Bisamidate and mixed amine/amidate NiN2S2 complexes as models for nickel-containing acetyl coenzyme A synthase and superoxide dismutase: An experimental and computational study

Article abstract of DOI:10.1021/ic9023053

The distal nickel site of acetyl-CoA synthase (Ni_d-ACS) and reduced nickel superoxide dismutase (Ni-SOD) display similar square-planar Ni^IIN₂S₂ coordination environments. One difference between these two sites,

Full text of DOI:10.1021/ic9023053

Inorg. Chem. 2010, 49, 5393–5406 5393
DOI: 10.1021/ic9023053
Bisamidate and Mixed Amine/Amidate NiN2S2 Complexes as Models for
Nickel-Containing Acetyl Coenzyme A Synthase and Superoxide Dismutase:
An Experimental and Computational Study
Vaidyanathan Mathrubootham,† Jason Thomas,† Richard Staples,‡ John McCraken,‡ Jason Shearer,*,§ and
Eric L. Hegg*,†
†Department of Biochemistry & Molecular Biology, and ‡Department of Chemistry, Michigan State University,
East Lansing, Michigan 48824, and §Department of Chemistry, University of Nevada, Reno, Nevada 89557
Received November 23, 2009
The distal nickel site of acetyl-CoA synthase (Nid-ACS) and reduced nickel superoxide dismutase (Ni-SOD) display
similar square-planar NiIIN2S2 coordination environments. One difference between these two sites, however, is that
the nickel ion in Ni-SOD contains a mixed amine/amidate coordination motif while the Nid site in Ni-ACS contains a
bisamidate coordination motif. To provide insight into the consequences of the different coordination environments on
the properties of the Ni ions, we systematically examined two square-planar NiIIN2S2 complexes, one with bisthiolate-
bisamidate ligation (Et4N)2(Ni(L1)) 2H2O (2) [H4L1 = N-(2-mercaptoacetyl)-N0-(2-mercaptoethyl)glycinamide] and
3
another with bisthiolate-amine/amidate ligation K(Ni(HL2)) (3) [H4L2 = N-(200-mercaptoethyl)-2-((20-mercap-
toethyl)amino)acetamide]. Although these two complexes differ only by a single amine versus amidate ligand, their
chemical properties are quite different. The stronger in-plane ligand field in the bisamidate complex (NiII(L1))2- (2)
results in an increase in the energies of the d f d transitions and a considerably more negative oxidation potential.
Furthermore, while the bisamidate complex (NiII(L1))2- (2) readily forms a trinuclear species (Et4N)2({Ni(L1)}2-
Ni) H2O (1) and reacts rapidly with O2, presumably via sulfoxidation, the mixed amine/amidate complex (NiII(HL2))-
3
(3) remains monomeric and is stable for days in air. Interestingly, the NiIII species of the bisamidate complex formed by
chemical oxidation with I2 can be detected by electron paramagnetic resonance (EPR) spectroscopy while the mixed
amine/amidate complex immediately decomposes upon oxidation. To explain these experimentally observed
properties, we performed S K-edge X-ray absorption spectroscopy and low-temperature (77 K) electronic absorption
measurements as well as both hybrid density functional theory (hybrid-DFT) and spectroscopy oriented configuration
interaction (SORCI) calculations. These studies demonstrate that the highest occupied molecular orbital (HOMO) of
the bisamidate complex (NiII(L1))2- (2) has more Ni character and is significantly destabilized relative to the mixed
amine/amidate complex (NiII(HL2))- (3) by ∼6.2 kcal mol-1. The consequence of this destabilization is manifested in
the nucleophilic activation of the doubly filled HOMO, which makes (NiII(L1))2- (2) significantly more reactive toward
electrophiles such as O2.
Introduction
ized.1-4 Nickel enzymes can be divided into two groups:
hydrolases and redox enzymes. In the hydrolase nickel
enzymes, the nickel ion is found to exist only in the NiII state,
and the Ni-center acts as a Lewis acid. The five known Ni
redox enzymes are [NiFe]-hydrogenases, CO dehydrogenase
(CODH), acetyl-CoA synthase (ACS), methyl-coenzyme
reductase (MCR), and nickel-dependent superoxide dismu-
tase (Ni-SOD).1-4 A common feature in redox active nickel
enzymes is the presence of thiolate coordination, which helps
the nickel centers access different oxidation states in their
catalytic cycles.4 In spite of several physical and biochemical
Nickel is an essential trace element for bacteria, plants,
animals, and humans. To date, several nickel-containing
enzymes are known and have been structurally character-
*To whom correspondence should be addressed. E-mail: erichegg@
msu.edu (E.L.H.), shearer@chem.unr.edu (J.S.).
(1) (a) Nickel and its surprising Impact in Nature; Sigel, A., Sigel, H., Sigel,
R. K. O., Eds.; John Wiley & Sons: Chichester, U.K., 2007; Metal ions in Life
Sciences, Vol. 2. (b) The Bioinorganic Chemistry of Nickel; Lancaster, J. R., Jr.,
Ed.; VCH: New York, 1988.
(2) (a) Ragsdale, S. W. Nickel Enzymes & Cofactors. In Encyclopedia of
Inorganic Chemistry [online]; King, R. B., Ed.; John Wiley & Sons: Chichester,
U.K., 2005; pp 1-16; http://mrw.interscience.wiley.com/emrw/9780470862100/
eic/article/ia149/current/pdf (accessed March 2, 2010). (b) Ragsdale, S. W.
J. Inorg. Biochem. 2007, 101, 1657–1666. (c) Ragsdale, S. W. J. Biol. Chem.
2009, 284, 18571–18575.
(3) Halcrow, M. A.; Christou, G. Chem. Rev. 1994, 94, 2421–2481.
(4) (a) Maroney, M. J.; Davidson, G.; Allan, C. B.; Figlar, J. Struct.
Bonding 1998, 92, 1–65. (b) Maroney, M. J. Curr. Opin. Chem. Biol. 1999, 3,
188–199.
r
2010 American Chemical Society
Published on Web 05/27/2010
pubs.acs.org/IC

5394 Inorganic Chemistry, Vol. 49, No. 12, 2010
Mathrubootham et al.
Figure 1. Active site structure of (A) Ni-ACS and (B) reduced Ni-SOD. (Structures prepared from the coordinates obtained from Protein Data Bank files
1OAO and 1T6U respectively for Ni-ACS and Ni-SOD, refs 9 and 26b, respectively.)
studies on these enzymes,1-5 it is still not fully understood
how the different ligand environments facilitate the catalysis
of such a wide range of reactions. The main goals of this study
are to elucidate how the NiN2S2 coordination environment
found in Ni-SOD and the distal Ni site of CODH/ACS
contributes to the different properties of these two nickel ions
and to quantify these effects.
The metalloenzyme carbon monoxide dehydrogenase/
acetyl coenzyme A synthase (CODH/ACS) is an unusual
bifunctionalenzymepresent in a number of acetogenic, meth-
anogenic, and sulfate-reducing bacteria.5-7 In the first step,
CO2 is reduced to CO at the C-cluster (CODH activity). The
subsequent formation of acetyl-CoA from CO, a methyl
group, and coenzyme A is catalyzed by the A-cluster (ACS
activity). Crystal structures of the A-cluster of ACS/CODH8,9
as well as the monomeric form of ACS10 have been reported.
The active site contains a unique dinuclear nickel center linked
to a [Fe4S4] cluster (Figure 1A).8-12 The proximal nickel (Nip)
is coordinated to three cysteine thiolates and an unidentified
exogenous ligand while the distal square-planar nickel (Nid) is
ligated in an unusual Cys-Gly-Cys motif consisting of two
deprotonated amides and two cysteine thiolates.8-10 Acetyl-
CoA is presumably synthesized at Nip; it is generally believed
that Nid is not involved in redox chemistry and that the nickel
ion remains in the NiII state throughout the catalytic cycle,
although this remains to be definitively verified.2,6b,13
Despite the availability of three high-resolution crystal
structures for ACS and nearly two decades of thorough
spectroscopic and biochemical studies, many questions
remain concerning the role of the endogenous ligands in tuning
the properties of the Ni ions.13 To address these questions,
we14 and others15-21 synthesized small metal complexes that
serve as well-defined mimics of the active site. Ironically,
22,23
Holm22 and Kruger
developed some of the earliest
€
models of the Nid(Cys-Gly-Cys) site when preparing models
for the active site of [NiFe]-hydrogenases well before the
structure of ACS was known. Since the elucidation of the
ACS crystal structure, considerable effort has been directed
toward modeling the dinuclear nickel active site (Nip-Nid
site), and several laboratories have reported on various
aspects of the dinuclear Ni-Ni site.14-21 Unfortunately,
attempts to synthesize discrete dinuclear ACS models often
(13) Hegg, E. L. Acc. Chem. Res. 2004, 37, 775–783.
(14) Hatlevik, Ø.; Blanksma, M. C.; Mathrubootham, V.; Arif, A. M.;
Hegg, E. L. J. Biol. Inorg. Chem. 2004, 9, 238–246.
(15) (a) Golden, M. L.; Rampersad, M. V.; Reibenspies, J. H.; Darens-
bourg, M. Y. Chem. Commun. 2003, 1824–1825. (b) Rampersad, M. V.; Jeffery,
S. P.; Golder, M. L.; Lee, J.; Reibenspies, J. H.; Darensbourg, D. J.; Darensbourg,
M. Y. J. Am. Chem. Soc. 2005, 127, 17323–17334. (c) Golder, M. L.; Whaley,
C. M.; Rampersad, M. V.; Reibenspies, J. H.; Hancock, R. D.; Darensbourg, M. Y.
Inorg. Chem. 2005, 44, 875–883.
(5) Ragsdale, S. W.; Kumar, M. Chem. Rev. 1996, 96, 2515–2539.
(6) (a) Lindahl, P. A. Biochemistry 2002, 41, 2097–2105. (b) Lindahl, P. A.;
Graham, D. E. In Nickel and Its Surprising Impact in Nature; Sigel, A., Sigel,
H., Sigel, R. K. O., Eds.; John Wiley & Sons: Chichester, U.K., 2007; Metal ions
in Life Sciences, Vol. 2, Chapter 9, pp 357-411.
(16) (a) Wang, Q.; Blake, A. J.; Davies, E. S.; McInnes, E. J. L.; Wilson,
€
C.; Schroder, M. Chem. Commun. 2003, 3012–3013. (b) Duff, S. E.; Barclay,
J. E.; Davies, S. C.; Evans, D. J. Inorg. Chem. Commun. 2005, 8, 170–173.
(17) (a) Evans, D. J. Coord. Chem. Rev. 2005, 249, 1582–1595. (b) Harrop,
T. C.; Mascharak, P. K. Coord. Chem. Rev. 2005, 249, 3007–3024.
(18) Linck, R. C.; Spahn, C. W.; Rauchfuss, T. B.; Wilson, S. R. J. Am.
Chem. Soc. 2003, 125, 8700–8701.
(7) (a) Ragsdale, S. W. Crit. Rev. Biochem. Mol. Biol. 2004, 39, 165–195.
(b) Ragsdale, S. W. Biofactors 1997, 6, 3–9.
(8) Doukov, T. I.; Iverson, T. M.; Servalli, J.; Ragsdale, S. W.; Drennan,
C. L. Science 2002, 298, 567–572.
(9) Darnault, C.; Volbeda, A.; Kim, E. J.; Legrand, P.; Vernede, X.;
Lindahl, P. A.; Fontecilla-Camps, J. C. Nat. Struct. Biol. 2003, 10, 271–279.
(10) Svetlitchnyi, V.; Dobbek, H.; Meyer-Klaucke, W.; Meins, T.; Thiele,
(19) (a) Harrop, T. C.; Olmstead, M. M.; Mascharak, P. K. J. Am. Chem.
Soc. 2004, 126, 14714–14715. (b) Harrop, T. C.; Olmstead, M. M.; Mascharak,
P. K. Chem. Commun. 2004, 1744–1745. (c) Harrop, T. C.; Olmstead, M. M.;
Mascharak, P. K. Inorg. Chem. 2006, 45, 3424–3436.
(20) (a) Krishnan, R.; Voo, J. K.; Riordan, C. G.; Zahkarov, L.;
Rheingold, A. L. J. Am. Chem. Soc. 2003, 125, 4422–4423. (b) Krishnan,
R.; Riordan, C. G. J. Am. Chem. Soc. 2004, 126, 4484–4485.
€
B.; Romer, P.; Huber, R.; Meyer, O. Proc. Natl. Acad. Sci. U.S.A. 2004, 101,
446–451.
(11) (a) Shin, W.; Anderson, M. E.; Lindahl, P. A. J. Am. Chem. Soc.
1993, 115, 5522–5526. (b) Bramlett, M. R.; Tan, X.; Lindahl, P. A. J. Am. Chem.
Soc. 2003, 125, 9316–9317.
(21) (a) Rao, P. V.; Bhaduri, S.; Jiang, J.; Hong, D.; Holm, R. H. J. Am.
Chem. Soc. 2005, 127, 1933–1945. (b) Rao, P. V.; Bhaduri, S.; Jiang, J.; Holm,
R. H. Inorg. Chem. 2004, 43, 5833–5849.
(12) Seravalli, J.; Xiao, Y.; Gu, W.; Cramer, S. P.; Antholine, W. E.;
Krymov, V.; Gerfen, G. J.; Ragsdale, S. W. Biochemistry 2004, 43,
3944–3955.
€
(22) (a) Kruger, H.-J.; Holm, R. H. Inorg. Chem. 1987, 26, 3645–3647.
(b) Kr€uger, H.-J.; Peng, G.; Holm, R. H. Inorg. Chem. 1991, 30, 734–742.
€
(23) Hanss, J.; Kruger, H.-J. Angew. Chem., Int. Ed. 1998, 37, 360–363.

