Engineering crystals of dendritic molecules

Oleg Lukina,b,1, Dirk Schuberta, Claudia M. Mullera, W. Bernd Schweizerc, Volker Gramlicha, Julian Schneidera, Grygoriy Dolgonosd, and Alexander Shivanyukb
of Polymers, Department of Materials, HCI G527, and cOrganic Chemistry Laboratory, HCI G301, Eidgenossische Technische Hochschule, 8093 Zurich, Switzerland; bNational Taras Shevchenko University, Volodymyrska Street 64, Kiev 01033, Ukraine; and dBremen Center for Computational Materials Science (BCCMS), Bremen University, Am Fallturm 1, 28359 Bremen, Germany Communicated by Julius Rebek, Jr., The Scripps Research Institute, La Jolla, CA, April 22, 2009 (received for review January 20, 2009)
aInstitute

A detailed single-crystal X-ray study of conformationally exible sulfonimide-based dendritic molecules with systematically varied molecular architectures was undertaken. Thirteen crystal structures reported in this work include 9 structures of the secondgeneration dendritic sulfonimides decorated with different aryl groups, 2 compounds bearing branches of both second and rst generation, and 2 representatives of the rst generation. Analysis of the packing patterns of 9 compounds bearing second-generation branches shows that despite their lack of strong directive functional groups there is a repeatedly reproduced intermolecular interaction mode consisting in an anchor-type packing of complementary second-generation branches of neighbouring molecules. The observed interaction tolerates a wide range of substituents in meta- and para-positions of the peripheral arylsulfonyl rings. Quantum chemical calculations of the molecule-molecule interaction energies agree at the qualitative level with the packing preferences found in the crystalline state. The calculations can therefore be used as a tool to rationalize and predict molecular structures with commensurate and non-commensurate branches for programming of different packing modes in crystal.
dendrimers single-crystal X-ray sulfonimides supramolecular chemistry

present study was stimulated by both basic and applicative reasons. Indeed, the majority of dendritic molecules have low or no tendency to form crystals. There are only a few papers reporting single-crystal X-ray analysis of some dendritic structures of first (1528) and second (2934) generations. Although these results are highly interesting, there is still the lack of a premeditated structural study of dendritic molecules, which might shed some light on the factors controlling self-assembly of these flexible architectures. Dendritic molecules capable of self-assembly have potential applications as liquid crystalline and electronic materials (3540). It is therefore important to be able to deliberately construct covalent and supramolecular architectures involving dendritic species. Recently, we demonstrated that sulfonimide-based dendrimers (4143) can be precisely tailored with respect to their structural details enabling programmed analysis of the influence of small structural variations on a number of their physicochemical properties. The sulfonimidebased dendritic molecules are mainly crystalline solids with relatively high melting points, which prompted us to undertake a detailed study toward growing single crystals suitable for the X-ray analysis. Results and Discussion Branched sulfonimides, such as 1-6 depicted in Scheme 1, bear multiple aromatic rings held together by sulfonimide branching points. The synthetic methodology affording this type of structures has already been reported by us (41) and synthetic procedures for the new compounds are collected in the SI. Crystals suitable for single-crystal X-ray analyses were grown from methanol (compounds 1 and 2, 2 to 3 days, slow evaporation), dichloromethane/methanol (compounds 3, 5a, 5b, 5d, 5e, 5g and 6a, 2 weeks, slow evaporation; compound 4, 5 days, vapor diffusion), dichloromethane/methanol/acetonitrile (compounds 5c and 5f, 3 to 4 days, slow evaporation), and chloroform (compound 6b, 4 weeks, slow evaporation). Following the order of increase of molecular complexity, Figs. 16 depict the singlecrystal X-ray structures of compounds 16 focusing on their packing in crystal. Parallel -stacking involving 2-naphthyl groups with interplanar distances of 3.5 is present in crystal structures of simple sulfonimides 1 and 2 (Fig. 1). This type of intermolecular interaction is often observed in crystal structures of different aromatic compounds (1) and in some cases is regarded as a directional interaction, e.g., in crystal structures involving combination of aromatic and perfluoroaromatic rings (44) and other combinations of electron-rich and electron-deficient aromatic compounds (45). The directional role of the parallel -stacking
Author contributions: O.L. designed research; O.L., D.S., C.M.M., W.B.S., V.G., J.S., G.D., and A.S. performed research; O.L. analyzed data; and O.L. wrote the paper. The authors declare no conict of interest. Data deposition: The atomic coordinates have been deposited in the Cambridge Structural Database, Cambridge Crystallographic Data Centre, Cambridge CB2 1EZ, United Kingdom (CSD reference nos. 641741 641747 and 685214 685219).
1To

