Abstract
The formation of complex superstructures via hydrogen bonding of two ditopic building blocks, diazadibenzoperylenes 1a,b and isophthalic acid 2, has been investigated. It was found that only the phenoxy-substituted diazadibenzoperylene 1a forms extended assemblies with 2 in complexes of a 1:1 stoichiometry, whereas for the 4-tert-butylphenoxy-substituted analogue 1b, no indications for superstructure formation with 2 were found. The different behavior is explained by the presence of additional π–π interactions, which are only observed for [1a⋅2], as revealed by concentration-dependent optical absorption and fluorescence spectroscopy. Based on variable temperature x-ray diffraction studies, a lamellar structure for [1a⋅2] is proposed that takes into account the concept of microphase-segregation.
The current demand in highly ordered solid-state π-conjugated materials for application in electronic devices, e.g., organic field-effect transistors and organic solar cells (1–3), constitutes a touchstone for supramolecular chemistry that stimulates several groups to synthesize receptor-functionalized chromophores and to control their solid-state organization by means of intermolecular forces (1–5). For most examples, this interaction relies on hydrogen-bonding, which is a directional–reversible interaction with adjustable binding strength (4, 5). Within the family of hydrogen-bonded systems, the association of carboxylic acids and pyridines has received special attention and includes work on defined molecular assemblies (6–9), supramolecular polymers and networks (4, 5, 10–14), and hydrogen-bonded liquid crystals (15–18). However, the low-binding constant between pyridines and carboxylic acids precludes high degrees of polymerization in solution between dicarboxylic acid and dipyridine building blocks, which, up to now, limited this supramolecular synthon to applications in solid-state materials. Nevertheless, as shown recently, the binding constants of hydrogen bonds increase considerably with decreasing competition of the solvent with the receptor groups, which is given, for example, in aliphatic solvents (5). If additional weaker interactions such as π–π interactions or microphase–segregation (19, 20) favorably contribute to the self-assembly process, then a hierarchical superstructure formation becomes likely in the solid state or even in dilute solution.
In the given contribution, highly fluorescent diazadibenzoperylene dyes 1a,b (21) were chosen as a ditopic hydrogen-bond acceptor, and their interaction with ditopic hydrogen-bond donating isophthalic acid 2 (22)§ was studied in solution and the solid state (Fig. 1). A prerequisite for the formation of extended hydrogen-bonded polymers is an exact 1:1 stoichiometry between the precursors 1a,b and 2. Special care had to be taken in the preparation of the complexes, because even the slightest deviation from the exact stoichiometry results in low degrees of polymerization and will shift the equilibria toward short oligomers (5). Already in the course of preparing the complexes (refer to supporting information, which is published on the PNAS web site, www.pnas.org), an interesting observation was made after dissolving complex [1a⋅2] in methylcyclohexane. The solution did not show the expected yellow-green color and fluorescence typical for the diazadibenzoperylene chromophore but, instead, displayed an orange-red color with a rather weak fluorescence, indicating aggregation processes. In contrast, the color change was much less pronounced for complex [1b⋅2], which still showed the typical yellow-green color of 1b in concentrated solution and turned orange only in the solid state. The formation of hydrogen-bonded complexes in the solid state could be verified by IR spectroscopy, as two new broad bands appeared around 2,500 and 1,900 cm−1 that are typical for hydrogen bonds between pyridines and benzoic acids (23).
Figure 1.

Ditopic diazadibenzoperylenes 1a,b and substituted isophthalic acid 2, which can self-assemble via hydrogen bonds to an extended polymeric chain.
To gain insight into the aggregation process, dilution experiments of the 1:1 complexes were carried out and monitored by UV/visible (Vis) absorption and fluorescence spectroscopy. Fig. 2 shows the concentration dependence of UV/Vis spectra recorded for solutions of [1a⋅2] in methylcyclohexane in the concentration range from 10−4 to 2.5 × 10−7 M. As the strong changes in the S0-S1 absorption band of the diazadibenzoperylene chromophore (λabs = 494 nm) cannot be explained on the basis of pure hydrogen-bonding between the two building blocks, π–π interactions between the chromophores have to be involved. Notably, the concentration range where the spectral changes occur is rather dilute (Fig. 2 Inset), considering the low binding constant of the pyridine–carboxylic acid hydrogen bonding interaction (21), which gives another argument for the presence of additional structure-directing intermolecular forces, i.e., π–π aggregation, which contributes to the stability of the supramolecular assembly in a cooperative way already at very low concentration.
Figure 2.
Concentration-dependent UV/Vis absorption spectra in methylcyclohexane for a stoichiometric mixture of diazadibenzoperylene 1a and isophthalic acid 2 (c = 2.5 × 10−7 to 10−4 M). Arrows indicate changes upon increasing concentration. (Inset) Change of the extinction coefficient at λ = 494 nm vs. concentration.
