supramolecular chemistry and crystal engineering...supramolecular chemistry and crystal engineering...

Supramolecular chemistry and crystal engineering

ASHWINI NANGIA

School of Chemistry, University of Hyderabad, Hyderabad 500 046, India.e-mail: [email protected]

Advances in supramolecular chemistry and crystal engineering reported from India withinthe last decade are highlighted in the categories of new intermolecular interactions, designedsupramolecular architectures, network structures, multi-component host–guest systems, cocrys-tals, and polymorphs. Understanding self-assembly and crystallization through X-ray crystalstructures is illustrated by two important prototypes – the large unit cell of elusive saccharinhydrate, Na16(sac)16 • 30H2O, which contains regular and irregular domains in the same struc-ture, and by the Aufbau build up of zinc phosphate framework structures, e.g. ladder motif in[C3N2H12][Zn(HPO4)2] to layer structure in [C3N2H12][Zn2(HPO4)3] upon prolonged hydrothermalconditions. The pivotal role of accurate X-ray diffraction in supramolecular and structural stud-ies is evident in many examples. Application of the bottom-up approach to make powerful NLOand magnetic materials, design of efficient organogelators, and crystallization of novel pharma-ceutical polymorphs and cocrystals show possible future directions for interdisciplinary researchin chemistry with materials and pharmaceutical scientists. This article traces the evolution ofsupramolecular chemistry and crystal engineering starting from the early nineties and projects acenter stage for chemistry in the natural sciences.

1. Introduction

The award of Nobel Prize to Charles J Pedersen,Donald J Cram and Jean-Marie Lehn in 1987marked the emergence of a new branch ofchemistry, namely supramolecular chemistry. Lehndefined supramolecular chemistry as ‘chemistrybeyond the molecule’, i.e. the chemistry of mole-cular aggregates assembled via non-covalent inter-actions [1,2]. Two decades later, supramolecularchemistry [3,4] is an important, interdisciplinarybranch of science encompassing ideas of physi-cal and biological processes. The roots of thisinterdisciplinary science lie in more than onefield. Host–guest chemistry goes back to the dis-covery of chlorine hydrate by Humphrey Davyin 1810 and Wohler’s H2S clathrate of β-quinolin 1849. Supramolecular chemistry in biologi-cal processes is nothing but molecular bind-ing recognized by Paul Ehlrich (1906) and Emil

Fischer’s lock-and-key principle (1894) brought incomplementarity and selectivity. Molecular recog-nition at the supramolecular level is mediated bycomplementarity – even for like molecules it is thedissimilar portions of functional groups that inter-act with one another. An electropositive hydro-gen bond donor approaches an electronegativeacceptor (Dδ− ––– Hδ+ · · ·Aδ−), cation· · · anion elec-trostatic interaction in salts and metal complexes(M+X−), and bumps in one part of the moleculefit into hollows of another portion (hydrophobicinteractions), and so on. Even as the fundamentalrecognition processes that guide supramolecularaggregation are governed by the same princi-ples and forces, the chemical systems studiedare broadly classified into two major categories(figure 1): molecular recognition in solution isgenerally referred to as supramolecular chemi-stry, and organized self-assembly in the solid-state as crystal engineering [5,6]. There have been

Keywords. Crystallization; hydrogen bond; materials; nanoscience; pharmaceutical; self-assembly.

35

36 ASHWINI NANGIA

Figure 1. Molecular recognition of molecules to give supermolecule and periodic arrangement of supermolecules in acrystal lattice. Note the complementary shape and bonding feature of interacting molecules.

Table 1. Strength scale of different intermolecular interactions and hydrogen bonds.

Interaction type Energy (kcal mol−1) Examples

Very strong H bonds 15–40 O––– H · · ·O−, F ––– H · · ·F−

Coordinative bonds 20–45 M ––– N, M––– OStrong hydrogen bonds 5–15 O––– H · · ·O, N––– H · · ·OWeak hydrogen bonds 1–4 C––– H · · ·O, O ––– H · · · πvan der Waals interactions 0.5–2 CH3 · · ·CH3, CH3 · · ·PhHeteroatom interactions 1–2 N · · ·Cl, I · · · I, Br · · ·Brπ-stacking 2–10 Ph · · ·Ph, nucleobases

significant advances in both these streams over thelast two decades, and this review presents somesalient developments and highlights.

2. Research overview

This review article is part of a special issue tocelebrate the Platinum Jubilee year of the IndianAcademy of Sciences. In this background, Indiancontributions in supramolecular chemistry andcrystal engineering are selected from those pub-lished in the current decade. Related developmentsin materials science and nanotechnology and thoseat the chemistry–biology interface are covered else-where in the same volume.

2.1 Intermolecular interactions

Systematic studies on the nature of hydro-gen bonds and intermolecular interactions lie atthe heart of directed self-assembly. From thevery strong negatively-charged hydrogen bondsand metal–heteroatom coordination bonds tostrong and weak hydrogen bonds and inter-heteroatom interactions span an energy range of50 kcal/mol (table 1) [7]. The esoteric organicchlorine group was examined as an acceptorfor OH and CH donors by a few groups.Intramolecular O ––– H · · ·Cl hydrogen bond in1-chloro derivatives of trans-9,10-diethynyl-9,10-dihydroanthracene-9,10-diols (DDA, figure 2) wasconfirmed by NMR spectroscopy and X-ray diffrac-tion [8]. The intramolecular O––– H · · ·Cl hydrogenbond (d = 2.25–2.40 A) persisted in five out of six

Figure 2. Intramolecular O––– H · · ·Cl interaction in1-chloro derivative of trans-9,10-diethynyl-9,10-dihydroan-thracene-9,10-diol (DDA). The persistence of this intramole-cular geometry in five chloro derivatives was confirmed byX-ray diffraction, 1H NMR spectroscopy and D2O exchangeexperiments.

structures with varying number of Cl groups onthe aromatic core. Of the two OH groups, the freeOH resonates at δ 2.8 ppm in the NMR spectrumwhereas the hydrogen bonded OH is significantlydownfield at 4.4–4.6 ppm. Moreover, the free OHexchanged with D2O immediately but the bondedOH is fully exchanged after 1 h. O ––– H · · ·Cl hydro-gen bond energy of 4.0 kcal mol−1 (DFT, GAMESS,B3LYP/6 ± 31G∗) is at the upper limit of weakhydrogen bond range (0.5–4.0 kcal mol−1). Interac-tions to the elecgtronegative Cl acceptor are acti-vated by metal and anion nature (Cl ––– M,Cl−)compared to organic chlorine (Cl ––– C). Coor-dination compounds of Co, Cu and Zn withbi/tridentate pyrazolyl/pyridyl ligands showed anew C ––– H · · ·Cl ––– M inorganic supramolecular

SUPRAMOLECULAR CHEMISTRY AND CRYSTAL ENGINEERING 37

Figure 3. Exponential dependence of ρb (e A−3

on Rij (A). Circles and triangles represent experimental and theoreticalvalues and solid and dashed black lines show the corresponding best fit curves. Vertical dashed lines demarcate the threeregions of strong H bonds, weak H bonds and van der Waals interactions. See table 1 for energies of these interactions.Reproduced with permission of Royal Society of Chemistry.

synthon [9]. An electron rich Cl can act as anacceptor from one, two and upto three CH aro-matic/methyl donors (mono, di and trifurcation)in C ––– H · · ·Cl ––– M hydrogen bonds (2.48–2.88 A).Involvement of both O ––– H · · ·Cl and C ––– H · · ·Clhydrogen bonds is seen in a hydrate coordi-nation compound. Aromatic π ––– π stacking ofpyridyl/pyrazolyl rings gave ladder networks inthese crystal structures. An interesting anomalyin O ––– H · · ·Cl/C ––– H · · ·Cl interactions comparedto O––– H · · ·O/C ––– H · · ·O hydrogen bonds is that,surprisingly, OH donors make longer contacts thanCHs, which is quite the opposite with O and Nacceptors in hydrogen bonds. Given the bifurcationin these interactions, the term hydrogen bridge [10]was recently resurrected instead of the often usedhydrogen bond term.

