Group 17 element fluorides EF3 (E = Cl, Br, I) are
well-known to undergo second-order Jahn−Teller symmetry
breaking toward a T-shaped
C
2
v
arrangement mainly
due to a1‘HOMO ⊗ e‘LUMO mixing at
the expected symmetric
trigonal planar D
3
h
state. For heavy elements, the a1‘ HOMO is
relativistically stabilized because of large
element s-orbital participation. Hence, relativistic effects
diminish the second-order Jahn−Teller term. This
results in a large relativistic change in the
Feq−E−Fax bonding angle of
αe
R − αe
NR =
5.5° in the case of AtF3
and causes an anomaly in the bond angle behavior down the group 17
compounds, α(ClF3) > α(BrF3)
>
α(AtF3) > α(IF3).
Furthermore, the difference between the symmetric
D
3
h
and the distorted
C
2
v
structure
of
AtF3 is only 10 kJ/mol at the coupled cluster level of
theory, indicating that the measured
Feq−At−Fax angle
αe will be very sensitive upon the temperature applied in
gas phase diffraction studies. Vibrational
frequencies
are predicted for all group 17 fluorides EF3. As a
consequence of the second-order Jahn−Teller distortion,
the A1 symmetric bending mode is strongly influenced by
relativistic effects and becomes much lower in
frequency compared to the B1 out of plane mode for the
heavier elements. With the exception of IF3,
the
symmetric D
3
h
structure
represents a (metastable) weak local minimum at the MP2 level, rather
than a transition
state as expected. The
D
3
h
point represents,
however, a second-order saddle point at the HF level, and
therefore,
electron correlation seems to be responsible for changing the nature of
the trigonal planar structure. Extended
basis sets at the MP2 level as well as coupled cluster calculations
were applied in order to obtain more
accurate information for the energetics and structure of
ClF3. These studies show, however, that the
nature
of the D
3
h
point is
critically dependent upon the basis set (and the electron correlation
procedure) applied.