Ligand-centered redox activity: Redox properties of 3d transition metal ions ligated by the weak-field tris(pyrrolyl)ethane trianion

Article abstract of DOI:10.1021/ic100028y

First-row transition metal complexes of the tris(pyrrolyl)ethane (tpe) trianion have been prepared. The tpe ligand was found to coordinate In a uniform n¹,n¹,n-coordination mode to the divalent metal series as revealed by X-ray diffraction studies. Magnetic and structural characterization for complexes of the type [(tpe)M^II(py)][Li(THF)₄] (M: Mn, Fe, Co, Ni) reveal each divalent ion to be high-spin and have a distorted trigonal-monopyramidal geometry in the solid state. The pyridine ligand binds significantly canted from the molecular C₃ axis due to a stabilizing -stacking interaction with a ligand mesityl substituent. Cyclic voltammetry on the [(tpe)M^II(py)]^- series reveals a common irreversible oxidation pathway that is entirely ligand-based, invariant to the divalent metal bound. This latter observation indicates that fully populated ligand-based orbitais from the tpe construct are energetically most accessible In the electrochemical experiments, akin to their dipyrromethane analogues. Chemical oxidation of [(tpe)Fe^II(py)]^- yields a product in which the ligand has dissociated one pyrrole (following tpe oxidation and H-atom abstraction) and binds a second equivalent of pyridine to form the neutral, tetrahedral Fe^II species (κ²-tpe)Fe(Py)₂. Similarly, chemical oxidation of the Zn(II) analogue shows evidence for tpe oxidation by electron paramagnetic resonance spectroscopy (77 K, toluene glass) with an isotropic signal for the organic radical at g = 2.002. Density functional theory analysis on this family of complexes reveals that the highest lying molecular orbitals are completely ligand-based, corroborating our proposed electronic structure assignment.

Full text of DOI:10.1021/ic100028y

2512 Inorg. Chem. 2010, 49, 2512–2524
DOI: 10.1021/ic100028y
Ligand-Centered Redox Activity: Redox Properties of 3d Transition Metal Ions
Ligated by the Weak-Field Tris(pyrrolyl)ethane Trianion
Graham T. Sazama and Theodore A. Betley*
Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street, 306E Mallinckrodt,
Cambridge, Massachusetts 02138
Received January 5, 2010
First-row transition metal complexes of the tris(pyrrolyl)ethane (tpe) trianion have been prepared. The tpe ligand was
found to coordinate in a uniform η1,η1,η1-coordination mode to the divalent metal series as revealed by X-ray
diffraction studies. Magnetic and structural characterization for complexes of the type [(tpe)MII(py)][Li(THF)4] (M: Mn,
Fe, Co, Ni) reveal each divalent ion to be high-spin and have a distorted trigonal-monopyramidal geometry in the solid
state. The pyridine ligand binds significantly canted from the molecular C3 axis due to a stabilizing π-stacking
interaction with a ligand mesityl substituent. Cyclic voltammetry on the [(tpe)MII(py)]- series reveals a common
irreversible oxidation pathway that is entirely ligand-based, invariant to the divalent metal bound. This latter
observation indicates that fully populated ligand-based orbitals from the tpe construct are energetically most
accessible in the electrochemical experiments, akin to their dipyrromethane analogues. Chemical oxidation of
[(tpe)FeII(py)]- yields a product in which the ligand has dissociated one pyrrole (following tpe oxidation and H-atom
abstraction) and binds a second equivalent of pyridine to form the neutral, tetrahedral FeII species (κ2-tpe)Fe(py)2.
Similarly, chemical oxidation of the Zn(II) analogue shows evidence for tpe oxidation by electron paramagnetic
resonance spectroscopy (77 K, toluene glass) with an isotropic signal for the organic radical at g = 2.002. Density
functional theory analysis on this family of complexes reveals that the highest lying molecular orbitals are completely
ligand-based, corroborating our proposed electronic structure assignment.
I. Introduction
platforms are an important subclass of the prototypical
redox-active ligand platforms. Porphyrin2 π electrons in
metal porphyrin complexes are typically inert with the
exception of compound I in cytochrome P450, wherein a
single oxidizing equivalent is stored in the porphyrin macro-
cyle.3 In the case of porphyrinogen systems,4 the pyrrole
subunits are unconjugated and thus have higher energy π
orbitals, which results in increased ligand participation in
molecular redox events. Transition metal complexes of both
Cooperative molecular redox comprised of metal and
ligand contributions requires both energetic and angular
overlap between the frontier ligand-based π orbitals and
the metal-based d orbitals.1 Pyrrole-based, oligomeric ligand
*To whom correspondence should be addressed. E-mail: betley@chemistry.
harvard.edu.
(1) Common redox active ligands: Catecholates :(a) Haga, M.; Dodsworth,
E. S.; Lever, A. B. P. Inorg. Chem. 1986, 25, 447–453. (b) Masui, H.; Lever,
A. B. P.; Auburn, P. R. Inorg. Chem. 1991, 30, 2402-2410. Dithiolates:
(c) Schrauzer, G. N.; Mayweg, V. J. Am. Chem. Soc. 1962, 84, 3221. (d) Stiefel,
E. I.; Waters, J. H.; Billig, E.; Gray, H. B. J. Am. Chem. Soc. 1965, 87, 3016-3017.
Amido-phenolates:(e) Chaudhuri, P.;Verani, C. N.; Bill, E.;Bothe, E.;Weyherm€uller,
T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 2213–2223. (f) Herebian, D.;Bothe,
E.; Bill, E.; Weyherm€uller, T.; Wieghardt, K. J. Am. Chem. Soc. 2001, 123, 10012–
10023. (g) Blackmore, K. J.; Ziller, J. W.; Heyduk, A. F. Inorg. Chem. 2005, 44,
5559–5561. (h) Haneline, M. R.; Heyduk, A. F. J. Am. Chem. Soc. 2006, 128,
8410–8411. (i) Blackmore, K. J.; Lal, N.; Ziller, J. W.; Heyduk, A. F. J. Am. Chem.
Soc. 2008, 130, 2728-2730. Phenolates:(j) Hockertz, J.; Steenken, S.; Wieghardt,
K.; Hildebrandt, P. J. Am. Chem. Soc. 1993, 115, 11222–11230. (k) Chaudhuri, P.;
Hess, M.; Muller, J.; Hildenbrand, K.; Bill, E.; Weyherm€uller, T.; Wieghardt, K.
J. Am. Chem. Soc. 1999, 121, 9599-9610. Imino-pyridines:(l) Bouwkamp,
M. W.; Bowman, A. C.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc. 2006,
128, 13340–13341. (m) Bart, S. C.; Chlopek, K.; Bill, E.; Bouwkamp, M. W.;
Lobkovsky, E.; Neese, F.; Wieghardt, K. J. Am. Chem. Soc. 2006, 128, 13901–
13912. (n) Bart, S. C.; Lobkovsky, E.; Bill, E.; Wieghardt, K.; Chirik, P. J. Inorg.
Chem. 2007, 46, 7055–7063.
(2) Kadish, K. M.; Smith, K. M.; Guilard, R. The Porphyrin Handbook;
Academic Press: San Diego, CA, 2003; Vols. 15-20.
(3) (a) Watanabe, Y. In The Porphyrin Handbook; Kadish, K. M., Smith,
K. M., Guilard, R., Eds.; Academic Press: New York, 2000; Vol. 4, pp 97-118. (b)
Schulz, C. E.; Devaney, P. W.; Winkler, H.; Debrunner, P. G.; Doan, N.; Chiang,
R.; Rutter, R.; Hager, L. P. FEBS Lett. 1979, 103, 102–105. (c) Groves, J. T.;
Haushalter, R. C.; Nakamura, M.; Nemo, T. E.; Evans, B. J. J. Am. Chem. Soc.
1981, 103, 2884–2886. (d) Penner-Hahn, J. E.; Smith Eble, K.; McMurry, T. J.;
Renner, M.; Balch, A. L.; Groves, J. T.; Dawson, J. H.; Hodgson, K. O. J. Am.
Chem. Soc. 1986, 108, 7819–7825.
(4) (a) Floriani, C.; Floriani-Moro, R. In The Porphyrin Handbook;
Academic Press: San Diego, CA, 2003; Vol. 3, pp 405-420. (b) Jubb, J.; Floriani,
C.; Chiesi-Villa, A.; Rizzoli, C. J. Am. Chem. Soc. 1992, 114, 6571–6573. (c)
Piarulli, U.; Solari, E.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C. J. Am. Chem.
Soc. 1996, 118, 3634–3642. (d) Crescenzi, R.; Solari, E.; Floriani, C.; Chiesi-
Villa; Rizzoli, C. J. Am. Chem. Soc. 1999, 121, 1695–1706. (e) Bachmann, J.;
Nocera, D. G. J. Am. Chem. Soc. 2004, 126, 2829–2837. (f) Bachmann, J.;
Nocera, D. G. J. Am. Chem. Soc. 2005, 127, 4730–4733.
r
pubs.acs.org/IC
Published on Web 01/26/2010
2010 American Chemical Society

