AmericanChemicalSociety.com

Context: The main objective of this project is to characterize the structural properties of organofluoride – boron trifluoride complexes (RF'–BF₃) in the gas-phase and in bulk, condensed-phase environments via computations and low-temperature infrared spectroscopy. These species are key intermediates in Friedel-Crafts reactions [1] - an important class of carbon-carbon forming processes that facilitate the conversion of petroleum feedstocks to commercially-viable compounds. One common example is the alkylation of benzene, viz.

   +   RF'   +   BF₃        +   HF'   +   BF₃    (1)

The first step is the formation of an RF–BF₃ intermediate [1], and it is known that the methyl complex is quite weak in the gas phase with an experimental B-F' distance of 2.42  [2], but reacts like an effective carbocation in solution [1]. Thus, its solution-phase reactivity can only be rationalized by invoking a substantial structural re-arrangement that results from interactions with the solvent.

Infrastructure: In the past year, we have purchased and assembled a 7-node, 56 CPU computational cluster, which has greatly expedited the computational part of this work. This acquisition was made possible not only with funds from this grant, but also startup funds from a recent hire, as well as generous hardware and administrative support from UWEC's Learning and Technology Services Department. We have been running almost all of our computational jobs on this system since last spring, and it has enabled us to use very high levels of theory.

N₂–BH₃, a very interesting test system: Some time ago we discovered that the N₂–BH₃ complex had a very short B-N (< 1.6 ) according to density functional theory, but the MP2 binding energy was only 7.1 kcal/mol. This past spring, we resumed our work on this complex since we were need of a small test system to run some performance benchmarks on our new cluster and experiment with Natural Bond Orbital [3] and Atoms in Molecules [4] analyses of the bonding that will be key for the RF'–BF₃ studies. In the end, we conducted an extensive study using both density functional theory (e.g. B3PW91, with basis sets up to aug-cc- pV5Z), and post-HF calculations (e.g. CCSD, with basis sets up to 6-311+(2df,2pd)). Select structural results are shown below, and it is apparent that there is marginal agreement between DFT and post-HF methods, and moreover, the binding energy is even less than that predicted by MP2, perhaps only a few kcal/mol. In spite of this, NBO [3] and AIM [4] analyses indicate that the extent of charge transfer (0.25 e^- from N₂ to BH₃) and covalent character in this B-N bond is comparable to systems with much larger binding energies (OC–BH₃ and H₃N–BH₃). Thus, the complex has a very short, yet weak B-N dative bond. This work will soon be in press in the Journal of Physical Chemistry A.

Gas-Phase Structures and Binding Energies of RF-BF₃ Complexes

Last year we reported the gas-phase structure of CH₃F–BF₃, which had a calculated (X3LYP/aug-cc-pVTZ) B-F' distance of 2.41 , and an MP2/aug-cc-pVTZ binding energy of 3.6 kcal/mol. We now have obtained X3LYP structures for several conformers of (CH₃)₂CHF–BF₃ and (CH₃)₃CF–BF₃, of which the most stable are shown below. These complexes have B-F' distances that are only 0.05  shorter than the methyl complex, but binding energies that are more than 2 kcal/mol larger than for CH₃F–BF₃.

B-F' potentials: Gas phase and dielectric media

Calculated (X3LYP/aug-cc-pVTZ) B-F' potential curves for CH₃F–BF₃, (CH₃)₂CHF–BF₃ and (CH₃)₃CF–BF₃, are shown in the figures below. The gas-phase potentials (top traces in each curve) show no major peculiarities, but are rather soft long the inner wall, such that the energy only rises by about 2.5 kcal/mol over several tenths of an  between the minima and the inner wall. The curves computed for dielectric media (via PCM [5]), do differ notably between the methyl complex and the two larger systems. As the dielectric increases, the potential of CH₃CF–BF₃ responds uniformly across all bond lengths, such that there is no change in shape, and no signs that a distinct structure with a shortened B-F' bond length is stabilized by the medium. For the larger complex however, the inner portion of the curve is preferentially stabilized. For (CH₃)₂HCF–BF₃ and the potential at 1.7 is only about 1 kcal/mol above the minimum.

CH₃F–BF₃

(CH₃)₂HCF–BF₃

(CH₃)₃CF–BF₃

Frequencies: CH₃F-BF₃

At this point we have also collected IR spectra CH₃F–BF₃ and (CH₃)₂HCF–BF₃, in both thin films and argon matrices. In the latter case, making spectral assignment has proved to be quite difficult. For the methyl complex, we have observed at least one key frequency shift - in the BF₃ "Umbrella" mode, which red shifts in the early stages of B-F' bond contraction. Moreover, this shift is systematic across various media including our measured and calculated frequencies, as well as those previously measured in cryogenic liquids [6]. This depicts a subtle yet significant medium effect on the complex, which should be more extreme in the larger systems according to the computational results.

1. Olah, G.A. Friedel-Crafts and Related Reactions; Wiley/Interscience: New York, 1963.

2. Leopold, K.R.; Canagaratna, M.; Phillips, J.A. Accts. Chem. Res. 1997, 30, 57.

3. Reed, A.E.; Curtiss, L.A.; Weinhold, F. Chem. Rev. 1988, 88, 899.

4. Bader, R.F.W. Atoms in Molecules - A Quantum Theory; Oxford University Press: Oxford, 1990.

5. Miertus, S.; Tomasi, J. Chem. Phys, 1982, 65, 239.

6. van der Veken, B.J.; Sluyts, E.J. J. Phys. Chem. A. 1997, 107, 9070.