A homologous series of first-row transition-metal complexes of 2,2'-bipyridine and their ligand radical derivatives: Trends in structure, magnetism, and bonding

Article abstract of DOI:10.1021/ic301587f

The organometallic first-row transition-metal complexes [M(2,2'-bipy)(mes)₂] (M = Cr (1), Mn (2), Co (4), Ni (5); 2,2'-bipy = 2,2'-bipyridine; mes = 2,4,6-Me₃C₆H₂) were reacted with potassium and a suitable alkali-metal sequestering agent to yield salts of the anionic species [M(2,2'-bipy)(mes)₂]^-. The neutral parent compounds and their corresponding anionic congeners were characterized by single-crystal X-ray diffraction in [Cr(2,2'-bipy)(mes) ₂]·1.5C₆H₆, [Mn(2,2'-bipy)(mes) ₂], [Co(2,2'-bipy)(mes)₂]·THF, [Ni(2,2'-bipy)(mes) ₂], [K(dibenzo-18-crown-6)·THF][Cr(2,2'-bipy)(mes) ₂]·2THF, [K(18-crown-6)][Mn(2,2'-bipy)(mes) ₂]·2THF, [K(18-crown-6)][Mn(2,2'-bipy)(mes) ₂]·0.67py·0.67tol, [K(2,2,2-crypt)][Co(2,2'-bipy)(mes) ₂], and [K(2,2,2-crypt)][Ni(2,2'-bipy)(mes)₂]. These species, along with the previously reported neutral and anionic iron complexes [Fe(2,2'-bipy)(mes)₂]^0/- (3/3^-), form a homologous series of compounds which allow for an in-depth study of the interactions between metals and ligands. Single-crystal X-ray diffraction data, DFT calculations, and various spectroscopic and magnetic measurements indicate that the anionic complexes (1^--5^-) can be best formulated as M(II) complexes of the 2,2'-bipyridyl radical anion. These findings complement recent studies which indicate that bond metric data from single-crystal X-ray diffraction may be employed as an important diagnostic tool in determining the oxidation states of bipyridyl ligands in transition-metal complexes.

Full text of DOI:10.1021/ic301587f

Article
pubs.acs.org/IC
A Homologous Series of First-Row Transition-Metal Complexes of
2,2′-Bipyridine and their Ligand Radical Derivatives: Trends in
Structure, Magnetism, and Bonding
Mark Irwin,† Laurence R. Doyle,† Tobias Kramer,† Radovan Herchel,‡ John E. McGrady,*,†
̈
and Jose M. Goicoechea*,†
†Department of Chemistry, Inorganic Chemistry Laboratory, University of Oxford, South Parks Road, Oxford OX1 3QR, U.K.
‡Regional Centre of Advanced Technologies and Materials, Department of Inorganic Chemistry, Faculty of Science, Palacky
University, 17. listopadu 12, CZ-77146 Olomouc, Czech Republic
́
S
* Supporting Information
ABSTRACT: The organometallic first-row transition-metal
complexes [M(2,2′-bipy)(mes)2] (M = Cr (1), Mn (2), Co
(4), Ni (5); 2,2′-bipy = 2,2′-bipyridine; mes = 2,4,6-Me3C6H2)
were reacted with potassium and a suitable alkali-metal
sequestering agent to yield salts of the anionic species
[M(2,2′-bipy)(mes)2]−. The neutral parent compounds and
their corresponding anionic congeners were characterized by
single-crystal X-ray diffraction in [Cr(2,2′-bipy)-
(mes)2]·1.5C6H6, [Mn(2,2′-bipy)(mes)2], [Co(2,2′-bipy)-
(mes)2]·THF, [Ni(2,2′-bipy)(mes)2], [K(dibenzo-18-crown-
6)·THF][Cr(2,2′-bipy)(mes)2]·2THF, [K(18-crown-6)][Mn-
(2,2′-bipy)(mes)2]·2THF, [K(18-crown-6)][Mn(2,2′-bipy)-
(mes)2]·0.67py·0.67tol, [K(2,2,2-crypt)][Co(2,2′-bipy)-
(mes)2], and [K(2,2,2-crypt)][Ni(2,2′-bipy)(mes)2]. These species, along with the previously reported neutral and anionic
iron complexes [Fe(2,2′-bipy)(mes)2]0/− (3/3−), form a homologous series of compounds which allow for an in-depth study of
the interactions between metals and ligands. Single-crystal X-ray diffraction data, DFT calculations, and various spectroscopic and
magnetic measurements indicate that the anionic complexes (1−−5−) can be best formulated as M(II) complexes of the 2,2′-
bipyridyl radical anion. These findings complement recent studies which indicate that bond metric data from single-crystal X-ray
diffraction may be employed as an important diagnostic tool in determining the oxidation states of bipyridyl ligands in transition-
metal complexes.
distances is observed for the interpyridyl ring bond, which
varies from approximately 1.49 Å for the neutral ligand to 1.43
Å for 2,2′-bipy•− and finally to 1.39 Å for 2,2′-bipy2−.
INTRODUCTION
■
The redox-active character of the bidentate ligand 2,2′-
bipyridine (2,2′-bipy) has been well documented since early
spectroscopic measurements demonstrated that it could be
chemically reduced to yield metal salts of the π-radical
monoanion (2,2′-bipy•−) or the diamagnetic dianion (2,2′-
bipy2−).1−5 All three oxidation states of the ligand have been
structurally characterized by single-crystal X-ray diffraction and
show variations in bond metrics which are consistent with the
occupation of an energetically low-lying π orbital upon
reduction.6−8 This orbital (the LUMO of neutral 2,2′-bipy;
Figure 1a) has net antibonding character and nodes located
between alternating C−N and C−C bonds. A comparison of
the structure of neutral bipyridine with alkali-metal salts of the
radical anion and dianion shows significant changes to C−C
and C−N distances, indicative of the population of the
aforementioned molecular orbital (Figure 1b).7 This can also
be rationalized as a loss of aromaticity of neutral 2,2′-bipyridine
to yield structures with localized double bonds such as those
pictured in Figure 1c. The most pronounced change in bond
To date, there have been several reports of structurally
authenticated complexes of bipyridyl radical anions and
dianions with main-group elements,9 lanthanides, and acti-
nides.10,11 In contrast, related “non-innocent” complexes of the
transition metals in which the oxidation state of the metal and
ligand have been determined unequivocally and proven to
contain reduced forms of 2,2′-bipyridine are much rarer. To our
knowledge, the first example of such a complex was
[(CO)4Cr(2,2′-bipy)]−, where EPR and UV-vis spectroscopy
as well as DFT calculations suggest formulation as a Cr(0)
complex of the radical anion 2,2′-bipy•−.12 However, this
anionic complex has never been characterized crystallo-
graphically. In contrast, there are several examples of
crystallographically characterized transition-metal complexes
Received: July 20, 2012
Published: October 30, 2012
© 2012 American Chemical Society
12301
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
EXPERIMENTAL SECTION
■
General Methods. All reactions and product manipulations were
carried out under an inert atmosphere employing standard Schlenk-
line or glovebox techniques. Mg turnings (Fisher) and 2-
bromomesitylene (99%, Alfa Aesar) were used as received. Potassium
metal (99.95%, Aldrich), CrCl2 (99.9%, Strem), MnCl2 (97%, Strem),
CoBr2 (Co 26.6% min, Alfa Aesar), NiBr2 (99+%, Strem), 2,2′-
bipyridine (>99%, TCI Europe), PMe3 (99%, Fluorochem), PEt2Ph
(98%, Alfa Aesar), 18-crown-6 (1,4,7,10,13,16-hexaoxacyclooctade-
cane; 99%, Alfa Aesar), dibenzo-18-crown-6 (2,3,11,12-dibenzo-
1,4,7,10,13,16-hexaoxacyclooctadeca-2,11-diene; 98% Aldrich), and
2,2,2-crypt (4,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8.8.8]-
hexacosane; ≥99% Merck) were stored under dinitrogen in an
MBraun UNIlab glovebox (<0.1 ppm H2O and <0.1 ppm O2) and
used as received. [Cr(PMe3)2(mes)2], Mn3(mes)6 (mes = 2,4,6-
Me3C6H2), [Co(PPhEt2)2(mes)2], and [Ni(2,2′-bipy)(mes)2] (5)
were synthesized according to previously reported literature
methods.18−21 [Cr(2,2′-bipy)(mes)2] was synthesized according to a
modified procedure using [Cr(PMe3)2(mes)2] as a precursor.22
Toluene (tol; 99.9%, Rathburn Chemicals, Ltd.), pentane (99.9%
Rathburn Chemicals, Ltd.), and diethylether (99.9%, Fisher, stabilized
with copper gauze) were dried using an MBraun SPS-800 solvent
purification system. Tetrahydrofuran (THF; 99%, Rathburn Chem-
icals, Ltd.) and pyridine (py; 99+%, Alfa Aesar) were distilled over
sodium/benzophenone and CaH2, respectively. All solvents were
stored in gastight ampules under argon. In addition, toluene, THF,
diethyl ether, and hexanes were stored over activated 3 Å molecular
sieves (Acros).
[Cr(2,2′-bipy)(mes)2]·1.5C6H6 (1·1.5C6H6). [Cr(PMe3)2(mes)2]
(100 mg, 0.226 mmol) and 2,2′-bipyridine (52.9 mg, 0.339 mmol)
were stirred in THF (5 mL) for 30 min, giving rise to a dark purple
solution. Following filtration, the solution was layered with hexanes,
yielding small, purple platelike crystals of [Cr(2,2′-bipy)(mes)2]·THF
after 4 days.22c Similar diffusion into a benzene solution yielded higher
quality crystals of [Cr(2,2′-bipy)(mes)2]·1.5C6H5 (1·1.5C6H5) in
greater crystalline yields (60% yield). The solvent of crystallization was
found to be removed when samples were sealed under vacuum; as a
result, we were unable to obtain a powder X-ray diffractogram due to a
loss of crystallinity. Anal. Calcd for C28H30N2Cr: C, 75.31; H, 6.77; N,
6.27. Found: C, 74.95; H, 6.60; N, 6.06. IR (cm−1): 650 (w), 733 (m),
756 (s), 767 (m), 800 (w), 842 (m), 1015 (m), 1059 (w), 1153 (m),
1171 (w), 1244 (w), 1260 (w), 1311 (w), 1377 (s), 1429 (w), 1441
(s), 1458 (s), 1562 (w), 1571 (w), 1600 (s). Raman (cm−1): 132 (w),
371 (w), 650 (w), 1016 (s), 1159 (m), 1170 (m), 1227 (m), 1282 (s),
1318 (m), 1414 (w), 1468 (m), 1487 (s), 1510 (w), 1554 (m), 1561
(w), 1572 (w). A solid-state EPR spectrum was recorded at room
temperature and exhibited a very weak resonance with a g factor of
1.986.