In a charge density based classification of hydro-gen bonds, topological parameters such as electrondensity (ρb), Laplacian (∇2ρr), interpenetration ofthe van der Waals spheres (ΔrD + ΔrA) correlatewell with the length of the interaction line Rij.Based on Rij (1.6–3.8 A) and ρr (0.3–0.005 e A

−3)

values, the continuum of HBs to vdW inter-actions was classified into three regions, shownin figure 3 [11]. The strong HB Region 1 ofO ––– H · · ·O/N ––– H · · ·O HBs has Rij < 2.2 A andρb > 0.1 e A

−3, Region 2 of C ––– H · · ·O/N ––– H · · · S

HBs in range 2.2 < Rij < 2.8 A and 0.08 > ρb >

0.02 e A−3

, and finally vdW region of Rij > 2.8 Aand ρb < 0.05 e A

−3containing C ––– H · · · π, π · · · π,

etc. interactions. There is an exponential depen-dence in ρb vs. Rij curve spanning the three regionsfor all types of hydrogen bonds and intermolecularinteractions, and a remarkable correlation betweenexperimental results and theoretical calculations.In a related study on O ––– H · · ·O hydrogen bonds[12] electron density at the bond critical point ρb

is in the range 0.03–0.4 e A−3

and its Laplacian is0.7–6.0 A

−5. H · · ·O bond CPs lie in the expected

range but the bond paths are often highly curvedand displaced away from the HB axis, by as muchas 0.4 A at the critical point. The hydrogen bondcharge density distribution is related not only tothe cores of the donor and acceptor atoms but alsoto the lone pairs as well.

Very short hydrogen bonds are important inenzyme catalysis, proton transfer, drug–receptorrecognition and binding, ice structures, andsupramolecular chemistry. The traditional view isthat short–strong HBs are stabilized by chargeor resonance or polarization assistance (CAHB,RAHB, PAHB). Neutron diffraction on a largesingle crystal of pyrazine-2,3,5,6-tetracarboxylicacid [13] revealed a new type of hydrogen bondshortening phenomenon (H · · ·O 1.5 A,O · · ·O <2.5 A), namely a cooperative, finite array of σ-and π-assistance in the hydrogen bond network.

38 ASHWINI NANGIA

Figure 4. Synthon assisted hydrogen bond (SAHB)shortening (H · · ·O 1.41 A) through cumulative π- andσ-cooperative hydrogen bond array in the neutron diffrac-tion crystal structure of 2,3,5,6-pyrazinetetracarboxylic acidat 20 K. Nuclear density is localized on the H atom andthere is no evidence of disorder in the calculated Fourier map(Fobs). Reproduced with permission of American ChemicalSociety.

The short O ––– H · · ·O hydrogen bond (figure 4) ofcarboxylic acid donor is activated by π-cooperativeRAHB and the water acceptor becomes strongerby σ-cooperative PAHB in the synthon assistedhydrogen bond (SAHB). That there is no disorderin the H atom position of short O––– H · · ·O bondwas confirmed in the neutron diffraction Fouriermap at 20 K.

2.2 Crystal engineering

Desiraju [6] defined crystal engineering as “theunderstanding of intermolecular interactions inthe context of crystal packing and the utiliza-tion of such interactions in the design of newsolids with desired physical and chemical prop-erties”. The subject was brought into the main-stream of organic chemistry through the conceptof supramolecular synthons [14], which are repeat-ing structural units in crystal structures that areable to guide the rational design of supramolecu-lar architectures based on a small number of recur-ring hydrogen bond patterns. Whitesides [15] gavea physical organic chemistry interpretation to crys-tal engineering as “the study of molecular andcrystal structure correlation in a family of com-pounds”. A seminal study on hydrogen bondingand crystal packing in amino-phenols highlightsthe importance of systematic molecular variationto understand crystal structure packing types.

In a series of homologous amino-phenol crystalstructures, odd and even methylene chain link-ers (n = 1–5) and CH2 → S isosteric replacementwere examined [16]. The expected β-arsenic sheetmotif of N ––– H · · ·O and O ––– H · · ·N hydrogenbonds is the dominant motif along with squareand infinite chain (figure 5) and even N ––– H · · · πinteraction. The even linker structures invariablycontained the stable β-As motif whereas the oddseries tends towards the sheet motif only when thelinker length is long. There is excellent structuralsimilarity in 4-aminophenol, 4,4′-aminobiphenylol,and the even linker compounds, with all of themcontaining the β-As sheet but with subtle differ-ences in terms of space group and the networkbeing diamondoid or wurtzite-like. On the otherhand, the odd series compound (n = 1) has thetetramer N(H)O synthon and more significantlyan NH donor that does not participate in conven-tional H bonding but instead makes N ––– H · · · πinteraction with a phenol ring. The approach geom-etry of the N ––– H · · · π interaction to the phenylC ==C bond was confirmed by neutron diffraction(2.39 A, 155.5◦). As the methylene linker becomeslonger (n = 3, 5) the N(H)O motif changes to infi-nite chain but the weak N––– H · · · π persists. A nearlinear disposition of OH and NH2 groups givesthe β-As prototype whereas a bent arrangement,either in meta- and ortho-aminophenols or in oddlinker amino-biphenylols, gave square or infiniteN(H)O together with the unexpected N ––– H · · · πinteraction.

The common carboxylic acid dimer synthonchanges to the rare catemer motif when substi-tuted mesitoic acids have halogen atoms in themeta-positions. Thus weak inter-halogen (Cl · · ·Cl,Br · · ·Br) and C––– H · · ·X interactions were shownto direct the helical assembly of strong O ––– H · · ·Ohydrogen bonds [17]. The structure-directing roleof halogen-driven interactions was confirmed inthe crystal structure of pentamethyl benzoic acidwhich formed the expected dimer synthon. In sub-stituted diaryl ureas with NO2 and halogen groups,the dominant hydrogen bond synthon is notthe expected urea α-network but urea· · · nitroN ––– H · · ·O synthon (X = Cl, Br, CN, H, Me, etc).The elusive N ––– H · · ·O urea tape in this fam-ily was directed by the soft and weak I · · ·O2Nand C ––– H · · ·O2N interactions [18]. The idea thatstrong (hard) and weak (soft) interactions bondpair-wise leads to synthon control and crystaldesign in multifunctional molecules.

The structural motif of amides and reverseamides in homolog series showed that the β-sheetprototype is sustained exclusively by N––– H · · ·Ohydrogen bonds in aromatic amides such as PhAm.However, introduction of the pyridyl moiety, e.g.as in 3-PyAm and 3-PyRevAm homologue series


Figure 5. (a) Hydrogen bond synthons in homolog series of amino-phenols. (b) β-As sheet in 4-(4-aminophenethyl)phenol(n = 2). (c) Square motif in 4-(4-aminophenmethylene)phenol (n = 1). N ––– H · · ·π interaction is not seen in this view.(d) Infinite NHO chain in 4-(4-aminophenpropyl)phenol (n = 3) and N ––– H · · ·π interaction. Reproduced with permissionof American Chemical Society.

(figure 6) [19], gave structures with N––– H · · ·Ohydrogen bonds and C ––– H · · ·N interactionsand N––– H · · ·O and N––– H · · ·N hydrogen bonds(n = 6), respectively. Thus, while the pyridyl groupplays an auxiliary role in 3-PyAm structuresit is definitely interfering via N ––– H · · ·N hydro-gen bond in 3-PyRevAm. 4-PyAm compoundsare similar to 3-PyAm but 4-PyRevAm struc-tures crystallized as hydrates. The above examples[17–19] convey that interference from weak halogen

or C ––– H · · ·O/N interactions to strong and robustO ––– H · · ·O/N ––– H · · ·O synthons is almost impos-sible to know prior to X-ray crystal structure analy-sis. The prediction and control of functional groupinterference, or cross-talk, in hydrogen-bondedstructures is a continuing challenge, further com-plicated by the conformational flexibility of organicmolecules.

Carbohydrates are important biomolecules oflife. Hydrogen bond patterns in crystal structures

40 ASHWINI NANGIA

Figure 6. 3- and 4-pyridyl amides have a β-sheet or square motif but the reverse amides show very different hydrogenbonding and molecular packing due to perturbation by N––– H · · ·N hydrogen bonding and introduction of water in thecrystal lattice.

of carbohydrates were summarized by Jeffrey andSaenger [20]. (1) maximization of the total num-ber of hydrogen bonds per molecule using asmany donor/acceptor oxygen atoms as possible,and (2) maximization of cooperativity by form-ing as many finite and infinite chains of hydrogenbonds as possible. Inositols and polyols representmanageable molecular systems to understand thecomplexities of hydrogen bonding possible in poly-hydroxylated molecules. Rigid poly-hydroxylatedcyclohexanes with trans ring fusions (polycycli-tols) have 1,3-syn OH groups that make an invari-ant intramolecular O ––– H · · ·O hydrogen bondedsix-member ring (figure 7), and hence the pack-ing of polycyclitols may be understood in termsof a limited number of intermolecular hydrogenbonds with neighboring molecules [21]. The forma-tion of infinite O ––– H · · ·O chains is predicted butthe number of crystallographic unique molecules(Z ′) is assumed to be 1 though this is not alwaysthe case for diols [22]. In the event, cooperativeO ––– H · · ·O chains and (OH)4 tetramers of inter-molecular hydrogen bonds connect 1,3-syn diaxialintramolecular H-bonded molecules. By lockingthe conformational flexibility of the OH groupin diaxial orientation through intramolecular Hbonds, the packing modes of rigid polyols are pre-dicted in a limited number of hydrogen-bondedarchitectures.

