfil this requirement. This review aims to give a summary of the recent developments in different fields of supramolecular chemistry and to encourage further research in fabricating functional materials based on melamine.

Acknowledgements

B.R and P.B acknowledge CSIR, New Delhi, for granting the fellowship.

Notes and references

1. Recent literature on supramolecular chemistry: (a) J.-M. Lehn, Supramolecular Chemistry: Concepts and Perspectives, VCH, Weinheim, 1995; (b) H.-J. Schneider and A. Yatsimirsky, Principles and methods in supramolecular chemistry, Wiley, Chichester, 2000; (c) J. W. Steed and J. L. Atwood, Supramolecular Chemistry, Wiley, Chichester, 2000; (d) Supramolecular Materials and Technologies, Perspectives in Supramolecular Chemistry, ed. D. N. Reinhoudt, Wiley, 1999, vol. 4; (e) J. L. Atwood and J.-M. Lehn, Comprehensive Supramolecular Chemistry, Pergamon, Oxford, 1996; (f) T. Aida, E. W. Meijer and S. I. Stupp, Science, 2012, 335, 813; (g) F. J. M. Hoeben, P. Jonkhejim, E. W. Meijer and A. P. H. Schenning, Chem. Rev., 2005, 105, 1491.
2. D. J. Cram and J.-M. lehn, Science, 1974, 183, 803.
3. A. Robertazzi, F. Krull, E.-W. Knapp and P. Gamez, CrystEngComm, 2011, 13, 3293.
4. (a) C. A. Hunter and J. K. M. Sanders, J. Am. Chem. Soc., 1990, 112, 5525; (b) C. R. Martinez and B. L. Iverson, Chem. Sci., 2012, 3, 2191.
5. (a) G. A. Jeffrey, An introduction to hydrogen bonding, Oxford university press, 1997; (b) T. Steiner, Angew. Chem., Int. Ed., 2002, 41, 48; (c) Supramolecular Assembly via Hydrogen Bonds I & II, ed. D. M. P. Mingos, Springer, 2004; (d) G. R. Desiraju, Angew. Chem., Int. Ed., 2011, 50, 52.
6. M. Arduini, M. Crego-Calama, P. Timmerman and D. N. Reinhaudt, J. Org. Chem., 2003, 68, 1097.
7. (a) Y. Wang, B. Wei and Q. Wang, J. Crystallogr. Spectrosc. Res., 1990, 20, 79; (b) J. A. Zerkowski, C. T. Seto, D. A. Wierda and G. M. Whitesides, J. Am. Chem. Soc., 1990, 112, 9025; (c) C. T. Seto and G. M. Whitesides, J. Am. Chem. Soc., 1990, 112, 6409.
8. (a) A. Ranganathan, V. R. Pedireddi and C. N. R. Rao, J. Am. Chem. Soc., 1999, 121, 1752; (b) J. P. Mathias, E. E. Simanek, C. T. Seto and G. M. Whitesides, Macromol. Symp., 1994, 77, 157; (c) J. A. Zerkowski, J. C. MacDonald, C. T. Seto, D. A. Wierda and G. M. Whitesides, J. Am. Chem. Soc., 1994, 116, 2382; (d) J. P. Mathias, C. T. Seto and G. M. Whitesides, Polym. Prepr., 1993, 34, 92; (e) C. T. Seto and G. M. Whitesides, J. Am. Chem. Soc., 1993, 115, 905; (f) J. A. Zerkowski, C. T. Seto and G. M. Whitesides, J. Am. Chem. Soc., 1992, 114, 5473; (g) M. G. J. ten Cate, J. Huskens, M. Crego-Calama and D. N. Reinhoudt, Chem.–Eur. J., 2004, 10, 3632; (h) S. Yagai, S. Kubota, K. Unoike, T. Karatsu and A. Kitamura, Chem. Commun., 2008, 4466.
9. J. Xu, G. Wu, Z. Wang and X. Zhang, Chem. Sci., 2012, 3, 3227.
10. (a) A. Ranganathan, V. R. Pedireddi, G. Sanjayan, K. N. Ganesh and C. N. R. Rao, J. Mol. Struct., 2000, 522, 87; (b) S. A. Barnett, A. J. Blake and N. R. Champness, CrystEngComm, 2003, 5, 134.
11. (a) A. Ranganathan, V. R. Pedireddi and C. N. R. Rao, J. Am. Chem. Soc., 1999, 121, 1752; (b) R. Thomas and G. U. Kulkarni, Beilstein J. Org. Chem., 2007, 3, 17.
