I analysed the bonding in chlorine trifluoride a few years back in terms of VSEPR theory. I noticed that several searches on this topic which led people to this post also included a query about the differences between it and the bromine analogue. For those who posed this question, here is an equivalent analysis.

The calculation is done at the same level as before (ωB97XD/6-311++D(d,p)) for consistency (DOI: 10.14469/hpc/2160)

Click for 3D

- Basins 8 and 9 have electron populations of 2.33e (2.07e for the chlorine analogue) with an angle subtended at Br of 159°. The greater electron population and hence electron pair repulsion has the effect of increasing the angle compared to Cl (154°). The coordination is even more square pyramidal than with Cl.
- Basin 7 has a population of 0.73e, this time less than Cl (0.87e).
- Basins 11 and 12 are 0.82e. With Cl, this single basin was replaced by a pair of split basins, each pair summing to 0.91e (the same effect happens with F-F). The angle 4-2-3 is 172° (174° for Cl) which suggests a slightly increased 2-electron-3-centre interaction between e.g. atoms 1-4 or 1-3 compared to Cl.
- The total basin count surrounding the Br is therefore 7.03e, compared to 6.84e with Cl, which suggests Br is slightly more electronegative in this context than Cl.

Bromine has a habit of springing surprises, but not so much in this example.

### Popular Posts:

To complete the series, here is IF

_{3}. This brings however a new variable, which is the now more significant contribution from relativistic core electrons on iodine.The below is the non-relativistic solution (DOI: 10.14469/hpc/2161), derived from a full-electron basis set for I (6-311G(d,p) as obtained from https://bse.pnl.gov/bse/portal).

Basins 8 and 9 have 2.21e each and subtend an angle of 168° at the central iodine. The other difference is that basins 10 and 15 for BrF

_{3}(the anomeric anti-periplanar effect to one Br-F bond) are not apparent for IF_{3}.A wavefunction where the all-electron basis set is replaced by one with a effective-core for the relativistic component (thus the relativistic contraction is absorbed into this calculation, Def2-TZVPP, DOI: 10.14469/hpc/2162) gives the 8/9 basin populations as 2.36e, which is more consistent with the trend Cl => Br => I and an angle subtended at I of 153°.