Article
Inorganic Chemistry, Vol. 49, No. 12, 2010 5395
leads to the formation of trinuclear or multinuclear complexes,
although some success has been achieved using phosphine
ligands.16,19-21 Recently, binuclear Ni complexes containing a
methylnickel moiety have been reported by Riordan and co-
workers.24 In addition, Tatsumi and co-workers25 recently
synthesized dinuclear models with exogenous monodentate
thiolates and thiocarbamate ligands.
ligands relative to neutral amine ligands.29,32 Thus, it is
hypothesized that the mixed amine/amidate ligation utilized
by Ni-SOD may allow the enzyme to stabilize the NiIII state
while at the same time avoiding sulfur-based oxidation.32-34
Relatively few examples of mixed amine/amidate nickel
complexes have been reported as models for Ni-SOD.34-38
Both Shearer34 and others35 have synthesized nickel-peptide
complexes as functional models of Ni-SOD. The first syn-
thetic mixed amine/amidate analogue of Ni-SOD was re-
ported by Shearer and Zhao.37 More recently, Grapperhaus
and co-workers38 reported a synthetic model with mixed
amine/amidate ligation and imidazole ligation. Other models
with a different ligand environment other than mixed amine/
amidate have also been reported.39 Interestingly, none of the
synthetic analogues except the nickel-peptide models show
SOD activity.
In this manuscript, we report the synthesis of related
NiIIN2S2 complexes, one with bisamidate ligation and the
other with mixed amine/amidate ligation (Scheme 1). Because
the ligands are identical except for a single amine/amidate,
this study provides a unique opportunity to establish the role
of the N donors in tuning the chemical and electrochemical
properties in NiN2S2 complexes, to assess the importance of
the N donors in controlling the reactivity of the metal center in
these complexes, and to quantify the relative molecular orbital
energies of amine/amidate versus bisamidate NiN2S2 com-
plexes. Specifically, we prepared the NiII complex ((NiII-
(HL2))- (3), which represents one of the very rare NiN2S2
model complexes containing NiII in a mixed amine/amidate
environment. For comparison, we have also synthesized the
analogous bisamidate complex, (NiII(L1))2- (2), which also
provides information on whether the unsymmetrical orienta-
tion of amide carbonyl groups of the ligand backbone
influences the properties of the Ni center. These complexes
were probed using X-ray crystallography, electronic absorp-
tion spectroscopy, X-ray absorption spectroscopy, electron
paramagnetic resonance (EPR) spectroscopy, and electro-
chemistry. Hybrid density functional theory (hybrid-DFT)
and spectroscopy oriented configuration interaction (SORCI)
Interestingly, the active site of the enzyme Ni-superoxide
dismutase26 (Ni-SOD) (Figure 1B) resembles the NiIIN2S2
coordination environment found in the distal nickel site of
Ni-ACS. In the reduced state, the NiII ion exhibits a square-
planar coordination geometry composed of two thiolates in a
cis arrangement, a deprotonated amide nitrogen, and the
terminal -NH2 group. Ni-SOD catalyzes the disproportio-
nation of superoxide to O2 and H2O2, accessing both the NiII
and NiIII states during the catalytic cycle. An intriguing
question concerning Ni-SOD is how the active-site avoids
sulfur-based oxidation chemistry and performs Ni-based
oxidation in the presence of superoxide, O2, and H2O2 as
reactants/products. Because Ni-bound thiolate groups are
susceptible to modification and oxidation,27-30 only a small
number of synthetic NiIII alkyl thiolate complexes have been
22,23
reported thus far.3,22,23,31 Holm22 and Kruger
reported
€
NiIII complexes with thiolate ligation in the early to mid
1990s, one of which is the only NiIII thiolate small molecule
complex crystallographically characterized to date. Fiedler
and Brunold31 performed detailed spectroscopic/computa-
tional studies on these complexes, revealing important insight
into the nature of NiIII-S bonding interactions. Signifi-
cantly, recent work has demonstrated that NiN2S2 complexes
in bisamine ligand environments are more stable toward O2
than the corresponding bisamidate complexes.14,28,29 It is
also known that NiIII is greatly stabilized by anionic amidate
(24) Dougherty, W. G.; Rangan, K.; O’Hagan, M. J.; Yap, G. P. A.;
Riordan, C. G. J. Am. Chem. Soc. 2008, 130, 13510–13511.
(25) (a) Ito, M.; Kotera, M.; Song, Y.; Matsumoto, T.; Tatsumi, K. Inorg.
Chem. 2009, 48, 1250–1256. (b) Song, Y.; Ito, M.; Kotera, M.; Song, Y.;
Matsumoto, T.; Tatsumi, K. Chem. Lett. 2009, 38, 184–185. (c) Ito, M.; Kotera,
M.; Matsumoto, T.; Tatsumi, K. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 11862–
11866.
(26) (a) Wuerges, J.; Lee, J.-W.; Yim, Y.-I.; Yim, H.-S.; Kang, S. -O.;
Carugo, K. D. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 8569–8574.
(b) Barondeau, D. P.; Kassmann, C. J.; Bruns, C. K.; Tainer, J. A.; Getzoff,
E. D. Biochemistry 2004, 43, 8038–8047.
(27) (a) Grapperhaus, C. A.; Maguire, M. J.; Tuntulani, T.; Darensbourg,
M. Y. Inorg. Chem. 1997, 36, 1860–1866. (b) Grapperhaus, C. A.; Darensbourg,
M. Y. Acc. Chem. Res. 1998, 31, 451–459. (c) Kaasjager, V. E.; Bouwman, E.;
Gorter, S.; Reedijk, J.; Grapperhaus, C. A.; Reibenspies, J. H.; Smee, J. J.;
Darensbourg, M. Y.; Derecskei-Kovacs, A.; Thomson, L. M. Inorg. Chem. 2002,
41, 1837–1844. (d) Green, K. N.; Brothers, S. M.; Jenkins, R. M.; Carson, C. E.;
Grapperhaus, C. A.; Darensbourg, M. Y. Inorg. Chem. 2007, 46, 7536–7544.
(28) Grapperhaus, C. A.; Mullins, C. S.; Kozlowski, P. M.; Mashuta,
M. S. Inorg. Chem. 2004, 43, 2859–2866.
(33) It has been suggested by Darensbourg and co-workers that the
kinetic control is responsible for preventing the sulfur oxygenation in the
enzyme active site (ref 27d).
(34) (a) Shearer, J.; Long, L. M. Inorg. Chem. 2006, 45, 2358–2360.
(b) Neupane, K. P.; Shearer, J. Inorg. Chem. 2006, 45, 10552–10556. (c)
Neupane, K. P.; Gearty, K.; Francis, A.; Shearer, J. J. Am. Chem. Soc. 2007,
129, 14605–14618. (d) Shearer, J.; Neupane, K. P.; Callan, P. E. Inorg. Chem.
2009, 48, 10560–10571.
(35) (a) Schmidt, M.; Zahn, S.; Carella, M.; Ohlenschlager, O.; Gorlach,
M.; Kothe, K.; Weston, J. ChemBioChem 2008, 9, 2135–2146. (b) Krause,
M. E.; Glass, A. M.; Jackson, T. A.; Laurence, J. S. Inorg. Chem. 2010, 49, 362–
364.
(29) Mullins, C. S.; Grapperhaus, C. A.; Kozlowski, P. M. J. Biol. Inorg.
Chem. 2006, 11, 617–625.
(36) Tietze, D.; Breitzke, H.; Imhof, D.; Kothe, E.; Weston, J.; Bunt-
kowsky, G. Chem.;Eur. J. 2009, 15, 517–523.
(30) (a) Colpas, G. J.; Kumar, M.; Day., R. O.; Maroney, M. J. Inorg.
Chem. 1990, 29, 4779–4788. (b) Mirza, S. A.; Pressler, M. A.; Kumar, M.; Day,
R. O.; Maroney, M. J. Inorg. Chem. 1993, 32, 977–987. (c) Mirza, S. A.; Day,
R. O.; Maroney, M. J. Inorg. Chem. 1996, 35, 1992–1995. (d) Chohan, B. S.;
Maroney, M. J. Inorg. Chem. 2006, 45, 1906–1908.
(31) Fiedler, A. T.; Brunold, T. C. Inorg. Chem. 2007, 46, 8511–8521.
(32) (a) Fiedler, A. T.; Bryngelson, P. A.; Maroney, M. J.; Brunold, T. C.
J. Am. Chem. Soc. 2005, 127, 5449–5462. (b) Margerum, D. W.; Anliker, S. L. In
The Bioinorganic Chemistry of Nickel; Lancaster, J. R., Jr., Ed.; VCH: New
York, 1988; pp 29-51. (c) Wang, J. F.; Kumar, K.; Margerum, D. W. Inorg.
Chem. 1989, 28, 3481–3484. (d) Lappin, A. G.; Murray, C. K.; Margerum, D. .W.
Inorg. Chem. 1978, 17, 1630–1634. (e) Bal, W.; Djuran, M. I.; Margerum, D. W.;
Gray, E. T., Jr.; Mazid, M. A.; Tom, R. T.; Nieboer, E.; Sadler, P. J. Chem.
Commun. 1994, 1889–1890.
(37) (a) Shearer, J.; Zhao, N. Inorg. Chem. 2006, 45, 9637–9639.
(b) Shearer, J.; Dehestani, A.; Abanda, F. Inorg. Chem. 2008, 47, 2649–2660.
(38) Mullins, C. S.; Grapperhaus, C. A.; Frye, B. C.; Wood, l. H.; Hay,
A. J.; Buchanan, R. M.; Mashuta, M. S. Inorg. Chem. 2009, 48, 9974–9976.
(39) Recently, Darensbourg and co-workers and Harrop and co-workers
prepared models with imidazole and pyridine ligations, respectively, to
model the histidine ligation of the Ni-SOD, while Jenson and co-workers
synthesized models with trispyrazolyl tripodal ligand and thiocabamate
ligands. (a) Jenkins, R. M.; Singleton, M. L.; Almaraz, E.; Reibenspies, J. H.;
Darensbourg, M. Y. Inorg. Chem. 2009, 48, 7280–7293. (b) Gale, E. M.; Patra,
A. K.; Harrop, T. C. Inorg. Chem. 2009, 48, 5620–5622. (c) Ma, H.; Chatto-
padhyay, S.; Petersen, J. L.; Jensen, M. P. Inorg. Chem. 2008, 47, 7966–7968. (d)
Ma, H.; Wang, G.; Yee, G. T.; Petersen, J. L.; Jensen, M. P. Inorg. Chim. Acta
2009, 362, 4563–4569.