evealing the interplay of noncovalent forces that direct processes of self-assembly and self-organization has been objective of numerous studies (14). This is central for the design and then the targeted synthesis of molecules capable of assembling into predefined supramolecular structures of practical significance. Although there are many examples in which the knowledge about the noncovalent interactions helped to arrive at predetermined complex molecular architectures (5, 6), the serendipity contributes to a large extent to supramolecular design. Once a new fascinating supramolecular structure is discovered, it is often difficult to decode the underlying principles of its assembly. In this context the crystallization of organic molecules is a specifically complicated case that remains poorly understood. Although it is now often possible with the aid of computations to predict a most favorable intermolecular interaction between molecules there is no guarantee that this particular intermolecular contact will be found in the crystal structure. Attempts at the rational design of organic crystals evolved into a nowadays well-established field of crystal engineering (711) that aims at general rules for crystal structure control. For the time being the field of crystal engineering has generated some knowledge that often helps to design organic crystals with desired properties. The main recipe is to use so-called tectons (12), which are small molecules with welldefined shape and strong directive functional groups, such as hydrogen-bonding units and metal-coordinating sites (13, 14). In case reproducible trends in crystal packing for a given group of compounds are observed the molecules are usually synthetically modified in a systematic way to influence the crystal structure. In this article, we report the first systematic structural study of flexible branched molecules lacking directive functionalities. This study relies on the single-crystal X-ray analyses of 13 structures of increasing branched complexity and is accompanied by theoretical calculations. The selection of objects for the
PNAS July 7, 2009 vol. 106 no. 27

whom correspondence should be addressed. E-mail: o.lukin@ukrorgsynth.com.

This article contains supporting information online at www.pnas.org/cgi/content/full/ 0904264106/DCSupplemental.

1092210927

www.pnas.org cgi doi 10.1073 pnas.0904264106

Fig. 1.

Fragments of packing of 1 (A) and 2 (B) in crystal.

overlap of the complementary branches (Fig. 3) in crystals of 3 and 4 reveals that arylsulfonyl and p-sulfonimidobenzenesulfonyl rings in both structures are not parallel, which would be a prerequisite for an effective - -stacking. Instead, the rings are considerably tilted (the angle between the aromatic rings in 4 is 40) that makes the intermolecular contact very compact and secures the efficient space filling. Symmetrical 5a5e bearing small substituent in para or meta positions of the peripheral arylsulfonyl rings are representatives of the next branching event. Figs. 4 and 5 show the interactions of molecular pairs in crystals of compounds 5a5c and 5d,e, respectively. Like in case of compounds 3 and 4, the most frequently observed molecule-molecule contacts in crystal structures of 5a5c and 5d,e involve overlapping of their complementary branches. Notably, the molecular structure of 5e (Fig. 5 Middle) has the unusual conformation of the central sulfonimide branching point differing significantly from those found in all other dendritic sulfonimides. This noticeable conformational change at the central sulfonimide unit in 5e reflects an ease of geometric adjustability of branched oligosulfonimides to the
CHEMISTRY

Scheme 1.