The same dilution experiment was carried out with complex [1b⋅2] in methylcyclohexane, but here, only very small changes in the diazadibenzoperylene absorption bands occurred, and the color of the solutions did not change (Fig. 3). These weak changes in the absorption bands are now attributed to simple hydrogen-bond formation. Although such hydrogen bonds are weak in chloroform (we earlier reported a binding constant K = 140 M−1 for hydrogen bonding between diazadibenzoperylene 1a and 3,4,5-tridodecyloxybenzoic acid 3 in CDCl3, ref. 21) there will be a strong increase of the binding constant in methylcyclohexane, a nonpolar and noncompetitive solvent. From our detailed studies on the solvent dependence of the binding constant for perylene bisimide-melamine triple hydrogen bonds (5), we can estimate a binding constant of K = 8,000 M−1 for hydrogen bonding between 1a and 3 in methylcyclohexane by assuming a linear-free energy relationship between these two hydrogen-bonding motifs. For such a situation, about 35% of the aza receptor sites will be complexed at a concentration of 10−4 M, giving rise to the observed small spectral changes. As the two diazadibenzoperylenes 1a and 1b differ only in the phenoxy substituents, their entirely different mode of self-assembly in the presence of 2 should be related to the bulkiness of the tert-butyl groups in 1b which prevent a close contact between the diazadibenzoperylene π-systems that are required to induce two-dimensional growth to extended assemblies already in dilute solution.¶
Figure 3.

Concentration-dependent UV/Vis absorption spectra in methylcyclohexane for a stoichiometric complex of diazadibenzoperylene 1b and isophthalic acid 2 (c = 10−6 to 10−4 M). Arrows indicate changes upon increasing concentration.
To shed more light on the self-organization process between 1a and 2, control experiments were carried out to ensure that structural ordering by both hydrogen bonding and π–π interactions are involved and to prove that aggregation of the two components is not a nonspecific random process like in (reversed) micelles of amphiphilic molecules. First, a UV/Vis dilution experiment for pure 1a in methylcyclohexane confirmed that aggregation of the chromophores is negligible in the given range of concentration (c = 10−6 to 10−4 M). Then, 3,4,5-tridodecyloxybenzoic acid (3, a monotopic carboxylic acid) was added to 1a to give a complex [1a⋅32] of 1:2 stoichiometry, and again concentration-dependent UV/Vis spectra were recorded in methylcyclohexane. In both complex [1b⋅2] and [1a⋅32], no significant changes in the absorption bands of the diazadibenzoperylene chromophore were observed (c = 10−6 to 10−4 M), pointing to the uniqueness of the hydrogen bond-directed chromophore aggregation for [1a⋅2].
In addition to pronounced changes in color and absorbance for concentrated solutions of complex [1a⋅2], fluorescence quenching also was noted. Therefore, a dilution experiment was carried out in methylcyclohexane that was monitored by steady-state fluorescence spectroscopy with excitation at λex = 450 nm. According to Fig. 4, the characteristic diazadibenzoperylene emission with maxima at 510 and 545 nm gradually disappears, and a very broad and unstructured new emission band evolves around 650 nm upon increasing concentration. The spectra in Fig. 4 are normalized with respect to an equal integrated emission which does not represent the actual emission, but shows the spectral changes in a clearer way. The extent of fluorescence quenching upon aggregation can be seen in the Fig. 4 Inset, where the emission spectra were corrected for the optical density at the excitation wavelength (λex = 450 nm), integrated over the whole wavelength range, and plotted vs. concentration. Thus, the initial fluorescence [ΦF(1a, CHCl3) = 75%] drops to ≈1% in the aggregated state. Because of the well separated emission bands of the aggregate and the free diazadibenzoperylene ligand, it can be assumed that 1a is almost quantitatively incorporated into the aggregate at a concentration of 10−4 M.
Figure 4.

Concentration-dependent fluorescence spectra of [1a⋅2] in methylcyclohexane (λex = 450 nm; c = 2.5 × 10−7 to 10−4 M). The corrected spectra were normalized to the same integrated emission, and the arrows indicate changes upon increasing concentration. (Inset) Fluorescence quenching in the course of aggregation. The integrated emission of the spectra after correction for the optical density at the excitation wavelength are plotted vs. concentration.