Article
Inorganic Chemistry, Vol. 49, No. 5, 2010 2513
Scheme 1
activity, attenuated N π donicity). We hypothesized that a
tris(pyrrolyl)ethane triply deprotonated and bound κ3 to a
divalent metal ion should be easier to oxidize than their
neutral dipyrromethane analogues and may confer greater
stability to the oxidized products.10 Furthermore, the tris-
(pyrrolyl)ethane platform, along with the dipyrromethene
and dipyrromethane systems, relaxes the structural rigidity
porphyrins and porphyrinogens require, opening the coordi-
nation sphere of the bound transition metal considerably.
II. Results
of these tetra-pyrrole macrocycle systems feature transition
metal and ligand π-orbital alignment such that coupling of
the two reservoirs is permitted.
II.a. Synthesis and Characterization of [tpe] and Its
Metal Complexes. The synthesis of the tris(pyrrolyl)ethane
ligand was accomplished using a modified acid-catalyzed9a
condensation reaction between 2-mesityl pyrrole, 1.5 equiv
of trimethylorthoformate, and ∼7% loading of pyridi-
nium p-toluenesulfonate acid catalyst in two portions.6,8
Refluxing this mixture in dichloromethane for several
days, followed by flash chromatography and trituration
with hexanes, cleanly affords 1,1,1-tris(2-mesitylpyrro-
lyl)ethane (tpeH3, 1) as a pale yellow-white solid in greater
than 69% isolated yield (Scheme 2). To assess the steric
and electronic properties that 1 imparts, transition metal
complexes of the tpe ligand were prepared in the following
manner: reaction of 1 with 3.1 equiv of phenyllithium
(LiN(SiMe3)2 works equally well) in tetrahydrofuran gen-
erates the trilithio complex (tpeLi3), which can be isolated
but was typically prepared in situ for the following synth-
eses. The in situ generated trilithio species was then reacted
with a stoichiometric amount of a divalent metal precursor
(i.e., MCl2(py)2; M=Mn, Co, Fe, Ni, Zn; py=pyridine) in
thawing THF solutions to afford the metalated salts of the
type [(tpe)M(py)][Li(THF)4] upon isolation via crystal-
lization (M=pale yellow Mn (2), bright orange Fe (3), blue
Co (4), green Ni (5), and yellow-orange Zn (6)). The tpe
ligand was found to uniformly bind in an η1,η1,η1-coordi-
nation mode, strictly enforcing a four-coordinate geo-
metry at the metal ion, as confirmed by X-ray diffraction
studies (Table 1). The solid state molecular structures for
the tetrahedral Mn (2), Fe (3), Co (4), Ni (5), and Zn (6)
complexes are shown in Figure 1.
Our recent interest in pyrrole-based ligands stems from our
desire to determine if the rich redox chemistry which features
prominently in metal-porphyrin and porphyrinogen com-
plexes would translate to more easily synthesized truncated
frameworks (Scheme 1). Dipyrromethenes (A, Scheme 1),5
like their porphyrin analogs, exhibit reversible, ligand-
based redox events which are almost entirely decoupled from
metal-based redox activity.6 Dipyrromethane complexes
(B, Scheme 1),7 in contrast, display surprisingly uniform
electrochemical behavior despite differences in metal valence
electron configuration, geometry, and spin state.8 The
energetically high-lying π electrons from the dipyrromethane
framework almost exclusively account for the observed redox
behavior of the metal complexes studied, indicating that fully
populated ligand-based orbitals from the dipyrromethane
construct lie above partially filled metal 3d orbitals. Whereas
the macrocylic nature of porphyrinogens stabilizes ligand-
centered radicals through the reversible formation of C-C
bonds within the macrocyle,4 the dipyrromethane metal
complexes do not exhibit similar reactivity. Chemical oxida-
tion of the metal-dipyrromethane complexes does not yield
stable ligand-based diradical/dicationic species or readily
identified metal-containing products. Thus, we sought to
develop a similar pyrrole-derived ligand platform that would
be more amenable to the study of the resulting oxidation
products.
Sterically encumbered, trianionic tris(pyrrolyl)ethane li-
gands (C, Scheme 1)9 are potential candidates to circumvent
the instability of the dipyrromethane analogues while main-
taining the attractive attributes of these platforms (e.g., redox
II.b. Structural Characterization of Complexes 2-6.
Complexes 2-6 are nearly isostructural, featuring an
intermediate geometry between trigonal-monopyramidal
and pseudotetrahedral at the metal center. The pyridine
ligand and two of the tpe pyrrolide substituents (Nb, Nc)
form the trigonal base with one pyrrolide ligand (Na)
capping the trigonal monopyramid. The sum of the angles
of the monopyramidal base (Npy, Nb, Nc) R þ β þ jbc sum
to nearly 360° for each complex (R þ β þ jbc (deg): Mn
347, Fe 350, Co 348, Ni 351, Zn 348). The pyridine ligand
is consistently positioned off the molecular 3-fold axis
(axis depicted d in the diagram for Table 2), positioning
the pyridine ligand in a distinct π-stacking interaction
with the Na pyrrolide mesityl substituent (ranging from
(5) Wood, T. E.; Thompson, A. Chem. Rev. 2007, 107, 1831–1861.
(6) King, E. R.; Betley, T. A. Inorg. Chem. 2009, 48, 2361–2363.
(7) (a) Lee, C.-H.; Lindsey, J. S. Tetrahedron 1994, 50, 11427–11440. (b)
Jones, D. J.; Gibson, V. C. Heterocycles 2006, 68, 1121–1127. (c) Swartz, D. L.;
Odom, A. L. Organometallics 2006, 25, 6125–6133. (d) Novak, A.; Blake, A. J.;
Wilson, C.; Love, J. B. Chem. Commun. 2002, 2796–2797. (e) Love, J. B.;
Salyer, P. A.; Bailey, A. S.; Wilson, C.; Blake, A. J.; Davies, E. S.; Evans, D. J.
Chem. Commun. 2003, 1390–1391.
(8) King, E. R.; Betley, T. A. J. Am. Chem. Soc. 2009, 131, 14374–14380.
(9) (a) Wang, Q. M.; Bruce, D. W. Synlett 1995, 1267–1268. (b) Barnard,
T. S.; Mason, M. R. Inorg. Chem. 2001, 40, 5001–5009. (c) Reese, C. B.; Yan, H.
Tetrahedron Lett. 2001, 42, 5545–5547. (d) Beer, P. D.; Cheetham, A. G.; Drew,
M. G. B.; Fox, O. D.; Hayes, E. J.; Rolls, T. D. Dalton Trans. 2003, 4, 603–611.
(e) Betley, T. A.; Surendranath, Y.; Childress, M. V.; Alliger, G. E.; Fu, R.;
Cummins, C. C.; Nocera, D. G. Philos. Trans. Royal Soc. B 2008, 363, 1293–
1303.
(10) (a) Byrne, E. K.; Theopold, K. H. J. Am. Chem. Soc. 1987, 109, 1282–
1283. (b) Barner, C. J.; Collins, T. J.; Mapes, B. E.; Santarsiero, B. D. Inorg.
Chem. 1986, 25, 4322–4323. (c) Collins, T. J.; Kostka, K. L.; M€unck, E.;
Uffelman, E. S. J. Am. Chem. Soc. 1990, 112, 5637–5639. (d) Collins, T. J.; Fox,
B. G.; Hu, Z. G.; Kostka, K. L.; M€unck, E.; Rickard, C. E. F.; Wright, L. J. J. Am.
Chem. Soc. 1992, 114, 8724–8725. (e) Collins, T. J. Acc. Chem. Res. 1994, 27,
279–285.
˚
3.5 to 4.0 A separation between the pyridine centroid and
the mesityl centroid).11 The angle θ describes the orienta-
tion of the pyridine ligand with the (tpe)M fragment C3
axis, such that a θ of 180° would describe a perfect C3
ꢀ
ꢀ
(11) (a) McGaughey, G. B.; Gagne, M.; Rappe, A. K. J. Biol. Chem. 1998,
273, 15458–15463. (b) Meyer, E. A.; Castellano, R. K.; Diederich, F. Angew.
Chem., Int. Ed. 2003, 42, 1210–1250.

2514 Inorganic Chemistry, Vol. 49, No. 5, 2010
Sazama and Betley
Scheme 2
Table 1. X-Ray Diffraction Experimental Details for 2-6a,b
2
3
4
5
6
empirical formula
fw
space group
C
62H79LiMnN4O4
C62H83FeLiN4O4
1011.11
C2/c
19.4776(12)
15.5448(10)
38.314(2)
90
102.561(2)
90
11322.8(12)
8
0.316
49291
99.9% (27.48°)
1.012
0.0614, 0.1399
0.1350, 0.1694
C66H87CoLiN4O5
1082.27
P1
C62H79LiN4NiO4
1009.94
P21/n
13.0737(17)
22.273(3)
19.323(5)
90
93.16(3)
90
5618.1(18)
4
0.395
C66H87LiN4O5Zn
1088.71
P1
1006.17
C2/c
18.6178(9)
16.7236(8)
38.7477(18)
90
102.5460(10)
90
11776.3(10)
8
0.271
53543
99.8% (26.66°)
1.041
0.0534, 0.1508
0.0663, 0.1597
˚
a (A)
12.0728(7)
13.1443(7)
21.9752(12)
77.5480(10)
88.9730(10)
63.4090(10)
3032.6(3)
2
0.334
51355
100.0% (25.03°)
1.034
12.0972(8)
13.2008(8)
21.9749(14)
77.6430(10)
89.1500(10)
62.8850(10)
3037.3(3)
2
0.455
50961
99.7% (28.01°)
1.035
˚
b (A)
˚
c (A)
R (deg)
β (deg)
γ (deg)
3
˚
V (A )
Z
μ (mm-1
reflections
)
78916
99.6% (27.31°)
1.019
0.0506, 0.1228
0.0817, 0.1429
completeness (to θ)
GOF on F2
R1, wR2c [I > 2σ (I)]
Rindices (all) (R1, wR2)
0.0433, 0.1187
0.0496, 0.1253
0.0526, 0.1464
0.0634, 0.1579
ꢀ
ꢁ
P
1=2
2
½wðFo2 -Fc2Þ ꢀ
P
λ (A) = 0.71073. b Collection temperature (T) = 193(2) K for 2-6, 100(2) for 7. c R1 ¼ jjF j -jF jj, wR2 ¼
.
a
o
c
˚
P
P
2
jFoj
½wðFo2Þ ꢀ
Figure 1. Solid-state molecular structure of the anions for complexes 2-6 with the thermal ellipsoids set at the 50% probability level (hydrogen atoms,
solvated Li cations, and positional disorder in 5 are omitted for clarity).
symmetric orientation of the M-Npyridine bond axis. The
θ angles listed in Table 2 show a deviation from linearity
(θ = 157-168°). For complexes 2-6, the bond lengths
within the tpe ligand framework do not vary considerably