It depicts structural formulae of compounds 16.

in the crystal formation in 1 and 2 is questionable because, as will be shown below, the -stacking is not observed in larger dendritic sulfonimides. It is rather reasonable to assume that the parallel arrangements of the aromatic rings in crystals of 1 and 2 favor more compact packing and, consequently, the gain in cohesion energy. Compounds 3 and 4 are the simplest expanded branched analogues of 1 and 2, respectively. As shown in Fig. 2, the packing of molecules of 3 and 4 in crystal involves the intermolecular anchor-type overlap of the complementary branches (shown in the dotted circles in Fig. 2 and schematically illustrated in Fig. 3). Additionally, like in the case of compounds 1 and 2, the parallel -stacking involving naphthalene rings is also observed in crystals of 3 and 4. A detailed inspection of the intermolecular
Lukin et al.

Fig. 2. Fragment of packing of 3. (Upper) and 4 (Lower) in crystal. Dotted circles denote the characteristic intermolecular contacts.

PNAS

July 7, 2009

vol. 106

no. 27

10923

Fig. 3. Schematic representation of the intermolecular interaction between the complementary branches. This is reproducibly observed in packing of the second generation sulfonimide dendrimers in crystals. Packing of molecular pairs in crystal of compounds 5a (Top), 5b (Middle), and 5c (Bottom).

crystal packing effects. Noteworthy, the unusual molecular conformation of 5e does not influence the characteristic intermolecular contact between the complementary branches (Fig. 5 Bottom), as in the case of 3, 4, and 5a5d. Fig. 6 Top and Middle show packing of molecular pairs of second-generation unsymmetrical dendritic sulfonimides, socalled Janus-dendrimers (46, 47) 6a and 6b in crystal. Similarly to the above discussed symmetrical second-generation dendritic sulfonimides the intermolecular interaction consists of the overlapped second-generation branches. An interesting difference in the crystal packing of 6a and 6b is that the unlike branches of the former and the like ones of the latter are involved in their intermolecular contacts. The intermolecular contacts involved in packing of 6b in crystal are only the interactions of the complementary branches. As shown in Fig. 6 Bottom, this leads to the formation of channel-like structure with channels occupied by solvent molecules. Considerable sizes and matching shapes of the secondgeneration branches participating in the molecular packing of 5a5e and 6a,b seemingly make the main contribution to the overall cohesion energy in their crystals. This type of intermolecular interaction, schematically depicted in Fig. 3, reproducibly observed in the 9 above mentioned crystal structures does not depend on the solvents that are either used for crystallization or present in the crystals. It therefore depends on the primary structure of the dendrimer. Although the rationalization of crystal structures on the basis of intermolecular interactions in molecular pairs is a rather dangerous venture, the repeated occurrence of the packing mode depicted in Fig. 3 in the 9 crystal structures prompted us to carry out a theoretical analysis of this type of interaction. In view of the recent comprehensive report by Dunitz and Gavezzotti (48) stressing the nonlocalized nature of intermolecular interactions, the following discussion of structural effects on the packing patterns in crystal structures of compounds 16 lacking strong directive sites relies on theoretically assessed moleculemolecule interaction energies rather than contacts between point atoms. The X-ray structural geometries were used as starting points in the full-energy minimization with a densityfunctional-derived self-consistent-charge density-functional tight-binding with dispersion term (SCC-DFTB-D) method (49, 50). The interaction energies were obtained by subtracting the sum of total energies of monomers from the total energy of the corresponding bimolecular complex. The calculations revealed that the interaction energy is sensitive to the nature of the overlapping branches. For instance, interaction energy calculated for the homodimeric complex of the interacting pair of 4-nitrophenyl-decorated branches, as in 4, is lower than that of the homodimeric complex of 2-naphthyl-decorated branches, as in 3, 5e, and 6b ( 27.3 vs. 23.8 kcal/mol, respectively). The heterodimeric complex involving both 2-naphthyl- and 4-nitrophenyl-decorated branches, as in 6a, is the most stable in this triad ( 28.0 kcal/mol). This agrees with the experimentally observed interaction of the unlike branches in the crystal
10924 www.pnas.org cgi doi 10.1073 pnas.0904264106