2.3 Supramolecular architectures andnetwork structures

As mentioned in the introduction, host–guest andclathrate structures are the original supramolec-ular structures, even before the term ‘supramole-cular chemistry’ was coined. A supramolecular

Figure 7. Conformationally locked cyclitols with anintramolecular O––– H · · ·O hydrogen bond crystallize in asmall number of H bonding motifs. Crystal structures ofthese model compounds help to understand the more com-plex and unpredictable structures of carbohydrates.

architecture of alternate open (4 × 4 A) and closedchannels sustained by C ––– H · · ·O interactions(2.4–2.7 A) was observed in the cubic symmetrycrystal structure of Cu(II) N-salicylidene-2-methoxyaniline coordinate complex [23]. Theclosed channels are filled with phenyl rings ofsalicylaldimine. The crystal structure of trigonalmolecule bis(4-hydroxyphenyl)(phenyl)methaneshows symmetry carry-over to supramolecular tri-angular and hexameric O ––– H · · ·O synthons inrhombohedral space group R3 [24]. Rhombohedraland monoclinic polymorphs of β-hydroquinonewere reproduced in phenyl-extended 2,2′,6,6′-tetramethyl-4,4′-terphenyldiol [25], illustrating afine example of network engineering in polymorphs.

Crystal-to-crystal guest inclusion/release reac-tions in the solid-state are as such rare. Thetrinuclear compound [Fe(μ3-O)(μ2-OAc)6(2-pyri-done)2(H2O)]ClO4 • 3H2O containing a Fe ––– OH2


coordinate bond transformed to Fe ––– O(H)Mein [Fe(μ3-O)(μ2-AOc)6(2-pyridone)2(MeOH)]ClO4•3H2O while still retaining lattice water and singlecrystallinity [26]. The methanolyated complexcould be regenerated to the hydrate by exposureto atmospheric water vapor without color loss orX-ray diffraction intensity in a reversible transfor-mation. The reversibility of states was confirmedby IR on MeOH and MeOD complexes. A double[2 + 2] photocycloaddition of alkenes to cyclobu-tanes was templated by phloroglucinol in thecrystalline environment of cocrystal [27]. Reactionof bis(pyridinecarboxamido)alkanes with Cu(II)resulted in open 1D chains containing solvent andcounterion molecules. 2D layers of (4,4) topologyhaving rhomboidal cavities are either filled withcounterions or interpenetrated to give close-packedcrystal structures. The exchange of ClO−

4 withPF−

6 anion resulted in the transformation of 1Dchain to 3D interpenetrated network in an irre-versible manner [28].

Apart from coordination polymers, organometal-lic clusters too can be used to build supramolecularassemblies via a variety of intermolecular inter-actions and metal coordination bonds. Hexamericorganostannanes [n-BuSn(O)OC(O)Ar]6, popu-larly called as drums, contain a structurally similarstannoxane unit made up of prismatic Sn6O6 core[29]. Weak C ––– H · · ·O, C ––– H · · · π interactionsand π-stacking of aromatic substituents direct thefinal supramolecular architecture of such organotindrums.

The assembly of lattice inclusion hosts orclathrate compounds has advanced rapidly in thelast decade. A three component host lattice madeup of 1,3cis,5cis-cyclohexanetricarboxylic acid(H3CTA) hydrogen bonded to 4,4′-bipy-eta and4,4′-bipy-bu (where eta and bu are (CH2)2and (CH2)4 methylenes separating 4,4′-bipyridinebase). A novel feature of self-assembled termole-cular organic host container [H3CTA • bipy-eta •(bipy-bu)0.5] is that bipy-eta in a gauche conforma-tion builds the closed 1D channels with H3CTA,while bipy-bu connects them. A variety of aro-matic guest molecules occupy the 10× 12 A squarecavities [30]. A hierarchic self-assembly model of1D helices to 2D hexagonal sheets is proposed.Starting from 1,2,4,5-benzenetetracarboxylic acida variety of aza-acceptors such as 1,10- and 1,7-phenanthroline, phenazine, bis-pyridines, etc. werecocrystallized to make sheet-like networks. Thesecomplexes crystallize in two broad categories –host–guest compounds with aza partner moleculesin channels created by the acid host and supramole-cular assemblies of infinite molecular tapes [31].The tetra-acid does not engage in the usual cen-trosymmetric COOH dimer but instead makessingle O ––– H · · ·O or O ––– H · · ·O− hydrogen bonds.

The adventitious inclusion of water invariablypromoted the formation of channels to give host–guest complexes. These 2D sheet structures adoptdifferent stacking modes.

Conglomerate crystallization, or the sponta-neous assembly of homochiral crystals as opposedto racemates, is as such a rare phenomenonbelieved to occur in no more than 5–10% cases[32]. Crystallization of phenols, alcohols anddiols/polyols is more likely in enantiomorphousspace groups such as P21 and P212121 because OHgroups often hydrogen bond via screw axis sym-metry [33]. An interesting and unusual exampleof C ––– H · · ·Cl interaction (2.7 A) and π-stacking(3.7 A) mediated chiral crystallization is the heli-cal channels in coordination polymer [(L)ZnIICl2](L = α,α′-bis(pyrazolyl)-m-xylene). Both left- andright-handed helical polymer chains along theb-axis (space group (P21) in different crystals fromthe same batch were identified [34]. Optical rota-tion of the bulk solution made up of equal amountsof M and P crystals is zero.

Hydrothermal synthesis of zinc phosphate frame-work structures starting from amine, phosphoricacid and ZnII in suitable solvents gave avariety of supramolecular architectures. Rao andNatarajan [35,36] demonstrated that self-assemblyprogresses in a hierarchical manner followingthe Aufbau principle – from 1D chains andladders to 2D sheets and finally to the 3Dframeworks. For example, 1,3-diaminopropanephosphate (DAPP) on reaction with Zn2+ ionsgives a ladder phosphate, [C3N2H12][Zn(HPO4)2],comprising edge-shared four membered ringswhereas prolonged reaction at 30–50◦C yieldeda layered structure, [C3N2H12][Zn2(HPO4)3]. Thelayers are formed from a zigzag chain of four-membered rings, constructed from two Zn and Patoms (Zn2P2O4 units), that are connected to eachother via two PO4 units, creating a bifurcationwithin the layer. Reaction of DAPP with Zn2+

ions in aqueous solution at 30◦C for 24 h gives aproduct whose XRD pattern shows lines due tothe ladder structure while the product obtainedfrom reaction at 50◦C (24 h) shows a reflectiondue to ladder and layer structure (d002) at 8.5 A.Transformation to the layer structure probablyoccurs through the ladder phase. These resultsare supported by in situ 31P NMR studies carriedout at 85◦C, which showed the disappearance of theamine phosphate signal followed by the immediateappearance of a signal due to the precursor phase,before the ladder phase is formed. A generalizedscheme of structural relationships and dimension-ality evolution was proposed (figure 8) in which thefour-membered ring appears to be the first unitformed in the process of building of these complex

42 ASHWINI NANGIA

open-framework structures, which initially form aone-dimensional chain or a ladder structure, andthen transform to 2D and 3D structures [35].In effect, the four membered rings or/and theone-dimensional structures are the synthons ofthe more complex structures. The formation ofsix-, eight-, and higher membered rings, commonlypresent in the open framework phosphates, mayfollow from the 0D/1D structures, with the four-membered rings themselves transforming to thehigher rings.

2.4 Water clusters

Science ranked the study of water among the top10 breakthroughs in 2004 [37]. Interest in waterclusters of different topologies and dimensionalitytrapped in small organic and metal-organic crystalstructures continued to grow in the current decade.Rare example of planar water hexamers, penta-gonal 2D sheets, 1D water helices, and macrocyclewater rings were reported by several groups [38].However, a critical review [39] showed that someof these water motifs were not as remarkable andnovel as made out by the original authors. Such anomission in the Google search engine internet ageis difficult to condone as mere oversight. Whereaswater has a prolific ability to hydrogen bond withpolar functional groups and metal centers in aplethora of motifs, the genuine new chemistrygenerated by water clusters, other than novel struc-tural motifs displayed as colorful images, is an openquestion.