Article
Inorganic Chemistry, Vol. 49, No. 5, 2010 2515
˚
Table 2. Selected Bond Lengths (A) and Angles (deg) for Complexes 2-6
˚
d (A)
˚
x (A)
˚
y (A)
M
M-Na
M-Nb
M-Nc
θ (deg)
R (deg)
β (deg)
γ (deg)
jab
jac
jbc
Mn (2)
Fe (3)
Co (4)
Ni (5)
Zn (6)
2.070
2.015
1.973
1.960
1.988
2.078
1.999
1.981
1.955
1.981
2.073
2.027
1.975
1.989
1.984
2.996
2.918
2.863
2.935
2.876
2.147
2.068
2.016
2.040
2.009
4.040
3.739
3.617
3.528
3.623
168.13
161.43
168.06
157.95
168.24
114.18
107.67
109.17
105.74
109.62
134.84
138.65
127.75
115.20
123.97
120.77
117.07
123.67
143.96
127.73
92.10
94.38
96.27
94.33
95.90
93.48
95.52
96.20
94.76
95.86
90.93
94.29
96.67
91.87
96.43
Table 3. Magnetic and Spectral Properties of Complexes 2, 3, 4, and 5
complex
S
μ
eff (BM)a
λ/nm (ε/M-1 cm-1
)
δ (ΔEq) (mm/s)
[(tpe)Mn(py)] [Li(THF)4] (2)
[(tpe)Fe(py)] [Li(THF)4] (3)
[(tpe)Co(py)] [Li(THF)4] (4)
[(tpe)Ni(py)] [Li(THF)4] (5)
5/2
2
3/2
1
6.2(1)
5.1(1)
4.8(1)
2.9(1)
331 (11000), 502 (140), 477 (110)
317 (16000), 516 (1100), 560 (920)sh, 1636 (250), 1911 (180)sh
0.84 (2.54)
325 (12000), 550 (850), 621 (730), 650 (860), 998 (200), 1250 (90)sh
319 (13000), 398 (1500)sh, 504 (500)sh, 631 (570), 800 (120), 928 (60)
a Room temperature (295 K) moment in solution by Evans method.12
Figure 2. (a) UV/vis molar absorptivity spectra of 2, 3 in benzene, 3 in THF, 4, 5 in benzene, and 5 in THF and (b) NIR molar absorptivity spectra of 3, 4,
and 5 in THF. Spectra were taken in THF (unless otherwise noted); molar absorptivities are based on measurements at a minimum of four different
concentrations.
(Table S15, Supporting Information). The metal pyrro-
lide nitrogen bond lengths decrease across the series from
region (NIR). In benzene solutions, the high spin d6 (S =
2, μB = 5.1(1)) Fe complex 3 maintains its bright orange
hue (λmax/nm (ε/M-1 cm-1): 505 (1500), 477 (1600)) but
turns an intense magenta color upon dissolution in THF
(λmax/nm (ε/M-1 cm-1): 516 (1100), 560 (920)). This
significant λmax shift (55 nm) could be attributable to
solvent binding in THF, but we have no evidence for
accessing five-coordinate structures (in solution or the
˚
Mn f Ni, with Zn being slightly elongated (davg (A):
Mn-Ntpe = 2.074, Fe-Ntpe =2.014, Co-Ntpe = 1.976,
Ni-Ntpe=1.968, Zn-Ntpe=1.984; see Table 2 for a more
comprehensive comparison of bond lengths).
II.c. Electronic and Magnetic Characterization of Com-
plexes 2-5. The magnetic and spectral data for complexes
2-5 are provided in Table 3. All four complexes feature
an intense absorption in the UV region centered around
320 nm (ε/(M-1 cm-1): 11 000-16 000) which dominate
all other transitions observed. The pale yellow, 1H NMR
silent, high spin d5 (S = 5/2, μB = 6.2(1)) Mn complex 2
has only two very weak transitions in the visible region
(Figure 2a, λmax/nm (ε/M-1 cm-1): 502 (140), 477 (110)),
and no transitions were observed in the near-infrared
€
solid state). The Mossbauer spectrum of 3 (Figure 6a) is
best modeled as a single quadrupole doublet and para-
meters consistent with the FeII assignment (δ = 0.84 mm
s
-1, |ΔEQ| = 2.54 mm s-1) in a tetrahedral ligand field.
3
The high spin d7 (S = /2, μB = 4.8(1)), dark blue Co
complex 4 shows several transitions in the visible region
(λmax/nm (ε/M-1 cm-1): 550 (850), 621 (730), 650 (860)),
but no solvent dependence like complex 3. While not as

2516 Inorganic Chemistry, Vol. 49, No. 5, 2010
Sazama and Betley
Figure 3. (a) Cyclic voltammograms ofsolutionsinTHF of3 mM[(tpe)M(py)]- (M= Mn (2), Fe(3), Co(4), Ni(5), Zn(6); Bu4NPF6 (0.3 M); scan rate 25
mV/songlassy Celectrode). (b) Differentialpulsevoltammogramsforsolutions in THF of10-4 M [(tpe)M(py)]- (M=Mn(2), Fe(3), Co (4), Ni(5), Zn(6);
Bu4NPF6 (0.3 M); scan rate 25 mV/s on glassy C electrode).
Scheme 3
significant as 3, the spectral features of the high spin d8
(S = 1, μB = 2.9(1)), green Ni complex 5 sharpen con-
siderably in benzene solutions, but the gross spectral
features are consistent with THF solutions (λmax/nm (ε/
M-1 cm-1): 504 (500), 631 (570), 800 (120)). All of the
spectral features for complexes 2-5 are of the appropriate
magnitude for spin-allowed d-d transitions for the high
spin complexes in pseudotetrahedral ligand fields. The Fe
(3), Co (4), and Ni (5) all show very weak absorptions in
the NIR region (Figure 2b) with molar absorptivities (ε =
50-250 M-1 cm-1) substantially less than expected for
charge transfer bands (e.g., LMCT).
(Figure 3b) shows that the peak anodic current for the Mn
(2), Ni (5), and Zn (6) complexes is identical (-690 mV),
whereas the Fe (3, -770 mV) and Co (4, -660 mV)
complexes are slightly shifted. We interpret the redox
inactive ZnII data to imply the observed oxidation step is
ligand-centered, originating from the tpe framework.
Although the oxidations observed appear to be ligand-
centered and vary little with divalent cation substitution,
the onset of oxidation for metalated (tpe) shifts nearly 1.1 V
from the free (tpe)H3 (Ep = þ375 mV).
II.e. Chemical Oxidation Products. Chemical oxida-
tion of [(tpe)Zn(py)][Li(THF)4] (6) with FcþPF6 did
-
II.d. Electrochemical Behavior of Tris(pyrrolide)ethane
Complexes. Cyclic voltammetry was performed on the
monopyridine complexes, 2-6 (0.3 M Bu4NPF6, Figure 3).
Only the Fe complex 3 showed a quasi-reversible redox
wave centered at -750 mV (Δ(EC-EA) = 190 mV). Com-
plexes 2-6 nearly exhibit the same onset of oxidation in the
cyclic voltammetry experiments, with all of the complexes
having the same peak anodic current at -655 mV (vs Fc/
Fcþ). Differential pulse voltammetry measurements scanning
oxidatively over the potential range of -500 to -900 mV
not lead to tractable products when warmed to room
temperature (Fc = (η5-H5C5)2Fe). However, maintain-
ing the reaction temperature at -42 °C in an acetonitrile/
dry ice bath, we were able to observe the consumption of
1
the ferrocenium oxidant by H NMR and a detectable
oxidized species by EPR (Scheme 3). Consumption of the
oxidant is also apparent, as the pale yellow color of 6
quickly gives way to an intense dark brown color of the
oxidized product. The EPR spectrum for this oxidation
product is shown in Figure 4, which shows an isotropic
signal virtually unperturbed for an organic radical (g =
2.0023).
(12) (a) Evans, D. F. J. Chem. Soc. 1959, 2003–2005. (b) Sur, S. K. J. Magn.
Reson. 1989, 82, 169–173.