Fig. 4. Packing of molecular pairs in crystal of compounds 5a (Top), 5b (Middle), and 5c (Bottom).

structure of 6a. Interestingly, the calculated interaction energies of the intermolecular parallel -stacking involving 2-naphthyl groups in structures 1 and 2 are 15.3 and 17.4 kcal/mol, respectively. These intermolecular contacts are considerably less stabilized compared with the one involving anchored branches. A reasonable explanation for these remarkably different calculated stabilities is that the interaction energy between charge distributions (48) in a molecular complex lacking directive functional groups seems to depend on the van der Waals area involved into intermolecular contact. The area of the intermolecular overlap of the second-generation branches is apparently larger than that of the -stacked pairs involving naphthalene and p-nitrophenyl rings. Therefore, the interaction between complementary branches in crystal is favorable not only due to its compactness resulting in the efficient space-filling. It is also beneficial from the energetic standpoint. Consequently, on the basis of the consistent experimental and theoretical results we reasoned that introduction of secondLukin et al.

Fig. 5. Packing of 5d in crystal (Top), and X-ray molecular structure (Middle) and packing in crystal of 5e (Bottom). The dotted circle denotes the characteristic intermolecular interaction of the complementary branches (see also Fig. 3).

generation arylsulfonyl groups that are not supportive of the repeatedly observed interaction depicted in Fig. 3, should change the packing in crystal. For example, analysis of molecular models of compounds 5f and 5g bearing 4-bromobiphenylsulfonyl and mesitylsulfonyl peripheral groups, respectively, shows that the overlap of second-generation branches in these compounds should be highly unfavorable, if not impossible. It is expected that in compound 5f the 4-bromobiphenyl units would stick far out of the branch-to-branch contact under study and lead to open arrangements in the crystal. The resulting voids would, however, be too small to be filled by some solvent molecules to compensate for the cohesion energy loss. In case of compound 5g, the interaction under study is impossible merely because of its sterically loaded second-generation branches. According to the model calculations, the o-methyl groups of the peripheral mesitylsulfonyl units should lead to significant changes in the conformation of the corresponding sulfonimide branching point changing at least one Car-S-N-S torsion angle from its equilibrium value of 90 (41) to 150.
Lukin et al.

Fig. 6. Packing of the molecular pair of 6a in crystal (Top), packing of the molecular pair of 6b in crystal (Middle), and a fragment of the crystal packing of 6b viewed along the crystallographic c axis (Bottom). CHCl3 molecules residing in the channels are omitted for clarity.

In line with the expectations, crystallization and the subsequent single-crystal X-ray analysis of compounds 5f and 5g revealed that their packing in crystal does not involve the overlap of the complementary branches, as depicted in Fig. 3, repeatedly
PNAS July 7, 2009 vol. 106 no. 27 10925

CHEMISTRY

Fig. 7. X-ray molecular structure of 5f (Upper) and a fragment of the crystal packing of 5f (Bottom).

Fig. 8. Compound 5g: X-ray molecular structure (Upper Left), packing of the molecular pair (Upper Right), and a fragment of the crystal packing viewed along the crystallographic c axis (Lower). CH2Cl2 molecules residing in the channels are rendered space-lled.

