Aggregation of direct dyes investigated by molecular modeling
Molecular modeling simulating aqueous conditions is used to elucidate the formation of aggregates of sulfonated chromophores, including Congo Red and Trypan Blue. The relative energies involved in the aggregation process are calculated and geometric features of the molecules and their aggregates are discussed. Parallel stacking facing one another occurs with slight shifts along the long axes of the chromophores to accommodate the sterically demanding sulfonate groups. The Connolly surfaces and volumes of the individual molecules as well as their aggregates are determined.
As part of a larger investigation using direct dyes as molecular sensors to characterize different cellulose substrates, we noted that the tendency of these dyes to form aggregates requires elucidation. Results from our dyeing experiments [3] appeared to indicate that, even when dilute solutions of the dyes were employed to obtain Langmuirian adsorption, aggregates of different sizes rather than single molecules might have been adsorbed onto the surfaces, depending on the dye. Thus we used molecular modeling to obtain a better understanding of aggregate formation under these conditions.
Molecular modeling has been successfully applied to a diversity of problems to characterize properties that are extremely difficult to assess by experiment [5, 6, 9, 11]. With respect to dyes, molecular modeling has been employed, for instance, to characterize the geometry of direct and disperse dyes [7, 10, 17]. Aggregation was modeled by Skowronek et aL [12, 13], who extensively used various modeling techniques and force fields to elucidate the geometry and self-assembly of direct dyes. The supramolecular organization of Congo Red was studied because of its unusual complexation characteristics associated with liquid crystal formation. Since liquid crystal formation requires high concentrations of electrolyte and dye, the dyes were modeled in their neutral form. A critical consideration is the fact that modeling calculations are typically based on a gas phase model at a dielectric constant of 1 [11]. However, this class of dyes is usually applied from aqueous solution. Of concern, then, is the question of how well such modeling calculations relate to the solution state. Modeling software addresses the problem in various ways. The molecular mechanics software (Amber 4.1) used by Skowronek et al. [13] permits placement of a water shell around the supramolecular system while maintaining a dielectric constant of 1.
Urukawa et al. [16] used Allinger’s MM2 force field to study aggregation of acid dye molecules, including Azorubin, in order to estimate intermolecular distances between the two molecules forming a dimer. These researchers reported a shifted, parallel-stacked geometry for the Azorubin dimer with an intermolecular distance of about 6A. In general acid dyes are smaller than direct dyes, but are similar to the latter in that they are also azo dyes and have negatively charged sulfonate groups. Azorubin consists of a naphthalene ring connected at its 4-position by an azo-linkage to the 7-position of a second naphthalene group.
For this study, we use molecular mechanics calculations to generate geometries and energies involved in aggregation processes of six direct dye molecules. The dyes are presented in Table I together with the codes used in the text. Five of the molecules under investigation for this study-C.I. Direct Red 28, C.I. Direct Red 2, C.I. Direct Blue 1, C.I. Direct Blue 14, and C.I. Direct Blue 53-are related in that they have the same backbone (Figure la, b). However, dyeing experiments have shown that they can differ significantly in dyeing behavior [3]. It appears that the type, number, and position of functional groups help determine their aggregation properties, which in turn, may affect their dyeing behavior.
Thus our goal in this work is to use molecular modeling as a tool to explore what kind of aggregate formation is possible, the geometries involved, and whether there is a limit beyond which aggregate growth is no longer energetically favorable. In addition, we determine the Connolly surfaces and volumes of the individual dye molecules and their aggregates to obtain an estimate of their size.
A Connolly surface is the van der Waals surface of a molecule that is accessible to a solvent molecule, in this case water. Such a surface is generated by rolling a probe sphere over the van der Waals surface of the molecule of interest. The surface consists of three components: contact, saddle, and concave surfaces. The contact surface comprises those atoms that are accessible to the probe sphere. The saddle-shaped and concave surfaces are denoted as reentrant surfaces and represent the boundary at the volume from which a probe sphere is excluded if it is not to experience van der Waals overlap with the atoms .
Cerius 2 by Molecular Simulations Inc. with the Universal Force Field was employed, as well as CAChe by Oxford Molecular Group. For all experiments performed with Cerius 2, a dielectric constant of 78 was used corresponding to water. The temperature default was absolute zero. Geometry optimizations of one to ten molecules per aggregate were performed for the six dyes under investigation. All dye molecules were treated in their anionic form to model the system as closely as possible to our experimental practices, i.e. aqueous solutions at low concentrations of dye and electrolyte. Under such conditions, the anionic chromophore and its counterion are separated by a large distance [81, such that in this exploratory investigation, the Na+ counterions were omitted. Geometry optimizations of all individual dye molecules were performed prior to assembling them into various starting positions for the aggregation experiments. Geometric optimization procedures on larger aggregates (those containing three to ten chromophores) were performed by manually assembling individually optimized molecules and/or dimers. Intermolecular distances for the manual placement of individual molecules forming dimers, two dimers forming aggregates containing an even number of molecules, as well as individual molecules and dimers to build aggregates with odd numbers of molecules per aggregate, were manually varied from approximately 2.5 to 20Angstrom. Only the most energetically favorable configurations of the dimers were stacked for aggregate assembly. Stacking occurred with the ring systems of the chromophores lined up face-to– face.
The Connolly surface areas and volumes were calculated for all single molecules and all aggregates. Dot displays of the van der Waals surfaces were generated to illustrate the solvent accessible surfaces of the molecules and aggregates. To calculate the Connolly surface areas of the molecules and aggregates, the probe sphere radius was set to 1.4Angstrom, corresponding to the van der Waals radius of water. Conformational analyses were performed with CAChe using the MM2 force field.