A solitary exception is crystal structures ofsodium saccharinate hydrates [40]. A dihydrate ofNa3(sac)3 • 2H2O (triclinic P 1, 0.66 water per Nasaccharinate) was well known as the only formwhose X-ray coordinates were accurately deter-mined. A monoclinic crystal structure (P21/n)was solved and refined to good R-factor (0.045)for Na16(sac)16 • 30H2O, which is equivalent toNa(sac) • 1.875H2O. This novel hydrate of Na(sac)has a large unit cell of 15, 614 A

3containing 362

atoms (238 non H atoms) in the asymmetric unit[41]. The 64 Na+ cations, 64 sac− anions, and120 water molecules in the unit cell make thiscrystal structure one of the largest and most com-plex ever for simple ions/molecules. The crystalstructure (figure 9) has ‘regular’ and ‘irregular’regions showing certain similarities and differences.In the regular domain, saccharinate anions arenearly parallel and stacked, Na+ ions are hexaco-ordinated with water and sac−, and water mole-cules are hydrogen bonded. In the irregular region,there is disorder of sac−, Na+ (some of which isnot necessarily hexacoordinated), and water (someof which is ill-resolved). The regular region ofthe structure consists of ten sac− anions, which

are arranged in a stack of five water-bridgedhydrogen-bonded pairs, with an average inter-planar distance of 3.69 A (figure 9, left side). Twowater molecules are involved in each saccharinatepair through strong O––– H · · ·N hydrogen bonds(2.85 A, θ 165.9◦). Cross-linking of the pairs occurswith octahedrally coordinated Na+ ions (meanNa+ · · ·O 2.39 A). The result is a compact, finitearrangement of sac−, Na+, and water molecules inthe form of three supramolecular cubes and twohalf-cubes. In the irregular region (figure 9, rightside), six sac− anions are not parallel and Na+ andwater are positionally/orientationally disordered.Based on structure determination at four differ-ent temperatures (100, 150, 200, and 298 K) it wasconcluded that the regular part of the structureresembles a conventional crystal whereas the adja-cent irregular region has solution-like characteris-tics. In effect, the structure represents a state of‘incipient crystallization’ somewhere between theanhydrate, dihydrate and water rich forms. Mostremarkably, and as a very rare case, it was shownthat this Na(sac) 1.875 hydrate crystal picks up

Figure 8(a). 1D ladder phosphate [C3N2H12][Zn(HPO4)2] (above) and 2D layer structure of [C3N2H12][Zn2(HPO4)3] (below).


Figure 8(b). Various types of transformations in open framework zinc phosphates. As discussed in the text, featuresof the 1D ladder phosphate in the 2D layer structure establish an evolutionary relationship in self-assembly. Structuraltransformations were monitored by XRPD and 31P NMR at different time intervals and temperature ranges. T is tetrahedralframework atom (Zn or P). Refer to original paper [36] for compound numbers. (Reproduced with permission of AmericanChemical Society).

Figure 9. Crystal structure of Na(sac)•1.875H2O (Na pink, O red, N blue, S yellow, C gray, H cream). The regular regionon the left side has 10 sac− residues and the irregular region on the right has six sac− ions. Note the finite supramolecularcube arrangement. Sac residues are numbered. (Reproduced with permission of Wiley-VCH).

and loses water equally easily. Another structuralreport on saccharinate hydrate appeared simulta-neously [42].

2.5 Polymorphism

Research papers on polymorphism dominated thecrystal engineering literature in the current decade.

The number of organic polymorphic sets has risen10 fold, from about 160 in 1995 to >1600 in 2007(table 2) [43]. The record for maximum numberof solved crystal structures for the same chemicalcompound are the seven polymorphs of 5-methyl-2-[(2-nitrophenyl)amino]-3-thiophenecarbonitrile, orROY, so named because of its red, orangeand yellow crystal colors arising from different

44 ASHWINI NANGIA

Table 2. Number of ‘organic’ polymorphs with ≥3 formsin the CSD. The total number of organic polymorph sets in1995 was 163, and this number rose 10-fold to ca. 1600 in2007.

1995 2000 2002 2005 2007

3 forms 13 27 42 102 1244 forms 3 3 3 14 205 forms 0 0 1 1 36 forms 0 0 1 1 07 forms 0 0 0 0 1

molecular conformations in different structures.Dimorphs of diphenyl ether (m.p. 20◦C) were crys-tallized by in situ cryo-crystallization in a sealedcapillary tube at 260–255 K to give a crystal whichsolved in centrosymmetric space group P21/n.Lowering of the temperature to 240 K and anneal-ing, to improve crystal quality, indeed gave a betterquality crystal but in non-centrosymmetric spacegroup P212121 [44]. An intramolecular C ––– H · · · πinteraction locks the molecular conformation inboth forms. The crystal structure of form I is medi-ated by a three-dimensional network of C ––– H · · · πinteractions whereas form II has a tetramer ofthe same interactions. An additional C ––– H · · ·Ointeraction in the latter modification is believedto provide the extra stability to the thermody-namic polymorph II. Subsequent to the isolationof the stable orthorhombic polymorph, the mono-clinic form became a disappeared species. Suchanecdotes are known for polymorphs.

Polymorphism has important implications inpharmaceutical solid-state formulation, dissolutionprofile, drug life-cycle management, and tablet-ing. Polymorphism became a major issue in thepharmaceutical industry in the mid-to-late 1990sbecause of litigation surrounding forms 1 and 2of Zantac (Glaxo vs. Novpharm) and the acci-dental appearance of a stable, less soluble form2 of Ritonavir (Abbott) in production batches.A second polymorph of the popular analgesicaspirin was accidentally discovered during cocrys-tallization [45] and the structural landscape ofaspirin polymorphs were revised [46]. Both form 1(known) and form 2 (new) crystal structurescontain the centrosymmetric O ––– H · · ·O dimersynthon between COOH groups. However, the dif-ference lies in the way in which these acid dimerlayers are connected. In form 1, they are connectedthrough C ––– H · · ·O dimers related by the inver-sion center whereas in form 2 they form C ––– H · · ·Ocatemers between screw axis related molecules. Ineffect, identical O––– H · · ·O layers are displacedwith respect to each other in the two structures.In crystal structure prediction of form 2, it wasmentioned that this low energy structure has a

low shear elastic constant and hence a low energybarrier to transformation.

To estimate the total domain ratio in a givenbatch of aspirin crystals, a batch scale factor wasintroduced into the crystallographic refinements,applied only to reflections with odd l. The refinedvalue of this scale factor gives the relative weightsof form 1 and form 2 reciprocal lattices and there-fore a direct estimate of the crystal composition.This procedure would be exact for two perfectlyordered domains with a single domain boundary,but becomes progressively approximate for realaspirin crystals since the extent of domain disorderincreases. The procedure gives a total compositionestimate but no direct information concerning thesizes of form 1 and 2 domains or their distributionwithin the crystal lattice. On an unsatisfactorilysolved crystal of aspirin, such a refinement againstform 2 data set gave R1 = 0.054 and wR2 = 0.132[46], a significant improvement on the standardrefinement to give refined batch scale factor ofroughly 75% form 2 domains. In comparison, anearlier data set (R1 0.162,wR2 0.308) [45] solved asform 2 was shown to contain an equal proportionof form 1 and 2 domains.