observed in the 9 above discussed structures. Fig. 7 shows that although the X-ray molecular structure of the compound 5f is very similar to those of 5a5d, its molecules pack in crystal in parallel stacks, which are in turn complementary to each other. The resulting structure is dense: It does not contain voids, and no solvent is present in the crystal. The calculated interaction energy of the X-ray revealed molecule-molecule contact of 2 superposed 5f is 49.3 kcal/mol. This value of the interaction energy is significantly lower than the value of 22.0 kcal/mol calculated for the visionary branch-branch contact involving 2 molecules of 5f. Similarly to the above discussed case of the different calculated stabilities of the branch-branch interaction and the -stacking of naphthalene rings, this is seemingly due to a considerably increased van der Waals area of the intermolecular contact in 5f compared with that of the repeatedly observed branch-branch contact (Fig. 3). Summarizing this tendency, the structures lacking strong directive functionalities (e.g., hydrogen-bonding centers) tend to maximize the area of mutual overlap in molecular pairing. In accord with the model considerations, X-ray molecular structure of 5g (Fig. 8 Top Left) has 2 conformationally distorted sulfonimide branching points. This makes the commonly observed packing mode (Fig. 3) physically inadmissible. As illustrated in Fig. 8, the packing of 5g in crystal generates voids occupied by CH2Cl2 solvent molecules. Although the packing of 5f and 5g in crystal could not be predicted on the basis of both available crystal structures of other second-generation dendritic sulfonimides and theoretical calculations, one can reliably calculate the molecular structure and conclude on the feasibility of the packing modes.
10926 www.pnas.org cgi doi 10.1073 pnas.0904264106

Finally, the n-octyl chains present in all of the structures under study do not seem to play any critical role in determining the packing in crystal. The chains exhibited significant disorder in most of the presented X-ray structures. The intermolecular contacts involving the aliphatic chains and aromatic rings revealed negligible calculated stabilization energies ( 2.0 to 3.0 kcal/mol) compared with those of the above discussed intermolecular interactions. Conclusions In stark contrast to the vast majority of dendritic molecules that have low or no tendency to crystallize dendritic oligosulfonimides have the advantageous ability to form single crystals with no regard to the peripheral substitution. This allowed carrying out the unprecedented correlation of molecular and crystal structures of flexible dendritic species. The conclusions drawn in this contribution rest upon single-crystal X-ray analyses of 13 dendritic sulfonimides with systematically modified molecular architecture and accompanied by theoretical calculations. The compounds showed reproducible and tunable packing modes in the crystalline state. An unlimited structural diversity of dendritic oligosulfonimides (4143) makes them highly promising flexible building blocks for crystal engineering. Properly shaped and functionalized dendritic sulfonimides could possibly be used in topochemistry and the rational design of materials, such as conducting organic crystals (51, 52). Our approach to the diversity of dendrimer crystals can be equally applied to other types of branched compounds, providing their molecular structures can be tailored in a systematic way. We are now focusing
Lukin et al.

on crystallization of larger generation dendrimers and the growing of dendrimer cocrystals. Materials and Methods
The synthesis of the compounds discussed in this work was either reported in refs. 41 43 or carried out following the published methods. Detailed synthetic procedures and characterization data for new compounds are in

the SI. Single-crystal X-ray analyses were performed on an Xcalibur PX CCD diffractometer. The details of the crystallographic data and summary of data collection and renement for the dendritic sulfonimides are also given in the SI. ACKNOWLEDGMENTS. We thank Dr. J. van Heijst (ETH Zurich) for his critical remarks on the manuscript.