12. (a) S. Yagai, S. Kubota, K. Unoike, T. Karatsu and A. Kitamura, Chem. Commun., 2008, 4466; (b) S. Yagai, Y. Monma, N. Kawauchi, T. Karatsu and A. Kitamura, Org. Lett., 2007, 9, 1137.
13. (a) N. Lu, G. Wang, W. Yang, Z. Du and T. Li, J. Colloid Interface Sci., 2002, 252, 222–225; (b) W. Yang, X. Chai, L. Chi, X. Liu, Y. Cao, R. Lu, Y. Tang, H. Fuchs and T. Li, Chem.–Eur. J., 1999, 5, 1144–1149.
14. (a) S. Yagai, M. Usui, T. Seki, H. Murayama, Y. Kikkawa, S. Uemura, T. Karatsu, A. Kitamura, A. Asano and S. Seki, J. Am. Chem. Soc., 2012, 134, 7983; (b) I. S. Choi, X. Li, E. E. Simanek, R. Akaba and G. M. Whitesides, Chem. Mater., 1999, 11, 684; (c) H. Schonherr, V. Paraschiv, S. Zapotoczny, M. Crego-Calama, P. Timmerman, C. W. Frank, G. J. Vancso and D. N. Reinhoudt, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 5024; (d) A. Piermattei, M. Giesbers, A. T. M. Marcelis, E. Mendes, S. J. Picken, M. Crego-Calama and D. N. Reinhoudt, Angew. Chem., Int. Ed., 2006, 45, 7543; (e) S. Yagai, T. Nakajima, K. Kishikawa, S. Kohmoto, T. Karatsu and A. Kitamura, J. Am. Chem. Soc., 2005, 127, 11134.
15. S. Yagai, T. Nakajima, T. Karatsu, K.-i. Saitow and A. Kitamura, J. Am. Chem. Soc., 2004, 126, 11500.
16. (a) J. L. Prins, K. A. Jolliffe, R. Hulst, P. Timmerman and D. N. Reinhoudt, J. Am. Chem. Soc., 2000, 122, 3617; (b) J. L. Prins, P. Timmerman and D. N. Reinhoudt, J. Am. Chem. Soc., 2001, 123, 10153; (c) L. J. Prins, J. Huskens, F. de Jong, P. Timmerman and D. N. Reinhoudt, Nature, 1999, 398, 498; (d) L. J. Prins, F. De Jong, P. Timmerman and D. N. Reinhouldt, Nature, 2000, 408, 181; (e) T. Ishi-i, M. Crego-Calama, P. Timmerman, D. N. Reinhoudt and S. Shinkai, J. Am. Chem. Soc., 2002, 124, 14631.
17. (a) X. Huang, C. Li, S. Jiang, X. Wang, B. Zhang and M. Liu, J. Colloid Interface Sci., 2005, 285, 680; (b) T. Kawasaki, M. Tokuhiro, N. Kimizuka and T. Kunitake, J. Am. Chem. Soc., 2001, 123, 6792.
18. (a) J. F. Constant, J. Fahy, J. Lhomme and J. E. Anderson, Tetrahedron Lett., 1987, 28, 1777; (b) V. M. Rotello, E. A. Viani, G. Deslongchamps, B. A. Murray and J. Rebek Jr, J. Am. Chem. Soc., 1993, 115, 797; (c) Y. Kato, M. Conn and J. Rebek Jr, Proc. Natl. Acad. Sci. U. S. A., 1995, 92, 1208; (d) N. Kimizuka, T. Kawasaki, K. Hirata and T. Kunitake, J. Am. Chem. Soc., 1998, 120, 4094.
19. (a) D. J. Owen and G. B. Schuster, J. Am. Chem. Soc., 1996, 118, 259; (b) M. Wang, G. L. Silva and B. A. Armitage, J. Am. Chem. Soc., 2000, 122, 9977.
20. J. Barbera, L. Puig, P. Romero, J. L. Serrano and T. Sierra, J. Am. Chem. Soc., 2006, 128, 4487.
21. (a) N. Kimizuka, T. Kawasaki, K. Hirata and T. Kunitake, J. Am. Chem. Soc., 1995, 117, 6360; (b) C. T. Seto and C. M. Whitesides, J. Am. Chem. Soc., 1993, 115, 905; (c) O. E. Sielcken, M. M. Tilhorp, M. F. M. Rocks, R. Hendriks, W. Drenth and R. J. M. Nolle, J. Am. Chem. Soc., 1987, 109, 4261.