5396 Inorganic Chemistry, Vol. 49, No. 12, 2010
Scheme 1. Structures of Ligands
Mathrubootham et al.
to þ15 °C over an hour, diluted with 200 mL of CH2Cl2,
washed with 1 M HCl (100 mL), H2O (100 mL), and brine
(100 mL), dried (Na2SO4), and concentrated via rotary eva-
poration. A sticky foam-like material was obtained, and this
material was dried under vacuum. Yield. 3.72 g, 85%.
Step 2. Preparation of N-[2-(2-Triphenlmethylthio)ethyl)amino)-
acetyl]-S-(triphenylmethyl)-2-aminoethanethiol. N-(2-Bromoac-
etyl)-S-(triphenylmethyl)-2-aminoethanethiol was converted to
N-[2-(2-triphenlmethylthio)ethyl)amino)acetyl]-S-(triphenylmethyl)-
2-aminoethanethiol using a previously reported procedure.40b
Step 3. Ligand H4L2. Triethylsilane (2-3 mL) was added to a
solution of N-[2-(2-triphenylmethylthio)ethyl)amino)acetyl]-S-
(triphenylmethyl)-2-aminoethanethiol (3.40 g, 5 mmol) in tri-
fluoroacetic acid (30 mL) cooled in an ice bath. A bright yellow
color formed initially followed by the formation a white pre-
cipitate. The mixture was diluted with hexanes (50 mL) and
water (50 mL), and the aqueous phase was separated, washed
with several portions of hexanes, and evaporated to a colorless
oil. The oil was triturated with diethyl ether to yield a white solid
(trifluoroacetate salt) that was collected, washed with ether, and
dried in vacuo. Yield.1.49 g, 96.8%, 1H NMR (DMSO-d6): δ 2.58
(m, 2H, CH2) merged with residual solvent peak, 2.74 (t, 2H,
CH2), 3.13 (t, 2H, CH2), 3.32 (q, 2H, CH2), 3.78 (s, 2H, CH2), 8.68
(t, 1H, NH), 9.02 (br, 2H, NH2).
calculations were performed to explain the experimentally
observed properties. These results and their implications with
respect to the possible role of mixed amine/amidate ligation in
Ni-SOD are discussed.
Experimental Section
General Methods. Solvents were purified according to stan-
dard procedures and were freshly distilled under N2 atmosphere
prior to use. All chemicals/solvents were obtained from Aldrich
or Acros Organics unless otherwise noted. All air-sensitive che-
mical manipulations were performed in a glovebox under an
atmosphereof N2 or using standard Schlenk-line techniques under
an Ar atmosphere. The ligand N-(2-mercaptoacetyl)-N0-(2-mer-
captoethyl)glycinamide H4(L1) was synthesized according to
literature procedure.40a
Preparation of (Et4N)2({Ni(L1)}2Ni) H2O (1). NaOMe (2.27 g,
3
Physical Methods. Optical spectra were obtained on a Hewlett-
Packard 8453 diode array spectrophotometer. Electrochemistry
was performed under anaerobic conditions in dimethylformamide
(DMF) on a BAS CV-50W voltammetric analyzer using the fol-
lowing conditions: ∼1 mM sample, 100 mV/s scan rate, 0.1 M
N-tetrabutylammonium perchlorate (TBAP) as supporting elec-
trolyte, Pt disk working electrode, Ag/Agþ (0.01 M AgNO3/0.1 M
TBAP in CH3CN) reference electrode, and Pt counter electrode.
Ferrocene was used as a standard, and all potentials are reported
relative to NHE (Fc/Fcþ = þ0.720 V versus NHE in DMF41).
Under our experimental conditions, the potential for the Fc/Fcþ
couple was þ0.083 V versus Ag/Agþ reference electrode. NMR
spectra were recorded on a Varian 300 MHz spectrometer at
the NMR facility, Department of Chemistry, Michigan State
University.
42 mmol) in 20 mL of CH3OH was added to a solution of ligand
H4L1 (2.08 g, 10 mmol) in 20 mL of CH3OH. To the clear yellow
solution was added a methanolic solution (5 mL) of Et4NCl
(3.31 g, 20 mmol) resulting in the appearance of a turbid yellow
solution. The solution was stirred for 30 min followed by the
dropwise addition of Ni(OAc)2.4H2O (2.49 g, 10 mmol) in 50 mL
of CH3OH. The initial pale yellow color of the solution changed to
red and then to brownish green. This mixture was allowed to stir
for 2 h, and the solvent was removed in vacuo. The residue was
extracted with CH3CN, and the extract was filtered, concentrated,
and layered with diethyl ether to cause the formation of green
crystals. Yield: 0.482 g (17% based on Ni). (ESI-): (m/z) 713.9
[M - Et4N]-.
Preparation of the Mononuclear Complex (Et4N)2(Ni(L1))
3
2H2O (2). NaOMe (0.24 g, 4.4 mmol) in 5 mL of CH3OH was
added to a solution of H4L1 (0.028 g, 0.132 mmol) in 20 mL of
CH3OH. This mixture was slowly added to a green solution of
1 (0.1 g, 0.12 mmol) in 50 mL of CH3CN resulting in the form-
ation of a red solution. The reaction mixture was stirred for 10 min,
and a solution of Et4NCl (0.08 g, 0.48 mmol) in CH3CN (10 mL)
was added, resulting in the precipitation of NaCl. The mixture was
stirred for 30 min, and the solvent was evaporated to dryness. The
red solid was extracted with CH3CN, and the extract was filtered
and layered with diethyl ether to obtain red crystals. The crystals
were filtered and washed with diethyl ether. Yield.0.094 g (47%
The EPR data were collected on a Bruker ESP-300E X-band
EPR spectrometer using an ST4102 cavity resonator. The
sample temperature was maintained at 10 K using an Oxford
ESR-900 liquid helium cryostat system equipped with an ITC-
502 temperature controller. The microwave frequency was
measured with an EIP model 25B frequency counter, and the
magnetic field was calibrated using weak pitch as the standard.
Conditions for the EPR measurement: microwave frequency,
9.4628 GHz; microwave power, 0.2 mW; field modulation,
0.8 mT at 100 kHz; time constant, 163 ms; receiver gain, 1.0ꢀ 106;
scans averaged, 4; sample temperature, 10 K.
Synthesis. N-(200-Mercaptoethyl)-2-[(20-mercaptoethyl)amino]-
acetamide, H4L2. This ligand was synthesized as the trifluoroa-
cetate salt in three steps using modified literature procedures.40b,c
Step 1. Preparation of N-(2-Bromoacetyl)-S-(triphenylmethyl)-
2-aminoethanethiol. A solution of bromoacetic acid (1.39 g,
10 mmol) in CH2Cl2 (25 mL) was treated with N-methylmor-
pholine (15 mmol, 1.70 mL) and cooled to -15 °C (salt-ice
bath). After dropwise addition of isobutylchloroformate
(11.5 mmol, 1.5 mL), the resulting mixture was stirred for 10 min
at -15 °C. A slurry of 2-(triphenylmethythio)ethylamine
(11.5 mmol, 3.70 g) in 25 mL of CH2Cl2 was added at -15 °C
in 5 aliquots to the reaction mixture. The solution was warmed
1
based on Ni). H NMR δ 1.30 (t-t, 24H, tetraethylammonium
cation-CH3), 2.02 (t, 2H, CH2), 2.56 (s, 2H, CH2), 2.78 (t, 2H, CH2),
3.26 (s, 2H, CH2), 3.40 (q, 12H, tetraethylammonium cation-CH2);
MS (ESI-): (m/z) 392.1 [M - Et4N]-.
Preparation of K(Ni(HL2)) (3). KOH (1.40 g, 25 mmol) in
20 mL of CH3OH was added to a solution of ligand H4L2 (1.54 g,
5 mmol) in 20 mL of CH3OH. The solution was stirred for 30 min
followed by the dropwise addition of a solution of Ni(OAc)2
3
4H2O (1.25 g, 5 mmol) in 50 mL of CH3OH. The solution
changed to red immediately and then reddish brown. This
mixture was allowed to stir for 2 h, and the solvent was removed
in vacuo. The resulting residue was dissolved in 50:50 CH3OH/
CH3CN, filtered, and layered with diethyl ether resulting in the
formation of red crystals. Yield. 1.1 g (75.9%) 1NMR (DMSO-
d6): δ 1.74 (t, 2H, CH2), 1.84 (t, 2H, CH2), 2.38 (t, 2H, CH2) br,
2.66 (t, 2H, CH2) br, 2.96 (s, 2H, CH2). MS (ESI-): (m/z) 248.9
[M - K]-. The same cationic potassium complex was obtained
even in the presence of Et4NCl, as confirmed by NMR and
X-ray crystallography.
(40) (a) Brenner, D.; Davison, A.; Lister-James, J.; Jones, A. G. Inorg.
Chem. 1984, 23, 3793–3797. (b) O'Neil, J. P.; Wilson, S. R.; Katzenellenbogen,
€
J. A. Inorg. Chem. 1994, 33, 319–323. (c) Oya, S.; Plossl, K.; Kung, M.-P.;
Stevenson, D. A.; Kung, H. F. Nucl. Med. Biol. 1998, 25, 135–140.
(41) Barrette, W. C., Jr.; Johnson, H. W., Jr.; Sawyer, D. T. Anal. Chem.
1984, 56, 1890–1898.