1. Muller-Dethlefs K, Hobza P (2000) Noncovalent interactions: A challenge for experi ment and theory. Chem Rev 100:143168. 2. Meyer EA, Castellano RK, Diederich F (2003) Interactions with aromatic rings in chemical and biological recognition. Angew Chem Int Ed 42:1210 1250. 3. Paulini R, Muller K, Diederich F (2005) Orthogonal multipolar interactions in structural chemistry and biology. Angew Chem Int Ed 44:1788 1805. 4. Dunitz JD, Schweizer WB (2006) Molecular pair analysis: C-H F interactions in the crystal structure of uorobenzene? And related matters. Chem Eur J 12:6804 6815. 5. Seeman NC (2003) DNA in a material world. Nature 421:427 431. 6. Rothemund WK (2006) Folding DNA to create nanoscale shapes and patterns. Nature 440:297302. 7. Schmidt GMJ (1971) Photodimerization in the solid state. Pure Appl Chem 27:647 678. 8. Desiraju GR (1989) Crystal Engineering: The Design of Organic Solids (Elsevier, Amsterdam). 9. Hollingsworth MD (2002) Crystal engineering: From structure to function. Science 295:2410 2413. 10. Braga D (2003) Crystal engineering, Where from? Where to? Chem Commun 2751 2754. 11. Zaworotko MJ (2007) Molecules to crystals, crystals to molecules. And back again? Cryst Growth Design 7:4 9. 12. Simard M, Su D, Wuest JD (1991) Use of hydrogen bonds to control molecular aggregation. Self-assembly of three-dimensional networks with large chambers. J Am Chem Soc 113:4696 4698. 13. Wuest JD (2005) Engineering crystals by the strategy of molecular tectonics. Chem Commun 5830 5837. 14. Hosseini MW (2005) Molecular tectonics: From simple tectons to complex molecular networks. Acc Chem Res 38:313323. 15. Mekelburger H-B, Vogtle F, Rissanen K (1993) Repetitive-synthesis of bulky dendrim ersA reversibly photoactive dendrimer with six azobenzene side chains. Chem Ber 126:11611169. 16. Lambert JB, Pug JL, Denari JM (1996) First-generation dendritic polysilanes. Organometallics 15:615 625. 17. Kriesel JW, et al. (1998) Synthesis of highly charged organometallic dendrimers and their characterization by electrospray mass spectrometry and single-crystal X-ray diffraction. J Am Chem Soc 120:1220712215. 18. Larre C, et al. (1998) Chemistry within megamolecules: Regiospecic functionalization after construction of phosphorus dendrimers. J Am Chem Soc 120:13070 13082. 19. Gossage RA, et al. (1999) A novel phenol for use in convergent and divergent dendrimer synthesis: Access to core functionalisable trifurcate carbosilane dendrimers The X-ray crystal structure of [1,3,5-tris{4-(triallylsilyl)phenylester}benzene]. Chem Eur J 5:21912197. 20. Nanjo M, Sunaga T, Sekiguchi A, Horn E (1999) Crystal structures of the rst generation of phenyl-substituted and permethyl-substituted dendritic polysilanes. Inorg Chem Commun 2:203206. 21. Ranganathan D, Kurur S, Gilardi R, Karle IL (2000) Design and synthesis of AB3-type (A 1,3,5-benzenetricarbonyl unit; B Glu diOME or Glu7 octa OMe) peptide dendrimers: Crystal structure of the rst generation. Biopolymers 54:289 295. 22. Brewis M, Clarkson GJ, Helliwell M, Holder AM, McKeown NB (2000) The synthesis and glass-forming properties of phthalocyanine-containing poly(aryl ether) dendrimers. Chem Eur J 6:4630 4636. 23. Bauer RE, Enkelmann V, Wiesler UM, Berresheim AJ, Mullen K (2002) Single-crystal structures of polyphenylene dendrimers. Chem Eur J 8:3858 3864. 24. Portner K, Nieger M, Vogtle F (2004) Dendritic EDA-schiff bases of the salen-type. Synlett 11671170. 25. Harder S, Meijboom R, Moss JR (2004) Selective lithiation and crystal structures of G1-carbosilane dendrimers with dimethoxybenzene functionalities. J Organomet Chem 689:10951101.

