60th Annual Report on Research 2015

We are interested in the intermediates in Friedel-Crafts alkylation reactions, which encompass a broad class of carbon-carbon bond forming processes that are key for the conversion of petroleum feedstocks to specialty chemicals and consumer products. The proposed intermediates in these processes are donor-acceptor complexes formed from alkyl halides and a Lewis acid (e.g., RCl–AlCl₃), which in some cases lead to ions (R⁺/MX_n^–) in solution. We have spent the past year completing a computational survey of structural and energetic properties in a broad range of these complexes, of the general form: RX’–MX₃ (M = B, Al, Ga; X = F, Cl; R= CH₃).

Overall, the metal atom appears to have the greatest impact on the structure and bonding in these systems, with M=Al complexes being the most strongly bonded, followed closely by Ga, while B-containing systems tend to be quite weak. For example, the binding energies (MP2/aug-cc-pVTZ) across an analogous series of X=X’=Cl complexes are: -2.7 kcal/mol for CH₃Cl–BCl₃, -14.2 kcal/mol for CH₃Cl–AlCl₃, -11.7 kcal/mol for CH₃Cl–GaCl₃. The differences between analogous Cl- and F-containing systems are less significant (with regard to either type of halogen, X or X’). For example, relative to the binding energy for CH₃F–AlF₃ (-21.5 kcal/mol), the value for CH₃Cl–AlF₃ (-17.5 kcal/mol) illustrates that the complexes are a bit more stable with X’=F; while a comparison to the value for CH₃F–AlCl₃ (-17.1 kcal/mol) indicates that the complexes are also somewhat more stable with X=F as well. The trends are less clear-cut among the relatively weak B-containing systems, of which CH₃F–BF₃ is clearly the most stable (-5.1 kcal/mol).

Charge analyses (NPA) have provided insight into the extent of charge transfer from donor (RX’) to acceptor (MX₃) in these systems, as well as the trends toward carbocation character in the R=CH₃ group. For the B-containing complexes, there is minimal charge transfer; less than 0.02e across the entire series. The Al- and Ga-containing systems exhibit much more charge transfer, and in these cases, it is critically dependent on the alkyl halogen (X’). For the CH₃Cl-containing systems with M=Ga or Al, the extent of charge transfer ranges from 0.22e in the GaCl₃ complex, to 0.16e in the AlF₃ complex. For the CH₃F-containing systems with M=Ga or Al, the extent of charge transfer ranges from 0.03e in the GaF₃ complex, to 0.11e in the AlCl₃ complex.

The trend toward carbocation character is reflected in the net change in the charge on the R=CH₃ substituent, before and after coordination to MX₃. For the X’=Cl systems, the net charge increases from +0.08e in free CH₃Cl to between +0.09e (CH₃Cl–BCl₃) and +0.17e (CH₃Cl–GaCl₃ and CH₃Cl–GaF₃). For the X’=F systems, the net charge increases from +0.38e in free CH₃F to between +0.39e (CH₃F–BCl₃) and +0.49e (CH₃Cl–AlCl₃ and CH₃Cl–AlF₃). As such, it appears that there is little tendency for substantial charge transfer or carbocation-like charge distributions for the B-containing complexes, while there is a minor trend towards additional charge transfer and carbocation character in the M=Al, Ga systems (+0.02 to +0.11e). Also, the CH₃F systems tend to exhibit more charge transfer and carbocation character with M=Al, while the CH₃Cl system systems tend to show more charge shift with M=Ga; a possible reflection of relative hardness/softness among the donor and acceptor atoms.

The bottom line is that not one of these systems exhibits any significant degree of ionic bonding (i.e., R⁺/MX₄^–) in their gas-phase structures. However, we are the process of revisiting these analyses, and also mapping the donor-acceptor bond potentials, in dialectic media (as represented in continuum solvation models), which will indicate the extent to which a solvent environment may affect structure and bonding in these systems, and reconcile these results with experimental observations. We are also investigating alternative ionization pathways involving second RX molecule, and preliminary results indicate that such pathways lie much lower in energy that direct ionization from the 1:1 complexes. In addition, low temperature IR experiments, aimed at identifying specific ionic species in low RX/MX₃ mixtures, will be conducted during the next year.