Article
Inorganic Chemistry, Vol. 49, No. 12, 2010 5397
Table 1. Details of Structure Determination, Refinement, and Experimental Parameters for Compounds 1, 2, and 3
1
2
3
chemical formula
formula weight
crystal system
space group
C28H58N6Ni3O5S4
863.17
monoclinic
P21/c
13.0861(4)
9.7990(3)
28.5967(9)
90
90.078(2)
90
3667.0(2)
4
1.797
C22H52N4NiO4S2
559.51
orthorhombic
Pna21
14.3804(4)
13.8008(4)
14.2473(3)
90
C6H11KN2NiOS2
289.10
monoclinic
P21/c
8.5857(2)
11.6498(3)
11.2093(3))
90
111.7573(15)
90
˚
a (A)
˚
b (A)
˚
c (A)
R (deg)
β (deg)
γ (deg)
90
90
3
˚
V (A )
2827.53(13)
4
0.867
1041.30(5)
4
2.624
Z
μ(Mo KR) (mm-1
crystal size (mm)
F (000)
)
0.36 ꢀ 0.34 ꢀ 0.08
0.363 ꢀ 0.332 ꢀ 0.233
0.25 ꢀ 0.20 ꢀ 0.13
1824
1216
592
2θ range (deg)
T (K)
number of data collected
number of unique data
observed data [I > 2σ(I)]
Rint (%)
1.56 to 25.36
173(2)
39471
6687
5713
4.40
435
3.46
6.87
2.05 to 25.34
173(2)
15683
5042
4464
6.08
307
3.70
8.68
2.55 to 27.46
150(1)
4414
2371
2042
1.80
158
3.35
8.18
number of parameters
R1 (%)a
wR2 (%)b
Sc
1.072
-0.326
0.495
0.992
-0.260
0.984
1.040
-0.756
0.846
-3
ΔFmin (e-
ΔFmax (e-
A
)
˚
-3
˚
A
)
P
P
P
P
P
a
R1 = ||Fo| - |Fc||/ |Fo|. bwR2 = { [w(Fo2 - Fc2)2/ [w(Fo2)2]}1/2. cGoodness-of-fit on F2 = [ [w(Fo2 - Fc2)2]/(n - p)]1/2, where n is the number
of reflections and p is the number of parameters refined.
X-ray Structure Determination. A green hexagonal plate
crystal of 1 with dimensions 0.08ꢀ0.34ꢀ0.36 mm was mounted
on a Nylon loop using a very small amount of paratone oil. Data
were collected using a Bruker CCD (charge coupled device)
based diffractometer equipped with an Oxford Cryostream low-
temperature apparatus operating at 173 K. Data were measured
using ω and j scans of 0.5° per frame for 30 s. The total number
of images was based on results from the program COSMO42
where redundancy was expected to be 4.0 and completeness of
instrument and methods as described for 1. The structure was
solved in the orthorhombic space group Pna21. All non-hydro-
gen atoms were refined anisotropically. Hydrogens were calcu-
lated by geometrical methods and refined as a riding model. The
Flack47 parameter was used to determine chirality of the crystal
studied. A value of zero or one indicates that the crystal is
composed of a single enantiomer, while a value of 0.5 indicates a
racemic crystal. The Flack parameter was refined by use of the
twin law to 0.456(12), indicating a racemic crystal.
˚
100% out to 0.83 A. Cell parameters were retrieved using APEX
A red prism-shaped crystal of 3 with dimensions 0.25ꢀ0.20ꢀ
0.13 mm was mounted on a glass fiber with traces of viscous oil
and then transferred to a Nonius Kappa CCD diffractometer
II software43 and refined using SAINT on all observed reflec-
tions. Data reduction was performed using the SAINT soft-
ware44 which corrects for Lorentz-polarization. Scaling and
absorption corrections were applied using SADABS45 multi-
scan technique, supplied by George Sheldrick. The structures
were solved by the direct method using the SHELXS-97 pro-
gram and refined by least-squares method on F2, SHELXL-97,
incorporated in SHELXTL-PC V 6.10.46 The structure was
solved in the space group P21/c. All non-hydrogen atoms were
refined anisotropically. Hydrogens were calculated by geome-
trical methods and refined as a riding model. The twin law was
applied and refined to a value of 0.262 for the twin law matrix
[1 0 0, 0 -1 0, 0 0 -1]. Crystallographic summary of data and
selected experimental information is given in Table 1, and
selected bond lengths and angles are given in Table 2.
˚
equipped with Mo KR radiation (λ = 0.71073 A). Ten frames
of data were collected at 150(1) K with an oscillation range of
1 deg/frame and an exposure time of 20 s/frame.48 Indexing and
unit cell refinement based on all observed reflection from those
10 frames indicated a monoclinic P lattice. A total of 4414 refl-
ections (Θmax=27.46°) were indexed, integrated, and corrected
for Lorentz-polarization and absorption effects using DENZO-
SMN and SCALEPAC.49 Post refinement of the unit cell gave
˚
˚
˚
a = 8.5857(2) A, b = 11.6498(3) A, c = 11.2093(3) A, and V =
3
1041.30(5) A . Axial photographs and systematic absences were
˚
consistent with the compound having crystallized in the mono-
clinic space group P21/c. The structure was solved by a combi-
nation of direct methods and heavy atom using SIR 97.50 All of
the non-hydrogen atoms were refined with anisotropic displace-
ment coefficients. Hydrogen atoms were either located and
refined isotropically or were assigned isotropic displacement
coefficients U(H)=1.2U(C), and their coordinates were allowed
to ride on their respective carbons using SHELXL-97.51 The
A red block crystal of 2 with dimensions 0.36ꢀ0.33ꢀ0.23 mm
was mounted on a Nylon loop using a very small amount of
paratone oil. The data collection was performed using the same
(42) COSMO V1.56, Software for the CCD Detector Systems for Deter-
mining Data Collection Parameters ; Bruker Analytical X-ray Systems:
Madison, WI, 2006.
(43) APEX2 V 1.2-0 Software for the CCD Detector System; Bruker
Analytical X-ray Systems, Madison, WI, 2006.
(44) SAINT V 7.34 Software for the Integration of CCD Detector System;
Bruker Analytical X-ray Systems: Madison, WI, 2001.
(47) Flack, H. D. Acta Crystallogr. 1983, A39, 876–881.
(48) COLLECT Data Collection Software; Nonius B.V.: Delft, The
Netherlands, 1998.
(49) Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276, 307–326.
(50) Altomare, A.; Burla, M.C.; Camalli, M.; Cascarano, G.; Giacovazzo,
C.; Guagliardi, A.; Moliteni, A.G. G.; Polidori, G.; Spagna, R. SIR97
(Release 1.02) - A program for automatic solution and refinement of crystal
structure.
(45) SADABS V2.10 Program for absorption corrections using Bruker-
AXS CCD based on the method of Robert Blessing; Blessing, R. H. Acta
Crystallogr. 1995, A51, 33-38.
(46) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112-122.

5398 Inorganic Chemistry, Vol. 49, No. 12, 2010
Mathrubootham et al.
˚
Table 2. Selected Bond Lengths (A) and Angles (deg) for the Anionic Complexes
character using the pre-edge feature of Ni(DACO) as a standard
(46(2)%),53 which by our fitting procedure corresponds to an
intensity of 1.22(1) units.
({NiII(L1)}2NiII)2- (1), (NiII(L1))2- (2), and (NiII(HL2))- (3)
1
2
3
77 K Electronic Absorption Spectroscopy. Under an atmo-
sphere of N2, solids of the monomeric NiIIN2S2 compounds
were finely ground in Nujol and placed between two quartz
plates. The plates were then mounted in a custom-built low-
temperature optical cryostat, which was cooled to 77 K. Data
were then recorded on an OLIS-CARY 14. Each spectrum
represents the average of 5 scans. Data were deglitched and smo-
othed prior to examining the second derivative of the averaged
spectra. The resulting transitions were deconvoluted into four
Gaussian line shapes (Supporting Information, Table S13). The
TT-multiplets software package was then used to generate
correlation diagrams for NiII contained within D4h symmetry
using previously determined values for 10Dq, Ds, and Dt to allow
for a fitting of the resulting Slater-Condon parameters.37,54
Electronic Structure Calculations. All electronic structure and
excited-state calculations were performed using the software
package ORCA 2.6.35.55 All calculations on S = 0 NiII com-
plexes were spin-restricted and all calculation on low-spin S =
1/2 NiIII complexes were spin-unrestricted. Geometry-opti-
mized (GO) structures were obtained for all complexes starting
from the available X-ray crystallographic coordinates of the
reduced NiII compounds, and utilized the following thresholds
in the GO procedure (in au): root-mean square (rms) and
maximum forces of 0.00003 and 0.0001, respectively; rms and
maximum gradients 0.002 and 0.0003, respectively. These cal-
culations utilized Becke’s three-parameter hybrid functional for
exchange along with the Lee-Yang-Parr correlation func-
tional (B3LYP),56 and utilized Ahlrichs’ triple-ζ valence basis
set with one set of first polarization functions (TZVP).57 All
geometry-optimized structures were subjected to vibrational
analyses to ensure only real vibrational modes were present.
Single-point calculations also utilized the B3LYP functional,
but all atoms were treated with Ahlrichs’ triple-ζ valence basis
set with two sets of first polarization functions (TZVPP)57
except Ni, S, and N, which were treated with Ahlrichs’ def2
triple-ζvalence basis set with two sets of first polarization functions
and diffuse functions (Def2-aug-TZVPP).57,58 Iso-surface plots
were generated with the visualization package gOpenMol59a,b or
Molekel.59c All population analyses were performed within
Ni(1)-N(1)
Ni(1)-N(2)
Ni(1)-S(1)
Ni(1)-S(2)
Ni(2)-S(1)
Ni(2)-S(2)
1.846(3)
1.841(4)
2.1504(12)
2.1405(11)
2.2420(10)
2.2419(10)
1.858(3)
1.874(3)
2.1812(10)
2.1886(9)
1.862(2)
1.937(3)
2.1671(8)
2.1711(7)
N(1)-Ni(1)-N(2)
N(1)-Ni(1)-S(1)
N(2)-Ni(1)-S(1)
N(1)-Ni(1)-S(2)
N(2)-Ni(1)-S(2)
S(2)#1-Ni(1)-S(1)
N(2)-Ni(1)-Ni(2)
N(1)-Ni(1)-Ni(2)
86.11(15)
91.19(11)
168.46(11)
169.27(11)
90.80(11)
89.80(4)
84.96(12)
173.44(9)
88.57(9)
87.87(8)
172.68(9)
84.46(12)
89.57(7)
172.37(10)
174.78(8)
90.62(9)
118.79(11)
119.62(10)
nitrogen atom N2 exhibits orientation disorder into two sites
(75:25). The weighting scheme employed was w = 1/[σ2(Fo2) þ
(0.0379P)2 þ 1.7429P] where P=(Fo þ 2Fc )/3. The refinement
converged to R1=0.0335, wR2=0.0818, and S=1.04 for 2042
reflections with I > 2σ(I), and R1=0.0419, wR2=0.0859, and
S= 1.04 for 2371 unique reflections and 158 parameters.52
Sulfur K-edge X-ray Absorption Spectroscopy. The nickel
complexes (NiII(L1))2- (2) and (NiII(HL2))- (3) were finely
ground under an atmosphere of N2, and the powders were
spread onto sulfur-free Kapton tape (verified by scanning the
tape over the energy region used in this study) forming thin
layers. These samples were then mounted onto polycarbonate
sample holders. Data were collected at room temperature on
beamline X19A at the National Synchrotron Light Source
(Brookhaven National Laboratories; Upton, NY). Spectra were
recorded in fluorescence mode using a passivated implanted
planar silicon detector (Canberra Industries). The X-ray beam
passed from an incident ionization chamber to the sample
chamber, which were both continually purged with He and
separated from one another by polyethylene windows (5 μm
thickness). Data were collected from 200 eV below the edge to
300 eV above the edge. In the pre-edge region (2270-2465 eV),
data were collected in 5 eV steps; in the edge region (2465-
2475 eV), data were collected in 0.1 eV steps, and in the near edge
region (2475-2765 eV) data were collected in 0.5 eV steps. All
spectra were calibrated against the spectrum of Na2SO4, which
was independently recorded in between each scan. The energy
drifted by less than 0.1 eV from scan to scan. All reported spectra
represent the average of 5 scans.
2
2
€
ORCA 2.6.35 and are reported to the nearest 0.1%. Lowdin
populations are reported in the text of the manuscript while
Mulliken populations are reported in the Supporting Informa-
tion and ref 66 for comparative purposes. Excited-state calcula-
tions were performed using the spectroscopy oriented con-
figuration interaction (SORCI) formalism of Neese.60 These
calculations employed Ahlrichs’ Def2-aug-TZVPP basis-set on
all atoms.56,57 For the SORCI calculations, the selection thresh-
old was set to 10-6 Eh, the prediagonalization threshold was set
to 10-6, and the natural orbital threshold was set to 10-5. All
orbitals between -3 to 5 Eh of the highest occupied molecular
Data were averaged and a baseline was applied to each
spectrum by fitting the pre-edge region to a polynomial func-
tion. This baseline was then subtracted from the whole spec-
trum. The region above the edge jump was fit to a two knot cubic
spline and the data normalized to the edge height. Pre-edge and
the rising-edge features were modeled as pseudo-Voigt line
shapes (a 1:1 sum of Gaussian and Lorentzian line shapes).
Each spectrum required the use of only one peak per feature, and
valid fits to the data were judged by matches to the second
derivatives of the spectra. The intensity values of the pre-edge
features, which represent the average of four different fits to the
data (differing by less than 2%), are the products of the peak
widths at half height and intensities of the pseudo-Voigt line
shapes. All pre-edge intensities were then converted into %S(3p)
(54) (a) de Groot, F. M. F. J. Electron Spectrosc. Relat. Phenom. 1994, 67,
529–622. (b) de Groot, F. M. F. Chem. Rev. 2001, 101, 1779–1808.
(55) Neese, F. ORCA Version 2.6.35; Universitat Bonn: Bonn, Germany,
2008.
(56) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Becke, A. D.
J. Chem. Phys. 1993, 98, 1372–1377. (c) Lee, C. T.; Yang, W. T.; Parr, R. G.
Phys. Rev. B: Condens. Matter 1988, 37, 785–789.
€
(57) (a) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571–
2577. (b) Ahlrichs, R. and coworkers, unpublished work.
(51) Sheldrick, G. M. SHELX97 [Includes SHELXS97, SHELXL97,
CIFTAB ] - Programs for Crystal Structure Analysis (Release 97-2); University
(58) (a) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007–1023. (b) Woon,
D. E.; Dunning, T. H., Jr. J. Chem. Phys. 1994, 100, 2975–2988. (c) Woon, D. E.;
Dunning, T. H., Jr. J. Chem. Phys. 1993, 98, 1358–1371.
(59) (a) Bergman, D. L.; Laaksonen, L.; Laaksonen, A. J. Mol. Graphics
Modell. 1997, 15, 301–306. (b) Laaksonen, L. J. Mol. Graphics 1992, 10, 33–34.
(c) Varetto, U. Molekel v. 5.4; Swiss National Supercomputing Center: Manno,
Switzerland.
€
€
of Gottingen: Gottingen, Germany, 1997.
P
P
P
P
(52) R1 = ||Fo| - |Fc||/ |Fo|, wR2 = [ w(Fo - Fc2)2/ w(Fo2)2]1/2
,
2
P
and S=Goodness-of-fit on F2=[ [w(Fo2 - Fc2)2]/(n - p)]1/2, where n is the
number of reflections and p is the number of parameters refined.
(53) Dey, A.; Jeffery, S. P.; Darensbourg, M.; Hodgson, K. O.; Hedman,
B.; Solomon, E. I. Inorg. Chem. 2007, 46, 4989–4996.
(60) Neese, F. J. Chem. Phys. 2003, 119, 9428–9443.