26. Saalfrank RW, et al. (2004) Synthesis, structure, and dynamics of six-membered metallacoronands and metallodendrimers of iron and indium. Chem Eur J 10:1899 1905. 27. Williams AA, et al. (2005) Long-chain alkyl dendrons form homologous thin lms on silver oxide surfaces. Chem Commun 50535055. 28. Yang H-B, et al. (2007) Coordination-driven self-assembly of metallodendrimers possessing well-dened and controllable cavities as cores. J Am Chem Soc 129:2120 2129. 29. Seyferth D, Son DY, Rheingold AL, Ostrander RL (1994) Synthesis of an organosilicon dendrimer containing 324 Si-H bonds. Organometallics 13:26822690. 30. Rajca A, Janicki S (1994) Synthesis of sterically hindered 1,3-connected polyarylmethanes. J Org Chem 59:7099 7107. 31. Sekiguchi A, Nanjo M, Kabuto C, Sakurai H (1995) Polysilane dendrimers. J Am Chem Soc 117:4195 4196. 32. Karakaya B, Claussen W, Gessler K, Saenger W, Schluter AD (1997) Toward dendrimers with cylindrical shape in solution. J Am Chem Soc 119:3296 3301. 33. Brewis M, et al. (1998) Silicon phthalocyanines with axial dendritic substituents. Angew Chem Int Ed 37:10921094. 34. Ma C-Q, Mena-Osteritz E, Wienk M, Janssen RAJ, Bauerle P (2007) Functionalized 3D oligothiophene dendrons and dendrimersNovel macromolecules for organic electronics. Angew Chem Int Ed 46:1679 1683. 35. Malenfant PRL, Jayaraman M, Frechet JMJ (1998) Well-dened triblock hybrid den drimers based on lengthy oligothiophene cores and poly(benzyl ether) dendrons. J Am Chem Soc 120:10990 10991. 36. Guillion D, Deschenaux R (2002) Liquid-crystalline dendrimers. Curr Opin Solid State Mat Sci 6:515525. 37. Percec V, et al. (2002) Self-organization of supramolecular helical dendrimers into complex electronic materials. Nature 419:384 387. 38. Campidelli S, et al. (2005) Supramolecular fullerene materials: Dendritic liquidcrystalline fulleropyrrolidines. Macromolecules 38:79157925. 39. Percec V (2006) Bioinspired supramolecular liquid crystals. Phil Trans R Soc A 364:2709 2719. 40. Kamikawa Y, Kato T (2006) One-dimensional chiral self-assembly of pyrene derivatives based on dendritic oligopeptides. Org Lett 8:24632466. 41. Lukin O, et al. (2006) Designer dendrimers: Branched oligosulfonimides with controllable molecular architectures. J Am Chem Soc 128:8964 8974. 42. Lukin O, Schubert D, Muller C, Corda M, Kandre R (2008) Persulfonylation of amines applied to the synthesis of higher generation dendrimers. J Org Chem 73:35623565. 43. Schubert D, et al. (2008) A topological view of isomeric dendrimers. Eur J Org Chem 4148 4156. 44. Williams JH, Cockcroft JK, Fitch AN (1992) Structure of the lowest temperature phase of the solid benzene-hexauorobenzene adduct. Angew Chem Int Ed 31:16551657. 45. Amabilino DB, Stoddart JF, Williams DJ (1994) From solid-state structures and superstructures to self-assembly processes. Chem Mater 6:1159 1167. 46. Percec V, et al. (2005) Self-assembly of semiuorinated Janus-dendritic benzamides into bilayered pyramidal columns. Angew Chem Int Ed 44:4739 4745. 47. Saez IM, Goodby JW (2003) Design and properties of Janus-like supermolecular liquid crystals. Chem Commun 1726 1727. 48. Dunitz JD, Gavezzotti A (2005) Molecular recognition in organic crystals: Directed intermolecular bonds or nonlocalized bonding? Angew Chem Int Ed 44:1766 1787. 49. Porezag D, Frauenheim T, Kohler T, Seifert G, Kaschner R (1995) Construction of tight-binding-like potentials on the basis of density-functional theory: Application to carbon. Phys Rev B 51:1294712957. 50. Elstner M, et al. (1998) Self-consistent-charge density-functional tight-binding method for simulations of complex materials properties. Phys Rev B 58:7260 7268. 51. Pope, M Swenberg CE (1999) Electronic Processes in Organic Crystals and Polymers (Oxford Univ Press, Oxford). 52. Reese C, Bao Z (2007) Organic single-crystal eld-effect transistors. Mat Today 3:20 27.

Lukin et al.

PNAS

July 7, 2009

vol. 106

no. 27

10927

CHEMISTRY