Article
Inorganic Chemistry, Vol. 49, No. 5, 2010 2517
Figure 5. Solid-state molecular structure of 7 with the thermal ellipsoids
set at the 50% probability level (hydrogen atoms are omitted for clarity).
€
complex 3. Mossbauer analysis of an authentic sample of
7 is provided in Figure 6c, and its isomer shift and
Figure 4. Toluene glass EPR of oxidation product of [(tpe)Zn-
-
(py)][Li(THF)4] þ FcþPF6 (dotted line is experimental data and red
quadrupole splitting (δ, |ΔEQ| (mm s-1): 0.81, 3.47)
3
line is data simulation: g = 2.0023, S = 1/2; 77 K, X-band, 9.3298 GHz;
fit with Gaussian and Lorentzian functions).
match the major product from the chemical oxidation
of 3 and are nearly identical for the related dipyrromethane
analogue (Mesdpma)FeII(py)2 (δ, |ΔEQ| (mm s-1): 0.80,
Encouraged bythe quasi-reversible nature of the electro-
chemical oxidation of the Fe complex 3, we canvassed the
reactivity of 3 with a range of chemical oxidants. Low
3
3.46).8 The secondary component apparent in the
€
Mossbauer spectrum of 7 (Figure 6c) is potentially due
temperature oxidation of 3 with FcþPF6- or NOþSbF6
-
to a geometric isomer of 7,13 similar to that observed in
the related dipyrromethane analogue (Mesdpma)FeII(py)2,
although the presence of two isomers could not be resolved
by 1H NMR spectroscopy.8 The presence of two isomers for
(frozen THF f room temperature) gave rise to a new
paramagnetic product(s) by 1H NMR. As with the oxida-
tion of 6, oxidation of 3 quickly changes from orange (in
arene solvents) to an intense brown color upon consump-
tion of the oxidant. This product mixture was unaffec-
ted by temperature (thawing THF, -42 °C, or room
temperature) or chemical oxidant employed during the
reaction. Following removal of volatiles in vacuo, extrac-
tion of the reaction mixture with diethyl ether and removal
of the ferrocene by washing with copious amounts of
hexanes separated the tpe-containing products. Storage
of this material in diethyl ether at -35 °C gave crystals of a
yellow brown material that were suitable for X-ray diffrac-
tion analysis. The molecular structure of this crystal is
shown in Figure 5. In the product, one of the pyrrolide
donors is not bound to the Fe, which now features two
pyridine molecules giving the formulation (κ2-tpeH)Fe(py)2
(7). The production of 7 could be maximized by running
the oxidation of 3 in a mixture of THF and pyridine as a
solvent (frozen THF f RT). Authentic samples of 7 were
prepared by deprotonation of (tpe)H3 (1) with 2 equiv of
LiN(SiMe3)2 in THF followed by the addition of FeCl2-
1
7 in solution is readily apparent in the H NMR spectra,
with two sets of proton resonances obvious in analytically
pure samples of 7, where the composition was ascertained
by combustion analysis. The spectrum shown in Figure 6c
€
was obtained quickly after preparation of 7. Mossbauer
analysis on samples of 7 where both isomers are present
1
in equal quantities by H NMR spectroscopy show both
sets of quadrupole doublets that comprise the spectrum
in Figure 6c.
II.f. Density Functional Theoretical Considerations. To
further probe the electronic structure of the [(tpe)M(py)]-
anions 2-6, geometry optimization calculations were
performed (M05-2X/6-31G).14,15 Use of the M05-2X
functional was necessary in order to reproduce the pyr-
idine/mesityl π-stacking effect observed in each of the
solid-state structures for complexes 2-6. Upon satisfac-
tory optimization of the molecular geometries, a single
point population analysis was performed to compute the
orbital energies for complexes 2-6 using standard func-
tionals and larger basis sets (B3LYP/M,N, TZV(P); all
other atoms, SV(P)).16-18 The frontier orbital energies for
complexes 2-6 are presented in Figure 7. The orbitals
1
(py)2. Comparing the H NMR spectrum of the crude
reaction mixture from the chemical oxidation of 3 with an
authentic sample of 7 confirms that 7 is the major species
formed in that product mixture. Furthermore, reaction of
insituprepared(tpe)Li3 withFeCl3 ina mixtureof pyridine
and THF produced large quantities of 7 as the dominant
product (Scheme 3).
(13) Long, G. J.; Whitney, D. L.; Kennedy, J. E. Inorg. Chem. 1971, 10,
1406–1410.
(14) (a) Zhao, Y.; Schultz, N. E.; Truhlar, D. G. J. Chem. Theory Comput.
2006, 2, 364–382. (b) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157–
167.
€
Mossbauer spectroscopy on the reaction product of 3
(15) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54,
724–728.
with chemical oxidants indicated that a mixture of species
was present (Figure 6b). The spectrum can be tentatively
modeled as a three-component mixture with one major
(16) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Vosko, S. H.;
Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200–1211. (c) Lee, C.; Yang, W.;
Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
species (δ, |ΔEQ| (mm s-1), component 1: 0.81, 3.27
3
€
(17) Schafer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829–
5835.
(18) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571–
2577.
(56.9%); component 2: 0.71, 2.21 (33.5%); component
3: 1.02, 1.51 (9.6%)). The isomer shifts and quadrupole
splitting parameters are similar in nature to those for
€

2518 Inorganic Chemistry, Vol. 49, No. 5, 2010
Sazama and Betley
Figure 6. Mossbauer spectrum of (a) [(tpe)Fe(py)][Li(THF)4] (3) at 77 K (δ, |ΔEQ| (mm s-1): 0.84, 2.54); (b) oxidation product of 3 with FcþPF6-, at 77 K
(δ, |ΔEQ| (mm s-1), component 1: 0.81, 3.27 (56.9%); component 2: 0.71, 2.21 (33.5%); component 3: 1.02, 1.51 (9.6%)); (c) (κ2-tpe)Fe(py)2 (7) at 77 K
(δ, |ΔEQ| (mm s-1), component 1: 0.81, 3.47 (80.6%); component 2: 0.94, 2.43 (19.4%)).
€
Figure 7. Calculated R (2) and β (1) orbital energies for [(tpe)M(py)]- complexes: Mn (2, 5 R), Fe (3, 5 R, 1 β), Co (4, 5 R, 2 β), and Ni (5, 5 R, 3 β).
Calculated restricted molecular orbitals for [(tpe)Zn(py)]- (6). Metal-based orbitals are gray, tris(pyrrolyl)ethane ligand-based orbitals are red (6 R, 6 β
for 2-5), and pyridine-based orbitals are black.
originating from the tpe ligand are shown in red, metal 3d
orbitals in gray, and pyridine orbitals in black. In every
instance, the Mulliken population analysis (MPA) calcu-
lations indicate the highest lying occupied molecular
orbitals (HOMO, HOMO-1) exclusively feature ligand
π-character on the pyrrole subunits, illustrated for com-
plex 3 in Figure 8. Across the series, the metal 3d orbitals
do not significantly contribute to the frontier orbitals
until HOMO-4, approximately 1.5 eV lower in energy
than the ligand-based HOMO. Only the Fe complex 3
shows a significant component of metal 3d orbital con-
tribution to the frontier orbitals in the β population. The
spin density plot (R-β) for the Fe complex (Figure 9d)
shows that all of the unpaired spin is localized on the
metal ion, which is consistent for all the open-shell
molecules examined (2, 4, and 5; Figure S2-S4, Support-
ing Information).
(S =1/2, 8) and (tpe)Fe(py) (S = 5/2, 9) were optimized
using the optimized structure data of 6 and 3 as starting
points, respectively. The geometry optimization of (tpe)-
3
Fe(py) with S = /2 was also examined but failed to
converge using standard convergence criteria. The total
energy of both optimized structures for the oxidation
products was higher in energy (8, þ2.531 eV; 9, þ2.465
eV) than the geometry minimized structures for the
corresponding anions. The calculated R and β orbital
energies presented in Figure 9b and e were offset by these
energy differences to correct for the overall structural
zero-point energy differences. In both instances, the new
HOMO-LUMO gaps (1.2 eV for oxidized Zn and Fe
products) were substantially lower than those calculated
for 6 (1.9 eV) or 3 (2.2 eV). The calculated spin densities
(R-β) for the oxidation products of 6 and 3 are shown in
Figure 9c and f, respectively. The resultant radical density
for the S = 1/2 (tpe)Zn(py) is contained entirely within the
tpe framework, consistent with organic-radical-like EPR
spectrum (Figure 4) observed for the oxidation product.
We also probed the structure of the oxidation products
for the Zn and Fe anions by density functional methods.
The geometries for the oxidized complexes (tpe)Zn(py)

Article
Inorganic Chemistry, Vol. 49, No. 5, 2010 2519
Figure 8. HOMO (R and β) and LUMO (R and β) from frontier orbital diagrams for complex 3 (Fe, S = 2) by DFT (B3LYP/TZV(P), SV(P); Gaussian
03).
Figure 9. Calculated spin population (R-β) for (a) [(tpe)Zn(py)]- (S = 0), (c) (tpe)Zn(py) (S = 1/2), (d) [(tpe)Fe(py)]- (S = 2), and (f) (tpe)Fe(py) (S =
5/2). (b) Molecularorbitalsfor[(tpe)Zn(py)]- andcalculatedR (2) andβ (1) orbitalenergiesfor the oxidationproduct(tpe)Zn(py) (S=1/2). (e)CalculatedR
(2) and β (1) orbital energies for [(tpe)Fe(py)]- (S = 2) and the oxidation product (tpe)Fe(py) (S = 5/2) by DFT (B3LYP/TZV(P), SV(P); Gaussian 03).
Similarly, the calculated spin density for the oxidation
product (tpe)Fe(py), assuming an S = /2 ground state,
now shows considerable ligand contribution (Figure 9f),
whereas the spin is localized completely on the Fe ion for 3
(Figure 9d).
“tetrahedral enforcer” ligands (e.g., tris(RL)borates19-21
)
5
but differs in that it coordinates transition-metals as a
trianionic species. The Ntpe-M bond lengths decrease
across the series from Mn f Ni, which is consistent with
the decreasing high-spin ionic radius for the series. The
pyridine ligand is not bound along the 3-fold axis of
symmetry (i.e., axis depicted “d” in the diagram for
Table 2) for any of the molecules reported, even for the
III. Discussion
III.a. [(tpe)M(py)]- Structural Considerations. The tpe
ligand framework is a rigid, tripodal framework that
sterically enforces a pseudotetrahedral coordination
environment about the bound transition metal ion. Ac-
cordingly, this class of complexes is related to other
(19) (a) Trofimenko, S. J. Am. Chem. Soc. 1966, 88, 1842–1844. (b)
Trofimenko, S. Chem. Rev. 1993, 93, 943–980.
(20) Byers, P. K.; Canty, A. J.; Honeyman, R. T. Adv. Organomet. Chem.
1992, 34, 1–65.