Article
Inorganic Chemistry, Vol. 49, No. 12, 2010 5399
Figure 2. ORTEP view of [Ni{Ni(L1)}2]2- (anion of 1) showing 50%
probability displacement ellipsoids. H atoms, counterions, and solvent
molecules are not shown.
Figure 3. ORTEP view of [Ni(L1)]2- (anion of 2) showing 50% prob-
ability displacement ellipsoids. H atoms, counterions, and solvent mole-
cules are not shown.
orbital (HOMO)/lowest unoccupied molecular orbital (LUMO)
gap were considered in the SORCI treatment. In addition, the
CAS-SCF(12,7) reference space was used,61 and we examined
the first seven spin allowed excitations.62 EPR g-values and
superhyperfine coupling constants were calculated by solving
the coupled-perturbed SCF equations,63 using the B3LYP
functional, the TZVP basis-set for all non-ligating atoms, the
Def2-aug-TZVPP basis set for Ni and S, and Kutzelnigg’s IGLO
III basis set for ligated N atoms.64 All orbitals from -100 to 100
Eh of the HOMO/LUMO gap were considered, with the center
of electronic charge defined as the origin of the g-matrix.
Results
Synthesis. The deprotonation of the ligand H4L1 with 4
equiv of NaOMe, followed by the addition of 1 equiv of
Ni(acetate)2 4H2O afforded the green trinuclear complex
3
({NiII(L1)}2NiII)2- (1) whichwasisolatedastheEt4Nþ salt.
The desired mononuclear bisamidate complex (NiII(L1))2-
(2) was obtained as the Et4Nþ salt by breaking apart the
trinuclear complex 1 by the addition of 1 equiv of the ligand
H4L1 and 4 equiv of NaOMe. The red mixed amine/amidate
complex (NiII(HL2))- (3) was prepared by deprotonating
the ligand H4L2 using KOH, followed by the addition of
Figure 4. ORTEP view of [Ni(HL2)]- (anion of 3) showing 50% prob-
ability displacement ellipsoids. H atoms, counterions, and solvent mole-
cules are not shown.
Ni(acetate)2 4H2O in CH3OH and isolated as the Kþ salt.
3
X-ray Crystal Structure Descriptions. The Oak Ridge
thermal ellipsoid plot (ORTEP) of complexes 1, 2, and 3 is
shown in the Figures 2, 3, and 4, respectively, and the
selected bond lengths and angles are given in Table 2.
Complexes 1 and 3 crystallize in the monoclinic space
group P21/c while complex 2 crystallizes in the orthor-
hombic space group Pna21 with two independent mole-
cules in the same asymmetric unit. The overall geometry
of the trinuclear anion of 1 can be described as a chair
consisting of three square planes in which a central Ni ion
sits at an inversion center and bridges two anionic bisa-
midate complexes (2). The dihedral angle formed between
these planes is 111.4°, although the deviation from the
NiN2S2 least-squares planes is appreciable for Ni (0.404
˚
˚
˚
˚
A), C2 (0.609 A), C6 (0.305 A), and C7 (0.420 A). In one
of the trimers in the asymmetric unit, all of the Ni-S
distances in the NiS4 plane refine to the same value by
coincidence, while in the second molecule the two inde-
pendent Ni-S distances are slightly different. In both
molecules of the asymmetric unit, the Ni-S bond lengths
in the NiS4 plane are similar to those reported by us14
and others.16b,19b,19c,21 Conversely, while the Ni-S bond
lengths in the NiN2S2 planes are similar to ({Ni-
(ema)}2Ni),16b where ema = N,N0-ethylenebis-2-mercap-
toacetamide, they are significantly lower than in the 6,5,6
chelated bis[NiIIN,N0-ethylenebis(3-mercaptopropiona-
mide)]NiII trimer complex14 and the 6,5,5 chelating ring
complex reported by Holm.21b
(61) CAS(n,m): complete active space composed of
n electrons in
m orbitals; for this case we consider the LUMO through HOMO-6 in our
reference space.
(62) Expanding the reference space does not in general lead to an
improvement in calculated transitions energies or strengths for the lowest
energy transitions in similar NiIIN2S2 complexes. Shearer, J., unpublished
results.
In the mononuclear bisamidate complex (NiII(L1))2- (2),
the asymmetric unit contains a nickel coordinated by two
amido nitrogens and two thiolate sulfurs in a square-planar en-
vironment. The Ni-NbonddistancesareintherangeforNi-
N amidate bonds in square-planar complexes.14,17b,21-23
Likewise, the Ni-S distances are in the range of other
NiN2S2 square-planar thiolate complexes.14,17b,21-23 Inter-
estingly, the asymmetrically positioned carbonyl groups
(63) (a) Neese, F. J. Chem. Phys. 2001, 115, 11080–11096. (b) Neese, F.
Curr. Opin. Chem. Biol. 2003, 7, 125–135.
(64) Kutzelnigg, W.; Fleischer, U.; Schindler, M. The IGLO-Method:
Ab-initio Calculation and Interpretation of NMR Chemical Shifts and
Magnetic Susceptibilities. In NMR, Basic Principles and Progress; Diehl,
P., Fluck, E., G€unther, H., Kosfield, R., Seelig, J., Eds.: Springer Verlag: Berlin,
1990; Vol. 23, pp 165-262.