[Mn(2,2′-bipy)(mes)2] (2). Mn3(mes)6 (150.0 mg, 0.171 mmol)
and 2,2′-bipy (80.0 mg, 0.512 mmol) were weighed out into a sample
vial and dissolved in THF (5 mL) inside an inert atmosphere
glovebox, and the solution was stirred for 6 h. The resultant dark blue
solution was filtered and layered with hexanes. After 3 days, very small,
flat, needlelike crystals of [Mn(2,2′-bipy)(mes)2] (2) had formed,
which were found to be suitable for single-crystal X-ray diffraction
(121 mg, 52% crystalline yield). A powder X-ray diffraction pattern
was collected on the bulk sample and found to match that calculated
from the single-crystal X-ray diffraction data. Anal. Calcd for
C28H30N2Mn: C, 74.82; H, 6.73; N, 6.23. Found: C, 74.62; H, 6.70;
N, 6.15. IR (cm−1): 627 (w), 645 (w), 665 (w), 704 (w), 707 (w), 737
(m), 743 (w), 763 (m), 799 (w), 839 (m), 850 (m), 892 (w), 934 (w),
963 (w), 1012 (s), 1043 (w), 1058 (w), 1102 (w), 1112 (w), 1157 (s),
1167 (w), 1210 (m), 1246 (m), 1257 (w), 1279 (w), 1314 (m), 1365
(w), 1374 (m), 1390 (w), 1485 (w), 1528 (w), 1562 (w), 1571 (w),
1588 (m), 1596 (s), 1600 (s), 1701 (w), 1726 (w). Raman (cm−1): 95
(w), 130 (s), 235 (m), 289 (2), 351 (w), 413 (w), 482 (w), 541 (w),
627 (w), 647 (w), 766 (w), 883 (w), 919 (w), 937 (w), 1013 (s), 1140
(w), 1168 (w), 1245 (w), 1280 (w), 1302 (w), 1311 (w), 1487 (s),
1563 (w), 1597 (w), 1613 (w), 1721 (w). A solid-state EPR spectrum
Figure 1. (a) LUMO of 2,2′-bipyridine. (b) Variations in interatomic
bond distances for 2,2′-bipyx (x = 0, 1−, 2−) as taken from ref 7
(bond numbering shown in (c)). (c) Lewis structures of the neutral,
radical anionic, and diamagnetic dianionic forms of 2,2′-bipyridine.
of 2,2′-bipyridine where bond metric parameters strongly
suggest the possibility of radical (or even dianionic) character
to the ligand, although the compounds were not formulated as
such.6,13,14 The principal reason such species have gone
unidentified is that shifts in the interatomic distances within
the bipyridyl moieties could, at least in principle, be rationalized
in terms of extensive back-bonding from electron-rich metal
centers to the bipyridine ligand. This mechanism would also
effectively populate the LUMO of 2,2′-bipy and hence lead to a
contraction of the C−C bond linking the two rings. Recently,
however, Wieghardt and co-workers have disputed this notion,
arguing that the π-acceptor ability of 2,2′-bipyridine is
negligible and that changes in bipyridyl bond metrics are an
indicator of ligand oxidation state.15 The authors invoke early
studies by Dieck which revealed negligible structural changes to
the ligand, regardless of the nature of the metal to which it was
bonded.16 Of course, bond metric data are not the only
indicators of the relative strength of π back-donation;
vibrational and UV−vis spectroscopy also offer a measure of
the level of the interaction arising between metal and ligand. It
is nevertheless striking that no significant structural changes are
observed in the bipyridyl ligand in homologous series of
complexes for which the degree of metal to ligand π* donation
might reasonably be expected to vary significantly.
We recently reported a bona fide example of a transition-
metal complex of the bipyridyl radical anion, which was
characterized in [K(2,2,2-crypt)][Fe(2,2′-bipy)(mes)2].17 To
our knowledge, 3− is the first transition-metal complex for
which single-crystal X-ray diffraction, Mossbauer, IR, Raman,
̈
and EPR spectroscopy, SQUID magnetometry, and density
functional theory combine to give an unequivocal formulation
as a complex of the 2,2′-bipyridyl radical anion. Herein we
extend this work by reporting the synthesis of a homologous
series of neutral [M(2,2′-bipy)(mes)2] (M = Cr (1), Mn (2),
Co (4), Ni (5)) complexes and their chemical reduction to
yield the closely related anionic species [M(2,2′-bipy)(mes)2]−
(1−, 2−, 4−, and 5−). These species were characterized by
single-crystal X-ray diffraction, IR and Raman spectroscopy,
EPR spectroscopy, SQUID magnetometry, and calculations at
the DFT level. The data converge to indicate that all of the
anionic species contain the radical anionic ligand 2,2′-bipy•−.
12302
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
of a sample of 2 diluted with [NBu4][PF6] was recorded at room
temperature and exhibited an extremely broad resonance centered at a
g factor of 2.009.
temperature and at 10 K, revealing a broadened resonance alongside a
much stronger sharp resonance with a g factor of 1.986.
[K(18-crown-6)][Mn(2,2′-bipy)(mes)2]·x(solv) ([K(18-crown-
6)][2−]·x(solv)). THF (5 mL) was added to a sample vial containing
2 (66.0 mg, 0.147 mmol), K metal (6.1 mg, 0.16 mmol), and 18-
crown-6 (39.0 mg, 0.148 mmol) in a glovebox, and the mixture was
stirred for 24 h. Filtration yielded a dark brown solution which when
layered with hexanes produced crystals of [K(18-crown-6)][Mn(2,2′-
bipy)(mes)2]·2THF ([K(18-crown-6)][2a−]·2THF; ca. 5 mg, 4%
crystalline yield). Due to the small size and weakly diffracting nature of
the crystalline samples, single-crystal X-ray diffraction data were
collected at Beam Station I19 at the Diamond Light Source
Synchrotron. Better quality crystalline samples suitable for single-
crystal X-ray diffraction on in-house instrumentation could be
obtained from pyridine solutions which were allowed to diffuse into
a toluene layer. These crystals were characterized as [K(18-crown-
6)][Mn(2,2′-bipy)(mes)2]·0.67py·0.67tol ([K(18-crown-6)]-
[2b−]·0.67py·0.67tol). Due to the low yields of this reaction and the
poor stability of the complex, suitable PXRD data and an elemental
analysis of the bulk product were unattainable. IR (cm−1): 600 (w),
619 (w), 639 (w), 665 (w), 680 (w), 725 (m), 755 (w), 797 (w), 832
(w), 843 (w), 930 (w), 945 (w), 951 (s), 990 (w), 1011 (m), 1055
(w), 1077 (w), 1101 (s), 1130 (m), 1144 (m), 1172 (w), 1209 (w),
1235 (w), 1260 (s), 1273 (w), 1289 (s), 1294 (w), 1352 (s), 1358 (w),
1410 (w), 1419 (m), 1493 (m), 1506 (w), 1540 (w), 1558 (w), 1583
(w), 1617 (w), 1655 (w), 1695 (w), 1744 (w), 1807 (w), 1866 (w).
Raman (cm−1): 127 (w), 237 (w), 287 (w), 343 (w), 524 (w), 632
(w), 738 (w), 963 (w), 984 (m), 1070 (w), 1159 (m), 1201 (m), 1268
(w), 1348 (w), 1458 (w), 1478 (s), 1500 (m), 1534 (w), 1559 (w),
1579 (w). Solid- and solution-phase EPR spectra were recorded at
room temperature, revealing a strong resonance centered at a g factor
of 2.003. This very strong resonance arising from the bipyridyl radical
was observed over a much broader resonance, similar to that observed
for 2.
[Co(2,2′-bipy)(mes)2]·THF (4·THF). [Co(PEt2Ph)2(mes)2] (301.1
mg, 0.478 mmol) and 2,2′-bipy (74.9 mg, 0.480 mmol) were weighed
out into a Schlenk tube in a glovebox. The reaction vessel was
transferred to a Schlenk line. The solids were dissolved in THF (∼5
mL), and the solution was stirred for 16 h. During this time the
brown-yellow solution turned dark red. The reaction mixture was
filtered and layered with hexanes. After 3 days, dark blocklike crystals
of [Co(2,2′-bipy)(mes)2]·THF (4·THF) had formed and were found
to be suitable for single-crystal X-ray diffraction (101.1 mg, 40%
crystalline yield). Variable-temperature single-crystal X-ray diffraction
experiments showed significant variations in the lattice parameters of
[Co(2,2′-bipy)(mes)2]·THF as a function of temperature (see the
Supporting Information for full details). A room-temperature powder
X-ray diffraction pattern was collected on the bulk sample and found
to match that calculated from single-crystal X-ray diffraction data
collected at 293 K. The elemental analyses recorded for this
compound indicate that the samples lose solvent of crystallization
under vacuum (the same effect was observed for 1·C6H6). Anal. Calcd
for C28H30N2Co: C, 73.12; H, 7.29; N, 5.33. Found: C, 72.68; H, 7.09;
N, 5.55. IR (cm−1): 408 (m), 415 (m), 440 (w), 465 (vs, broad), 648
(w), 652(w), 668 (w), 712 (w), 721 (m), 728 (m), 736 (w), 754 (s),
874 (w), 889 (w), 917 (w), 935 (m), 962 (w), 984 (w), 1008 (m),
1018 (w), 1030 (w), 1042 (w), 1061 (m), 1069 (w), 1116 (w), 1152
(m), 1161 (m), 1215 (w), 1245 (w), 1262 (w), 1272 (m), 1291 (w),
1308 (w), 1309 (m), 1553 (w), 1564 (w), 1592 (w), 1600 (w). Raman
(cm−1): 104 (w), 126 (w), 229 (w), 365 (m), 419 (w), 544 (w), 558
(w), 650 (m), 733 (w), 934 (w), 1004 (m), 1029 (w), 1101 (w), 1163
(m), 1259 (m), 1268 (m), 1317 (m), 1477 (s), 1554 (s), 1600 (s). A
solid-state EPR spectrum was recorded at room temperature and
exhibited an extremely broad resonance with a g factor of 2.428.