2520 Inorganic Chemistry, Vol. 49, No. 5, 2010
Sazama and Betley
d5 Mn (2) and d10 Zn (6) complexes where steric minimiza-
tion should dominate bonding. Instead, the pyridine is
bound well off-axis, positioned in a π-stacking orientation
with one of the ligand mesityl fragments, resulting in a
trigonal monopyramidal geometry instead of a tetrahedral
arrangement. The pyridine-mesityl separation (3.5-4.0
Figure S1, Supporting Information). Furthermore, calcu-
lations on the oxidation product of 6 ((tpe)Zn(py), 8)
provide an estimate of the energetic cost in excess of 2 eV
(ΔE = Etot(6) - Etot(8)) toward oxidation of the tpe
framework, suggesting an intramolecular redox transfer
should be energetically disfavored.
Scenario 2. Each of the complexes from Mn (2) to Ni (5)
feature high-spin configurations which are evident by the
low temperature X-ray diffraction studies (193 K), low
˚
A) is within π-stacking distances reported in the literature
and is estimated to provide an additional 1-2 kcal/mol of
energy to the stabilization of the ground state.11 While the
pyridine binding off-axis could maximize the π overlap
between the metal d orbitals (i.e., 2e set for a ML3
fragment) and the pyridine π-accepting orbitals,22 the π
back-bonding contribution to the M-pyridine binding
should be quite minimal for the high-spin, 3d M2þ ions
examined, and therefore the π-stacking interaction is the
most likely cause for deviation from the expected pseudo-
C3v geometry.
III.b. Electronic Structure. In our studies of the redox
properties of the related dipyrromethane metal com-
plexes of the type (Mesdpma)M(py)2, we observed match-
ing of the ligand-based orbitals with the frontier metal
orbitals similar to the tpe system presented here.8 We
considered three possible scenarios regarding potential
intramolecular redox for those systems which are relevant
to consider when evaluating the electronic structure of the
current tpe frameworks: (1) high-energy pyrrole-based
electrons reduce the bound metal ion with n e-, leaving n
holes on the ligand platform (i.e., (L0)M2þ f (Lnþ)M2-n);
(2) the ligand and metal 3d orbitals are close enough in
energy such that spin-state tautomerism is possible;23 (3)
no intramolecular redox occurs, leaving each metal in the
divalent state and the ligand in a fully reduced, closed-
shell configuration.
€
temperature Mossbauer for 3 and 7, and the room
temperature solution magnetic moment determinations
for complexes 2-5 (Table 3). This is especially notable for
M = Ni (5), which in the dipyrromethane system is found
to be square planar and diamagnetic.8 The DFT spin
density analyses of 2-5 suggest that all spin is localized on
the corresponding 3d ion without significant contribution
from the ligand (Figure 9; Figures S2-4, Supporting
Information). These data imply that no spin-state tauto-
merism is occurring. If the oxidation of the tpe frame-
work is as energetically disfavored as estimated by DFT,
this suggests a considerable barrier for spin-state tauto-
merism.
Scenario 3. On the basis of the evidence we have
collected, we have determined that metal complexes of
the tpe trianion fall into scenario 3, where the available
redox reservoirs (i.e., tpe, 3d metal ion) are well separated
and do not show any evidence for intramolecular redox
transfer in the outer-sphere electron transfer reactions, as
also observed in the dipyrromethane system we pre-
viously reported.8 This feature is most obvious in the
electrochemical behavior of complexes 2-6, both in the
cyclic and differential pulse voltammetry experiments
(Figure 3), which suggest that the [(tpe)M(py)]- species
traverse a similar oxidation pathway regardless of the
metal ion present or its spin-multiplicity. The common-
ality of the electrochemical oxidations, EPR on the
oxidation product of 6, and the DFT results indicate the
kinetic products of the one-electron oxidation reactions
are ligand-centered radicals, rather than metal-centered
oxidations (eq 1):
Scenario 1. If scenario 1 was operative, we would
anticipate a stepwise elongation of the pyrrole subunits
within the tpe framework across the 3d series from
oxidation of the ligand framework. No ligand metrical
perturbations to suggest this are apparent when compar-
ing the structural data for complexes 2-6 (Table S15,
Supporting Information). In addition, the UV-vis ab-
sorption spectra do not reveal any readily assignable
ligand-to-metal charge-transfer bands that would suggest
the presence of excited states consistent with this scenario.
The only absorption band of appropriate intensity (λ ≈
320 nm, ε > 104 M-1 cm-1) corresponds to a ligand-to-
ligand charge transfer band that is present in the deproto-
-e-
H-atom abstractionf
-
½ðtpeÞMIIðpyÞꢀ f ½ðtpe•þÞMIIðpyÞꢀ
þ L
½ðK2-tpeÞMIIðpyÞðLÞꢀ
ð1Þ
Both cyclic and differential pulse voltammetry exhibit
multiple oxidation events following the primary oxida-
tion highlighted in Figure 3. Unlike the dipyrromethane
analogues previously reported, we do not observe the
multielectron redox events attributable to a one-electron
oxidation of each of the pyrrolide subunits. The primary,
irreversible oxidation of the [(tpe)M(py)]- anions pro-
duces a dipyrromethane-like product (κ2-tpe)M(py)(L)
following H-atom abstraction from the (tpe•þ)M(py)
intermediate. The source of H atoms is likely THF, either
bound to lithium or in the solvent. Experiments with
dihydroanthracene as a sacrificial H-atom donor showed
no anthracene formation. The onset of oxidation for
the dipyrromethane complexes of the formulation
natedligand (tpe)Li3 (λmax = 313 nm, ε = 9000 M-1 cm-1
;
(21) (a) Tris(alkylthiol)borate: Ge, P.; Haggerty, B. S.; Rheingold, A. L.;
Riordan, C. G. J. Am. Chem. Soc. 1994, 116, 8406-8407. (b) Schebler, P. J.;
Mandimutsira, B. S.; Riordan, C. G.; Liable-Sands, L. M.; Incarvito, C. D.;
Rheingold, A. L. J. Am. Chem. Soc. 2001, 123, 331–332. (c) Tris(alkyl-
phosphino)borates: Peters, J. C.; Feldman, J. D.; Tilley, T. D. J. Am. Chem. Soc.
1999, 121, 9871-9872. (d) Barney, A. A.; Heyduk, A. F.; Nocera, D. G. Chem.
Commun. 1999, 2379–2380. (e) Betley, T. A.; Peters, J. C. Inorg. Chem. 2003,
42, 5074–5084. (f) Thomas, C. M.; Mankad, N. P.; Peters, J. C.
J. Am. Chem. Soc. 2006, 128, 4956-4957. Tris(carbeno)borates:(g) Nieto, I.;
Cervantes-Lee, F.; Smith, J. M. Chem. Commun. 2005, 3811–3813. (h) Cowley,
R. E.; Bontchev, R. P.; Duesler, E. N.; Smith, J. M. Inorg. Chem. 2006, 45, 9771–
9780.
(
Mesdpma)M(py)2 does not occur until -240 mV (vs Fc/
(22) (a) Albright, T. A.; Burdett, J. K.; Whangbo, M.-H. Orbital Inter-
actions in Chemistry; John Wiley & Sons, Inc: New York, 1985; pp 381-389.
Fcþ).8 Thus, the secondary redox event for the
[(tpe)M(py)]- anions is not anticipated to occur until a
similar potential, separated from the initial oxidation by
nearly 450 mV.
ꢁ
(b) Detrich, J. L.; Konecnꢀy , R.; Vetter, W. M.; Doren, D.; Rheingold, A. L.;
Theopold, K. H. J. Am. Chem. Soc. 1996, 118, 1703–1712.
(23) Storr, T.; Verma, P.; Pratt, R. C.; Wasinger, E. C.; Shimazaki, Y.;
Stack, T. D. P. J. Am. Chem. Soc. 2008, 130, 15448–15459.