Finally, information on the structure and the thermotropic properties of the aggregates in the solid state was sought from differential scanning calorimetry (DSC) and x-ray scattering. The DSC trace of the first heating of [1a⋅2] shows a transition at 183°C which is characterized by an enthalpy change of 2.3 J/g and a transition to the isotropic liquid state at 262°C with an associated enthalpy change of 34.4 J/g. The latter phase transition into the isotropic liquid state also could be confirmed by optical microscopy at crossed polarizers, whereas no morphological changes could be observed for the former transition. However, temperature-dependent small (SAXS) and middle (MAXS) angle x-ray scattering experiments between room temperature and 250°C confirmed the transition at 183°C and related it to a major structural change (Fig. 5). In the diffractograms that were taken from room temperature to 183°C, four sharp reflections can be observed at 0.88°, 1.66°, 3.03°, and 4.70°. It is remarkable that the reflection at 0.88° corresponds to a periodic distance of ≈98 Å, which is noticeably large compared with the size of the individual molecules, and, thus, indicates long-range order in the given supramolecular system. Unfortunately, we could not derive a reasonable packing model owing to the absence of a simple mathematical relationship between the given reflections.
Figure 5.
Variable temperature (a) small-angle x-ray scattering (SAXS) and (b) middle-angle x-ray scattering (MAXS) between 30 and 250°C taken at the polymer beam line A2 at Deutsches Elektronen Synchrotron in Hamburg.
At temperatures higher than 183°C, the SAXS and MAXS diffractograms change, and the reflections in the small-angle region at 0.88° and 1.66° disappear. At the same time, new reflections in the middle-angle region appear at 3.2°, 6.4°, 9.6°, and 11.3°, with the reflection at 3.2° being the most intense (Fig. 5b). The first three reflections are equidistant, which is typical for a layered structure, and the lamellar period could be calculated to 28 Å (24).‖ Furthermore, wide angle x-ray scattering (WAXS) indicates the presence of well defined π–π-aggregated dyes at a distance of 3.76 Å, as the diffractogram exhibits a sharp reflection at 2 ϑ = 24.18° on top of the broad halo of the disordered alkyl chains.
For the proposal of a structural model for [1a⋅2], these x-ray data and some additional experimental facts were taken into account. From the concentration-dependence of the UV/Vis absorption and fluorescence spectra, strong excitonic interactions between the diazadibenzoperylene chromophores are evident. This finding implies that the dyes are tightly packed with a close distance of their π-conjugated systems, which is 3.76 Å, according to WAXS. Furthermore, from the structural point of view [1a⋅2] assemblies are expected to give rise to microphase-segregation (19, 20) because of strong contrasts in the polarity of the polar diazadibenzoperylene and isophthalic acid core units and the nonpolar aliphatic chains of 2. The lamellar period of 28 Å found in x-ray scattering at temperatures higher than 183°C fits to the size of the assemblies with extended alkyl chains. Therefore, the model depicted in Fig. 6 seems likely, as it represents hydrogen-bonded zigzag chains of repeating diazadibenzoperylene and isophthalic acid units that are stacked on top of each other, driven by π–π contacts, thereby forming layers. The alkyl chains of the isophthalic acid units point between the layers and most likely interdigitate with the alkyl chains of the neighboring layers, resulting in an alternating polar-nonpolar structure. Corey–Pauling–Kolton (CPK) models support this structure, although it seems that the lamellar period of 28 Å is a lower limit, already imposing some difficulties in accommodating the interdigitating alkyl chains (25). However, the lamellar period can easily be reduced by a small tilt of the chains within the layers, as observed recently in columnar liquid-crystalline perylene bisimide dyes (26).
Figure 6.
Packing model for complex [1a⋅2] and CPK model from molecular modeling (MM-MM2 force field; ref. 25) of the arrangement of the building blocks.
In summary, we showed that hydrogen-bonding between the ditopic building blocks 1a and 2 can initiate the formation of extended dye aggregates in low-polarity solvents and the pristine state. In the solid state, complex [1a⋅2] forms a layered structure with a lamellar period of 28 Å according to x-ray diffraction studies. If, however, small structural changes are introduced in one of the building blocks, i.e., in 1b with four bulky tert-butyl groups, extended assemblies do not form anymore because of the reduced strength of the π−π interactions.
Supplementary Material
Acknowledgments
We thank G. Dörfner and U. Thewalt (Sektion Röntgen- und Elektronenbeugung, Universität Ulm) for room temperature WAXS experiments. Financial support from the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie is gratefully acknowledged.
Abbreviation
- Vis
visible
Footnotes
This paper was submitted directly (Track II) to the PNAS office.
Compound 2 was synthesized following the literature procedure described for the analogous compound which only differs in the alkyloxy chain length (C10 instead of C12, ref. 22). For synthetic details and characterization see supplementary information.
The size of these assemblies in solution is still an open question which we intend to address in the future by dynamic light-scattering experiments. Preliminary TEM and SEM experiments could not provide any information which is, however, not surprising, if we consider the lamellar structure observed in the solid state.
