Article
OMe
compounds [(Me TPyA) Fe ( L)](BF )·2.0MeCN (2-
3
2
2
4
Cl
OMe·2.0MeCN), [(Me TPyA) Fe ( L)](BF )·0.5Et O (2-
3
2
2
2
4
2
NO
Cl·0.5Et O), [(Me TPyA) Fe ( L)](BF ) (2-NO ), and
2
3
2
2
4
2
SMe
2
[
0
(Me TPyA) Fe ( L)](BF ) ·0.9MeCN·0.5Et O (2-SMe ·
3 2 2 4 3 2 2
.9MeCN·0.5Et O). Subsequent drying of these crystals
2
under reduced pressure gave the desolvated forms 2-R (R =
OMe, Cl, NO , SMe ) in moderate yields of 57−88%. IR
2
2
spectroscopy for solid samples reveals changes in peak
−
1
frequencies and intensities in the 1200−1600 cm spectral
2
2
and/or C−O bond order for the tetraoxolene bridging ligand
in these two series of compounds.
OMe
L)]+,
The cationic complexes [(Me TPyA) Fe (
3
2
2
2
Cl
+
NO
+
[
[
(Me TPyA) Fe ( L)] , [(Me TPyA) Fe ( L)] , and
3
2
2
3
2
2
SMe
2
3+
(Me TPyA) Fe ( L)] in 2-R·solvent feature very similar
3
2
2
II
structures to those in 1-R·solvent. The two Fe sites in each
molecule are related through a crystallographic inversion
II
center, with the exception of slightly inequivalent Fe centers
Cl
+
in [(Me TPyA) Fe ( L)] stemming from crystal packing of
3
2
2
−
39
sum
the (BF ) counterion. The nearly identical values of ∑
4
for 1-R·solvent and 2-R·solvent illustrate that the coordination
II
geometry at Fe is not significantly affected by the substituents
and redox state of the bridging ligand. In contrast, close
comparison of the bond distances between the two series of
compounds reveals several key differences. First, the mean C−
C bond distance decreases slightly by 0.6−1.5%, from
1
1
.436(4)−1.443(5) to 1.422(1)−1.430(2) Å, in moving from
-R·solvent to 2-R·solvent. Moreover, the mean C−O bond
distance for 2-R·solvent varies from 1.285(2) to 1.304(2) Å
across the series, which represents a 2.7−3.1% increase as
compared to the values obtained for the oxidized analogues.
These structural changes upon reduction reflect a net increase
in C−C bond order and a net decrease in C−O bond order, in
Figure 3. Cyclic voltammograms for solutions of 1-R in MeCN
containing 100 mM (Bu N)(PF ) supporting electrolyte, collected at
4
6
R
x−
2
98 K; R = OMe (red), Cl (green), NO (blue), SMe (gold). Scan
2 2
agreement with a ligand-centered reduction from
L
to
−
1
−1
rate = 100 mV s for R = OMe, Cl, NO ; 25 mV s for R = SMe2.
R (x+1)−•
2
L
(x = 2: R = OMe, Cl, NO ; x = 0: R = SMe ), as has
2
2
Black vertical lines and arrows denote the open-circuit potentials and
scan directions, respectively.
been observed for similar benzoquinoid-bridged metal
2
8d,29a,b
complexes.
Furthermore, the mean Fe−O bond
distance of 2.052(1)−2.084(1) Å in the 2-R·solvent series is
1.5−2.8% shorter than corresponding distances in 1-R·solvent,
and the mean intramolecular Fe···Fe distance decreases to a
similar degree in moving from 1-R·solvent to 2-R·solvent (R =
pronounced change in E1/2 for RLx−/(x+1)−• (x = 2: R = OMe,
Cl, NO ; x = 0: R = SMe ) is consistent with the substituents
2
2
primarily affecting the energy levels of the ligand; however, the
clear variation in the metal-based redox potential indicates that
the electronic properties of the substituents significantly
modulate the metal−ligand interactions as well. The remaining
OMe, Cl, NO
, SMe ). Together, these observations highlight
2
2
the stronger Fe−O interactions in the radical-bridged
R
x−• n+
complexes [(Me
OMe, Cl, NO ; x = 1, n = 3: R = SMe
d i a m a g n e t i c l i g a n d - b r i d g e d a n a l o g u e s
TPyA) Fe
( L )] (x = 3, n = 1: R =
3
2 2
0
/1+
oxidation event at +0.56 V versus [Cp Fe]
for 1-OMe is
2
2
) than in the
2
II III
III III
assigned to the metal-based Fe Fe /Fe Fe oxidation,
R
x− n+
whereas the additional reversible redox event at E = −1.17
[(Me TPyA) Fe ( L )] (x = 2, n = 2: R = OMe, Cl,
1
/2
3 2 2
0
/1+
V versus [Cp Fe]
for 1-SMe is assigned to the ligand-
NO ; x = 0, n = 4: R = SMe ) owing to the increase in anionic
2 2
2
2
SMe
2
−•/2−
II
based
L
couple.
charge. As a result of stronger interactions between the Fe
Together, the cyclic voltammetry measurements suggest that
centers and the bridging ligand in 2-R·solvent, the average Fe−
N bond distance increases slightly by 1.0−1.8% in moving
from 1-R·solvent to 2-R·solvent (R = OMe, Cl, NO , SMe ).
I I
t h e r a d i c a l - b r i d g e d F e
c o m p l e x e s
2
R
x−• n+
[
(Me TPyA) Fe ( L )] (x = 3, n = 1: R = OMe, Cl,
3
2
2
2
2
NO ; x = 1, n = 3: R = SMe ) should be chemically accessible.
Mossbauer Spectroscopy. To confirm the presence of a
̈
2
2
Toward this end, dark green MeCN solutions of 1-R were
bridging ligand-centered reduction and further probe the
treated with stoichiometric quantities of the reductant Cp Co
effects of tetraoxolene substitution on the electronic structures
2
5
7
to give red-brown (R = OMe, Cl, NO ) or green-brown (R =
of 1-R and 2-R, zero-field Fe Mo
collected for polycrystalline samples at 80 K. The Mo
̈
ssbauer spectra were
ssbauer
2
1
̈
2
more broad peaks relative to the oxidized analogues.
spectra for 1-R exhibit a single sharp doublet (see Figure 4,
top). Lorentzian fits to the data give an isomer shift of δ =
1.076(3)−1.111(3) mm s and a quadrupole splitting of ΔE
= 2.36(2)−2.61(2) mm s across the series (see Table 2).
−
1
Subsequent diffusion of Et O vapor into these solutions
2
Q
−1
afforded red-orange (R = OMe, Cl, NO ) or orange (R =
2
II
SMe ) plate-shaped crystals of the one-electron reduced
These parameters are consistent with high-spin Fe centers in
2
E
Inorg. Chem. XXXX, XXX, XXX−XXX