[Ni(2,2′-bipy)(mes)2] (5). This sample was prepared following a
previously published method.21 A THF solution of 5 was layered with
pentane, yielding dark purple crystals of the title compound. The bulk
sample was characterized by powder X-ray diffraction and found to
match the calculated diffractogram for the single crystal structure. Anal.
Calcd for C28H30N2Ni: C, 74.20; H, 6.67; N, 6.18. Found: C, 74.42; H,
6.56; N, 6.05. IR (cm−1): 713 (w), 722 (m), 729 (m), 737 (w), 756
(s), 843 (m), 874 (w), 889 (w), 940 (w), 1015 (m), 1021 (m), 1063
(w), 1153 (w), 1168 (w), 1218 (w), 1242 (w), 1250 (w), 1256 (w),
1278 (w), 1313 (w), 1340 (w), 1362 (m), 1377 (m), 1394 (w), 1405
(w), 1419 (w), 1430 (m), 1443 (s), 1562 (w), 1569 (w), 1604 (w).
Raman (cm−1): 258 (w), 550 (w) 566 (m), 635 (w), 772 (w), 943
(w), 1016 (s), 1170 (s), 1279 (s), 1325 (s), 1488 (s), 1564 (s), 1606
(m). 1H NMR (d6-benzene): δ (ppm) 8.05 (d, 2H, 3J = 6 Hz,
[K(2,2,2-crypt)][Co(2,2′-bipy)(mes)2] ([K(2,2,2-crypt)][4−]). A
mixture of 4 (50.0 mg, 0.110 mmol), K metal (5.2 mg, 0.13 mmol),
and 2,2,2-crypt (41.5 mg, 0.110 mmol) were dissolved in
approximately 5 mL of THF, and the mixture was stirred for 16 h
in a glovebox to yield a dark brown solution. Filtration and layering
with hexanes yielded crystals of [K(2,2,2-crypt)][Co(2,2′-bipy)-
(mes)2] after 10 days (25.4 mg, 26% crystalline yield). A powder X-
ray diffraction pattern was found to match that calculated from the
single-crystal structure. Anal. Calcd for C46CoH66KN4O6: C, 63.57; H,
7.65; N, 6.45. Found: C, 63.26; H, 7.49; N, 6.45. IR (cm−1): 608 (w),
638 (w), 667 (w), 682 (w), 714 (w), 729 (s), 756 (w), 782 (w), 832
(w), 841 (m), 864 (w), 934 (s), 943 (w), 956 (s), 1003 (w), 1019 (w),
1027 (w), 1076 (m), 1088 (w), 1107 (s), 1130 (m), 1162 (w), 1173
(w), 1211 (w), 1231 (w), 1258 (m), 1276 (w), 1289 (s), 1295 (m),
1341 (w), 1351 (s), 1359 (m), 1419 (w), 1496 (s), 1556 (m), 1564
(m), 1661 (w), 1699 (w), 1726 (w), 1754 (w), 1783 (w). Raman
(cm−1): 72 (m), 80 (m), 88 (m), 96 (m), 103 (m), 112 (m), 120 (m),
126 (w), 136 (m), 225 (w), 365 (m), 635 (w), 803 (m, broad), 998
(m), 1143 (m), 1207 (w), 1271 (w), 1445 (s), 1506 (m), 1552 (m),
1809 (w), 2328 (s), 2440 (w, broad), 2587 (w, broad), 2885 (w,
broad). A solid-state EPR spectrum was recorded at room temperature
and exhibited a broad resonance at g = 2.189 with a small, sharper
resonance at ca. g = 2.016.
3
bipyH3,3′), 6.92 (s, 4H, HMes), 6.75 (dd, 2H, J = 6 Hz, bipyH5,5′),
6.65 (d, 2H, bipyH6,6′), 6.14 (dd, 2H, bipyH4,4′), 3.04 (s, 12H, o-CH3),
2.41 (s, 6H, p-CH3). 13C NMR (d6-benzene): δ (ppm) 157.0, 153.9,
150.3, 144.3, 135.6, 131.7, 126.8, 126.7, 120.0, 27.9, 21.5.
[K(dibenzo-18-crown-6)·THF][Cr(2,2′-bipy)(mes)2]·2THF ([K-
(dibenzo-18-crown-6)·THF][1−]·2THF). A mixture of 1 (75.0 mg,
0.15 mmol), KC8 (22.7 mg, 0.17 mmol), and dibenzo-18-crown-6
(60.5 mg, 0.17 mmol) were weighed out into a sample vial in a
glovebox and dissolved in approximately 5 mL of THF. The reaction
mixture was stirred for 4 h. Subsequent filtration and layering with
hexanes yielded dark brown crystals of [K(dibenzo-18-crown-
6)·THF][1−]·2THF suitable for single-crystal X-ray diffraction (40
mg, 26% crystalline yield). Solid samples were found to lose
crystallinity upon grinding, and subsequently a satisfactory PXRD
diffractogram could not be collected. Due to the high air and moisture
sensitivity of the sample suitable elemental analyses could not be
obtained. IR (cm−1): 726 (s), 744 (w), 841 (w), 943 (s), 958 (m),
1061 (m), 1087 (w), 1126 (s), 1209 (m), 1251 (s), 1290 (m), 1377
(s), 1419 (w), 1489 (m), 1503 (m), 1593 (w). Raman (cm−1): 209
(m), 269 (w), 360 (s), 411 (w), 462 (w), 578 (w), 603 (m), 643 (s),
748 (w), 945 (w), 995 (s), 1011 (w), 1080 (m), 1161 (s), 1203 (s),
1273 (m), 1288 (w), 1354 (w), 1416 (m), 1480 (s), 1507 (s), 1536
(m), 1568 (m). Solid-state EPR spectra were recorded at room
[K(2,2,2-crypt)][Ni(2,2′-bipy)(mes)2] ([K(2,2,2-crypt)][5−]).
THF (5 mL) was added to a sample vial containing 5 (50.0 mg,
0.110 mmol), KC8 (14.9 mg, 0.11 mmol), and 2,2,2-crypt (41.5 mg,
0.110 mmol), and the mixture was stirred in a glovebox for 4 h.
Filtration and layering with hexanes yielded dark purple crystals of
[K(2,2,2-crypt)][Ni(2,2′-bipy)(mes)2], suitable for single-crystal X-ray
diffraction (13% crystalline yield). A powder X-ray diffraction pattern
matched that calculated from the single-crystal structure. Due to the
high air and moisture sensitivity of this sample, an entirely satisfactory
elemental analysis could not be obtained. Anal. Calcd for
C46H66N4NiKO6: C, 63.57; H, 7.66; N, 6.45. Found: C, 61.02; H,
7.29; N, 6.50. IR (cm−1): 674 (w), 717 (m), 756 (w), 805 (w), 822
(w), 841 (w), 935 (m), 959 (m), 1010 (m), 1018 (w), 1077 (m), 1107
12303
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
(s), 1129 (m), 1159 (w), 1203 (w), 1261 (m), 1275 (m), 1285 (w),
1294 (w), 1351 (m), 1360 (m), 1378 (s), 1416 (w), 1488 (m), 1569
(w). Raman (cm−1): 155 (w), 366 (w), 413 (w), 544 (w), 561 (w),
647 (w), 746 (m), 940 (w), 999 (s), 1016 (m), 1086 (w), 1156 (s),
1201 (s), 1272 (m), 1314 (w), 1353 (m), 1413 (m), 1493 (s), 1549
(s), 1597 (w). Solid-state EPR spectra were recorded at room
temperature and at 10 K, revealing a strong resonance with a g factor
of 2.010. A solution-phase spectrum was also collected, exhibiting
hyperfine coupling indicative of a ligand radical, as previously reported
in ref 21.
X-ray Diffraction. Single-crystal X-ray diffraction data were
collected using an Enraf-Nonius Kappa-CCD diffractometer or using
synchrotron radiation on a CrystalLogic Kappa (three-circle)
diffractometer at beamline I19 (EH1) at the Diamond Light Source,
Didcot, Oxfordshire, U.K. Crystals were selected under Paratone-N oil,
mounted on MiTeGen loops, and quench-cooled using an open flow
N2 cooling device.23 The laboratory data were collected at 150 K using
graphite-monochromated Mo Kα radiation and processed using the
DENZO-SMN package, including unit cell parameter refinement and
interframe scaling (which was carried out using SCALEPACK within
DENZO-SMN).24 The synchrotron data were collected at the
zirconium edge (λ = 0.6889 Å) and were processed using
CrystalClear-SM Expert 2.0 r4.25 Structures were subsequently solved
using direct methods and refined on F2 using the SHELXL 97-2
package.26
Transmission powder X-ray patterns were recorded using a Siemens
D5000 diffractometer in modified Debye-Scherrer geometry equipped
with an MBraun position-sensitive detector. The instrument produced
Cu Kα1 radiation (λ = 1.540 56 Å) using a germanium
monochromator and a standard Cu source. Data were recorded on
samples in flame-sealed capillaries under dinitrogen. The capillaries
were mounted on a goniometer head and aligned so that rotation
occurred along the long central axis of the capillary. During a
measurement the capillary was rotated at ∼60 rpm in order to
minimize any preferred orientation effects that might occur.
SQUID Magnetometry. Magnetic properties were measured on
unanchored polycrystalline samples with a Quantum Design MPMS
XL SQUID magnetometer. Due to the extreme air and moisture
sensitivity of the complexes, they were flame-sealed under vacuum in
Suprasil tubes. Samples 1 and 2 were cooled to 2 K in zero applied
field and their magnetic susceptibilities measured between 2 and 300 K
in a 0.01 T field. The samples were subsequently field-cooled to 2 K
and magnetic susceptibility measurements collected upon warming at
regular intervals in a 0.01 T field. The resulting zero-field-cooled and
field-cooled results were virtually identical and have been merged.
Magnetic measurements on 4 were conducted in a similar fashion but
in a 0.02 T magnetic field. Diamagnetic corrections were applied to the
measured molar magnetic susceptibilities of all of the complexes.27
The absence of ferromagnetic impurities in samples of 1 and 2 was
confirmed by variable field measurements of molar magnetization at
300 K: nonlinearity of the magnetization vs field plot is indicative of
such impurities (see Figure S27 in the Supporting Information).