MAXS data were also measured for a cooling run from 250°C to room temperature. Here, no transition at 183°C was observed, and no reflections reappeared in the small angle region. This observation suggests that the initial low-temperature structure, where large periodic distances are observed, is a metastable mesophase which formed upon evaporation of the solvent.
References
- 1.Cornil J, Beljonne D, Calbert J-P, Brédas J-L. Adv Mater. 2001;13:1053–1067. [Google Scholar]
- 2.Würthner F. Angew Chem Int Ed Engl. 2001;40:1037–1039. [PubMed] [Google Scholar]
- 3.Brabec C J, Sariciftci N S, Hummelen J C. Adv Funct Mater. 2001;11:15–26. [Google Scholar]
- 4.Schenning A P H J, Jonkheijm P, Peeters E, Meijer E W. J Am Chem Soc. 2001;123:409–416. [Google Scholar]
- 5.Würthner F, Thalacker C, Sautter A, Schärtl W, Ibach W, Hollricher O. Chem Eur J. 2000;6:3871–3886. doi: 10.1002/1521-3765(20001103)6:21<3871::aid-chem3871>3.3.co;2-w. [DOI] [PubMed] [Google Scholar]
- 6.Kuroda Y, Kawashima A, Hayashi Y, Ogoshi H. J Am Chem Soc. 1997;119:4929–4933. [Google Scholar]
- 7.Vreekamp R H, Verboom W, Reinhoudt D N. J Org Chem. 1996;61:4282–4288. doi: 10.1021/jo9603082. [DOI] [PubMed] [Google Scholar]
- 8.Koh K, Araki K, Shinkai S. Tetrahedron Lett. 1994;35:8255–8258. [Google Scholar]
- 9.Garcia-Tellado F, Goswami S, Chang S-K, Geib S J, Hamilton A D. J Am Chem Soc. 1990;112:7393–7394. [Google Scholar]
- 10. Carr, A. J., Melendez, R., Geib, S. J. & Hamilton, A. D. (1998) Tetrahedron Lett., 7447–7450.
- 11.Garcia-Tellado F, Geib S J, Goswami S, Hamilton A D. J Am Chem Soc. 1991;113:9265–9269. [Google Scholar]
- 12. Krishnamohan Sharma, C. V. & Zaworotko, M. J. (1996) Chem. Commun., 2655–2656.
- 13.Kayne J J, Liao R-F, Lauher J W, Fowler F W. J Am Chem Soc. 1995;117:12003–12004. [Google Scholar]
- 14. Valiyaveettil, S., Enkelmann, V. & Müllen, K. (1994) J. Chem. Soc. Chem. Commun., 2097–2098.
- 15.St. Pourcain C B. J Mater Chem. 1999;9:2727–2730. [Google Scholar]
- 16. Kotera, M., Lehn, J.-M. & Vigneron, J.-P. (1994) J. Chem. Soc. Chem. Commun., 197–199.
- 17.Kato T, Kihara H, Kumar U, Uryu T, Fréchet J M J. Angew Chem Int Ed Engl. 1994;33:1644–1646. [Google Scholar]
- 18.Sautter A, Thalacker C, Würthner F. Angew Chem Int Ed Engl. 2001;40:4425–4428. doi: 10.1002/1521-3773(20011203)40:23<4425::aid-anie4425>3.0.co;2-z. [DOI] [PubMed] [Google Scholar]
- 19.Tschierske C. J Mater Chem. 1998;8:1485–1508. [Google Scholar]
- 20.Pegenau A, Hegmann T, Tschierske C, Diele S. Chem Eur J. 1999;5:1643–1660. [Google Scholar]
- 21.Würthner F, Sautter A, Thalacker C. Angew Chem Int Ed Engl. 2000;39:1243–1245. doi: 10.1002/(sici)1521-3773(20000403)39:7<1243::aid-anie1243>3.0.co;2-z. [DOI] [PubMed] [Google Scholar]
- 22.Yang J, Marendaz J-L, Geib S J, Hamilton A D. Tetrahedron Lett. 1994;35:3665–3668. [Google Scholar]
- 23.Johnson S L, Rumon K A. J Phys Chem. 1965;69:74–86. [Google Scholar]
- 24.Ziessel R, Douce L, El-ghayoury A, Harriman A, Skoulios A. Angew Chem Int Ed Engl. 2000;39:1489–1493. doi: 10.1002/(sici)1521-3773(20000417)39:8<1489::aid-anie1489>3.0.co;2-2. [DOI] [PubMed] [Google Scholar]
- 25. CACHE (1999) (Oxford Molecular Group, Oxford), Version 3.2.
- 26.Würthner F, Thalacker C, Diele S, Tschierske C. Chem Eur J. 2001;7:2245–2253. doi: 10.1002/1521-3765(20010518)7:10<2245::aid-chem2245>3.0.co;2-w. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.