Article
Chemical oxidation of both 3 and 6 initially produces
Inorganic Chemistry, Vol. 49, No. 5, 2010 2521
tris(pyrrolyl)amine systems.26 Work is currently under-
way to determine the ability of [(tpe)MLn]- anions to
function as atom or group transfer accepting platforms
via metal-centered redox chemistry without being short-
circuited by intramolecular redox transfer.
highly colored intermediates consistent with the de-
creased HOMO-LUMO gap predicted for the oxidized
products by DFT. The rapidity with which the unshielded
tpe ligand radical abstracts H• from solvent to produce
the (κ2-tpe)M(py)(L) species (L = pyridine or solvent)
prevented isolation of the putative (tpe•þ)MII(py) oxida-
tion product. The resulting coordination change at the
metal accounts for the irreversibility of the cyclic voltam-
metry experiments. Attempted metalation of the tpe
trianion with FeIII sources leads to formation of the same
κ2 species in 7. While we cannot definitively rule out
intramolecular electron transfer from the tpe to FeIII
following ligation, the tpe trianion is a strong enough
reductant to reduce the FeIII precursor, and this process
likely precedes halide metathesis. Extending our analysis
of these observations to the dipyrromethane system, the
ligand-centered oxidation processes we observed likely
produced pyrrole-dissociated species, making identifica-
tion of chemically oxidized materials impossible.8
IV. Conclusions
We have investigated transition metal complexes of the
tris(pyrollyl)ethane ligand and found them to exhibit unusual
electronic properties attributable to the pyrrolide ligand
framework. Electrochemical investigations on the [(tpe)-
MII(py)]- series reveal a common irreversible oxidation
pathway that is entirely ligand-based, invariant to the diva-
lent metal bound. This observation indicates that fully
populated, ligand-based π orbitals from the tpe construct
are energetically most accessible, akin to their dipyrro-
methane analogues. DFT calculations on the anions suggest
electrons pair preferentially in the higher-energy, delocalized
ligand π orbitals rather than pairing within the lower-energy
metal-ion 3d orbitals (e.g., in the event of an intramolecular
redox transfer). This can be attributed to larger electron-
pairing energy costs for the metal-based 3d orbitals. While
the energetic overlap between the tpe ligand-based π orbitals
and the d-orbital manifold is favorable for cooperative redox,
the two redox reservoirs display totally disparate redox
behavior. A major criterion for coupling ligand and transi-
tion metal redox is to have a pathway by which the two
reservoirs can communicate, typically through conjugation
of the ligand π orbital with d orbitals of appropriate sym-
metry. DFT studies reveal the tpe platform, and by extension
the dipyrromethane ligand systems, do not feature frontier
π-orbital density on the ligand N termini, shielding the
ligand-based, highest occupied molecular orbitals from con-
jugation with the metal 3d orbitals. As a result, the ligand
redox behavior observed is devoid of transition metal influ-
ence. In the context of outer-sphere electron transfer chem-
istry, the pyrrolic π electrons become a liability to the
complex stability. These ligand platforms, however, could
find significant utility in the stabilization of inner-sphere
electron transfer reactivity and the resulting high-oxidation
state metal complexes if L f M π contribution can be
maximized.
The orbital analysis (Figure 7) for the tpe complexes
shows that the energetically most accessible orbital den-
sity is localized on the tpe scaffold. The frontier orbitals
consist entirely of the pyrrolide π system. This produces
an apparent non-Aufbau arrangement of electrons,
where fully populated ligand orbitals are higher in energy
than the partially filled metal 3d orbitals. We interpret
this to suggest that electron pairing in the delocalized,
ligand-based π orbitals incurs less Coulombic repulsion
than pairing electrons in the comparatively smaller 3d
orbitals. However, the corresponding 3d β orbitals may,
in fact, lie higher in energy than the ligand-based orbitals,
which would not violate the Aufbau principle.24 From a
qualitative standpoint, the electronic structures predicted
by DFT corroborate our experimental results, wherein
outer-sphere electron transfer is dominated by the ligand
π electrons.
As shown by Floriani and Floriani-Moro4a and Swartz
and Odom,7c more electrophilic, early transition metal
complexes bind pyrrolide subunits through an η5
π
association preferentially to terminal N ligation, even in
porphyrinogen complexes. For iron-porphyrinogens,4c
these same π electrons cause an intimate coupling bet-
ween the metal and ligand oxidation states when orbitals
of appropriate π symmetry are allowed to mix but become
a liability due to their energetic accessibility when this
conjugation is broken.8 For dipyrromethane and tris-
(pyrrolyl)ethane systems, the highest-lying ligand orbitals
do not feature any π density on the pyrrole N, making the
ligands poor candidates to function as π donors to metals
that could function as π acceptors. This property, coupled
with the high-spin, pseudotetrahedral nature of the di-
pyrromethane and tris(pyrrolyl)ethane complexes, func-
tions to make the pyrrolide and 3d ions completely
decoupled redox entities. This does not rule out the two
redox reservoirs engaging in cooperative redox transfer in
oxidations involving two or more electrons, a process
which plays a key role in stabilizing the high-valent ferryl
species in cytochrome P45025 and potentially in related
V. Experimental Section
All manipulations were carried out in the absence of water
and dioxygen using standard Schlenk techniques, or in an
MBraun inert atmosphere drybox under a dinitrogen atmo-
sphere. All glassware was oven-dried for a minimum of 1 h
and cooled in an evacuated antechamber prior to use in the
drybox. Benzene, diethyl ether, dichloromethane, n-hexane,
tetrahydrofuran, and toluene were dried and deoxygenated
on a Glass Contour System (SG Water USA, Nashua, NH)
˚
and stored over 4 A molecular sieves (Strem) prior to use.
Benzene-d6 was purchased from Cambridge Isotope Labora-
˚
toriesanddegassedandstoredover4 Amolecularsievesprior
to use. Pyridinium p-toluenesulfonate, ferrocenium hexa-
fluorophosphate, dihydroanthracene, and anhydrous pyri-
dine were purchased from Aldrich and used as received.
Anhydrous manganese(II), iron(II), cobalt(II), and nickel(II)
chlorides were purchased from Strem and used as received.
Phenyllithium was prepared by lithiation of iodobenzene
(24) Wang, L.-P.; Van Voorhis, T. Personal communication.
(25) Shaik, S.; Kumar, D.; De Visser, S. P.; Altun, A.; Thiel, W. Chem.
Rev. 2005, 105, 2279-2328 and references therein.
(26) Harman, W. H.; Chang, C. J. J. Am. Chem. Soc. 2007, 129, 15128–
15129.

2522 Inorganic Chemistry, Vol. 49, No. 5, 2010
Sazama and Betley
with n-butyllithium. 2-(2,4,6-Trimethylphenyl)-1H-pyrrole
was synthesized according to literature procedures.27 Metal
precursors MCl2(py)2 (M = Mn, Fe, Co, Ni) were synthe-
sized on the basis of literature procedures.13,28 Celite 545
(J. T. Baker), silica gel 32-63 μ (Dynamic Adsorbents,
placed at idealized positions and refined using a riding model.
The isotropic displacement parameters of all hydrogen atoms
were fixed to 1.2 times the U value of the atoms they are linked to
(1.5 times for methyl groups). In all structures with a Liþ
countercation, the ligated solvent THF and ether often exhib-
ited positional disorder, and similarity restraints were used to
refine the model. Except in certain cases (as noted below), the
major component of the disordered THF refined to between
54% and 95%. Further details on several structures are noted
below. Note on 2: Similarity restraints and distance and angle
restraints were insufficient for refinement of residual electron
density, most likely a heavily disordered free THF molecule,
found in the crystal structure. The SQUEEZE function of
PLATON32 was used to remove the electron density. Details
can be found in the .cif file for the structure of 2. Note on 5: The
pyridine bound to nickel exhibited positional disorder, which
was modeled using similarity and flat restraints. The major
component refined to 90%. Notes on 6: One THF-ligated to
Liþ was modeled using three positional disorders, with compo-
nents of 40%, 37%, and 23%, again using similarity restraints to
aid in refinement. In addition, a free THF molecule was
disordered over two positions and refined with the aid of
similarity restraints, with the major contributor refining to
55%. Note on 7: The hydrogen atom bound to the nitrogen of
the free pyrrole arm was explicitly located as a Q peak in the
refinement and refined semifreely using restraints.
˚
Atlanta, GA), 4 A molecular sieves, tetra-n-butylammonium
hexafluorophosphate (Alfa Aesar), and zinc chloride
(Aldrich) were dried in a Schlenk flask for 24 h under a
dynamic vacuum while heating to at least 150 °C.
Characterization and Physical Measurements. UV/visible
spectra were recorded on a Varian Cary 50 UV/Visible spectra,
with a scan rate of 300 nm/min. NIR spectra were recorded on a
Varian Cary model 5000 UV-vis-NIR spectrophotometer. 1H
and 13C NMR spectra were recorded on Varian Mercury 400
1
MHz or Varian Unity/Inova 500 MHz spectrometers. H and
13C NMR chemical shifts are reported relative to SiMe4 using
the chemical shift of residual solvent peaks as a reference.
Solution magnetic susceptibilities were determined by Evans’s
method12 using hexamethyldisiloxane as an internal reference.
Mass spectrometry was performed at the Harvard University
FAS Center for Systems Biology Mass Spectrometry and
Proteomics Resource Laboratory on an Agilent 6210 TOF
LC/MS with a dual nebulizer ESI source. Elemental analyses
were carried out at Complete Analysis Laboratories, Inc.
(Parsippany, NJ).
Electrochemical experiments were carried out using a CH
Instruments CHI660C Electrochemical Workstation. The elec-
trolyte used was 0.3 M (nBu4N)(PF6) in THF. The concentration
of all analytes was 3 mM. The working electrode was glassy
carbon. A platinum wire was used as the counter electrode, and
nonaqueous Ag/Agþ reference electrodes (10 mM AgNO3 in
THF) were used with the corresponding electrolyte solution.
Cyclic voltammetry was performed with a scan rate between
500 mV/s and 25 mV/s.
5,50,500-(Ethane-1,1,1-triyl)tris(2-mesityl-1H-pyrrole) (tpeH3)
(1). A 100 mL bomb flask was charged with 2-(2,4,6-
trimethylphenyl)-1H-pyrrole (3.468 g, 18.72 mmol), trimethyl-
orthoacetate (1.194 g, 9.94 mmol), and 20 mL of CH2Cl2. After
stirring for 5 min, pyridinium p-toluenesulfonate (0.170 g, 0.675
mmol) was added, resulting in a color change from tan to
magenta. The reaction was stirred at 55 °C for 41 h. Additional
pyridinium p-toluenesulfonate (0.168 g, 0.668 mmol) was added,
and the reaction was stirred at 50 °C for an additional 72 h. The
reaction mixture was then filtered through a silica gel plug,
washing with excess CH2Cl2 (∼50 mL). Solvent and volatiles
were removed in vacuo, and the resulting green-brown, sticky
solid was redissolved in minimal CH2Cl2. Hexanes were added
until a precipitate was observed, and the solution was placed
into the freezer for 2 h. The precipitate was collected on a glass
frit. The recrystallization procedure was repeated twice, afford-
X-band EPR spectroscopic measurements were performed
using a Bruker EMX spectrometer equipped with 13 in. mag-
nets, an ER 4102ST cavity, and a Gunn diode microwave
source. Samples in a toluene glass were cooled using a liquid
nitrogen coldfinger dewar. Simulation and fitting of EPR data
was done with the EasySpin 3.0.029 ToolBox for MATLAB
R2008a.30 57Fe Mossbauer spectra were measured with a con-
€
stant acceleration spectrometer (SEE Co, Minneapolis, MN).
Isomer shifts are quoted relative to Fe metal at room tempera-
ture. Data were analyzed and simulated with Igor Pro 6 software
(WaveMetrics, Portland, OR) using Lorentzian fitting func-
tions.
1
ing 1 as an off-white powder (2.5204 g, 70%). H NMR (300
MHz, C6D6): δ/ppm 7.33 (br s, 3 H), 6.82 (s, 6 H), 6.17 (t, J = 3.2
Hz, 3 H), 6.05 (t, J = 2.9 Hz, 3 H), 2.19 (s, 9 H), 2.16 (s, 16 H),
1.81 (s, 3 H). 13C {1H} NMR (125 MHz, C6D6): δ/ppm 138.8,
137.7, 136.7, 131.8, 129.9, 128.8, 108.5, 107.1, 41.3, 28.5, 21.5,
X-Ray Crystallography Procedures. All structures were col-
lected on either a Siemens or Bruker three-circle platform
goniometer equipped with either a Bruker Apex I or Apex II
CCD and an Oxford cryostream cooling device. Radiation was
21.2. HRMS (ESIþ) m/z for C41H46N3 [M þ H]þ, calcd
þ
580.36915. Found: 580.36933.
[(tpe)Mn(py)][Li(THF)4] (2). In separate 20 mL scintilla-
tion vials, 1 (448 mg, 0.772 mmol) and PhLi (202 mg, 2.413
mmol) were each dissolved in 3 mL of THF and cooled in
a -35 °C freezer. The cold solution of 1 was added to the PhLi
solution and stirred at room temperature for 30 min. In another
20 mL vial, MnCl2(py)2 (220 mg, 0.775 mmol) was slurried in
3 mL of THF with ∼0.5 mL of pyridine. The vials containing
deprotonated 1 and MnCl2(py)2 were placed in a liquid nitrogen
cooled cold well until frozen. The solution of 1 was thawed and
added to the stirring slurry of MnCl2(py)2. The reaction mixture
became a yellow solution as it warmed. The reaction was stirred
at room temperature overnight. Removal of the solvent in vacuo
gave a dark yellow residue. The residue was extracted into 5 mL
of benzene and filtered through a Celite plug to remove lithium
chloride. Removal of the solvent in vacuo gave a yellow powder
(522.4 mg, 67%). Crystals suitable for X-ray diffraction were
grown from a vapor diffusion of Et2O into a THF solution of 2
˚
from a graphite fine focus sealed tube Mo KR (0.71073 A)
source. Crystals were mounted on a cryoloop using Paratone-N
oil. Structures were collected at 193 K (2-6) and 100 K (7). Data
were collected as a series of j and/or ω scans. Data were
integrated using SAINT (Bruker AXS) and scaled with a multi-
scan absorption correction using SADABS (Bruker AXS). The
structures were solved by direct methods or Patterson maps
using SHELXS-9731 and refined against F2 on all data by full
matrix least-squares with SHELXL-97. All non-hydrogen
atoms were refined anisotropically. Hydrogen atoms were
(27) Reith, R. D.; Mankad, N. P.; Calimano, E.; Sadighi, J. P. Org. Lett.
2004, 6, 3981–3983.
(28) Allan, J. R.; Brown, D. H.; Nuttall, R. H.; Sharp, D. W. A. J. Chem.
Soc. A 1966, 1031–1034.
(29) Stoll, S.; Schweiger, A. J. Magn. Reson. 2006, 178, 42–55.
(30) MATLAB R2008a; The MathWorks: Natick, MA, 2008.
(31) (a) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467–473. (b)
€
€
Sheldrick, G. M. SHELX-97; University of Gottingen: Gottingen, Germany,
1997.
(32) The Platon Homepage. http://web.mit.edu/platon_v40505/platon/
docs/platon/pl000000.html (accessed Jan 2010).