Similar room-temperature isothermal magnetization measurements
were also performed for the four novel compounds containing a 2,2′-
bipyridyl radical, 1−, 2−, 4−, and 5− (see Figure S28 in the Supporting
Information). A linear plot of magnetization vs field in the case of 1−
indicated that there were no ferromagnetic impurities, and the
magnetic measurements were performed as described for 1. For the
remainder of the compounds, by contrast, the S-shaped dependence of
the magnetization vs field curve is a clear indication of the presence of
ferromagnetic impurities. In these cases, we distinguish the intrinsic
susceptibility of 2−, 4−, and 5− from that of trace ferromagnetic
impurities by measuring the magnetization at five applied fields
between 4 and 5 T following an initial zero-field cooling to 2 K. The
samples were warmed to 300 K in small increments and the
magnetization measured at each step at the same five fields. The slope
of the resulting linear plot of the molar magnetization as a function of
applied field was used to obtain the molar magnetic susceptibility at
each temperature. The procedure is illustrated in Figure S29 of the
Supporting Information. Diamagnetic corrections were applied to the
measured molar magnetic susceptibilities.27 In some cases we were
unable to measure magnetic data below 20 K due to technical issues.
In order to ensure a balanced treatment of all systems, only data above
20 K were used in the fitting of the magnetic data, even where lower
temperature data were available.
Computational Methods. All calculations described in this paper
were performed using density functional theory as implemented in the
ORCA program.28 Unrestricted geometry optimizations on all
[M(2,2′-bipy)(mes)2]x− (x = 0, 1) species were carried out using
the B3LYP functional,29 in combination with the TZVP basis set of
Ahlrichs and co-workers on M and all coordinating atoms (N(1),
N(2), C(11), C(21))30 and the SVP basis set elsewhere.31 Initial
atomic coordinates were taken from single-crystal X-ray diffraction
experiments. Stationary points were confirmed to be minima by the
absence of imaginary frequencies.
Additional Characterization Techniques. IR data were
recorded on solid samples in Nujol mulls. The mulls were made up
inside an inert-atmosphere glovebox and the KBr plates placed in an
airtight container prior to data collection. Spectra were recorded on a
Nicolet Magna-IR 560 spectrometer in absorbance mode (Happ-
Genzel FT apodization) with a Ge/CsI beam splitter and liquid
nitrogen cooled mercury cadmium telluride (MCT) detector.
Raman spectra were recorded on solid samples under dinitrogen in
flame-sealed Pyrex capillaries using a Dilor Labram 300 spectrometer.
The excitation radiation was produced by a 20 mW helium-neon laser
operating at a wavelength of 632.817 nm. Optical density filters could
be inserted into the beam to reduce photon flux, decreasing the
likelihood that photochemical reactions would take place during the
measurement. Typically, measurements were obtained at 0.1% of full
intensity with a counting time of 100 s. Calibration of the
spectrometer was performed before each measurement by referencing
to the 520.7 nm line of a silicon wafer.
CW EPR experiments were performed using an X-band Bruker
BioSpin GmbH EMX spectrometer equipped with a high-sensitivity
Bruker probe head and a low-temperature Oxford Instruments CF935
helium-flow cryostat. Experiments were conducted with 2−10 mW
microwave power, 0.1 mT modulation amplitude, and a modulation
frequency of 100 kHz. The magnetic field was calibrated at room
temperature with an external 2,2-diphenyl-1-picrylhydrazyl standard (g
= 2.0036). Solid-state spectra were recorded on approximately 20 mg
of sample in airtight EPR tubes under argon.
CHN elemental analyses were performed on 5 mg samples
submitted under vacuum in flame-sealed Pyrex ampules.
RESULTS AND DISCUSSION
■
The neutral complexes [M(2,2′-bipy)(mes)2] (M = Cr (1), Mn
(2), Co (4)) were synthesized by direct reaction of
[Cr(PMe3)2(mes)2], Mn3(mes)6, or [Co(PEt2Ph)2(mes)2]
with 2,2′-bipyridine. [Ni(2,2′-bipy)(mes)2] (5) was synthesized
following the stepwise procedure previously reported by
Klein.21 All four complexes were isolated as crystalline samples
in 1·1.5C6H6, 2, 4·THF, and 5. The structures of 1, 2, and 4 as
determined by single-crystal X-ray diffraction are pictured in
Figure 2. Crystallographic data and experimental parameters for
the structures are presented in Table 1. To the best of our
knowledge, 2 and 4 have not been structurally characterized to
date, while the structure we obtained for 1 is a different solvate
of a complex first characterized by single-crystal X-ray
diffraction by Gambarotta, Chiang, and co-workers.22c The
neutral iron analogue 3 was previously reported by our research
groups.17
The torsion angle between the planes defined by M1−C11−
C21 and M1−N1−N2 indicates that the coordination geo-
metries about the metal centers vary from almost square planar
(1, 4, and 5, ϕ = 8.1, 22.0. and 12.4°, respectively) to pseudo-
tetrahedral (2 and 3, 70.6 and 70.2°, respectively). The
preference for a pseudo-square-planar geometry for d4 (1) and
12304
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
Table 1. Selected X-ray Data Collection and Refinement
Parameters for 1·1.5C6H6, 2, and 4·THF
compd
1·1.5C6H6
2
4·THF
formula
fw
C37H39CrN2
563.70
C28H30MnN2
449.48
C32H38CoN2O
525.57
cryst syst
space group
a (Å)
monoclinic
C2/c
monoclinic
P21/c
orthorhombic
Pccn
27.8996(2)
13.1692(1)
18.3471(2)
12.0845(5)
13.8358(6)
14.8944(6)
10.7358(1)
14.5102(1)
17.9942(1)
b (Å)
c (Å)
α (deg)
β (deg)
γ (deg)
V (Å3)
110.830(1)
102.445(2)
6300.41(10)
8
2431.81(18)
4
2803.11(4)
4
Z
radiation, λ (Å)
Mo Kα,
0.710 73
Mo Kα,
0.710 73
Mo Kα,
0.710 73
T (K)
150(2)
1.189
150(2)
1.228
150(2)
1.245
ρcalcd (g cm−3
)
μ (mm−1
)
0.390
0.559
0.638
no. of rflns collected
no. of indep rflns
no. of params
R(int)
13 529
7151
7742
60 428
3192
4207
387
280
174
0.0226
4.68/12.53
5.85/13.46
1.037
0.0513
5.16/13.20
7.40/14.34
1.032
0.0400
4.37/12.50
5.04/13.16
1.064
a
R1/wR2, I ≥ 2σI (%)
a
R1/wR2, all data (%)
GOF
a
R1 = [∑||Fo| − |Fc||]/∑|Fo|; wR2 = {[∑w[(Fo)2 − (Fc)2]2]/
2
[∑w(Fo )2]}1/2; w = [σ2(Fo)2 + (AP)2 + BP]−1, where P = [(Fo)2 +
2(Fc)2]/3 and the A and B values are respectively 0.0711 and 3.82 for
1·1.5C6H6, 0.0731 and 0.92 for 2, and 0.0762 and 2.07 for 4·THF.
antibonding character, as a result of which both M−C and
M−N bonds elongate. The increase of effective nuclear charge
from 2 to 3 (Fe) then results in a contraction of ∼0.07 Å in
both M−C and M−N bonds within a constant pseudo-
tetrahedral environment. The sharp contraction for 4 and 5
reflects a switch back to pseudo-square-planar coordination, the
values being lower than those for 1 due to the substantial
contraction in metal radius from Cr to Co and Ni. In marked
contrast, the bond lengths within the 2,2′-bipy ligand units are
almost invariant across the series, suggesting that back-bonding
is indeed limited, as argued by Wieghardt.15
The chemical reduction of 1−5 with 1 equiv of potassium
metal in the presence of suitable cation sequestering agents was
found to yield salts of the anionic species [M(2,2′-bipy)-
(mes)2]− (M = Cr (1−), Mn (2−), Fe (3−), Co (4−), Ni (5−))
in varying crystalline yields. These species were crystallo-
graphically characterized in [K(dibenzo-18-crown-6)·THF]-
[1−]·2THF, [K(18-crown-6)][2a−]·2THF, [K(18-crown-6)]-
[2b−]·0.67py·0.67tol, [K(2,2,2-crypt)][4−], and [K(2,2,2-
crypt)][5−] (Figure 4). The iron analogue was previously
characterized in [K(2,2,2-crypt)][3−].17 Crystallographic data
and experimental parameters for the structures are presented in
Table 2. The coordination geometries of the anionic species
[M(2,2′-bipy)(mes)2]− are very similar to those of their neutral
analogues. Thus, 1−, 4−, and 5− (M = Cr, Co, Ni) retain
distorted-square-planar geometries (ϕ = 22.7, 4.2, and 5.3°,
respectively) while 2− and 3− have a pseudo-tetrahedral
geometry (81.9 or 82.3° for the 2a− and 2b− solvates,
respectively, and 78.6° for 3−).17 In 2 and 3, reduction
therefore appears to shift the structure toward a more
tetrahedral geometry. The trends in M−C and M−N bond
Figure 2. Thermal ellipsoid plots of (a) 1·1.5C6H6, (b) 2, and (c)
4·THF (only crystallographically unique positions are labeled;
positions labeled with a prime symbol are related by 0.5 − x, 0.5 −
y, z). All solvent and hydrogen atoms have been omitted for clarity.
Anisotropic displacement ellipsoids are pictured at the 50% probability
level.
d8 (5) configurations would be anticipated on the basis of
ligand-field arguments, as would the adoption of pseudo-
tetrahedral structures for d5 (2) and d6 (3).32 DFT calculations
(vide infra) suggest that the pseudo-square-planar and pseudo-
tetrahedral geometries are very close in energy for the cobalt
system 4, the marginal preference for the former being due to
the relatively strong ligand field imposed by the mesityl groups.