2.6 Cocrystals and salts

Cocrystals are a relatively recently studied class ofsolid-state structures compared to salts which arewell known in the pharmaceutical industry. A veryearly example of a cocrystal is the 1:1 molecularcomplex between benzoquinone and hydroquinone,named quinhydrone, reported by Wohler in 1844.It is the first cocrystal structure in the CambridgeStructural Database [47] with reported coordi-nates for two polymorphic forms, a monoclinicform in space group P21/c (QUIDON02) and atriclinic structure in P 1 space group (QUIDON).There is a resurgence of interest in cocrystals inthe last decade or so, particularly because recentexperiments suggest that they represent new solidstate pharmaceutical forms for solving solubility,hydration, stability and even toxicity issues indrugs [48]. If the supramolecular synthon concept[14] provided rational approaches to crystal syn-thesis, the classification of synthons as homosyn-thons (those between like functional groups) andheterosynthons (unlike functional groups) made itpossible to dissect cocrystals as being built upfrom molecules containing complementary func-tional groups. Thanks to the Cambridge Struc-tural Database [47], which contains over 450 000crystal structures and user-friendly fragment andmotif search protocols, it is possible to estimatethe probability of various synthons in the globalarchive (figure 10). The hierarchy of homo- andheterosynthon probability in turn becomes the


Figure 10. Probability of supramolecular homosynthonsand heterosynthons in the CSD. The higher the probabil-ity of a synthon, the greater is its likelihood of occurrence,or predictability, in crystal structures. A rationally designednovel amide-pyridine-N-oxide heterosynthon has high occur-rence probability of 80%. In contrast, amide-pyridine syn-thon having <5% probability has little predictability incocrystal synthesis.

guide to systematic cocrystal design and engineer-ing. The higher probability synthons are morereliable and robust in giving the expected hydro-gen bond motif in the designed crystal struc-ture. However, the main issue in carboxylic acidand pyridine cocrystals, i.e., whether the prod-uct will be neutral or ionic, remains far fromsolved.

Cocrystallization of nucleobases with aromaticcarboxylic acids [49] gave diffraction qualitysingle crystals of adenine •benzoic acid, cyto-sine • benzoic acid, cytosine • isophthalic acid, andcytosine • phthalic acid. Nucleobase self-recogni-tion is very strong to give hydrogen-bonded dimers,which are in turn connected via the carboxylicacid. In a related study, trimesic and pyromel-litic acid were used as cocrystal formers withcytosine. Now the base dimers were disrupted togive carboxylate· · · pyridinium hydrogen bondingin both adducts. In the opinion of this author, thefirst set of crystal structures [49a] were somewhatunsurprising since the main intent of exploit-ing directed hydrogen bonding to make cocrys-tals was not achieved, whereas the second study[49b] successfully achieved the target heterosyn-thon. A main question that remains unansweredis why the mono- and diacid coformers gave onestructure prototype whereas tri- and tetra-acidsafforded a different structure type. The fact thatpKa decreases as successive COOH groups areadded (towards more acidic) in the series benzoic,isophthalic, phthalic, trimesic and pyromellitic acid(pKa 4.17, 3.46, 2.98, 3.12 and 1.92), could bea factor in going from neutral to ionic synthon.

This latter point was brought out in a recent pairof studies on carboxylic acid–pyridine cocrystalswherein the presence of phenol OH group gave neu-tral cocrystals with O ––– H · · ·N hydrogen bond-ing whereas when OH and NH2 groups were bothpresent the ionic N+ ––– H · · ·O− was consistentlyfound [50]. These studies alert us to the limitationsof the ΔpKa rule in predicting the location ofproton in acid–base complexes. It is likely thatthe presence of additional functional groups inthe molecule and their location in the supramole-cular environment of the crystal structure modifiesthe acidity and basicity of functional groups com-pared to native values for the functional groups.In the absence of accurate pKa in the precisemolecular and/or supramolecular environment, theΔpKa rule should be applied with caution to knowneutral–salt states.

The discovery of new synthons adds to thecrystal engineering ‘building kit’. Carboxamide-carboxylic acid heterosynthon is well known in theliterature but a limitation with this motif is thatits probability of occurrence is modest at 50%, i.e.other motifs might occur in competition, notablythe parent homodimers. With the idea of optimi-zing hydrogen bond acceptor strength for theamide functional group, the pyridine N acceptorwas oxidized to the N-oxide resulting in a dramaticincrease in acceptor strength. ESP charge in isoni-cotinamide N = −43.7 kcal mol−1, isonicotinamide-N-oxide O− = −53.3 kcal mol−1) and pKHB ofpyridine N, amide O and N-oxide O− are1.86, 1.96 and 2.70 (increasing basicity). Theoccurrence probability of rationally designedcarboxamide· · · pyridine-N-oxide heterosynthon is80% (figure 10) in diverse crystal structures of APIsand co-formers [51].

2.7 Supramolecular materials

Supramolecular chemistry and crystal engineeringpractised at the A and nm scale are the meet-ing point of ‘top down’ chiseling and ‘bottomup’ construction of nanostructures for materialsscience and technology. Understanding structure–property correlation and finding the optimal mate-rial for a particular application is the goal in thesestudies. Third order (χ3) nonlinear optical pro-perties of core-modified porphyrins were shown todepend on the structure of the macrocycle, itsmolecular conformation, the number of π-electrons,and the extent of conjugation [52]. These factorswere evaluated by comparing the structures ofmodified 34π octaporphyrins with reference 26πhexaphyrins (figure 11) and their σ2 values mea-sured by the two-photon absorption process (whichare a measure of cubic susceptibility coefficientγ). Regular porphyrins generally have small σ2

46 ASHWINI NANGIA

Figure 11. Core modified hexaphyrin and octaphyrin analogues for 3rd order NLO materials. σ2 values increase from2–10,000 GM in 26π hexaphyrins to 80–90,000 GM for 34π octaphyrins due to enhanced electronic interactions between theinner porphyrin pockets of the ‘figure eight’ macrocycle. Various R groups studied are shown. Reproduced with permissionof American Chemical Society.

(absorption cross section) in the range 1–10 GM(1GM = 10−50 cm4 sec photon−1) in near-IR wave-length and 100–1600 GM in Soret band region.High σ2 values are due to extended conjugation,which increase further with electron-donating sub-stituents in the same macrocycle. The 34π octa-phyrins adopt a figure eight conformation thatenhances electronic interactions between the thioand seleno linkers compared to the planar 26πplanar hexaphyrins. Their exceptionally large GMvalues (80–90,000 GM) will lead to their applica-tions as organic NLO supramolecular materialswhose properties are tunable by rational molecularperturbation.

Polar growth of the non-centrosymmetric poly-morph of (4-pyrrolidinopyridyl)bis (acetylaceto-nato) zinc(II) (ZNPPA) in space group Fdd2 isfavored when the nucleation is done on chiral inor-ganic surfaces of KDP, KBrO3 and NaBrO3 to theextent of 27%, 43% and 59%, respectively [53]. Onthe other hand, there is no preference for chiralcrystals compared to solution crystallization upontemplating on K2SO4 and Ba(BrO3)2 hydrate sur-faces. Among NaBrO3 crystals of cubic and tetra-hedral morphology, the tetrahedral morphologygave higher preferential growth. Most likely, theinorganic template supports heterogeneous nuclea-tion with epitaxial control and oriented growth ofZNPPA crystals.

In order to develop a new family of organo-gelators, molecular salts of a series of cinnamicacids and n-alkyl primary amines were prepared,their X-ray crystal structures analyzed, and theirgelation behavior studied [54]. 4-halo-cinnamatesalts are gelators and furthermore the chain lengthof the primary amine has a profound effect on

the gelation of 4-Br derivative. The non-gelatorsalts belonged to space group P 1 whereas thegelators are in P21/c crystal setting. All saltstructures display an invariant 1D hydrogen bondchain of ammonium· · · carboxylate bonding. Withchain length variation, n = 3–6 are non-gelators,n = 7–15 are gelators and among the latter setn = 11–15 are better in their gelation property.The gelation behavior was measured in petrol,kerosene and diesel as test liquids. Alkyl-alkylinteractions in the longer chain gelator salts areresponsible for the specific property in this familyof salts. However, the 1D fiber in xerogel is differ-ent from the single crystal structures as indicatedby XRPD patterns.

Fatty acid amide, n-lauroyl-L-alanine, is aneffective gelator for both aliphatic and aromatichydrocarbon solvents [55]. Its gelation efficiencyincreased with the addition of Me groups in theseries of solvents benzene, toluene, p-xylene andmesitylene. The supramolecular association in thegel is COOH dimer and hydrogen bonding alongthe NHC ==O group such that the alkyl chainsadopt a bilayer assembly.

The bile acid template was used for a varietyof structural investigations and materials appli-cations such as in molecular recognition, ionreceptors/sensors, low molecular mass organo andhydrogelators and gel–nanoparticle composites.Bile acids having π-donor pyrene rings gelledorganic solvents in the presence of a π-acceptorfluorenone as a 1:1 composite (figure 12) [56].No gelation was observed with either componentarguing for a donor–acceptor interaction mediatedorganogelator. Helical supramolecular architec-tures through hydrogen bonding and π-stacking


Figure 12. Tris cholamide derived hydrogel is colorless but turns blue in long-wave UV light. This luminescent hydroge-lator immobilizes water molecules at extremely low 0.15 mM concentration. Reproduced with permission of Royal Societyof Chemistry.

interactions are being studied towards makingchiral organogelators. A tripodal derivative func-tions as a hydrogelator at extremely low concen-tration of 0.02% w/v, with one gelator effectivelyimmobilizing >105 water molecules.