5400 Inorganic Chemistry, Vol. 49, No. 12, 2010
Mathrubootham et al.
Figure 5. Room temperature electronic absorption spectra of [NiII-
(L1)]2- (2) (dashed blue line) and [NiII(HL2)]- (3) (solid red line) in
methanol.
Figure 6. Sulfur K-edge X-ray absorption spectra of (NiII(L1))2- (2)
(dashed blueline) and(NiII(HL2))- (3) (solidgreen line). There isa shiftin
the pre-edge feature of (NiII(HL2))- (3) versus (NiII(L1))2- (2) to higher
energy by ∼0.3 eV.
in the ligand backbone impose slight differences in the
bond lengths and angles compared to the analogous sym-
metric 5,5,5 chelate ring NiN2S2 square-planar bisami-
date complex previously reported.14,22b
feature results from the promotion of a S(1s) electron into
the unfilled LUMO, which is formally a Ni(3d) orbital. In
order for this transition to have any intensity, the orbital
must have some degree of S(3p)-character from Ni/S
covalency. Therefore, integration of the area under this
peak is a direct measure of the Ni/S covalent character of
this orbital.
The X-ray structure of the mixed amine/amidate com-
plex (NiII(HL2))- (3) reveals a nickel ion in a square-
planar coordination geometry, with the plane defined by
two thiolato sulfurs, one carboxamido nitrogen, and an
amine nitrogen. The Ni ion deviates from the least-
˚
squares N2S2 plane by only 0.097 A. Predictably, the
˚
Ni-N (amide) distance (Ni1-N1, 1.862(2) A) is shorter
The bisamidate compound (NiII(L1))2- (2) has a nor-
malized peak intensity of 0.57(2) relative to the edge,
while that of the mixed amine/amidate compound (NiII-
(HL2))- (3) has a normalized peak intensity of 0.88(2)
relative to the edge. This corresponds to a %S(3p)
character of 21(1) and 31(1)% for (NiII(L1))2- (2) and
(NiII(HL2))- (3), respectively. These compare well with
the previously determined37 %S(3p) character of the
structurally related (NiII(BEAAM))- (BEEAM = N-{2-
[benzyl(2-mercapto-2-methylpropyl)amino]ethyl-2-mer-
capto-2-methylpropionamide) and (NiII(emi))2- (emi =
N,N0-ethylenebis(2-(mercapto)propionamide)) complexes
suggesting that the absence of the gem-dimethyl groups is
not having a dramatic influence on the LUMO wave
functions. Also similar to (NiII(BEAAM))- and (NiII-
(emi))2-, there is a shift in the pre-edge feature of
(NiII(HL2))- (3) versus (NiII(L1))2- (2) to higher energy
by ∼0.3 eV. As one would not expect the core S(1s) orbital
to be dramatically influenced by the change in N-donor
fromanaminetoanamidate, thissignifiesa destabilization
of the LUMO in the bisamidate complex versus the mixed
amine/amidate complex because of ligand-field effects.
Therefore, the Ni(3dx2-y2) LUMO of (NiII(L1))2- (2) is
situated ∼2400 cm-1 higher in energy than that of
(NiII(HL2))- (3).
˚
than the Ni-N(amine) (Ni1-N2, 1.937(3) A) distance.
˚
As expected, the Ni-S bond (Ni-S1, 2.1671(8) A) trans
to the amine N is slightly shorter than the Ni-S bond
˚
(Ni-S2, 2.1711(7) A) trans to the amidate N because
N(amide) ligands are stronger σ donors than N(amine)
ligands and display more trans influence. The difference,
however, is not appreciable.
Room Temperature Solution Electronic Absorption
Properties. The trinuclear complex ({NiII(L1)}2NiII)2-
(1) displays two absorption bands at 21,881 cm-1 (8400
M
-1 cm-1) and 14,749 cm-1 (9600 M-1 cm-1) in CH3CN
similar to the other reported trinuclear NiN2S2
complexes.14-23 The UV-vis absorption spectrum of
complex (NiII(L1))2- (2) in CH3OH (Figure 5) exhibits
two ligand field transitions at 22,779 cm-1 (400 M-1
cm-1) and 18,382 cm-1 (100 M-1 cm-1), which is typical
for NiII complexes with dicarboxamido-dithiolato (N2S2)
coordination.14-23 The corresponding mixed amine/ami-
date complex (NiII(HL2))- (3) also exhibits two ligand
field transitions at 22,272 cm-1 (340 M-1 cm-1) and
17,544 cm-1 (a shoulder with ε ∼ 70 M-1 cm-1) in
CH3OH (Figure 5). The ligand field absorption bands
of complex (NiII(HL2))- (3) are shifted slightly to lower
energy compared to those of the bisamidate complex
(NiII(L1))2- (2). By replacing an amide with an amine in
(NiII(HL2))- (3), the anionic charge decreases from -2
to -1, exerting a weaker in-plane ligand field, thereby
decreasing the energies of the d f d transitions.
77 K Electronic Absorption Spectroscopy. Both (NiII-
(L1))2- (2) and (NiII(HL2))- (3) were subjected to low-
temperature (77 K) electronic absorption measurements
to aid in resolving some of the low energy ligand field
transitions from one another (Figure 7). As with the room
temperature samples, we see a low-energy transition at
Sulfur K-edge X-ray Absorption Spectroscopy. The
sulfur K-edge absorption spectra for (NiII(L1))2- (2) and
(NiII(HL2))- (3) are displayed in Figure 6. The pre-edge

Article
Inorganic Chemistry, Vol. 49, No. 12, 2010 5401
Figure 8. X-band EPR spectrum of the [NiIII(L1)]- in DMF solution
made by the chemical oxidation with I2 at 10 K.
oxygen in an CH3CN solution, turning color from red to
green within a few minutes. In the UV-visible spectrum,
a new peak appears around 350 nm after (NiII(L1))2- (2)
was exposed to air. Conversely, the UV-visible spectrum
of the mixed amine/amidate complex (NiII(HL2))- (3)
exhibited little change even 5 h after being exposed to air
in a CH3OH solution.
The electrochemical data of complex (NiII(L1))2- (2)
suggested the possibility of trapping the NiIII oxidation
state. We attempted to generate the NiIII species by
chemical oxidation in DMF using I2. The red bisamidate
NiII complex (NiII(L1))2- (2) turned dark green immedi-
ately upon oxidation. The low-temperature EPR spec-
trum of this oxidized complex (Figure 8) is nearly axial
with the experimental g values of g1 = 2.012, g2 = 2.195,
and g3 = 2.228 (gav = 2.145). These values are indica-
tive of an S = 1/2 NiIII species, consistent with those
reported previously.22,23,27d,31 Furthermore, the lack of
observable superhyperfine coupling along g1 is consistent
with a four-coordinate NiIII species. Both experimental
and computational (see Supporting Information) results
suggest the magnitude of the superhyperfine coupling
constant would be too small to observed under normal
X-band conditions.
Figure 7. Low temperature (77 K) UV/vis spectra of (NiII(L1))2- (2)
(top) and (NiII(HL2))- (3) (bottom).
∼22,000 cm-1 and
a corresponding lower-energy
shoulder for both compounds (Supporting Information).
For bisamidate ligated (NiII(L1))2- (2), the second deri-
vative of the lower-energy transition shows significant
asymmetry. This pronounced asymmetry suggests that
this peak is composed of two nearly irresolvable bands at
17,770 and ∼17,300 cm-1, which had to be estimated
from peak-fitting of the absorption spectrum. In contrast,
this band is resolvable into two transitions in the spec-
trum obtained for amine/amidate ligated (NiII(HL2))-
Trapping the NiIII species is challenging, presumably
because of complex degradation and the formation of
EPR-inactive multimeric species, which are also green in
color. Thus, it seems that the ability to trap the NiIII
complexes depends on various factors and further study is
needed to understand more fully the stability of NiIII
species. Interestingly, the mixed amine/amidate complex
(NiII(HL2))- (3) immediately turned from red to bluish
purple upon the addition of I2, but this complex rapidly
decomposed followed by the formation of a precipitate.
Attempts to trap oxidized (NiII(HL2))- (3) at low tem-
peratures have been unsuccessful.
(3), which appear at 17,075 and 18,300 cm-1
.
Electrochemistry. The cyclic voltammograms of the
mononuclear complexes recorded in DMF solvent are
shown in Supporting Information, Figures S4 and S5. A
quasireversible NiIII/NiII redox couple is observed for com-
plex (NiII(L1))2- (2) at -0.129 V versus NHE (-0.849 V vs
Fc/Fcþ) with a ΔEp of 123 mV. Conversely, the mixed
amine/amidate complex (NiII(HL2))- (3) displays an irre-
versible oxidation wave at þ0.262V versusNHE (-0.458 V
vs Fc/Fcþ) and the corresponding reduction wave is not
discerned. One possible explanation is that ligand-based
oxidation of (NiII(HL2))- (3) results in disulfide formation
and is followed by complex decomposition. This is in
contrast to the mixed amine/amidate complex reported
by Shearer and co-workers37 which exhibits a quasi-rever-
sible NiIII/NiII redox couple at approximately 0.320 V
versus NHE (-0.330 V vs Fc/Fcþ) in CH3CN.
Ground-State Electronic Structure of (NiII(L1))2- (2)
and (NiII(HL2))- (3). DFT geometry optimizations were
performed on (NiII(L1))2- (2) and (NiII(HL2))- (3) using
the B3LYP hybrid density functional level of theory
(TZVP basis set; Supporting Information, Figure S2).
Similar to other computational studies on NiN2S2 com-
plexes we find that the geometry optimized (GO) struc-
tures overestimate the Ni-S and Ni-N bond lengths
Oxygenation and Chemical Oxidation Studies. The
bisamidate complex (NiII(L1))2- (2) reacts quickly with

5402 Inorganic Chemistry, Vol. 49, No. 12, 2010
Mathrubootham et al.
Table 3. Selected Computationally-Derived Metric Parameters for (NiII(HL2))-
(3), (NiII(L1))2- (2), NiIII(HL2), and NiIII(BEAAM)
(calculated)
(NiII(HL2))- (NiII(L1))2-
NiIII
(HL2) (BEAAM)
NiIII
(3)
(2)
˚
˚
Ni-S(1) (A)
2.192
2.215
1.962
1.886
172.1
2.221
2.205
1.884
1.891
173.2
2.153
2.118
1.979
1.860
172.3
2.159
2.147
1.998
1.857
168.7
Ni-S(2) (A)
˚
˚
Ni-N(1) (A)
Ni-N(2) (A)
Ni-S(1)-N(1)
(deg)
Ni-S(1)-N(2)
(deg)
Ni-S(2)-N(1)
(deg)
Ni-S(2)-N(1)
(deg)
89.6
172.2
92.5
88.6
172.8
90.0
89.3
172.1
91.7
88.6
170.7
91.6
Ni-S(1)-S(2) (deg)
Ni-N(1)-N(2)
(deg)
96.7
83.7
98.3
84.2
94.5
85.1
96.5
86.3
relative to the crystal structures. The most significant
deviations are found in the Ni-thiolate bond lengths. In
the case of the bisamidate (NiII(L1))2- (2), we calculate
˚
Ni-S bond lengths of 2.221 and 2.205 A, while the ex-
perimental Ni-S bond lengths are only 2.181 and 2.189
˚
A. The Ni-N bond lengths, while still overestimated, are
more in line with the experimental data (1.884calc vs
Figure 9. Molecular orbital energy level diagram and iso-surface plot of
the LUMO to HOMO-2 of (NiII(HL2))- (3) (left) and (NiII(L1))2- (2)
(right). The energies have been normalized to the HOMO, which is the
higher-energy Ni(3d(π)/N/S(π)* orbital. Doubly filled orbitals are de-
picted in black while the unfilled LUMO is depicted in red.
˚
1.858X-ray and 1.891calc vs 1.874X-ray A) (Table 3). Similar
trends are observed for the calculated Ni-S bond lengths
for (NiII(HL2))- (3) (Table 3). All of the calculated bond
lengths are therefore overestimated by no more than
˚
0.045 A. Previous work has demonstrated that this rela-
Ni-character), while in (NiII(L1))2- (2), the percentage of
S(3p)-character in the HOMO is 52.5% (compared to
36.4% Ni-character) (Figure 9). The trends are similar to
those observed for the mixed amine/amidate NiIIN2S2
compound (NiII(BEAAM))- and the bisamidate NiIIN2S2
tively minor difference in geometries between computa-
tionally derived models and the experimental data does
not have a large impact on the electronic and energetic
properties of the computational versus experimental
structures.65 Thus, the use of more exact computational
methods coupled with a larger basis sets is typically not
warranted considering the high computational cost, espe-
cially for a comparative study.
compound (NiII(emi))2- 37b
However, the percentage of
.
S(3p)-character in the HOMO of (NiII(emi))2- and (NiII-
(BEAAM))- is approximately 20% lowerthan found in the
corresponding complexes investigated in this study. This is
likely responsible for the decreased stability of oxidized
(NiII(HL2))- (3) relative to Ni(BEAAM) upon oxidation
(vide supra).
The LUMOs of (NiII(L1))2- (2) and (NiII(HL2))- (3)
are Ni(3dx2-y2)/S(σ)/N(σ)* antibonding in character,
while the HOMO and HOMO-1 are both Ni(3d(π))/
S(π)* in character (Figure 9). Comparison of the degree
of S(3p) obtained from the S K-edge studies and the com-
The decrease in S(3p)-character in HOMO and HOMO-
1 of amine/amidate ligated (NiII(HL2))- (3) versus bisami-
date ligated (NiII(L1))2- (2) is a result of two factors. One
(minor) factor is the increase in Ni-amide covalency of the
bisamidate ligated 2 versus amine/amidate ligated 3; the
HOMO N-amide character increases from 0.2% in (NiII-
(HL2))- (3) to 2.6% in (NiII(L1))2- (2). The major factor
impacting the HOMO and HOMO-1 S(3p)-character is an
increase in the overall energy of the 3d-manifold of
(NiII(L1))2- (2) relative to (NiII(HL2))- (3). This can be
thought of as an increase in the zeroth-order crystal-field
strength because of the presence of two anionic amidate
donors in 2 versus one anionic amidate and one neutral
amine donor in 3. While this will lead to an overall dest-
abilization of the entire 3d-manifold in (NiII(L1))2- (2), the
zeroth-order increase in energy will have a greater influence
on the HOMO and HOMO-1. These orbitals are both
covalent Ni(3d(π))/S(π)* molecular orbitals that result
from the overlap of the Ni(3dxz/yz) orbitals and the S(3pz)
bonding and anti-bonding combinations (Scheme 2).
€
putational results (Lowdin population analysis) shows
good agreement between the experimentally and compu-
tationally derived values. We observed 31(1)% S(3p)-
character in the S K-edge studies for the LUMO of (NiII-
(HL2))- (3), while the B3LYP calculations for 3 yielded a
value of 27.1%. The S K-edge studies demonstrated a
reduction in the degree of S(3p)-character for the LUMO
of (NiII(L1))2- (2) to 21(1)%, while the B3LYP calcula-
tions yielded a value of 19.6% S(3p)-character.66
In addition to a decrease in the percentage of S(3p)-
character comprising the LUMO of (NiII(HL2))- (3)
versus (NiII(L1))2- (2), there is also a decrease in the per-
centage of S(3p)-character comprising the HOMO and
HOMO-1. For (NiII(HL2))- (3), the percentage of S(3p)-
character in the HOMO is 68.3% (compared to 32.5%
(65) Siegbahn, P. E. M. J. Comput. Chem. 2001, 22, 1634–1645.
(66) For comparison to other Sulfur K-edge X-ray absorption studies, a
Mulliken population analysis gives 27.1% S(3p) character in the LUMO for
(3) and 25.2% for (2).