The adoption of square-planar geometries for complexes of
Co(II) has precedent in the literature.33
Across the series of complexes [M(2,2′-bipy)(mes)2] (M =
Cr (1), Mn (2), Fe (3), Co (4), Ni (5)) there are significant
variations in both the M−C and M−N bond distances (Figure
3). Moving from 1 (Cr) to 2 (Mn), the switch from pseudo-
square-planar to pseudo-tetrahedral geometry results in
population of orbitals with significant M−C and M−N
12305
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
SQUID magnetometry in an effort to shed further light on the
formal oxidation states of ligands and metal ions. During the
synthesis of the anionic complexes we found it was very
difficult, despite extensive precautions, to avoid formation of
trace amounts of ferromagnetic decomposition products. The
presence of such impurities was detected by performing
variable field isothermal magnetization measurements. Where
present, the response of the impurities was distinguished from
that of the complex itself by measuring the gradient of the
linear portion of the magnetization vs field curve.
Samples of the neutral complexes 1−4 were cooled to 2 K in
zero field and the magnetic susceptibility measured between 2
and 300 K in an applied field of 0.01 or 0.02 T. The samples
were subsequently cooled to 2 K in an applied field, and the
magnetic susceptibility was measured. The zero-field-cooled
(ZFC) and field-cooled (FC) results were virtually identical
and have been merged. The magnetic behavior of the neutral
complexes 1, 2, and 4 is shown in Figure 5a and is largely in
line with what might be expected for paramagnetic organo-
metallic complexes with weak or no spin−orbit coupling.
Fitting to the Curie−Weiss law
N μ μ 2S(S + 1)
g2
T − Θ
A
0
B
χmol
=
+ χ
TIM
(1)
3k
where χTIM is the temperature-independent magnetism term,
leads to the values of g, Θ, and χTIM summarized in Table 4.
Fitting to the three-parameter equation (1) (g, θ, χTIM) was
performed using the simplex algorithm to locate the global
minimum in the error functional defined as F = ∑|μio − μic|,
where μio and μic are the observed and calculated effective
magnetic moments, respectively. Due to a combination of
factors, the fits are not exact, and the term χTIM in eq 1 probably
captures a number of contributions, including (i) the
contribution of excited electronic statesthe intrinsic temper-
ature-independent paramagnetism (TIP), (ii) approximations
in the estimation of the diamagnetic component from Pascal’s
constants, (iii) the uncompensated contribution of the sample
container, and (iv) the contribution of traces of ferromagnetic
impurities. Nevertheless, the best-fit parameters are sufficiently
accurate to give an indication of the number of unpaired
electrons and therefore, in the case of the anions, the nature of
the coupling between metal- and ligand-based spin densities.
The g values for chromium(II) (1) and manganese(II) (2)
lie within the expected ranges for related complexes, while that
for the cobalt(II) complex 4 is rather large, consistent with a
significant contribution of the orbital angular momentum. An
additional complicating factor may arise due to the fact that
crystals of 4 were found to undergo rather extensive changes to
the lattice parameters on varying the temperature (see the
Supporting Information for details). While this phenomenon
has no major effect on the coordination environment of the
cobalt center, there are notable changes to intermolecular
distances which could give rise to small variations to exchange
interactions between adjacent molecules. That being said, the
small and negative values of Weiss constants observed for all
three complexes are in agreement with expected negligible
intermolecular interactions.
Figure 3. Variations in interatomic distances for the neutral complexes
1−5 (top) and anionic complexes 1−−5− (bottom). Data for 3/3− and
5 are taken from refs 17 and 21, respectively.
lengths across the anionic series follow closely those noted
above for the neutral parents (Figure 3). Thus, the M−C and
M−N bond lengths elongate from 1− to 2−, reflecting the
switch in coordination geometry, and then contract from 2− to
3−, reflecting the differing ionic radii of Mn(II) and Fe(II). A
sharp contraction from 3− to 4− and 5− marks the return to
pseudo-square-planar coordination.
On comparison of the neutral and anionic series (Table 3), it
is clear that, while the M−C bond lengths are relatively
unaffected by reduction (0.032 Å for 4 to 4− being the largest
shift), the M−N bond lengths vary more significantly. In the
earlier transition metals (Cr, Mn, Fe), reduction causes a
contraction of ∼0.07 Å, while in the Co and Ni analogues the
changes are less striking. Indeed in the case of 5 and 5−, the
M−N bond lengths are almost identical. The bond lengths
within the bipyridyl ligand are again relatively constant across
the anionic series but are dramatically different from those of
the neutral analogues. The C−C bond linking the two rings, in
particular, is ∼0.06 Å shorter in the anions, a clear indication of
the population of the LUMO of neutral bipyridine. We return
to these structural trends following the discussion of electronic
structure.
The magnetic measurements for the anionic systems
[K(dibenzo-18-crown-6)][1−] and [K(18-crown-6)][2−] are
consistent with the presence of M(II) metal centers (S = 2 and
Magnetism and Spectroscopy. Complexes 1−4 and all of
the anionic moieties 1−−5− were found to be paramagnetic.
The magnetic behavior of these species was studied using
5
S = /2, respectively) that are strongly antiferromagnetically
1
coupled to a ligand radical, S = /2 (Figure 5b). The isotropic
12306
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
Figure 4. Thermal ellipsoid plots of (a) [K(dibenzo-18-crown-6)·THF][1−]·2THF, (b) [K(18-crown-6)][2a−]·2THF, (c) [K(2,2,2-crypt)][4−],
(positions labeled with a prime symbol are related by +y, +x, 2−z), and (d) [K(2,2,2-crypt)][5−] (positions labeled with a prime symbol are related
by +y, +x, −z). All solvent molecules and hydrogen atoms have been omitted for clarity. Anisotropic displacement ellipsoids are pictured at the 50%
probability level.
exchange can be described by the Heisenberg spin Hamil-
tonian:
DMn, suggesting that intermolecular exchange interactions are
the likeliest source of the deviations from ideal Curie behavior.
The magnetic behavior of the anionic cobalt complex 4− is
rather more complex than that of either 1− or 2−. The moment
varies between ∼1.15 (at 20 K) and 1.88 μB (at 300 K) across
the accessible temperature range, values that lie intermediate
between spin-only values for singlet and triplet states, probably
reflecting again the contribution of orbital angular momentum
in the cobalt(II) ion. Finally, no magnetic exchange is present
in complex 5− between diamagnetic nickel(II) and bipy radical.
Accordingly, the low-temperature value of μeff/μB = 1.51 (at 20
2
̂
̂
μ Bg S
B
i
H = −J(S ·S ) +
∑
1
2
i,z
(2)
i=1
An antiferromagnetic interaction between a M(II) center with
spin S1 and a 2,2′-bipy radical spin, S2 = 1/2, results in a ground
spin state Smin = |S1 − S2| with an excited state Smax = S1 + S2
lying Δ = −J/2[Smax(Smax + 1) − Smin(Smin + 1)] higher in
energy. For 1−, 2−, and 4−, Δ = −2.5J, −3J, and −2J,
respectively. For 1− and 2−, the antiferromagnetic interaction is
so strong (J < −350 cm−1) that the Smin ground state is
effectively isolated and the effective magnetic moment remains
almost constant over the 50−300 K temperature range. These
results are supported by DFT calculations (vide infra), which
gave computed J values of −508 and −492 cm−1 for 1− and 2−,
respectively (using the spin Hamiltonian in eq 2). Therefore,
the magnetic data were fitted to the Curie−Weiss law for S =
Smin (eq 1), yielding the parameters collected in Table 4. The
Weiss constants are somewhat smaller than those of the
corresponding neutral species. Attempts to fit the data to a spin
Hamiltonian containing a single-ion zero-field splitting
anisotropy term led to unphysically large values of DCr and
1
K) is close to the theoretical value of 1.73 for S = /2 and g =
2.0: the decrease at high temperatures may again be due to the
presence of unsaturated trace impurities.
Electronic Structure Trends across the [M(2,2′-bipy)-
(mes)2]0/− (M = Cr−Ni) Series. Following our published
discussion of the electronic structure of the [FeII(2,2′-
bipy)(mes)2]0/− complexes,17 we are now in a position to
analyze periodic variations in the electron density distribution
across the entire series of the neutral parent complexes
[M(2,2′-bipy)(mes)2] and their anions (M = Cr, Mn, Fe, Co,
Ni). The optimized geometries of all of the complexes are
summarized in Table 5 (the iron complexes described in ref 17
are included for completeness). Although absolute values of the
metal−ligand bond lengths are somewhat larger than the
12307
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
Table 2. Selected X-ray Data Collection and Refinement Parameters for [K(dibenzo-18-crown-6)·THF][1−]·2THF, [K(18-
crown-6)][2a−]·2THF, [K(18-crown-6)][2b−]·0.67py·0.67tol, [K(2,2,2-crypt)][4−], and [K(2,2,2-crypt)][5−]
[K(dibenzo-18-crown-6)·THF]
[1−]·2THF
[K(18-crown-6)]
[K(18-crown-6)][2b−] ·
0.67py·0.67tol
[K(2,2,2-crypt)]
[K(2,2,2-crypt)]
compd
[2a−]·2THF
[4−]
[5−]
formula
fw
C60H78CrK N2O9
1062.34
C48H70KMnN2O8
897.10
C48H62.67KMnN2.67O6
867.05
C46H66CoKN4O6
869.06
C46H66KN4NiO6
868.84
cryst syst
space group
a (Å)
monoclinic
P21/c
orthorhombic
Fdd2
monoclinic
C2/c
tetragonal
P41212
tetragonal
P43212
14.2207(2)
13.7416(2)
29.9804(5)
13.0840(2)
32.1910(4)
23.1510(3)
13.3060(1)
24.1421(2)
43.8747(4)
12.8822(1)
12.9305(2)
b (Å)
c (Å)
27.3451(2)
27.1526(8)
α (deg)
β (deg)
γ (deg)
V (Å3)
107.561(1)
94.833(1)
5585.59(15)
4
9750.9(2)
8
14044.0(2)
12
4537.95(6)
4
4539.86(17)
4
Z
radiation, λ (Å)
Mo Kα, 0.710 73
150(2)
1.263
synchrotron, 0.688 90
150(2)
Mo Kα, 0.710 73
150(2)
1.230
Mo Kα, 0.710 73
150(2)
1.272
Mo Kα, 0.710 73
150(2)
1.271
T (K)
ρcalcd (g cm−3
μ (mm−1
)
1.222
)
0.337
0.407
0.419
0.520
0.569
no. of rflns collected
no. of indep rflns
no. of params
R(int)
18 045
10 294
20 840
10 324
5171
10 334
9745
4039
12 268
5176
695
300
794
263
263
0.0399
0.0630
0.0246
0.0158
2.81/7.18
3.29/7.38
1.025
0.0228
a
R1/wR2, I ≥ 2σI (%)
5.04/11.28
8.89/12.46
0.996
4.32/10.89
5.07/11.56
4.44/9.31
7.52/10.40
1.017
3.38/7.98
4.23/8.39
1.027
a
R1/wR2, all data (%)
GOF
1.063
a
2
R1 = [∑||Fo| − |Fc||]/∑|Fo|; wR2 = {[∑w[(Fo)2 − (Fc)2]2]/[∑w(Fo )2]}1/2; w = [σ2(Fo)2 + (AP)2 + BP]−1, where P = [(Fo)2 + 2(Fc)2]/3 and the
A and B values are respectively 0.0556 and 2.96 for [K(dibenzo-18-crown-6)·THF][1−]·2THF, 0.0621 and 3.54 for [K(18-crown-6)][2a−]·2THF,
0.0387 and 12.93 for [K(18-crown-6)][2b−]·0.67py·0.67tol, 0.0427 and 0.79 for [K(2,2,2-crypt)][4−], and 0.0431 and 0.99 for [K(2,2,2-crypt)][5−].