Transition metal compounds with a Kagomelattice are important because of their magneticproperties. Different metal atoms such as Fe,V, Co, etc. have been studied to give ferro-magnetic, antiferromagnetic and ferrimagneticinteractions. Following theoretical models thatshowed ferro/ferrimagnetic interactions for inte-gral spins and spin frustration for half-integerspins, a Ni+2(S = 1) Kagome compound (figure 13)having the formula [C6N2H8][NH4]2[Ni3F6(SO4)2]was prepared with 1,4-diazacubane ligand [57].Anionic layers of vertex-sharing NiIIF4O2 octa-hedra and SO4 tetrahedra fused together byNi ––– F––– Ni and Ni ––– O ––– S bonds. The high-temperature inverse susceptibility data give aWeiss temperature of 60 K and an effective mag-netic moment per nickel atom of 3.02μB, whichis comparable to the value in V+3 jarosites(3.02–3.16μB). The susceptibility (χ) decreasesdue to antiferromagnetic superexchange couplingbetween spin moments (S) above 15 K. Thecoupling is primarily angle-dependent (canted)antiferromagnetic (AFM) induced below 15 K.At even lower temperatures (<10K), χ decreasesdue to weak AFM coupling between the layers. Theoccurrence of such diverse interactions of differentmagnitudes and signs is due to interplay betweenthe frustrated Kagome geometry and the integerspins of Ni2+ ions.

The focus so far was on crystalline solids.Compounds that do not crystallize, or amorphousmaterials, are important in several applicationssuch as soluble drug forms and in optoelectro-nics. Molecules with rigid structural features finda difficulty in crystallization which at the firstlevel may be reconciled by having multiple mole-cules in the asymmetric unit [22]. When this

Figure 13. (a) Asymmetric unit and hexagonal Kagomelayer in [C6N2H8][NH4]2[Ni3F6(SO4)2]. (b) Temperaturedependence of magnetic susceptibility at 100 Oe under field–cooled (FC) and zero-field-cooled (ZFC) conditions and tem-perature variation of inverse susceptibility at 1000 Oe weremeasured (insert). The material shows magnetic hysteresisat 5K (insert). Reproduced with permission of AmericanChemical Society.

too is difficult, e.g., as in large tetrahedral,spiro fused, star burst and dendrite structures,then the result is amorphous materials. Tetraaryl

48 ASHWINI NANGIA

Figure 14. (a) Twisted bimesityls of D2d symmetrysynthesized by Sonogashira coupling as amorphous OLEDmaterials. (b) Electroluminescence spectra of the best dis-ubstituted bimesityl compound recorded for three types ofmultilayer devices, A, B and C are blue-green to blue emit-ters. Reproduced with permission of American ChemicalSociety.

bimesityls represent such a class of hinderedtetrahedral molecules [58] for molecular engi-neering as amorphous OLED materials (organiclight emitting diodes). The UV-Vis absorption ofthese compounds differs significantly dependingon the anthracene substituent groups. Moreover,the molar extinction coefficients (ε) of 4-foldfunctionalized derivatives are almost twice that of2-fold derivatives (57 200 vs. 123 820M−1 cm−1).Photoluminescence spectroscopy studies in dilutesolutions (μM conc.) showed that the emissionmaxima occur in the blue (430 nm), blue-green(450 nm) and green (485 nm) regions, which werered shifted by 20–50 nm in thin films. Thermalproperties indicate good stability (decompositiontemperature Td 420–520◦C) and high glass tran-sition temperatures (Tg 200–300◦C). Three typesof multilayer devices A, B and C were fabricated.

Electroluminescence spectra recorded for the bestmaterial showed blue and blue-green emissions(figure 14). Small-molecule-based OLEDs are con-sidered more valuable than polymer counter-parts because of difficulty in controlling thefilm thickness with polymers and the possibilitythat the first layer may dissolve during spin-coating of the second layer. Further, the diversity oforganic molecules with varying molecular weightsand wide-ranging properties is simply unmatchedby inorganic materials.

3. Conclusions

3.1 Icons and Holy Grail

The debate on Icons in Chemistry is not new.Many chemists argue that unlike mathematicsand physics, chemistry lacks popular statespersons.Tracing the growth of nanotechnology, Philip Ball[59] points out how the origin of this subject goes toRichard Feynman. His famous talk There’s Plentyof Room at the Bottom on the atom by atomassembly at the American Physical Society meet-ing in Caltech [60] kick-started the new disciplineof nanotechnology exactly fifty years ago (1959).He envisioned back then “that we could arrangeatoms one by one, just as we want them.” Orga-nized assembly of molecules or ions as we know ofin Lehn’s description of supramolecular chemistry[1,2] is just that. There is a similar chronology tocrystal engineering [6]. The chemical leanings ofcrystal engineering are surely traced to GerhardSchmidt [5] from his pioneering work on con-trolled photochemical reactions in the solid-state.The term was, however, first coined by RaymondPepinsky at the 1955 meeting of the AmericanPhysical Society in Mexico City [61]. “Crystal-lization of organic ions with metal-containingcomplex ions of suitable sizes, charges and solu-bilities results in structures with cells and sym-metries determined chiefly by packing of complexions. These cells and symmetries are to a goodextent controllable: hence crystals with advanta-geous properties can be ‘engineered’.” This defi-nition in effect encompasses the modern scope ofcrystal engineering as it is practised 50 years later.Even as this article celebrates the contributions ofchemists in the last decade covering modern topicsof supramolecular chemistry, crystal engineering,and nanoscience and nanotechnology, the blueprintof these disciplines were outlined by physicists inthe 1950s. Does this mean that chemists must learnto start thinking on a Grand Scale into the Future[62]? A related issue that in part answers whychemists often think and work on multiple prob-lems and not a single Holy Grail [63] is perhaps to


do with the very nature of the subject [64]. Accor-ding to George Whitesides, “I don’t think thereis a single thing that would turn all of chemistryon its ear, since one of chemistry’s strengths is itsdiversity [63].”

3.2 Looking forward

It should be clear to the reader that bothsupramolecular chemistry and crystal engineeringare interdisciplinary subjects that define a meetingpoint for organic, inorganic, physical, and compu-tational chemists along with biologists and mate-rials scientists. The last two decades were theheydays for seamless sciences. This author wastrained for his PhD in natural product synthesisand started thinking in the supramolecular direc-tion inspired by a review entitled ‘Organic synthe-sis – where now?’ that appeared in the early daysof his independent research (1990s) [65]. The twofacets discussed in this article developed in a timesequence. The nineties saw the rapid emergenceof supramolecular chemistry in many fascinatingmanifestations, described in the Perspectives inSupramolecular Chemistry series published from1994 to 2004 [66]. Thanks to the CCD X-raytechnology becoming commercial around 1995, thecurrent decade is witnessing a dramatic rise in thespread and popularity of crystal engineering as anindependent subject [67]. Given the unique abilityof chemists to create their own objects, i.e. mole-cules, it is certain that chemistry will continue toplay a central role in the sciences.

Acknowledgments

I thank several of my students and coworkerswhose names appear in the publications citedin this article. I thank present members of myresearch group in collating data and referencesand assistance with graphics during the writingof this article. The Department of Science andTechnology, Council of Scientific and IndustrialResearch, and University Grants Commission areprofusely thanked for providing financial assistanceand infrastructure facilities.

References

[1] Lehn J-M 1978 Cryptates: Inclusion complexes ofmacrocyclic receptor molecules; Pure Appl. Chem. 50871–892.

[2] Lehn J-M 1995 Supramolecular chemistry: Conceptsand perspectives (Weinheim: VCH).

[3] Steed J W and Atwood J L 2000 Supramolecularchemistry (Chichester: Wiley).

[4] Schalley C A (ed.) 2007 Analytical methods insupramolecular chemistry (Weinheim: Wiley-VCH).

[5] Schmidt G M J 1971 Photodimerization in the solidstate; Pure Appl. Chem. 27 647–678.

[6] Desiraju G R 1989 Crystal engineering: The design oforganic solids (Amsterdam: Elsevier).

[7] Desiraju G R and Steiner T 1999 The weak hydro-gen bond in structural chemistry and biology (Oxford:OUP).