22. (a) F. Würthner, C. Thalacker and A. Sautter, Adv. Mater., 1999, 11, 754; (b) R. F. M. Lange, F. H. Beijer, R. P. Sijbesma, R. W. W. Hooft, H. Kooijman, A. L. Spek, J. Kroon and E. W. Meijer, Angew. Chem., Int. Ed. Engl., 1997, 36, 969.
23. (a) G. M. Whitesides, J. P. Mathias and C. T. Seto, Science, 1991, 254, 1312; (b) M. C. Etter, Acc. Chem. Res., 1990, 23, 120.
24. S. De Feyter and F. C. De Schryver, Chem. Soc. Rev., 2003, 32, 139.
25. (a) F. Silly, A. Q. Shaw, M. R. Castell and G. A. D. Briggs, Chem. Commun., 2008, 1907; (b) P. A. Staniec, L. M. A. Perdigao, B. L. Rogers, N. R. Champness and P. H. Beton, J. Phys. Chem. C, 2007, 111, 886; (c) W. Xu, M. Dong, H. Gersen, E. Rauls, S. Vazquez-Campos, M. Crego-Calama, D. N. Reinhoudt, I. Stensgaard, E. Laegsgaard, T. R. Linderoth and F. Besenbacher, Small, 2007, 3, 854.
26. L. M. A. Perdigao, N. R. Champness and P. H. Beton, Chem. Commun., 2006, 538.
27. J. A. Theobald, N. S. Oxtoby, M. A. Phillips, N. R. Champness and P. H. Beton, Nature, 2003, 424, 1029.
28. (a) S. SeethaLekshmi and T. N. Guru Row, CrystEngComm, 2011, 13, 4886; (b) X.-L. Zhang and X.-M. Chen, Cryst. Growth Des., 2005, 5, 617; (c) K. Udaya Lakshmi, S. Thamotharan, K. Ramamurthia and B. Varghese, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2006, 62, 455; (d) J. Janczaka and G. J. PerpeAtuo, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 2001, 57, 123.
29. (a) A. Cousson, B. Nicolai and F. Fillaux, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2005, 61, 222; (b) J. N. Varghese, A. M. O'Connell and E. N. Maslen, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1977, 33, 2102; (c) G. J. Perpetuo and J. Janczak, J. Mol. Struct., 2008, 891, 429.
30. A. Delori, E. Suresh and V. R. Pedireddi, Chem.–Eur. J., 2008, 14, 6967.
31. J. Hamblin, S. P. Argent, A. J. Blake, C. Wilson and N. R. Champness, CrystEngComm, 2008, 10, 1782.
32. (a) R. G. Weiss and P. Terech, Molecular Gels: Materials with Selfassembled Fibrillar Network, Springer, Dordrecht, 2006; (b) A. R. Hirst, D. K. Smith, M. C. Feiters, H. P. M. Geurts and A. C. Wright, J. Am. Chem. Soc., 2003, 125, 9010; (c) A. Y.-Y. Tam and V. W.-W. Yam, Chem. Soc. Rev., 2013, 42, 1540; (d) L. E. Buerklea and S. J. Rowan, Chem. Soc. Rev., 2012, 41, 6089; (e) M. D. Segarra-Maset, V. J. Nebot, J. F. Miravet and B. Escuder, Chem. Soc. Rev., 2013, 42, 7086; (f) N. M. Sangeetha and U. Maitra, Chem. Soc. Rev., 2005, 34, 821; (g) A. Noro, M. Hayashi and Y. Matsushita, Soft Matter, 2012, 8, 6416; (h) D. K. Smith, Molecular Gels – Nanostructured Softmaterials in Organic Nanostructured, ed. J. W. Steed and J. L. Atwood, Wiley-VCH, Weinheim, 2008; (i) M. de Loos, B. L. Feringa and J. H. van Esch, Eur. J. Org. Chem., 2005, 3615.