Article
Inorganic Chemistry, Vol. 49, No. 5, 2010 2523
at room temperature. The compound gave no signal by 1H
NMR. UV/vis (THF) λmax/nm (ε/M-1 cm-1): 331 (11 000), 502
(140), 477 (110). μeff (C6D6, 295 K): 6.2(1) μB. Comb. Anal. for
[C46H47MnN4]-[LiC16H32O4]þ, Calcd: C, 74.01; H, 7.91; N,
5.57. Found: C, 73.93; H, 7.74; N, 5.46.
benzene and filtered through a Celite plug to remove lithium
chloride. Removal of the solvent in vacuo gave a gray powder
(37.0 mg, 76%). Crystals suitable for X-ray diffraction were
grown from a vapor diffusion of hexanes into a THF solution of
5 at room temperature. 1H NMR (500 MHz, C6D6): δ/ppm
118.62 (br s), 109.37 (s), 43.39 (s), 27.73 (s), 13.97 (s), 6.27 (s),
[(tpe)Fe(py)][Li(THF)4] (3). In separate 20 mL scintillation
vials, 1 (246.2 mg, 0.425 mmol) and PhLi (114 mg, 1.36 mmol)
were each dissolved in 3 mL of THF and cooled in a -35 °C
freezer. The cold solution of 1 was added to the PhLi solution
and stirred at room temperature for 30 min. In another 20 mL
vial, FeCl2(py)2 (121.3 mg, 0.426 mmol) was slurried in 3 mL of
THF. The vials containing deprotonated 1 and FeCl2(py)2 were
placed in a liquid nitrogen cooled cold well until frozen. The
solution of 1 was thawed and added to the stirring slurry of
FeCl2(py)2. The reaction mixture became a dark red solution as
it warmed. The reaction was stirred at room temperature over-
night. Removal of the solvent in vacuo gave an orange residue,
which was extracted with Et2O. The Et2O solution was filtered
over Celite, approximately 20 drops of THF were added, and a
precipitate formed. The bright orange solid 3 was collected on a
glass frit (0.445 g, 103% yield due to remaining Et2O). Crystals
suitable for X-ray diffraction were grown from a vapor diffusion
of Et2O into a THF solution of 3 at room temperature. 1H NMR
(500 MHz, C6D6): δ/ppm 151.71 (s), 77.68 (s), 71.94 (s), 30.47
(s), 15.19 (s), 9.76 (s), 4.16 (s), 1.65 (s), 1.07 (s), 0.32 (s), -6.72 (br
s). UV/vis (THF) λmax/nm (ε/M-1 cm-1): 317 (16 000), 516
(1100), 560 (920, sh), 1636 (250), 1911 (180, sh). μeff (C6D6,
295 K): 5.1(1) μB. Comb. Anal. for [C46H47FeN4]-[LiC16-
H32O4]þ, Calcd: C, 73.94; H, 7.91; N, 5.55. Found: C, 73.80;
H, 7.81; N, 5.79.
5.76 (s), 3.24 (s), 1.78 (s), 1.27 (s), -1.36 (s). UV/vis (THF) λmax
/
nm (ε/M-1 cm-1): 319 (13 000), 398 (1500, sh), 504 (500, sh), 631
(570), 800 (120), 928 (60). μeff (C6D6, 295 K): 2.91(5) μB. Comb.
Anal. for [C46H47NiN4]-[LiC16H32O4]þ, Calcd: C, 73.73; H,
7.88; N, 5.55. Found: C, 73.67; H, 7.72; N, 5.65.
[(tpe)Zn(py)][Li(THF)4] (6). In separate 20 mL scintillation
vials, 1 (0.288 g, 0.497 mmol) and PhLi (130 mg, 1.54 mmol)
were each dissolved in 3 mL of THF and cooled in a -35 °C
freezer. The cold solution of 1 was added to the PhLi solution
and stirred at room temperature for 30 min. In another 20 mL
vial, ZnCl2 (70 mg, 0.51 mmol) was slurried in 3 mL of THF with
∼0.5 mL of pyridine. The vials containing deprotonated 1 and
ZnCl2 were placed in a liquid nitrogen cooled cold well until
frozen. The solution of 1 was thawed and added to the stirring
slurry of ZnCl2. The reaction mixture became a bright yellow
solution as it warmed. The reaction was stirred at room tem-
perature for 3 h. Removal of the solvent in vacuo gave a dark
yellow residue. The residue was extracted into 5 mL of benzene
and filtered through a Celite plug to remove lithium chloride and
lyophilized. The resultant bright yellow powder was dissolved in
Et2O, 10 drops of THF were added, and the solution was cooled
to -35 °C. A yellow precipitate, 6, formed and was collected on a
glass frit (0.234 g). The filtrate was dried in vacuo, dissolved in
Et2O, and THF was added. A second crop of 6 was collected
(32 mg; total yield 266 mg, 55%). Crystals suitable for X-ray
diffraction were grown from a solution of 6 in a 1:1 mixture of
Et2O and THF solution with 10 drops of TMEDA at -35 °C. 1H
NMR (500 MHz, C6D6): δ/ppm 6.70-6.68 (m, 2H, o-C5H5N),
6.61-6.58 (m, 4H, p-C5H5N, pyrrole C-H), 6.53 (s, 6H, -C6H2-
(CH3)3), 6.24 (br s, 3H, pyrrole C-H), 6.17-6.14 (m, 2H,
m-C5H5N), 3.46 (m, 12H, Li(OCH2(CH2)2CH2)4), 2.67 (s, 3H,
(CH3)C(C5H2MesN)3), 2.15 (s, 18H, o-C6H2(CH3)3), 2.02 (s,
[(tpe)Co(py)][Li(THF)4] (4). In separate 20 mL scintillation
vials, 1 (482 mg, 0.831 mmol) and PhLi (220 mg, 2.61 mmol)
were each dissolved in 3 mL of THF and cooled in a -35 °C
freezer. The cold solution of 1 was added to the PhLi solution
and stirred at room temperature for 30 min. In another 20 mL
vial, CoCl2(py)2 (241 mg, 0.835 mmol) was slurried in 3 mL of
THF. The vials containing deprotonated 1 and CoCl2(py)2 were
placed in a liquid nitrogen cooled cold well until frozen. The
solution of 1 was thawed and added to the stirring slurry of
CoCl2(py)2. The reaction mixture became a very dark purple-
blue solution as it warmed. The reaction was stirred at room
temperature for 3 h. Removal of the solvent in vacuo gave a
black-blue residue, which was dissolved in benzene. The benzene
solution was filtered over Celite and lyophilized. The resulting
dark blue solid was dissolved in Et2O, 10 drops THF were
added, and the solution was cooled to -35 °C. A dark blue
precipitate, 4, formed and was collected on a glass frit (0.581 g,
69% yield). Crystals suitable for X-ray diffraction were grown
from a vapor diffusion of hexanes into a THF solution of 4 at
room temperature. 1H NMR (500 MHz, CD2Cl2): δ/ppm 165.25
(s), 93.99 (s), 46.56 (s), 24.55 (s), 7.04 (br s), 6.51 (s), 5.97 (s), 3.86
(s), 1.55 (s), 1.17 (s), -4.38 (s). UV/vis (THF) λmax/nm (ε/M-1
cm-1): 325 (12 000), 550 (850), 621 (730), 650 (860), 998 (200),
1250 (90, sh). μeff (C6D6, 295 K): 4.84(5) μB. Comb. Anal. for
[C46H47CoN4]-[LiC16H32O4]þ, Calcd: C, 73.72; H, 7.88; N,
5.55. Found: C, 73.61; H, 7.78; N, 5.62.
9H, p-C6H2(CH3)3), 1.35 (m, 12H, Li(OCH2(CH2)2CH2)4). 13
C
NMR (101 MHz, C6D6): δ/ppm 149.3, 146.9, 138.4, 137.9,
136.5, 135.3, 124.3, 106.7, 102.5, 68.4, 43.3, 26.0, 22.0, 21.3.
MS (ESI-) m/z C46H47N4Zn- [M]-, Calcd: 719.3. Found: 719.3
(observed isotope pattern matches predicted isotope pattern).
(K2-tpeH)Fe(py)2 (7). A thawing 3 mL of THF solution of 1
(67.3 mg, 0.12 mmol) was added to solid LiN(TMS)2 (38.6 mg,
0.23 mmol) and stirred for 1.5 h at room temperature, during
which time the solution turned from clear pale yellow to clear
magenta. This solution was frozen in a liquid nitrogen cooled
cold well. FeCl2(py)2 (33.1 mg, 0.12 mmol) was slurried in 5 mL
of THF and frozen. The thawing solution of deprotonated 1 was
added to the thawing slurry, rinsing with room temperature
THF, and was stirred for 4 h. Solvent was removed in vacuo, and
the resulting residue was dissolved in Et2O and filtered through a
Celite plug. Volatiles were removed in vacuo, affording 7 (29.5
mg, 32%) as an orange-yellow powder. A sample containing
€
134.0 mg of 7 was analyzed by Mossbauer spectroscopy. Crys-
tals suitable for X-ray diffraction were grown from an Et2O
solution of 7. 1H NMR (500 MHz, C6D6): δ/ppm 114.86, 100.30,
41.18, 34.46, 28.77, 24.87, 12.09, 9.99, 6.95, 5.20, 2.39, 1.43,
-0.68, -8.40, -16.46. Comb. Anal. for C51H53FeN5, Calcd: C,
77.36; H, 6.75; N, 8.84. Found: C, 77.24; H, 6.67; N, 8.73.
Procedure for Oxidation of [(tpe)Fe(py)][Li(THF)4]. A solu-
tion of 3 (100.4 mg, 0.0997 mmol) in 5 mL of toluene was
cooled to -40 °C in an acetonitrile/dry ice cooled cold well. The
solution of 3 was added to cold, solid ferrocenium hexafluoro-
phosphate (33 mg, 0.10 mmol) and stirred at -40 °C for 2 h. The
solution was allowed to warm to room temperature and stirred
overnight. Toluene was removed in vacuo, and approximately
2.5 mL of benzene was added. The solution was frozen and
[(tpe)Ni(py)][Li(THF)4] (5). In separate 20 mL scintillation
vials, 1 (27.9 mg, 0.0481 mmol) and PhLi (12.9 mg, 0.154 mmol)
were each dissolved in 3 mL of THF and cooled in a -35 °C
freezer. The cold solution of 1 was added to the PhLi solution
and stirred at room temperature for 30 min. In another 20 mL
vial, NiCl2(py)2 (14.4 mg, 0.0500 mmol) was slurried in 3 mL of
THF with ∼0.5 mL of pyridine. The vials containing deproto-
nated 1 and NiCl2(py)2 were placed in a liquid nitrogen cooled
cold well until frozen. The solution of 1 was thawed and added to
the stirring slurry of NiCl2(py)2. The reaction mixture became a
green solution as it warmed. The reaction was stirred at room
temperature overnight. Removal of the solvent in vacuo gave
a dark gray residue. The residue was extracted into 5 mL of

