3.4 Second-order donor-acceptor charge transfer delocalization
from NBO analyses
Consistent with the work of others,11 the results of
the second-order perturbative estimates of donor-acceptor
(bond-antibond) interactions obtained from an NBO analysis suggests that
several complexes of Fig. 1 are the result of significant charge
transfer (CT) delocalization between the σ* anti-bonding orbital of the
chalcogen bond donor fragment and the lone-pair bonding orbitals
associated with the bases A–. For example, the
complexes F2O2···A–(A = F, Cl, Br) are predominately due to the result of the n (A) →
σ*(O–O) CT delocalization, with E(2) of 112.9,
159.9 and 198.44 kcal mol-1 for
F2O2···F–,
F2O2···Cl–, and
F2O2···Br–,
respectively, where n represents the lone-pair bonding orbital(s)
of the anionic species. Similarly, the complexes
F2O···F–,
F2O···Cl–, and
F2O···Br– are the results of then (A) → σ*(O–F) CT, with E(2) of 184.4,
141.9 and 127.9 kcal mol-1, respectively.
The complex Cl2O···F– is due to the
combined effect of CT delocalizations such as n(2) (F) →
σ*(O–Cl1)/σ*(O–Cl2) (E(2) = 8.3/1.0 kcal
mol-1), n(3) (F) → σ*(O–Cl1)/ σ*(O–Cl2)
(E(2) = 0.3/1.8 kcal mol-1)
and n(4) (F) → σ*(O–Cl1)/ σ*(O–Cl2)
(E(2) = 139.2/5.5 kcal
mol-1). For the complex
Cl2O···Cl–, theE(2) for the CT interactions, viz .n(1) (Cl) → σ*(O–Cl1)/ σ*(O–Cl2) and n(3) (Cl) →
σ*(O–Cl1)/σ*(O–Cl2) are 2.3/2.3 and 2.3/2.3 kcal
mol-1, respectively. The corresponding CT
delocalizations responsible for
Cl2O···Br– are n(1) (Br) →
σ*(O–Cl1)/σ*(O–Cl2) and n(3) (Br) → σ*(O–Cl1)/σ*(O–Cl2), withE(2) of 2.2 and 2.1 kcal
mol-1, respectively. These results indicate that
several lone-pair bonding orbitals on the halide ion
X– facilitate CT interactions with the anti-bonding
σ* orbitals of the C–X bonds, providing evidence of the formation of
chalcogen bonding in these complexes. The many-fold interaction
topologies between the donor and acceptor orbitals are apparently due to
the non-linear nature of the F···O, Cl···O and Br···O interactions in
these complexes, respectively (∠F···O–Cl = 149.9o in
Cl2O···F–, ∠Cl···O–Cl =
120.0o in
Cl2O···Cl–, ∠Br···O–Cl =
118.5o in
Cl2O···Br–).
The angle of interaction in the complexes
Br2O···F–,
Br2O···Cl–, and
Br2O···Br– is also significantly
non-linear. For instance, ∠F···O–Br, ∠Cl···O–Br and ∠Br···O–Br are
124.0, 121.2 and 120.0o in the respective complexes,
resulting in pseudo-windmill type geometries between the O and X atoms.
There is, therefore, an expectation of orbital interaction in these
complexes that could be associated with the two σ* anti-bonding orbitals
of the two Br–O bonds in the OBr2 moiety and the
lone-pair bonding orbitals of the halide anions. Indeed, the second
order analysis suggests such a possibility; the
Br2O···F– complex is the result of
(n(4) (F) → σ*(O–Br1)) and (n(4) (F) → σ*(O–Br2))
delocalizations, each with an E(2) of 61.2 kcal
mol-1. The strongest orbital interaction occurs
between n(4) on F and each of two σ*-orbitals of the O–Br bond;E(2) for analogous CT interactions involving
the lone-pair bonding orbitals 2 and 3 on F and σ*(O–Br) were 5.4 and
2.0 kcal mol-1, respectively. Similarly, for the
complex Br2O···Cl–, the CT
interactions were (n(1) (Cl) → σ*(O–Br1)/σ*(O–Br2) and
(n(3) (Cl) → σ*(O–Br1)/σ*(O–Br2), each withE(2) of 2.4/1.9 kcal mol-1.
For the complex Br2O···Br– the
corresponding CT delocalizations were (n(1) (Br) →
σ*(O–Br1)/σ*(O–Br2) and (n(3) (Br) → σ*(O–Br1)/σ*(O–Br2), each
with an E(2) of 2.4/1.8 kcal
mol-1.
Along similar lines, the complexes of OX2 (X = Cl, Br)
and the O end of OCN– (22, 33) are described byn (O) → σ*(O–X) CT delocalizations. TheE(2) for these are 1.76 and 2.36 kcal
mol-1, respectively. The trend inE(2) is consistent with the trend in
ΔE (BSSE) of these complexes. The O···Br secondary interaction in
Br2O···OCN– is characterized by anE(2) of 0.38 kcal mol-1.
The complex,
F2O2···NC–, is a
result of the π (C≡N) → σ*(O–O) and n (N) → σ*(O–O) CT
delocalizations, with E(2) of 0.63 and 1.20
kcal mol-1, respectively. Although the former provides
evidence of the involvement of a π···σ interaction, the latter supports
the formation of lone-pairσ* type chalcogen bonding. In addition, and
because the O atom in O2F2 is not
entirely positive, the lone-pair orbital of O facilitates back-bonding
CT interactions with the σ* orbital of the C≡N fragment (viz.n (O) → σ*(C≡N)), with E(2) of 1.42 kcal
mol-1. Similarly, the complex
(CN)2O···O(NO2)–features CT delocalizations of n (O)
(NO3–) → σ*(O–C)
((CN)2O) (E(2) = 0.9 kcal
mol-1) and n (O)
(NO3–) → π*(C≡N)
(E(2) = 6.3 kcal mol-1),
showing that in addition to the formation of the lone-pairσ* chalcogen
bond, the σπ* interaction plays a crucial role in determining the
equilibrium geometry of the complex. Most of the complexes discussed
above do involve several other weak CT interactions; detailed discussion
of these is beyond the scope of this study.
QTAIM’s bond path topologies discussed already above revealed that one
of the two O···X interactions in the complexes of OX2with Br3– (14, 25 and 34 of Fig. 1)
is marginally stronger than the other. The CT delocalizations
corresponding to these two interactions in
F2O···Br3– are
described by n (terminal Br) → σ*(O–F2) and n (middle Br) →
σ*(O–F3), , with E(2) of 1.06 and 0.86 kcal
mol-1, respectively. The analogous delocalizations in
Cl2O···Br3– weren (terminal Br) → σ*(O–Cl2) and n (middle Br) → σ*(O–Cl3),
with E(2) of 0.32 and 1.12 kcal
mol-1, respectively. Similarly, these aren (terminal Br) → σ*(O–Br2) and n (middle Br) → σ*(O–Br3)
for Br2O···Br3–, withE(2) of 0.73 and 1.58 kcal
mol-1, respectively.