Article
Inorganic Chemistry, Vol. 49, No. 12, 2010 5403
Scheme 2. Simplified Qualitative MO Diagram Depicting the Forma-
tion of the HOMO and HOMO-1
multiconfigurational character of the final state wave
functions explains why TD-DFT, which is a single-
determinantal method, cannot adequately reproduce the
electronic absorption spectra for these complexes.
Upon the addition of the second amidate ligand, the
energy of the ligand-field transitions all increase in en-
ergy. The lowest energy transition, the “3dxz/yz f 3dx2-y2
”
transition, occurs ∼1450 cm-1 higher in energy in bis-
amidate ligated 2 than in amine/amidate ligated 3. The
next highest energy transition, the “3dyz/xz f 3dx2-y2
”
transition, is also higher in energy in (NiII(L1))2- (2)
than (NiII(HL2))- (3). However, because of the increased
amide character in the excited-state wave function of this
transition in the amine/amidate ligated 3, the 3dyz/xz
f
3dx2-y2 transition occurs only 600 cm-1 higher in energy
in (NiII(L1))2- (2) versus (NiII(HL2))- (3). Therefore, the
splitting in energy of the 3dxz/yz f 3dx2-y2 versus 3dyz/xz
f
Unlike what is found in amine/amidate ligated 3, the
Ni(3dxz/yz) atomic orbitals (AOs) in bis-amidate ligated 2
are higher in energy than the S(π)-fragments. Despite the
increased energy of the Ni 3dxz/yz AOs relative to the S(π)-
fragments, there is actually a better energy match between
the two in (NiII(L1))2- (2) than in (NiII(HL2))- (3). The
consequence of this is 3-fold. First, there is a better balance
of Ni- and S(3p)-character in the HOMO and HOMO-1 of
the bisamidate complex 2 relative to the mixed amine/
amidate complex 3 because of better Ni/ligand orbital
mixing (i.e., there is higher Ni-S covalency in (NiII-
(L1))2- (2)). Second, the HOMO of (NiII(L1))2- (2) will
have less S(3p)-character and more Ni-character relative to
(NiII(HL2))- (3) owing to the increased Ni-S covalency.
Third, and most importantly, the HOMO of the bisamidate
complex 2 will be energetically activated relative to the
mixed amine/amidate complex 3 because of higher Ni-
S(3p) covalency. Because the nucleophilic HOMO of (NiII-
(L1))2- (2) is activated relative to (NiII(HL2))- (3), it
should be significantly more reactive toward electrophilic
O2, which is what is observed experimentally.
3dx2-y2 transition in (NiII(L1))2- (2) is calculated to be
much smaller in magnitude than in (NiII(HL2))- (3)
(∼780 vs 1600 cm-1). In both cases the most prominent
ligand-field transition at ∼22,000 cm-1 is predicted to be
the “3dxy f 3dx2-y2” transition.32a,34,37 Because of the
higher degree of ligand character in this transition for
(NiII(L1))2- (2) versus (NiII(HL2))- (3) from mixing of
lower-energy ligand centered states, this transition is
more charge-transfer like in (NiII(L1))2- (2) than in (NiII-
(HL2))- (3). Thus, it has larger calculated oscillator
strength in (NiII(L1))2- (2) than in (NiII(HL2))- (3)
(0.0077 vs 0.0055).
Energy Level Description of (NiII(L1))2- (2) vs (NiII-
(HL2))- (3). Using the S K-edge data, electronic ab-
sorption measurements, and excited-state computa-
tional results we can create an accurate picture of the
valence orbitals in (NiII(L1))2- (2) and (NiII(HL2))- (3)
(Scheme 3). Descending from the totally symmetric free
ion into D4h symmetry, the lowest energy transitions
correspond to the 1Eg r 1A1g, 1B2g r 1A1g, and 1A2g
r
1A1g transitions (see Scheme 3 for orbital descriptions
Excited-State Calculations of (NiII(L1))2- (2) and (NiII-
(HL2))- (3). To determine the origin of the features in the
electronic absorption spectra of (NiII(L1))2- (2) and (NiII-
(HL2))- (3), we performed excited-state electronic struc-
ture calculations on the two transition metal compounds.
Excited-state calculations were carried out using Neese’s
SORCI methodology, which has yielded excellent results in
calculating the electronic absorption spectra of transition
metal complexes in cases where methods such as time
dependent DFT (TD-DFT) have failed. A complete discus-
sion of the excited-state calculations and their results can be
found in the Supporting Information. For both complexes
the four lowest energy transitions can be thought of as
arising from within the 3d manifold (i.e., are d f d
transitions) and their leading configurations are attribu-
of these states). These all further descend into A r A
transitions because of the asymmetry of the ligand en-
vironment. In (NiII(L1))2- (2) and (NiII(HL2))- (3), the
1
1
1
1
weak A(B2g) r A transition is obscured by the other
ligand field transitions, while the 1Eg r 1A1g transition is
1
1
split into two A(Eg) r A transitions as a result of the
lower symmetry presented by the ligand field environ-
ment. As expected, the splitting of the 1Eg state is larger
for amine/amidate ligated (NiII(HL2))- (3) than it is for
bisamidate ligated (NiII(L1))2- (2); there is approximately a
1
1,225 cm-1 splitting of the Eg state in (NiII(HL2))- (3)
compared to a less than 470 cm-1 splitting of the 1Eg state in
(NiII(L1))2- (2).
The S K-edge X-ray absorption spectra indicate that
the Ni(3dx2-y2) LUMO of bisamidate ligated (NiII(L1))2-
(2) is approximately 2400 cm-1 higher in energy than that
of (NiII(HL2))- (3). Using the energies obtained from the
low-temperature (77 K) electronic absorption spectra, we
can arrive at reasonable estimates of the relative energies
of the HOMO for both of these complexes. The corre-
sponding difference of the lowest energy 1A(Eg) r 1A for
bisamidate ligated (NiII(L1))2- (2) versus amine/amidate
ted to: 3dxz/yz f 3dx2-y2, the 3dyz/xz f 3dx2-y2, 3dz2
f
3dx2-y2, and the 3dxy f 3dx2-y2 transitions. Because the
“3dz2 MO” is mostly AO in character (98.2% 3dz2 in
character), the intensity of the 3dz2 f 3dx2-y2 is low, and
one would not expect that it could be readily resolved
from the experimental data, which is what is observed
experimentally (vide infra). In contrast, we find that the
other ligand-field transitions arise from final state wave
functions that have (a) significant multiconfigurational
character, and (b) significant ligand character. The large
ligated (NiII(HL2))- (3) is approximately 225 cm-1
.
Considering that a ligand-field analysis suggests that
the Slater-Condon-Shortley parameters for the two