2524 Inorganic Chemistry, Vol. 49, No. 5, 2010
Sazama and Betley
lyophilized, then triturated in hexane and collected and dried on
a glass frit. The resultant powder (74.4 mg) was analyzed by
paramagnetic systems were optimized with multiplicities
based on their magnetic moments from solution NMR
determination (Evans’ method:7 Mn - sextet, Fe - quintet,
Co - quartet, Ni - Triplet). Intermediate-spin calculations
were also performed, but the high-spin structures were of
lower energy in all cases. Single point population analyses
were carried out on the optimized structures using the unrest-
rictedB3LYPfunctional with “tight” convergencecriteriafor
all complexes. The TZV(P) basis set17 was used for transition
metal atoms and the atoms within the first coordination
sphere (all nitrogen), and SV(P) basis set18 for all other main
group atoms. The singlet [(tpe)Zn(py)]- structure optimiza-
tion and population analysis calculations were performed
using both restricted and unrestricted functionals, and the
difference in energy was found to be negligible.
€
Mossbauer spectroscopy. X-ray quality crystals formed from an
Et2O solution stored at -35 °C. The major product was
identified as 7 by comparison with an authentic sample as
prepared above.
Procedure for Oxidation of [(tpe)Zn(py)][Li(THF)4] and EPR
Sample Preparation. A sample of 6 (18.7 mg, 0.0184 mmol) was
dissolved in 5 mL of toluene and cooled in an acetonitrile/
dry ice cooled cold well (-40 °C). The toluene solution was
added to cooled solid ferrocenium hexafluorophosphate (6.1
mg, 0.018 mmol) and stirred at -40 °C for 2 h. A sample of the
reaction mixture was added to a cooled EPR tube and frozen
into a glass in liquid nitrogen.
Spin density plots and molecular orbital pictures were
generated from Gaussian cube files using Visual Molecular
Dynamics 1.8.6.34 and rendered using POV-Ray v3.6 for
Windows.35 Orbital parentage Mulliken population calcula-
tions, which can be found in the Supporting Information,
were calculated using QMForge.36
VI. Computational Methods
The electronic structures of 2-6 were examined using
Gaussian 03.33 Geometry optimizations were performed
using the atomic coordinates of the crystal structures as a
starting point. Removal of the mesityl methyl groups resulted
in collapse of the ligand structure, and thus no truncations to
the ligand system were made. No symmetry was imposed
upon the molecules. All optimizations were carried out un-
restricted with the M05-2X2 hybrid functional, as the
B3LYP3 functional was insufficient for estimating the ligand
π-stacking effects. The 6-31G Pople basis set15 was used for
all structures, and “tight” convergence criteria were em-
ployed. In cases where the SCF failed to converge to the
“tight” criteria, more reliable but more computationally
expensive quadratic convergence was employed. The zinc
structure was optimized as a singlet, and the remaining
Acknowledgment. The authors thank the American Chemi-
cal Society (PRF Grant Type G) and Harvard University for
financial support. G.T.S. thanks the NSF for a predoctoral
fellowship. The authors would like to thank Lee-Ping Wang and
Troy Van Voorhis for insightful discussions.
Supporting Information Available: UV-vis spectra for tpeH3
and tpeLi3; calculated spin densities for 2, 4, and 5; geometry
optimized coordinates for complexes 2-6; Mulliken population
analyses for 2-6; selected X-ray crystallographic data for 2-6;
full reference for ref 32. This material is available free of charge
via the Internet at http://pubs.acs.org.
(33) Frisch, M. J. et al. Gaussian 03, revision E.01; Gaussian, Inc.: Wallingford,
CT, 2004.
(34) Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graphics 1996, 14, 33.
(35) Persistence of Vision Raytracer, v3.6; Persistence of Vision Pty. Ltd.:
Williamstown, Victoria, Australia, 2004.
(36) Tenderholt, A. L. QMForge, version 2.1; Stanford University: Stanford,
CA.