experimental lengths, the optimized structural parameters
reproduce the key trends identified previously in the structural
data. The Cr, Co, and Ni complexes (1/1−, 4/4−, and 5/5−,
respectively) are all approximately square planar, while their
Mn and Fe counterparts approach the idealized tetrahedral
limit. The much shorter C−C bonds in the reduced species are
apparent in the computed structures, as is the tendency for
both M−N and C−C bonds to contract upon reduction. This
trend is most striking in the earlier transition metals, although
the contraction in M−N bonds is smaller in magnitude than
that observed experimentally while the M−C bonds elongate
somewhat upon reduction (in comparison to the almost
invariant bond lengths observed in the X-ray data). This
suggests a somewhat unbalanced description of M−N and M−
C bonding, but nevertheless the basic trends in electronic
structure appear well reproduced. The net Mulliken spin
densities on the 2,2′-bipy ligands in the anionic species confirm
the presence of 2,2′-bipy radicals anions in all cases (see Figure
6 for spin density plots for 1/1− and 2/2−). In the case of
antiferromagnetic coupling (1−, 2−, 3−, 4td−), spin contami-
nation leads to values of ⟨S2⟩ of approximately S(S + 1) + 1.0
(in the limit of weak overlap, S2 = S(S + 1) + Nβ, where Nβ is
the number of unpaired electrons of β spin (in these cases
1)).34
metal dxz orbital, which also carries significant M−C
antibonding character. The computed exchange coupling
constants, J, are both very large and negative, consistent with
the experimental measurements that identify an upper limit of
∼−350 cm−1. We note here that the Heisenberg Hamiltonian is
not strictly applicable in the Fe case, where the iron(II) ion, in
the limit of perfect Td symmetry, is orbitally degenerate.
Nevertheless, it is clear that the antiferromagnetically coupled S
3
5
=
/ state is significantly more stable than its S = /
2 2
counterpart.
In the approximately square planar case of 1−, all four low-
lying metal d orbitals are singly occupied, and so the electron in
the π* orbital of the 2,2′-bipy radical is again antiferromagneti-
cally coupled to the metal, this time through the approximately
nonbonding dxz orbital. In the Ni analogue 5−, all four low-lying
orbitals on the metal are now doubly occupied, and so the
exchange coupling is necessarily zero. The structural prefer-
ences for the d7 metal center in 4 and 4− are less clear-cut, and
there is precedent in the literature for both tetrahedral and
square-planar geometries. For the neutral compound 4 we have
identified minima corresponding to both pseudo-tetrahedral
3
and pseudo-square-planar isomers (with S =
/
and 1/2,
2
respectively), the latter being preferred by 4 kcal mol−1,
consistent with the X-ray structure. The anion presents a more
complicated picture, due to the possibility of ferro- or
antiferromagnetic coupling between Co- and bipy-based
The trends in magnetic susceptibility can be understood
within the framework of two structurally distinct classes, the
approximately square planar complexes 1, 4, and 5 together
with their anionic derivatives, and the approximately tetrahedral
2/2− and 3/3−. For 2− and 3−, the very strong
antiferromagnetic coupling arises from the strong overlap of
the singly occupied π* orbital on the 2,2′-bipy radical with the
−
electrons. In the pseudo-tetrahedral case 4td the dominant
exchange coupling pathway is qualitatively identical with that in
2− and 3− (bipy π* with singly occupied metal dxz), resulting in
strong antiferromagnetic coupling and therefore a triplet
−
ground state. In the low-spin pseudo-square-planar form 4sp
,
12308
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
Table 3. Selected Interatomic Distances and Angles for the Neutral and Anionic Species Crystallographically Characterized in
1·1.5C6H6, 2, 3, 4·THF, 5, [K(dibenzo-18-crown-6)·THF][1−]·2THF, [K(18-crown-6)][2a−]·2THF, [K(18-crown-
6)][2b−]·0.67py·0.67tol, [K(2,2,2-crypt)][4−], and [K(2,2,2-crypt)][5−]
M = Cr
M = Mn
2a−/2b−
Bond Lengths (Å)
M = Fe
M = Co
M = Ni
a
a
3−
b
1
1−
2
3
4
4−
5
5−
C5−C6
1.473(3)
1.342(av)
1.359(av)
1.384(av)
1.379(av)
1.380(av)
1.388(av)
2.147(av)
2.118(av)
1.410(4)
1.361(av)
1.390(av)
1.361(av)
1.403(av)
1.357(av)
1.414(av)
2.065(av)
2.122(av)
1.480(4)
1.344(av)
1.349(av)
1.383(av)
1.378(av)
1.379(av)
1.390(av)
2.222(av)
2.144(av)
1.422(6)
1.479(3)
1.418(3)
1.356(av)
1.386(av)
1.372(av)
1.400(av)
1.359(av)
1.419(av)
2.087(av)
2.099(av)
1.473(4)
1.345(3)
1.363(2)
1.382(3)
1.382(3)
1.382(3)
1.382(2)
1.975(2)
1.939(2)
1.429(4)
1.364(2)
1.382(2)
1.369(3)
1.401(3)
1.364(3)
1.408(2)
1.937(1)
1.971(2)
1.465(4)
1.343(av)
1.359(av)
1.382(av)
1.372(av)
1.380(av)
1.386(av)
1.967(av)
1.914(av)
1.416(4)
1.354(3)
1.390(2)
1.378(3)
1.400(4)
1.360(3)
1.419(3)
1.961(2)
1.935(2)
/1.409(av)
N1−C1/N2−C10
N1−C5/N2−C6
C1−C2/C9−C10
C2−C3/C8−C9
C3−C4/C7−C8
C4−C5/C6−C7
M−N1/M−N2
M−C11/M−C21
1.337(4)
/1.352(av)
1.343(av)
1.351(av)
1.380(av)
1.381(av)
1.380(av)
1.393(av)
2.155(av)
2.085(av)
1.381(4)
/1.390(av)
1.374(5)
/1.362(av)
1.396(5)
/1.409(av)
1.349(5)
/1.354(av)
1.416(4)
/1.422(av)
2.141(2)
/2.150(av)
2.151(3)
/2.173(av)
Angles (deg)
C11−M−C21
N1−M−N2
99.3(1)
98.6(1)
77.0(1)
12.7
127.0 (1)
73.5(1)
15.2
106.3(2)/
108.9(av)
126.7(1)
113.6(1)
77.0(1)
9.2
90.7(1)
81.6(1)
5.8
90.4(1)
81.2(1)
4.4
91.5(1)
81.6(1)
3.2
89.3(1)
82.1(1)
4.7
75.7 (1)
76.9(1)/
76.0(av)
75.1(2)
interpyridyl ring torsion 11.3
angle
3.2/1.2(av) 13.8
torsion angle
M1,N1,N2/
M1,C11,C12
8.1
22.7
70.6
81.9/
82.3(av)
70.2
78.6
22.0
4.2
12.4
5.3
a
b
Data taken from ref 17. Data taken from ref 21.
(4sp−) is too small to allow a confident assignment of the
ground state. While these calculations do not provide a clear
rationale for the unusual magnetic susceptibility of 4−, it is clear
that of all the systems studied this is electronically the most
complex, with a number of electronic states with different
multiplicities very close to that of the ground state.
The body of data contained in Tables 3 (X-ray) and 5 (DFT)
offers a unique opportunity to explore the structural
consequences of the interaction between the metal and a
neutral or anionic bipyridyl ligand. We note immediately that
the C(5)−C(6) bond length in the bipyridyl unit is
indisputably determined almost entirely by the ligand oxidation
state, as proposed by Wieghardt and co-workers: the bond
lengths fall into two very narrow bands centered on 1.48 and
1.41 Å for the neutral and anionic derivatives, respectively.15
The M−N bond lengths, in contrast, show considerable
variation across the series. Interpretation of these variations
in terms of specific bonding interactions between the d orbitals
and the π* orbital of bipy is complicated by the normal periodic
variations in ionic radius across the periodic table and also the
change in geometry from pseudo-square-planar to pseudo-
tetrahedral, but the M−C bond lengths provide a convenient
internal reference point. Thus, in either the neutral or anionic
tetrahedral pairs (2/3 and 2−/3−) a decrease of ∼0.07 Å in both
M−N and M−C bond lengths on moving from Mn to Fe
reflects the increase in effective nuclear charge. In contrast, we
see that the one-electron reduction of either 2 or 3 causes a
∼0.07 Å contraction of M−N but only a marginal (∼0.01 Å)
elongation of the M−C bonds. Similar trends are apparent in
the computed structures of 4td/4td−, although in this case there
Figure 5. (a) Effective magnetic moments as a function of temperature
for (a) complexes 1, 2, and 4 and (b) complexes [K(dibenzo-18-
crown-6)][1−], [K(18-crown-6)][2−], [K(2,2,2-crypt)][4−], and [K-
(2,2,2-crypt)][5−]. In each instance the open symbols represent the
experimental data and the lines represent the calculated fit.