[8] Banerjee R, Desiraju G R, Mondal R andHoward J A K 2004 Organic chlorine as a hydrogen-bridge acceptor: Evidence for the existence ofintramolecular O––– H · · ·Cl ––– C interactions in somegem-alkynols; Chem. Eur. J. 10 3373–3383.

[9] Balamurugan V, Hundal M S and Mukherjee R 2004First systematic investigation of C ––– H · · ·Cl hydro-gen bonding using inorganic supramolecular synthons:Lamellar, stitched stair-case, linked-ladder, and helicalstructures; Chem. Eur. J. 10 1683–1690.

[10] Desiraju G R 2002 Hydrogen bridges in crystal engi-neering: Interactions without borders; Acc. Chem. Res.35 565–573.

[11] Munshi P and Guru Row T N 2005 Charge densitybased classification of intermolecular interactions inmolecular crystals; CrystEngComm 7 608–611.

[12] Ranganathan A, Kulkarni G U and Rao C N R 2003Understanding the hydrogen bond in terms of the loca-tion of the bond critical point and the geometry of thelone pairs; J. Phys. Chem. A 107 6073–6081.

[13] Vishweshwar P, Babu N J, Nangia A, Mason S A,Puschmann H, Mondal R and Howard J A K 2004Variable temperature neutron diffraction analysis ofa very short O ––– H · · ·O hydrogen bond in 2,3,5,6-pyrazinetetracarboxylic acid dihydrate: Synthon-assisted short Oacid ––– H · · ·Owater hydrogen bondsin a multicenter array; J. Phys. Chem. A 1089406–9416.

[14] Desiraju G R 1995 Supramolecular synthons in crystalengineering – a new organic synthesis; Angew. Chem.,Int. Ed. Engl. 34 2311–2327.

[15] (a) Zerkowski J A, Mathias J P and Whitesides G M1994 New varieties of crystalline architecture pro-duced by small changes in molecular structure intape complexes of melamines and barbiturates; J. Am.Chem. Soc. 116 4304–4315. (b) Palacin S, Chin D N,Simanek E E, MacDonald J C, Whitesides G M,McBride M T and Palmore T R 1997 Hydrogen-bondedtapes on symmetrically substituted diketopiperazines.A robust structural motif for the engineering of mole-cular solids; J. Am. Chem. Soc. 119 11807–11816.

[16] (a) Vangala V R, Bhogala B R, Dey A, Desiraju G R,Broder C K, Smith P S, Mondal R, Howard J A K andWilson C C 2003 Correspondence between molecularfunctionality and crystal structures: Supramolecularchemistry of a family of homologated aminophenols;J. Am. Chem. Soc. 125 14495–14509. (b) Dey A,Kirchner M T, Vangala V R, Desiraju G R, Mondal Rand Howard J A K 2005 Crystal structure predictionof aminols: Advantages of a supramolecular synthonapproach with experimental structures; J. Am. Chem.Soc. 127 10545–10559.

[17] Moorthy J N, Natarajan R, Mal P and Venugopalan P2002 Helical self-assembly of substituted benzoic acids:Influence of weaker X · · ·X and C ––– H · · ·X interac-tions; J. Am. Chem. Soc. 124 6530–6531.

[18] (a) George S, Nangia A, Lam C-K, Mak T C W andNicoud J-F 2004 Crystal engineering of urea α-networkvia I · · ·O2N synthon and design of SHG active

50 ASHWINI NANGIA

crystal N-4-iodophenyl-N′-4′-nitrophenylurea; Chem.Commun. 1202–1203. (b) Reddy L S, Chandran S K,George S, Babu N J and Nangia A 2007 Crystalstructures of N-Aryl-N′-4-nitrophenyl ureas: Mole-cular conformation and weak interactions direct thestrong hydrogen bond synthon; Cryst. Growth Des. 72675–2690.

[19] (a) Sarkar M and Biradha M 2006 Amide-to-amide hydrogen bonds in the presence ofa pyridine functionality: Crystal structures ofbis(pyridinecarboxamido)alkanes; Cryst. Growth Des.6 202–208. (b) Rajput L, Singha S and Biradha K2007 Comparative structural studies on homologuesof amides and reverse amides: Unprecedented 4-foldinterpenetrated quartz network, new β-sheet, and two-dimensional layers; Cryst. Growth Des. 7 2788–2795.

[20] Jeffrey G A and Saenger W 1991 Hydrogen bonding inbiological structures (Berlin: Springer-Verlag).

[21] Mehta G, Sen S and Ramesh S S 2007 Crystal struc-tures of conformationally locked cyclitols: An analysisof hydrogen-bonded architectures and their impli-cations in crystal engineering; Eur. J. Org. Chem.423–436.

[22] (a) Lehmler H-J, Robertson L W, Parkin S andBrock C P 2002 Z′ = 4 structure without obviouspseudosymmetry: Implications for the formation ofsolid-state compounds; Acta Crystallogr. Sect. B 58140–147. (b) Brock C P 2002 Crystal packing invicinal diols CnHm(OH)2; Acta Crysallogrt. Sect. B 581025–1031.

[23] Reddy P A N, Nethaji M and Chakravarty A R2003 Supramolecular self assembly of a monomericcopper(II) complex of N-salicylidene-2-methoxyanilineforming three-dimensional closed and open channels;Inorg. Chem. Comm. 6 698–701.

[24] Sarma R J and Baruah J B 2007 Supramolecularand host-guest chemistry of bis-phenol and analogues;Cryst. Growth Des. 7 989–1000.

[25] Aitipamula S and Nangia A 2005 Concomitant poly-morphs of 2,2′,6,6′-tetramethyl-4,4′-terphenyldiol: Theβ-quinol network reproduced in a metastable poly-morph; Chem. Commun. 3159–3161.

[26] Supriya S and Das S K 2007 Reversible single crystal tosingle crystal transformation through Fe-O(H)Me/Fe-OH2 bond formation/bond breaking in a gas-solid reac-tion at an ambient condition; J. Am. Chem. Soc. 1293464–3465.

[27] Santra R and Biradha K 2008 Stepwise dimeriza-tion of double [2 + 2] reaction in the co-crystals of1,5-bis(4-pyridyl)-1,4-pentadiene-3-one and phloroglu-cinol: A single-crystal to single-crystal transformation;CrystEngComm 10 1524–1526.

[28] Sarkar M and Biradha K 2007 Crystal engineering ofmetal-organic frameworks containing amide functional-ities: Studies on network recognition, transformations,and exchange dynamics of guests and anions; Cryst.Growth Des. 7 1318–1331.

[29] Chandrasekhar V, Gopal K, Nagendran S, Steiner Aand Zacchini S 2006 Influence of aromatic substituentson the supramolecular architectures of monoorganoox-otin drums; Cryst. Growth Des. 6 267–273.

[30] (a) Bhogala B R, Basavoju S and Nangia A 2005 Three-component carboxylic acid–bipyridine lattice inclusionhost: Supramolecular synthesis of ternary cocrystals;Cryst. Growth Des. 5 1683–1686. (b) Bhogala B Rand Nangia A 2008 Ternary and quaternary cocrystalsof 1,3cis,5cis-cyclohexanetricarboxylic acid and 4,4′-bipyridines; New J. Chem. 32 800–807.

[31] Arora K K and Pedireddi V R 2003 A rational studyof crystal engineering of supramolecular assemblies of1,2,4,5-benzenetetracarboxylic acid; J. Org. Chem. 689177–9185.

[32] Jacques J, Collet A and Wilen S H 1981 Enan-tiomers, racemates and resolution (New York: Wiley-Interscience).

[33] Brock C P and Duncan L L 1994 Anomalous space-group frequencies for monoalcohols CnHmOH; Chem.Mater. 6 1307–1312.

[34] Balamurugan V and Mukherjee R 2005 Homochiral1D-helical metal–organic frameworks from achiral com-ponents: Formation of chiral channel via C ––– H · · ·Clinteraction; CrystEngComm 7 337–341.

[35] (a) Mandal S and Natarajan S 2002 Hydrothermalsynthesis and structures of three-dimensional zincphosphates built-up from two-dimensional layers andone-dimensional chains and ladders; Cryst. GrowthDes. 2 665–673. (b) Natarajan S, Mandal S, Mahata P,Rao V K, Ramaswamy P, Banerjee A, Paul A K andRamya K V 2006 The use of hydrothermal methodsin the synthesis of novel open-framework materials;J. Chem. Sci. 118 525–526.

[36] Rao C N R, Natarajan S, Choudhary A, Neeraj Sand Ayi A A 2001 Aufbau principle of complex open-framework structures of metal phosphates with differ-ent dimensionalities; Acc. Chem. Res. 34 80–87.