33. B. Roy, A. Saha, A. Esterrani and A. K. Nandi, Soft Matter, 2010, 6, 3337.
34. (a) G. Cravotto and P. Cintas, Chem. Soc. Rev., 2009, 38, 2684; (b) K. Isozaki, H. Takaya and T. Naota, Angew. Chem., Int. Ed., 2007, 46, 2855; (c) D. Bardelang, F. Camerel, J. C. Margeson, D. M. Leek, M. Schmutz, M. B. Zaman, K. Yu, D. V. Soldatov, R. Ziessel, C. I. Ratcliffe and J. A. Ripmeester, J. Am. Chem. Soc., 2008, 130, 3313; (d) C. Dou, D. Li, H. Gao, C. Wang, H. Zhang and Y. Wang, Langmuir, 2010, 26, 2113; (e) N. Komiya, T. Muraoka, M. Iida, M. Miyanaga, K. Takahashi and T. Naota, J. Am. Chem. Soc., 2011, 133, 16054.
35. (a) S. Lena, P. Neviani, S. Masiero, S. Pieraccini and G. P. Spada, Angew. Chem., Int. Ed., 2010, 49, 3657; (b) S. Yagai, T. Nakajima, K. Kishikawa, S. Kohmoto, T. Karatsu and A. Kitamura, J. Am. Chem. Soc., 2005, 127, 11134; (c) X. Liao, G. Chen, X. Liu, W. Chen, F. Chen and M. Jiang, Angew. Chem., Int. Ed., 2010, 49, 4409.
36. (a) S. Banerjee, R. K. Das and U. Maitra, J. Mater. Chem., 2009, 19, 6649; (b) M.-O. M. Piepenbrock, G. O. Lloyd, N. Clarke and J. W. Steed, Chem. Rev., 2010, 110, 1960; (c) S.-Y. Hsueh, C.-T. Kuo, T.-W. Lu, C.-C. Lai, Y.-H. Liu, H.-F. Hsu, S.-M. Peng, C.-H. Chen and S.-H. Chiu, Angew. Chem., Int. Ed., 2010, 49, 9170; (d) J. W. Sadownik and R. V. Ulijn, Chem. Commun., 2010, 46, 3481; (e) J. Wang, Y. Cao, Y. Feng, F. Yin and J. Gao, Adv. Mater., 2007, 19, 3865.
37. (a) P. Mukhopadhyay, Y. Iwashita, M. Shirakawa, S. I. Kawano, N. Fujita and S. Shinkai, Angew. Chem., Int. Ed., 2006, 45, 1592; (b) B. Roy, P. Bairi, A. Saha and A. K. Nandi, Soft Matter, 2011, 7, 8067.
38. (a) S. Manna, A. Saha and A. K. Nandi, Chem. Commun., 2006, 4285; (b) S. Manna, A. Saha and A. K. Nandi, Chem. Commun., 2008, 3732; (c) A. Saha, B. Roy, A. Garai and A. K. Nandi, Langmuir, 2009, 25, 8457.
39. K. M. Anderson, G. M. Day, M. J. Paterson, P. Byrne, N. Clarke and J. W. Steed, Angew. Chem., Int. Ed., 2008, 47, 1058.
40. A. Saha, B. Roy, A. Garai and A. K. Nandi, Langmuir, 2009, 25, 8457.
41. (a) B. Roy, A. Saha, A. Esterrani and A. K. Nandi, Soft Matter, 2010, 6, 3337; (b) P. Bairi, B. Roy and A. K. Nandi, J. Phys. Chem. B, 2010, 114, 11454; (c) A. Saha, B. Roy, A. Esterrani and A. K. Nandi, Org. Biomol. Chem., 2011, 9, 770; (d) B. Roy, A. Saha, A. Esterrani and A. K. Nandi, Soft Matter, 2010, 6, 3337; (e) B. Roy, P. Bairi, A. Saha and A. K. Nandi, Soft Matter, 2011, 7, 8067.
42. S. Chatterjee and A. K. Nandi, Chem. Commun., 2011, 47, 11510.
43. R. de Gelder, R. Wehrens and J. A. Hageman, J. Comput. Chem., 2001, 22, 273.
44. P. K. Sukul, D. Asthana, P. Mukhopadhyay, D. Summa, L. Muccioli, C. Zannoni, D. Beljonne, A. E. Rowan and S. Malik, Chem. Commun., 2011, 47, 11858.
45. R. Maffezzoni and M. Zanda, Tetrahedron Lett., 2008, 49, 5129.
46. (a) K. Hanabusa, T. Miki, Y. Taguchi, T. Koyama and H. Shirai, J. Chem. Soc., Chem. Commun., 1993, 1382; (b) K. Hanabusa, J. Tange, Y. Taguchi, T. Koyama and H. Shirai, J. Chem. Soc., Chem. Commun., 1993, 390.