by contrast, the single vacancy in the low-lying 3d manifold is
2
localized in the dz orbital, approximately orthogonal to the π*
orbital of the 2,2′-bipy radical. The orthogonality imposes
ferromagnetic coupling, leading to a triplet ground state for
−
4sp−, the same multiplicity as for 4td but with a very different
electron density distribution. The pseudo-tetrahedral triplet
4td− is computed to be lower in energy than the pseudo-square-
planar isomer, apparently at odds with the X-ray structure, but
the separation of only 2 kcal mol−1 to the square-planar triplet
12309
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
Table 4. Calculated g, Θ, and χTIM Values for the Neutral and Anionic Complexes Obtained from Fitting the Variable-
a
Temperature Magnetic Data of the Complexes to the Curie−Weiss Law
complex (metal)
mol wt
diamagnetic correction/10−4 emu mol−1
μeff/μB (300 K)
spin state S
g
Θ/K
χTIM/10−4 emu mol−1
1 (Cr)
446.24
449.18
450.09
453.17
845.49
752.53
865.44
868.53
868.29
−2.95
−2.94
−2.93
−2.92
−5.35
−4.79
−5.60
−5.59
−5.59
4.55
5.74
5.12
2.62
3.87
4.84
4.98
1.88
1.27
2
1.84
1.91
2.09
2.67
1.88
1.81
−0.4
−0.3
−0.9
−0.4
−1.0
−5.4
2.3
4.7
5
2 (Mn)
3 (Fe)
/
2
2
0.0
1
4 (Co)
1− (Cr)
2− (Mn)
3− (Fe)
4− (Co)
5− (Ni)
/
6.4
2
1
3
|2 − /2| = /2
7.4
|5/2 − /2| = 2
19.6
1
1
3
|2 − /2| = /2
|1/2 + /2| = 1
1
1
1
|0 − /2| = /2
1.83
−1.5
−3.7
a
Data for 3 from ref 17.
Table 5. Optimized Geometric and Electronic Parameters for [M(2,2′-bipy)(mes)2]0/− (M = Cr−Ni) Complexes
Cr
Mn
Fe
Co
Ni
1−
2
2−
3
3−
4sp
4sp
4td
4td
5
5−
−
−
1
a
3
5
3
1
3
1
MS
SM; Sbipy
2
/
/
2
2
/
/
2
1
/
2
1
0
/
2
2
1
2
2
1
b
1
1
1
1
2; 0
2.19
2.12
1.48
18.9
2; /2
2.14
5/2; 0
2.22
2.13
1.48
63.9
5/2; /2
2; 0
2.20
2.09
1.48
67.5
2; /2
2.13
1/2; 0
2.04
1.96
1.48
17.1
1/2; /2
3/2; 0
2.12
2.04
1.48
69.5
3/2; /2
0; 0
2.01
1.93
1.48
17.0
0; /2
1.99
1.95
1.43
16.8
M−N/Å
M−C/Å
C(5)−C(6)/Å
2.19
2.02
2.07
2.17
2.20
2.14
1.98
2.09
1.43
1.43
1.43
1.43
1.43
c
Φ /deg
20.1
75.9
73.8
17.3
74.8
d
J /cm−1
−508
4.69
−492
6.89
−656
4.62
+396
2.03
−1050
2.80
⟨S2⟩
6.05
4.16
8.84
4.78
6.05
3.73
0.78
1.25
3.80
2.63
0.00
0.00
0.76
0.03
0.98
ρM
4.19
4.87
3.74
1.21
2.66
ρbipy
−1.00
−0.88
−0.86
0.92
−0.80
a
b
c
d
MS value. Local spin S on metal and ligand centers. Defined as the angle between the C11−M−C21 and N3−M−N4 planes. Coupling constant
J calculated using Yamaguchi’s formula: J = −2[EHS − EBS]/[S2HS − SB2S].35
Figure 7. Singly occupied spin β orbitals in 2− (Mn−N bonding) and
−
4sp (Co−N antibonding). In both cases the orbital is composed
primarily of bipy π*, with some residual metal dxz character.
Turning to the pseudo-square-planar complexes, a contrac-
tion of M−N relative to M−C is apparent in the 1/1− pair,
where again the coupling is antiferromagnetic and the bipy π*
electron has some M−N bonding character. For 4sp and 5−
−
the coupling is either ferromagnetic (4sp−) or zero (5−),
because the overlapping metal-based d orbital is doubly
occupied, and so the singly occupied spin β orbital has M−N
antibonding character. As a result, the redox-induced changes in
M−N bond lengths in 4 and 5 are much smaller and are less
distinct from the changes in M−C. Before leaving the question
of the relationship between structure and bonding, we can
make one further observation on the controversial issue of
back-bonding in the metal bipyridyl. Our analysis set out above
supports the proposal that changes in back-bonding across this
Figure 6. Total Mulliken spin densities for 1, 1−, 2, and 2−.
are no available X-ray data for comparison. The relative
shortening of the M−N bonds in the reduced species correlates
with the strong antiferromagnetic coupling in the ground state
and reflects the fact that the singly occupied spin β orbital
(primarily bipy π*) carries some M−N bonding character
(Figure 7).
12310
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
entire series, whether neutral or anionic, are limited, or at least
have minimal impact on bond metrics. The periodic variations
in M−N and M−C bond lengths appear entirely under-
standable in the framework of trends in ionic radius
superimposed on a small degree of delocalization of the
electron in the bipy π* orbital, where present. In the course of
our survey of the potential energy surface of 4−, we have also
the Operational Program Research and Development for
Innovations (CZ.1.05/2.1.00/03.0058) of the Ministry of
Education, Youth and Sports of the Czech Republic. Finally,
we also thank Diamond Light Source for beam time on I19
(MT1858) used for data collection on sample 2a−.
REFERENCES
1
■
located a closed-shell singlet state, 4−, which is isoelectronic
(1) Comprehensive reviews of the early coordination chemistry of
2,2′-bipyridine and its redox chemistry are available in: (a) Creutz, C.
Comments Inorg. Chem. 1982, 1, 293. (b) Constable, E. C. Homoleptic
Complexes of 2,2′-Bipyridine. In Advances in Inorganic Chemistry;
Sykes, A. G., Ed.; Academic Press: San Diego, CA, 1989; Vol. 34, p 1.
and isostructural with the neutral nickel complex 5. The spin-
restricted nature of the calculation means that this species can
be formulated unambiguously as a d8 Co(I) center coordinated
1
to a neutral bipy ligand. 4− is almost 20 kcal mol−1 higher in
energy than the triplet ground state and so is clearly not
relevant to the interpretation of the structural or magnetic data.
Nevertheless, the rather short C(5)−C(6) bond length of 1.44
Å (full details in the Supporting Information) in comparison to
1.48 Å in 5 confirms that conventional back-bonding of a pair
of metal-based electrons can have a structural impact on the C−
C bond of bipyridyl in cases where the metal center is
extremely electron rich. It seems unlikely, however, that such a
system will be realizable experimentally, because in cases where
the metal is sufficiently electron rich to afford strong back-
bonding, electron transfer to form a bipy− radical will be even
more favorable, as it is in 4−.
Our research has shown that transition-metal complexes
featuring anionic radical ligands are readily accessible via the
chemical reduction of their neutral parent compounds.
Furthermore, it appears that bond metric data in the bipyridyl
moiety can be used as strong indicators of the charge associated
with the ligand, although subtle fluctuations in M−N and M−C
bonds are complementary to the magnetic measurements in
offering insights into the nature of the interaction between
metal and ligand. To date there have been several high-profile
publications in the chemical literature in which metal centers
are bridged by radical ligands (for the synthesis of solids with
interesting magnetic properties).36 The results detailed in this
paper may provide a route toward such materials; our research
groups have already begun related studies on the bridging 4,4′-
bipyridine isomer and its radical and dianionic congeners.37
(2) UV spectroscopy: (a) Konig, E.; Herzog, S. J. Inorg. Nucl. Chem
̈
1970, 32, 585. (b) Konig, E.; Kremer, S. Chem. Phys. Lett. 1970, 5, 87.
̈
(c) Mahon, C.; Reynolds, W. L. Inorg. Chem. 1967, 6, 1927.
(3) IR spectroscopy: Saito, Y.; Takemoto, J.; Hutchinson, B.;
Nakamoto, K. Inorg. Chem. 1972, 11, 2003.
(4) Raman spectroscopy: (a) Kawashima, H.; Kato, T.; Shida, T. J.
Raman Spectrosc. 1991, 22, 187. (b) Castella-Ventura, M.; Kassab, E.;
̀
Buntinx, G.; Poizat, O. Phys. Chem. Chem. Phys. 2000, 2, 4682.
(c) Danzer, G. D.; Golus, J. A.; Strommen, D. P.; Kincaid, J. R. J.
Raman Spectrosc. 1990, 21, 3. (d) Bradley, P. G.; Kress, N.;
Hornberger, B. A.; Dallinger, R. F.; Woodruff, W. H. J. Am. Chem.
Soc. 1981, 103, 7441.
(5) EPR spectroscopy: (a) Zahlan, A.; Heineken, F. W.; Bruin, M.;
Bruin, F. J. Chem. Phys. 1962, 37, 683. (b) Dos Santos-Veiga, J.;
Reynolds, W. L.; Bolton, J. R. J. Chem. Phys. 1966, 44, 2214.
(c) Henning, J. C. M. J. Chem. Phys. 1966, 44, 2139. (d) Van Voorst, J.
D. W.; Zijlstra, W. G.; Sitters, R. Chem. Phys. Lett. 1967, 1, 321.
(e) Konig, E.; Fischer, H. Z. Naturforsch. 1962, 17a, 1063. (f) Kaim,
̈
W. Chem. Ber. 1981, 114, 3789. (g) Herzog, S.; Gustav, K. Z. Anorg.
Allg. Chem. 1966, 346, 150. (h) Wulf, E.; Herzog, S. Z. Anorg. Allg.
Chem. 1972, 387, 81. (i) Herzog, S.; Taube, R. Z. Chem. 1962, 2, 208.
(j) Herzog, S.; Luhder, K. Z. Chem. 1966, 6, 475. (k) Kaim, W.; Lubitz,
̈
W. Angew. Chem., Int. Ed. Engl. 1983, 22, 892.
(6) Chisholm, M. H.; Huffman, J. C.; Rothwell, I. P.; Bradley, P. G.;
Kress, N.; Woodruff, W. H. J. Am. Chem. Soc. 1981, 103, 4945.