[37] The News Staff 2004 Splish, splash; Science 3062013–2017.

[38] Nangia A 2007 Water clusters in crystal hydrates;In: Encyclopaedia of supramolecular chemistry (eds)Atwood J L and Steed J W (New York: Marcel Dekker)1 1–9. www.informaworld.com/10.1081/E-ESMC-120023838.

[39] Mascal M, Infantes L and Chisholm J 2006 Wateroligomers in crystal hydrates: What’s news and whatisn’t? Angew. Chem. Int. Ed. 45 32–46.

[40] (a) Jovanovski G and Kamenar B 1982 Two ionicsaccharinates: (1a) sodium saccharinate 2/3 hydrateC7H4NO3SNa.2/3H2O, (1b) magnesium disacchari-nate heptahydrate, (C7H4NO3S)2Mg.7H2O; Cryst.Struct. Commun. 11 247–255. (b) Naumov P,Jovanovski G, Abbrent S and Tergenius L-E 2000Thermal behavior of the saccharinates of K+, Na+,Rb+, Cs+ and NH4+: Structural inferences; Ther-mochim. Acta 359 123–130.

[41] Banerjee R, Bhatt P M, Kirchner M T andDesiraju G R 2005 Structural studies of the systemNa(saccharinate)·nH2O: A model for crystallization;Angew. Chem. Int. Ed. 44 2515–2520.

[42] Naumov P, Jovanovski G, Grupce O, Kaitner B,Rae A D and Ng S W 2005 Solid-state structureand temperature/evacuation-induced dehydration ofsodium saccharinate 1.875 hydrate; Angew. Chem. Int.Ed. 44 1251–1254.

[43] Roy S, Banerjee R, Nangia A and Kruger G J2006 Conformational, concomitant polymorphs of4,4-diphenyl-2,5-cyclohexadienone: Conformation andlattice energy compensation in kinetic and thermody-namic forms; Chem. Eur. J. 12 3777–3788.

[44] Choudhury A R, Islam K, Kirchner M T, Mehta Gand Guru Row T N 2004 In situ cryocrystallizationof diphenyl ether: C ––– H · · ·π mediated polymorphicforms; J. Am. Chem. Soc. 126 12274–12275.

[45] Vishweshwar P, McMahon J A, Oliveira M,Peterson M L and Zaworotko M J 2005 The pre-dictably elusive form II of aspirin; J. Am. Chem. Soc.127 16802–16803.


[46] Bond A D, Boese R and Desiraju G R 2007 On thepolymorphism of aspirin: Crystalline aspirin as inter-growths of two polymorphic domains; Angew. Chem.Int. Ed. 46 618–622.

[47] Cambridge Crystallographic Data Center,www.ccdc.cam.ac.uk. The current release of the CSD(Ver. 5.30, November 2008, ConQuest 1.11) containsover 450 000 entries.

[48] (a) Almarsson O and Zaworotko M J 2004 Crystal engi-neering of the composition of pharmaceutical phases:Do pharmaceutical co-crystals represent a new pathto improved medicines? Chem. Commun. 1889–1896.(b) Aakeroy C B and Salmon D J 2005 Building co-crystals with molecular sense and supramolecular sen-sibility; CrystEngComm 7 439–448. (c) Trask A Vand Jones W 2005 Crystal engineering of organicco-crystals by the solid state grinding approach;Top. Curr. Chem. 254 41–70. (d) Meanwell N A2008 The emerging utility of co-crystals in drugdiscovery and development; Ann. Rep. Med. Chem. 43373–404.

[49] (a) Perumalla S R, Suresh E and Pedireddi V R2005 Nucleobases in molecular recognition: Molecularadducts of adenine and cytosine with COOH func-tional groups; Angew. Chem. Int. Ed. 44 7753–7757.(b) Thomas R and Kulkarni G U 2008 Hydrogenbonding in proton-transfer complexes of cytosine withtrimesic and pyromellitic acids; J. Mol. Struct. 873160–167.

[50] (a) Sarma B, Nath N K, Bhogala B R and Nangia A2009 Synthon competition and cooperation in molecu-lar salts of hydroxybenzoic acids and aminopyridines;Cryst. Growth Des. 9 1546–1557. (b) Shattock T R,Arora K K, Vishweshwar P and Zaworotko M J2008 Hierarchy of supramolecular synthons. Persis-tent carboxylic acid· · ·pyridine in cocrystals that alsocontain a hydroxyl moiety; Cryst. Growth Des. 84533–4545.

[51] Babu N J, Reddy L S and Nangia A 2007 Amide–N-oxide heterosynthon and amide dimer homosynthonin cocrystals of carboxamide drugs and pyridine-N-oxides; Mol. Pharmaceutics 4 417–434.

[52] (a) Rath H, Sankar J, PrabhuRaja V,Chandrashekar T K, Nag A and Goswami D 2005Core-modified expanded porphyrins with large third-order nonlinear optical response; J. Am. Chem.Soc. 127 11608–11609. (b) Gokulnath S and Chan-drashekar T K 2008 Expanded porphyrins as thirdorder non-linear toptical materials: Some structure–function correlations; J. Chem. Sci. 120 137–142.

[53] Prakash M J, Raghavaiah P, Krishna Y S R andRadhakrishnan T P 2008 Growing molecular crystalson inorganic crystals: Polar structure induced by non-centrosymmetric templates; Angew. Chem. Int. Ed. 473969–3972.

[54] Ballabh A, Trivedi D R and Dastidar P 2006 New seriesof organogelators derived from a combinatorial libraryof primary ammonium monocarboxylate salts; Chem.Mater. 18 3795–3800.

[55] Bhattacharya S and Pal A 2008 Physical gela-tion of binary mixtures of hydrocarbons mediatedby n-lauroyl-L-alanine and characterization of theirthermal and mechanical properties; J. Phys. Chem. B112 4918–4927.

[56] (a) Nonappa and Maitra U 2008 Unlocking thepotential of bile acids in synthesis, supramolecu-lar/materials chemistry and nanoscience; Org. Biomol.Chem. 6 657–669. (b) Mukhopadhyay S, Maitra I,Ira, Krishnamoorthy G, Schmidt J and Talmon Y2004 Structure and dynamics of a molecular hydrogelderived from a tripodal cholamide; J. Am. Chem. Soc.126 15905–15914.

[57] Behera J N and Rao C N R 2006 A Ni2+(S = 1)Kagome compound templated by 1,4-diazacubane;J. Am. Chem. Soc. 128 9334–9335.

[58] (a) Moorthy J N, Venkatakrishnan P, Natarajan P,Huang D-F and Chow T J 2008 De novo design forfunctional amorphous materials: Synthesis and thermaland light-emitting properties of twisted anthracene-functionalized bimesitylenes; J. Am. Chem. Soc. 13017320–17333. (b) Moorthy J N, Venkatakrishanan P,Huang D-F and Chow T J 2008 Blue light-emittingand hole-transporting amorphous materials basedon diarylaminobiphenyl-functionalized bimesitylenes;Chem. Commun. 2146–2148.

[59] Ball P 2009 Fenynman’s fancy; Chem. World 58–62.

[60] Feynman R 1959 There’s plenty of room at thebottom (Caltech: American Physical Society)http://www.its.caltech.edu/∼feynman/plenty.html.

[61] Pepinsky R 1955 Crystal engineering: A new conceptin crystallography; Phys. Rev. 100 971.

[62] These thoughts came up at an informal discussion ofthe School of Chemistry faculty with Prof. GoverdhanMehta, Indian Institute of Science, Bangalore duringhis visit to University of Hyderabad in January 2009.

[63] Halford B 2009 A crusade against the holy grails;Chem. Engg. News March 30, 34.

[64] Desiraju G R 2005 Chemistry – the middle kingdom;Curr. Sci. 88 374–380.

[65] Seebach D 199 Organic synthesis – where now? Angew.Chem. Int. Ed. Engl. 29 1320–1367.

[66] Lehn J-M (Founding ed.) 1994–2004 Perspectives insupramolecular chemistry, Vols. 1–8 (Chichester: JohnWiley)

[67] (a) Tiekink E R T and Vittal J J 2006 Frontiers in crys-tal engineering (Chichester: John Wiley). (b) A secondvolume is currently in publication and a third volume isbeing planned for 2010. Private communication, Prof.Edward Tiekink, University of Texas at San Antonio.

supramolecular chemistry and crystal engineering...supramolecular chemistry and crystal engineering...

Documents