AN INTERPRETATION OF THE CL AGGREGATION PHENOMENON BASED ON ENERGY CALCULATIONS

O.V.Grineva, P.M.Zorky

Faculty of Physical Chemistry, Chemistry Department, Moscow State University, 117234 Moscow, Russia

Energy calculations for a series of crystalline chlorohydrocarbons (mostly chloroaromatic ones) have been carried out. The results confirm the hypothesis proposed in a previous work [1] to explain the causes that lead to Cl-aggregation. This phenomenon consists in the tendency of chlorine atoms to form various aggregates: chains, rods, layers.

It has been shown previously [2, 3] that specific intermolecular contacts (H-bonds, contacts of benzene rings, contacts Cl...Cl and others) often form finite or infinite aggregates. Systematic studies of chlorine-containing organic crystals performed by us show that usually Cl-aggregates present in such crystals and, as a rule, they can be distinguished unequivocally. The reason is that relatively small distances Cl...Cl that occur in a given crystal are usually in the range from d₁ to d₂, and the larger distances, the minimal of which we shall denote as d₃, are separated from the smaller ones by a pronounced interval. The value of d₂ is usually close to 4.2 - 4.3 A, but it can vary according to particular features of the crystalline structure.

To calculate the energy of intermolecular interaction usual atom-atom potentials of the 6-exp type were used:

E_ij = A exp (-Br_ij ) - Cr_ij^-6

This formula is often appended by an electrostatic term q_iq_jr_ij^-1, which gives the 6-exp-1 potential, but there are no distinct criteria for the choice of the charges q. The possibility of assigning an exact physical meaning of atomic charges to individual parameters of this potential (or any other empirical potential) is also questionable [4]. Since it has been shown in work [1] that basic results obtained for different systems of atomic charges (including zero charges) are close to one another, in the present study we used 6-exp potential without taking into account effective atomic charges.

The objects of the study were chlorobenzene, three modifications of p-dichlorobenzene, 1,4- and 2,6-dichloronaphthalenes, 4,5-dichlorophenanthrene and others (about 10 substances).

A new approach was used in energy calculations. It consists in the determination of the efficiency of molecular subsystems, i.e. finite or infinite (e.g. layers or rods) molecular agglomerates. The efficiency is defined as the part of the total energy of the crystal that is contained within the given molecular subsystem.

It has been found that the efficiency of molecular subsystems containing Cl-aggregates is substantially lower than that of similar subsystems without short distances Cl...Cl. The efficiency of the latter, in its turn, is close to the efficiency of similar subsystems in crystalline aromatic hydrocarbons that do not contain chlorine. This gives reasons to believe that the cause of Cl-aggregation (at least in some cases) is the chlorophobic effect. In other words, the interaction of benzene rings with one another turns out to be more advantageous in energy terms than their interaction with chlorine, and chlorine atoms are driven out from the areas of benzene interactions. At the same time, even though the arising and the aggregation of relatively short Cl...Cl distances is not advantageous in energy terms such drawing together of chlorine atoms proves to be necessary to increase the energy advantageousness of the structure as a whole.

It is obvious that the reliability of results of such kind depends on the adequacy of the description of interactions between different molecules in the crystal given by potential functions. At present mainly two criteria for assessing the adequacy of the calculation of the energy of intermolecular interactions are used: 1) the calculated energy of the crystal must be roughly in accordance with the heat of sublimation; 2) the experimentally determined structure of the crystal must at least approximately correspond to a minimum of the potential surface (this minimum can be not the deepest one). We have carried out a search for the nearest minimum of energy for some of the studied structures (the space group was fixed, the molecule was regarded as a rigid body). As a rule, the values characterizing the lattice parameters and the orientation of the reference molecule varied insignificantly, and this practically did not reflect on the efficiency of subsystems.

We believe that the energy calculation of intermolecular interactions in crystals provides radically new possibilities for the interpretation of Cl-aggregation. It should be noted that the commonly used analysis for the most preferable (in energy terms) arrangement of molecules in dimers by no means lead to this goal.

1. O.V.Grineva, P.M.Zorky, Russ. J. Phys. Chem. 72 (1998) (in press).
2. P.M.Zorky, O.N.Zorkaya, 16th European Crystallographic Meeting, Lund, Sweden, 6-11 August 1995, 87.
3. P.M.Zorky, 10th International Symposium on Organic Crystal Chemistry, Collected Abstracts, Pozna-Rydzyna, Poland, 17-21 August 1997, 24.
4. A.Gavezzotti, G.Filippini, Acta Chim. Hung. 130 (1993), 205-220.