(7) Gore-Randall, E.; Irwin, M.; Denning, M. S.; Goicoechea, J. M.
Inorg. Chem. 2009, 48, 8304.
(8) Bock, H.; Lehn, J.-M.; Pauls, J.; Holl, S.; Krenzel, V. Angew.
Chem., Int. Ed. 1999, 38, 952.
(9) (a) Bellavance, P. L.; Corey, E. R.; Corey, J. Y.; Hey, G. W. Inorg.
Chem. 1977, 16, 462. (b) Nikiforov, G. B.; Roesky, H. W.; Noltemeyer,
M.; Schmidt, H.-G. Polyhedron 2004, 23, 561. (c) Mansell, S. M;
Adams, C. J.; Bramham, G.; Haddow, M. F.; Kaim, W.; Norman, N.
C.; McGrady, J. E.; Russell, C. A.; Udeen, S. J. Chem. Commun. 2010,
46, 5070. (d) Mansell, S. M.; Norman, N. C.; Russell, C. A. Dalton
Trans. 2010, 39, 5084.
ASSOCIATED CONTENT
* Supporting Information
■
S
Figures, text, tables, and CIF files giving X-ray crystallographic
data for all complexes, detailed synthesis of [Cr-
(PMe3)2(mes)2], PXRD data, IR, Raman, and EPR spectra,
computational details, and magnetic data. This material is
available free of charge via the Internet at http://pubs.acs.org.
(10) (a) Evans, W. J.; Drummond, D. K. J. Am. Chem. Soc. 1989, 111,
3329. (b) Evans, W. J.; Gonzales, S. L.; Ziller, J. W. J. Am. Chem. Soc.
1994, 116, 2600. (c) Schultz, M.; Boncella, J. M.; Berg, D. J.; Tilley, T.
D.; Andersen, R. A. Organometallics 2002, 21, 460. (d) Roitershtein,
AUTHOR INFORMATION
Corresponding Author
̂
D.; Domingos, A.; Pereira, L. C. J.; Ascenso, J. R.; Marques, N. Inorg.
■
Chem. 2003, 42, 7666. (e) Walter, M. D.; Berg, D. J.; Andersen, R. A.
Organometallics 2006, 25, 3228. (f) Booth, C. H.; Walter, M. D.;
Kazhdan, D.; Hu, Y.-J.; Lukens, W. W.; Bauer, E. D.; Maron, L.;
Eisenstein, O.; Andersen, R. A. J. Am. Chem. Soc. 2009, 131, 6480.
(g) Williams, B. N.; Huang, W.; Miller, K. L.; Diaconescu, P. L. Inorg.
Chem. 2010, 49, 11493.
*E-mail: john.mcgrady@chem.ox.ac.uk (J.E.M.); jose.
goicoechea@chem.ox.ac.uk (J.M.G.).
Notes
The authors declare no competing financial interest.
(11) (a) Wiley, R. O.; Von Dreele, R. B.; Brown, T. M. Inorg. Chem.
1980, 19, 3351. (b) Kraft, S. J.; Fanwick, P. E.; Bart, S. C. Inorg. Chem.
2010, 49, 1103. (c) Ren, W.; Zi, G.; Walter, M. D. Organometallics
2012, 31, 672.
(12) (a) Kaizu, Y.; Kobayashi, H. Bull. Chem. Soc. Jpn. 1970, 43, 2492.
(b) Kaizu, Y.; Kobayashi, H. Bull. Chem. Soc. Jpn. 1972, 45, 470.
ACKNOWLEDGMENTS
■
We thank the EPSRC and the University of Oxford
(studentships to M.I. and T.K.) for financial support of this
research. We also thank Steve Boyer (London Metropolitan
University) for performing all elemental analyses and the
Oxford Crystallographic Service, the Oxford Supercomputing
Centre, and CAESR for access to instrumentation. R.H. thanks
(c) Vlce
Chem. Soc., Dalton Trans. 1998, 215. (d) Zal
J. Chem. Soc., Dalton Trans. 1999, 3081.
̌
k, A.; Baumann, F.; Kaim, W.; Grevels, F.-W.; Hartl, F. J.
́
is, S.; Daniel, C.; Vlcek, A.
̌
̌
12311
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312

Inorganic Chemistry
Article
(13) (a) McPherson, A. M; Fieselmann, B. F.; Lichtenberger, D. L.;
McPherson, G. L.; Stucky, G. D. J. Am. Chem. Soc. 1979, 101, 3425.
(b) Radonovich, L. J.; Eyring, M. W.; Groshens, T. J.; Klabunde, K. J. J.
NOTE ADDED AFTER ASAP PUBLICATION
■
Due to a production error, this paper was published on the
Web on October 30, 2012, with formatting errors. The
corrected version was reposted on October 31, 2012.
Am. Chem. Soc. 1982, 104, 2816. (c) Per
C.; Echegoyen, L. Angew. Chem., Int. Ed. 1997, 36, 137. (d) Rosa, P.;
Mezailles, N.; Richard, L.; Mathey, F.; Le Floch, P. Angew. Chem., Int.
́
ez-Cordero, E. E.; Campana,
́
Ed. 2000, 39, 1823. (e) Mork, B. V.; McMillan, A.; Yuen, H.; Tilley, T.
D. Organometallics 2004, 23, 2855.
(14) Lenges, C. P.; White, P. S.; Marshall, W. J.; Brookhart, M.
Organometallics 2000, 19, 1247.
(15) (a) Scarborough, C. C.; Wieghardt, K. Inorg. Chem. 2011, 50,
9773. (b) Scarborough, C. C.; Sproules, S.; Weyhermuller, T.; DeBeer,
̈
S.; Wieghardt, K. Inorg. Chem. 2011, 50, 12446. (c) Bowman, A. C.;
Sproules, S.; Wieghardt, K. Inorg. Chem. 2012, 51, 3707.
(16) Dieck, T.; Franz, K.-D.; Hohmann, F. Chem. Ber. 1975, 108,
163.
(17) Irwin, M.; Jenkins, R. K.; Denning, M. S.; Kramer, T; Grandjean,
̈
F; Long, G. J.; Herchel, R; McGrady, J. E.; Goicoechea, J. M. Inorg.
Chem. 2010, 49, 6160.
(18) Hermes, A. R.; Morris, R. J.; Girolami, G. S. Organometallics
1988, 7, 2372.
(19) Solari, E.; Musso, F.; Gallo, E.; Floriani, C.; Re, N.; Chiesi-Villa,
A.; Rizzoli, C. Organometallics 1995, 14, 2265.
(20) Chatt, J.; Shaw, B. L. J. Chem. Soc. 1961, 285.
(21) Klein, A. Z. Anorg. Allg. Chem. 2001, 627, 645.
(22) (a) Tsutsui, M.; Zeiss, H. J. Am. Chem. Soc. 1960, 82, 6255.
(b) Stolze, G.; Hahle, J. J. Organomet. Chem. 1967, 7, 301. (c) Edema,
J. J. H.; Gambarotta, S.; van Bolhuis, F.; Smeets, W. J. J.; Spek, A. L.;
Chiang, M. Y. J. Organomet. Chem. 1990, 389, 47.
(23) Cosier, J.; Glazer, A. M. J. Appl. Crystallogr. 1986, 19, 105.
(24) Otwinowski, Z.; Minor, W. Processing of X-ray Diffraction Data
Collected in Oscillation Mode. In Methods in Enzymolology; Academic
Press: New York, 1997.
(25) CrystalClear-SM Expert 2.0 r4; Rigaku, Tokyo, Japan, 2009.
(26) (a) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.
(b) Sheldrick, G. M. SHELX97-Programs for Crystal Structure Analysis
(Release 97-2); Institut fur Anorganische Chemie der Universitat,
̈
̈
Tammanstrasse 4, D-3400 Gottingen, Germany, 1998. (c) Sheldrick,
̈
G. M. Acta Crystallogr. 2008, A64, 112.
(27) Bain, G. A.; Berry, J. F. J. Chem. Educ. 2008, 85, 532.
(28) Neese, F. ORCA-an ab initio, density functional and semiempirical
program package, version 2.6, revision 35; University of Bonn, Bonn,
Germany, Feb 2008 (http://www.thch.uni-bonn.de/tc/orca/).
(29) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Stephens, P.
J.; Devlin, J. F.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994,
98, 11623. (c) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. Rev. B 1988,
37, 785.
(30) Schafer, A.; Huber, C; Ahlrichs, R. J. Chem. Phys. 1994, 100,
̈
5829.
(31) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571.
̈
(32) Ghosh, M.; Sproules, S.; Weyhermuller, T.; Wieghardt, K. Inorg.
̈
Chem. 2008, 47, 5963.
(33) See for example: (a) Soda, T.; Kitagawa, Y.; Onishi, T.; Takano,
Y.; Shigeta, Y.; Nagao, H.; Yoshioka, Y.; Yamaguchi, K. Chem. Phys.
Lett. 2000, 319, 223. (b) Lu, C. C.; Bill, E.; Weyhermuller, T.; Bothe,
̈
E.; Wieghardt, K. J. Am. Chem. Soc. 2008, 130, 3181. (c) Chatt, J;
Shaw, B. L. J. Chem. Soc. 1961, 285. (d) Cibian, M.; Derossi, S.;
Hanan, G. S. Dalton Trans. 2011, 40, 1038.
(34) Szabo, A.; Ostlund, N. S. Modern Quantum Chemistry:
Introduction to Advanced Electronic Structure Theory; Dover: New
York, 1996.
(35) Yamaguchi, K.; Takahara, Y.; Fueno, T. In Applied Quantum
Chemistry, Smith, V. H., Ed.; Reidel: Dordrecht, The Netherlands,
1986; p 155.
(36) For recent reviews see: (a) Dechambenoit, P.; Long, J. R. Chem.
Soc. Rev. 2011, 40, 3249. (b) Murray, K. Aust. J. Chem. 2009, 62, 1081.
(37) (a) Irwin, M; Kramer, T; McGrady, J. E.; Goicoechea, J. M.
̈
Inorg. Chem. 2011, 50, 5006. (b) Denning, M. S.; Irwin, M.;
Goicoechea, J. M. Inorg. Chem. 2008, 47, 6118.
12312
dx.doi.org/10.1021/ic301587f | Inorg. Chem. 2012, 51, 12301−12312