47. S. Yagai, T. Karatsu and A. Kitamura, Langmuir, 2005, 21, 11048.
48. S. Mahesh, R. Thirumalai, S. Yagai, A. Kitamura and A. Ajayaghosh, Chem. Commun., 2009, 5984.
49. (a) J.-S. Shen, Q.-G. Cai, Y.-B. Jiangb and H.-W. Zhang, Chem. Commun., 2010, 46, 6786; (b) S. T. Huang, G. L. Wang, N. B. Li and H. Q. Luo, RSC Adv., 2012, 2, 10948; (c) K. Fan, L. Niu, J. Li, R. Feng, R. Qu, T. Liu and J. Song, Soft Matter, 2013, 9, 3057; (d) L. Niu, J. Song, J. Li, N. Tao, M. Lu and K. Fan, Soft Matter, 2013, 9, 7780.
50. (a) P. Bairi, B. Roy and A. K. Nandi, J. Mater. Chem., 2011, 21, 11747; (b) K. Sivashankar, A. Ranganathan, V. R. Pedireddi and C. N. R. Rao, J. Mol. Struct., 2001, 559, 41; (c) A. A. Massoud and V. Langer, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 2009, 65, 198; (d) S. Zhang, S. Yang, J. Lan, S. Yang and J. You, Chem. Commun., 2008, 6170; (e) J. Yuan and M. Liu, J. Am. Chem. Soc., 2003, 125, 5051; (f) J. Fei, L. Gao, J. Zhao, C. Du and J. Li, Small, 2013, 9, 1021.
51. (a) L. Lang, Gastroenterology, 2007, 133, 5; (b) K. Burns, J. Am. Vet. Med. Assoc., 2007, 230, 1285; (c) K. Burns, J. Am. Vet. Med. Assoc., 2007, 230, 1600; (d) H. S. Lam, P. C. Ng, W. C. W. Chu, W. Wong, D. F. Y. Chan, S. S. Ho, K. T. Wong, A. T. Ahuja and C. K. Li, Br. Med. J., 2008, 337, 2991; (e) E. Y. Y. Chan, S. M. Griffiths and C. W. Chan, Lancet, 2008, 372, 1444.
52. (a) J. Litzau, G. Mercer and K. Mulligan, http://www.fda.gov/cvm/CMSMelamine.htm, 2007; (b) J. P. Toth and P. C. Bardalaye, J. Chromatogr., A, 1987, 408, 335; (c) M. S. Fligenzi, E. R. Tor, R. H. Poppenga, L. A. Aston and B. Puschner, Rapid Commun. Mass Spectrom., 2007, 21, 4027; (d) H. A. Cook, C. W. Klampfl and W. Buchberger, Electrophoresis, 2005, 26, 1576; (e) J. A. Campbell, D. S. Wunschel and C. E. Petersen, Anal. Lett., 2007, 40, 3107; (f) E. A. E. Garber, J. Food Prot., 2008, 71, 590.
53. (a) K. Ai, Y. Liu and L. Lu, J. Am. Chem. Soc., 2009, 131, 9496; (b) B. Roy, A. Saha and A. K. Nandi, Analyst, 2011, 136, 67; (c) J. Lee, E. J. Jeongb and J. Kim, Chem. Commun., 2011, 47, 358.
54. (a) K. Aslan, Z. Leonenko, J. R. Lakowicz and C. D. Geddes, J. Phys. Chem. B, 2005, 109, 3157; (b) R. L. Zong, J. Zhou, Q. Li, B. Du, B. Li, M. Fu, X. W. Qi, L. T. Li and S. Buddhudu, J. Phys. Chem. B, 2004, 108, 16713; (c) N. R. Jana, L. Gearheart and C. J. Murphy, Chem. Commun., 2001, 617.
55. N. N. Yazgan, I. H. Boyaci, A. Topcu and U. Tamer, Anal. Bioanal. Chem., 2012, 40, 2009.
56. J. Pang, L. Yang, B. F. McCaughey, H. Peng, H. S. Ashbaugh, C. J. Brinker and Y. Lu, J. Phys. Chem. B, 2006, 110, 7221.
57. (a) J. Lee, H. Jun and J. Kim, Adv. Mater., 2009, 21, 3674; (b) J. Lee, H.-J. Kim and J. Kim, J. Am. Chem. Soc., 2008, 130, 5010.
58. (a) K. Tashiro, H. Nishimura and M. Kobayashi, Macromolecules, 1996, 29, 8188; (b) R. W. Carpick, D. Y. Sasaki and A. R. Burns, Langmuir, 2000, 16, 1270.
59. Q. Tang, Y.-T. Nie, C.-B. Gong, C.-F. Chow, J.-D. Peng and M. H.-W. Lam, J. Mater. Chem., 2012, 22, 19812.
60. (a) D. C. Sherrington and K. A. Taskinen, Chem. Soc. Rev., 2001, 30, 83; (b) F. Vera, J. L. Serrano and T. Sierra, Chem. Soc. Rev., 2009, 38, 781.
61. M. Shellaiah, Y. C. Rajan and H.-C. Lin, J. Mater. Chem., 2012, 22, 8976.
62. (a) T. Sanji, M. Nakamura, S. Kawamata, M. Tanaka, S. Itagaki and T. Gunji, Chem.–Eur. J., 2012, 18, 15254; (b) Y. Hong, J. W. Y. Lam and B. Z. Tang, Chem. Commun., 2009, 4332.
63. Q. Tang, Y.-T. Nie, C.-B. Gong, C.-F. Chow, J.-D. Penga and M. H.-W. Lam, J. Mater. Chem., 2012, 22, 19812.
64. (a) J. W. Goodby, Chem. Soc. Rev., 2007, 36, 1971; (b) T. Kato, Science, 2002, 295, 2414; (c) S. Kumar, Chem. Soc. Rev., 2006, 35, 83.
65. (a) E. Beltrán, J. L. Serrano, T. Sierra and R. Giménez, J. Mater. Chem., 2012, 22, 7797; (b) A. R. A. Palmans and E. W. Meijer, Angew. Chem., Int. Ed., 2007, 46, 8948; (c) C.-H. Lee and T. Yamamoto, Tetrahedron Lett., 2001, 42, 3993; (d) G. G. Barclay, C. K. Ober, K. I. Papathomas and D. W. Wang, Macromolecules, 1992, 25, 2947.
66. M. Xu, L. Chen, Y. Zhou, T. Yi, F. Li and C. Huang, J. Colloid Interface Sci., 2008, 326, 496.
67. X. Cheng, J. Jin, Q. Li and X. Dong, Chin. J. Chem., 2010, 28, 1957.
68. (a) A. A. Vieira, H. Gallardo, J. Barbera, P. Romero, J. L. Serrano and T. Sierra, J. Mater. Chem., 2011, 21, 5916; (b) F. Vera, R. M. Tejedor, P. Romero, J. Barbera, M. B. Ros, J. L. Serrano and T. Sierra, Angew. Chem., Int. Ed., 2007, 46, 1873; (c) J. Barbera, L. Puig, P. Romero, J. L. Serrano and T. Sierra, J. Am. Chem. Soc., 2006, 128, 4487.
69. (a) A. Piermattei, M. Giesbers, A. T. M. Marcelis, E. Mendes, S. J. Picken, M. Crego-Calama and D. N. Reinhoudt, Angew. Chem., Int. Ed., 2006, 45, 7543; (b) L. J. Prins, C. Thalacker, F. Wurthner, P. Timmerman and D. N. Reinhoudt, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 10042.
70. (a) H. Asanuma, T. Hishiya, T. Ban, S. Gotoh and M. Komiyama, J. Chem. Soc., Perkin Trans. 2, 1998, 1915; (b) Z. Cao, W. Liu, D. Liang, G. Guo and J. Zhang, Adv. Funct. Mater., 2007, 17, 246; (c) H. Asanuma, T. Ban, S. Gotoh, T. Hishiya and M. Komiyama, Macromolecules, 1998, 31, 371; (d) S.-W. Kuo and H.-T. Tsai, Macromolecules, 2009, 42, 4701; (e) R. Banerjee, N. J. Pace, D. R. Brown and E. Weerapana, J. Am. Chem. Soc., 2013, 135, 2497; (f) S. Perrone, M. Usai, P. Lazzari, S. J. Tucker, H. M. Wallace and M. Zanda, Bioconjugate Chem., 2013, 24, 176.
71. K. Kurihara, K. Ohto, Y. Honda and T. Kunitake, J. Am. Chem. Soc., 1991, 113, 5077.
72. (a) L. P. Stubbs and M. Weck, Chem.–Eur. J., 2003, 9, 992; (b) K. P. Nair, V. Breedveld and M. Weck, Macromolecules, 2008, 41, 3429.

---