5404 Inorganic Chemistry, Vol. 49, No. 12, 2010
Mathrubootham et al.
Scheme 3. Energy Level Correlation Diagram for the Square-Planar
forcing the Ni-S bond to elongate to relieve steric
repulsion.
Low-Spin NiII Ion
Single point hybrid-DFT calculations of these NiIII
compounds point toward the reason for the reduced
stability upon oxidation on the CV time scale for (NiII-
(HL2))- (3) compared to (NiII(BEAAM)), which displays
a quasi-reversible NiIII/NiII couple.37 A Lowdin popula-
€
tion analysis shows that the singly occupied molecular
orbital (SOMO)68 of NiIII(HL2) is composed of 71.4%
S(3p)-character and only 8.3% Ni-character. This can be
contrasted with NiIII(BEAAM), which has a SOMO
that is composed of 59.7% S(3p)-character and 26.6%
Ni-character, indicating that less electron density resides
on the metal in NiIII(HL2) relative to NiIII(BEAAM). This
would lead to a less stable oxidized species in NiIII(HL2)
compared to NiIII(BEAAM), as it would contain more
S character. As sulfur-centered radicals are highly reac-
3
tive species, the increased S character in the SOMO of
3
NiIII(HL2) versus NiIII(BEAAM) would lead to a much
more chemically reactive species, and thus irreversibility in
the CV of NiIII(HL2).
complexes are approximately the same (a 78 vs 76% redu-
ction from the atomic values for 2 vs 3), we can assume a
similar degree of e-/e- repulsion within the d-orbitals.
Thus, when this difference in energy is subtracted from
the energy difference of the LUMO, we surmise that the
HOMO in bisamidate ligated (NiII(L1))2- (2) is activated
by ∼2175 cm-1 (∼6.2 kcal mol-1) relative to amine/
amidate ligated (NiII(HL2))- (3).67 The activation of the
HOMO in the bisamidate ligated (NiII(L1))2- (2) is largely
responsible for its increased O2 sensitivity relative to amine-
amidate ligated (NiII(HL2))- (3). The nucleophilic doubly
filled Ni(3dπ)/S(π)* HOMO is electronically tuned to be
more reactive toward electrophiles like O2 relative to that of
(NiII(HL2))- (3).
Discussion
The active site of reduced Ni-SOD and the Nid-site of ACS
both contain a Ni ion ligated in an N2S2 coordination
environment, and yet the properties of the two NiII ions are
very different. In Ni-ACS, the NiN2S2 site does not seem to
be involved in the catalytic redox cycle,2,6b while in Ni-SOD
the Ni ion switches between the NiII and NiIII state during the
catalytic cycle.26,69 Interestingly, the Ni-SOD active site
utilizes a mixed amine/amidate ligation as opposed to the bis-
amidate ligation found in the Nid(Cys-Gly-Cys) site of ACS.
Thus, a small difference in coordination environment around
nickel apparently leads to very different properties of the
NiN2S2 site in these two enzymes. It is possible that the sul-
furs in Ni-SOD are protected by second coordination sphere
effects, hydrogen bonding, and/or steric interactions rather
than inner sphere coordination effects. However, all bisami-
date NiIIN2S2 complexes reported to date are exceedingly air-
sensitive, while the active site of Ni-SOD is apparently stable
to O2, superoxide, and hydrogen peroxide, providing further
evidence that slight differences in coordination environment
can have a profound impact on the properties of the NiII ions.
Our results can be added to a small but growing body of
evidence indicating that electronic tuning of the Ni-SOD
active site is at least partially responsible for the increase the
stability of the Ni-SOD active site toward reactive oxygen
species.27b,29,34,37
Electronic Structure of NiIII(BEAAM) and NiIII(HL2).
To gain better insight into the irreversibility observed in
the CV of (NiII(HL2))-, we performed electronic struc-
ture calculations on oxidized NiIII(HL2) at the B3LYP
level of theory to compare with those performed on NiIII
-
(BEAAM). GO structures (Supporting Information,
Figure S3) for both NiIII(HL2) and NiIII(BEAAM) are
nearly identical to one another, which is expected con-
sidering that both are NiIIIN2S2 complexes with mixed
amine/amidate ligation with 5,5,5-chelate rings (Table 3).
Predictably, the Ni-ligand bond lengths are all contracted
relative to the reduced NiII GO structures. NiIII(HL2) has
calculated Ni-S bond lengths of 2.153 (trans amine) and
˚
To understand more fully the reason for this difference in
reactivity, we synthesized and characterized a set of model
complexes with NiN2S2 coordination that differ only in ami-
date versus amine ligation. These complexes correlate well
with the reduced active site of Ni-SOD and distalnickel site of
ACS and are therefore able to provide unique insight into
how the enzymes tune the Ni center to achieve very different
properties for the similar NiN2S2 site.
2.118 A (trans amide), while the calculated Ni-N bond
III
lengths are 1.979 (amine) and 1.860 A. Similarly, Ni -
˚
(BEAAM) possesses calculated Ni-S bond lengths of
2.159 (trans amine) and 2.147 A (trans amide) and
˚
calculated Ni-N bond lengths of 1.998 (amine) and
˚
1.857 A. The elongation of the Ni-S bond trans to the
amide nitrogen in NiIII(BEAAM) relative to NiIII(HL2) is
due to steric crowding; the gem-dimethyl groups of NiIII
-
(BEAAM) are in close contact with both the amine
substituent and the hydrogens of the ethylene backbone
(68) We refer to the highest filled spin-up MO as the SOMO, which is
largely 3d(π)-S(π) in character.
(69) Bryngelson, P. A.; Maroney, M. J. In Nickel and Its Surprising Impact
in Nature; Sigel, A., Sigel, H., Sigel, R. K. O., Eds.; John Wiley & Sons:
Chichester, U.K., 2007; Metal ions in Life Sciences, Vol. 2; Chapter 9,
pp 417-444.
(67) If the differences in e-/e- repulsion are considered then the HOMO
activation of (NiII(L1))2- (2) vs NiII(HL2)- (3) could be as high as ∼3000
cm-1. However, we consider the 78 vs 76% reduction in the Slater-
Condon-Shortley parameters to be within the error of the data fitting.

Article
The X-ray structures of the mononuclear complexes
Inorganic Chemistry, Vol. 49, No. 12, 2010 5405
NiIII(BEAAM), which has a SOMO that is composed of
59.7% S(3p)-character and 26.6% Ni-character. This im-
plies that NiIII(HL2) has a significant increase in S(3p)-
character relative to NiIII(BEAAM). Thus, upon oxida-
tion there will be more sulfur-based radical character in
oxidized NiIII(HL2) than in NiIII(BEAAM), which should
(NiII(L1))2- (2) and (NiII(HL2))- (3) show that the NiII ions
are in a square-planar coordination geometry. In complex
(NiII(L1))2- (2), the asymmetrically positioned carbonyl
groups in the ligand backbone impose slight differences in
the bond lengths and angles compared to the analogous 5,5,5
chelated NiN2S2 planar complexes.14,22b In the mixed amine/
amidate complex (NiII(HL2))- (3), the Ni-N(amide) bond
distance is shorter than the Ni-N(amine) bond distance
while the Ni-S (trans to the amide nitrogen) bond distance
is longer than the Ni-S(trans to amine nitrogen). This is as
expected because the N(amide) ligands are stronger σ-donors
than N(amine) ligands, and thus exhibit a larger trans
influence. In comparison to the mixed amine/amidate com-
plex Ni(BEAAM) reported by Shearer and co-workers,37 the
difference in the bond lengths of the Ni-S bonds trans to the
N(amide) versus N(amine) is smaller in (NiII(HL2))- (3).
This difference may be due to steric crowding from the gem-
dimethyl groups in Ni(BEAAM) and/or the presence of the
secondary amine nitrogen in (NiII(HL2))- (3) as opposedto a
tertiary amine nitrogen.
make NiIII(HL2) significantly more unstable than NiIII
-
(BEAAM). Also, the increased steric bulk of the gem-
dimethyls of NiIII(BEAAM) R to the thiolate ligands will
lead to additional stability of the oxidized complex rela-
tive to NiIII(HL2). This is manifested in the irreversibility
of NiIII(HL2) on the CV time scale. We note, however,
that the NiIII forms have not been successfully isolated for
either complex, suggesting that the reduced amount of
sulfur based radical character in NiIII(BEAAM) versus
NiIII(HL2) is still not sufficient to stabilize adequately the
complex on longer time-scales.
The electrochemical data of the complex (NiII(L1))2- (2),
suggest that the NiIII formed by chemical oxidation should be
detectable by EPR. The EPR spectrum of the chemical
oxidation product of (NiII(L1))2- (2) in DMF (Figure 8)
displays a nearly axial EPR signal with g values of g1=2.012,
g2 =2.195, and g3 =2.228 (gav=2.145), indicative of a NiIII
formulation with S = 1/2 and with the unpaired electron
residing in a Ni(dz2) orbital. This spectrum is consistent with
A comparison of the optical spectra of the bisamidate com-
plex (NiII(L1))2- (2) and the mixed amine/amidate complex
(NiII(HL2))- (3) reveals that absorption bands for the latter
are at slightly lower energy. This can be easily rationalized
from a simple crystal-field point of view. The additional
anionic amidate in (NiII(L1))2- (2) versus (NiII(HL2))- (3) in
the equatorial plane leads to a significant destabilization of
the Ni(3dx2-y2) orbital, which in turn increases the energy of
the ligand-field transitions in the bisamidate complex. This
simplistic view is fully supported by our electronic structure
calculations. Although the energy of the Ni 3d(π)-type
HOMO and HOMO-1 of (NiII(L1))2- (2) are significantly
destabilized relative to those of (NiII(HL2))- (3), the LUMO
rises even higher in energy in (NiII(L1))2- (2) because of
the stronger σ-bonding interaction between the two amide-
nitrogens and the Ni(3dx2-y2), resulting in a higher energy
transition.
The influence of amine versus amide bonding is also
reflected in the electrochemical properties of the complexes.
As expected, the bisamidate complex (NiII(L1))2- (2),
displays a more negative NiIII/NiII oxidation potential than
the mixed amine/amidate complex (NiII(HL2))- (3). This can
be easily explained by the fact that the electron-rich bisami-
date ligand better stabilizes the NiIII oxidation state. This
is supported by DFT calculations which indicate that the
redox active molecular orbital (RAMO) is destabilized in the
bisamidate complex.
the NiIII EPR spectra reported by Holm,22 Kruger,22,23 and
€
Brunold,31 and with our DFT calculations at the B3LYP
level (Supporting Information). In the case of complex
(NiII(HL2))- (3), the chemical oxidation with I2 forms a
purple species which is not stable and decays within seconds.
Attempts to trap the NiIII species even at low temperature
have not been successful, consistent with the electrochemical
irreversible behavior. Again from the DFT calculations, a
III
€
Lowdin population analysis shows that the SOMO of Ni -
(HL2) is composed of 71.4% S(3p)-character and only 8.3%
Ni-character. This implies NiIII(HL2) has a significant inc-
rease in highly reactive radical S(3p)-character and may
preclude its isolation because of sample decomposition.
The bisamidate complex (NiII(L1))2- (2) is highly reactive
toward O2 in CH3CN, while the mixed amine/amidate
complex (NiII(HL2))- (3) is considerably less reactive and
is stable for at least a week in solution. This agrees with our
predictions from the DFT studies that the bisamidate com-
pound (NiII(L1))2- (2) should be more O2 sensitive than
(NiII(HL2))- (3). This is because the filled Ni(3d(π))/S(π)*
HOMO in (NiII(L1))2- (2) acts as a nucleophile toward the
electrophilic O2 molecule.27b,28,29,37 Increasing the energy of
the HOMO via an increase in Ni-S covalency in (NiII(L1))2-
(2) effectively activates this species toward electrophiles
relative to (NiII(HL2))- (3). These results are consistent with
previously reported Ni-SOD model compounds containing
mixed amine/amidate NiIIN2S2 coordination that are also
air-stable.37,38
Similar reasoning also explains the formation of the trinuclear
complex ({NiII(L1)}2NiII)2- (1) and the fact that (NiII(HL2))-
(3) remains mononuclear in solution. The thiolates of (NiII
(L1))2- (2) are far more nucleophilic than observed in (NiII-
(HL2))- (3). Thus, (NiII(L1))2- (2) rapidly forms a stable
trinuclear complex in solution with any “free” Ni ions. In
contrast, (NiII(HL2))- (3) reacts sluggishly with Ni ions, pre-
cluding the formation of the corresponding trinuclear species.
In summary, we synthesized and characterized two new
NiN2S2 complexes which serve as models for the Nid site of
The DFT calculations also provide insight into the stability
of the NiIII complexes. Experimentally we observe that the
bisamidate complex (NiII(L1))2- (2) is more stable with res-
pect to one-electron oxidation relative to mixed amine/
amidate complex (NiII(HL2))-. This can be explained by
the fact that the RAMO in (NiII(L1))2- (2) has less S(3p)-
character. Thus, the resulting radical will contain more Ni
character and less S character. Interestingly, the oxidation of
the complex (NiII(HL2))- (3) is completely irreversible while
the one-electron oxidation of a similar mixed amine/amidate
complex, Ni(BEAAM), reported by Shearer and co-workers37
is quasi-reversible. From the single-point DFT calculations,
a Lowdin population analysis reveals that the SOMO68 of
€
NiIII(HL2) is composed of 71.4% S(3p)-character and
only 8.3% Ni-character. This can be contrasted with

5406 Inorganic Chemistry, Vol. 49, No. 12, 2010
Mathrubootham et al.
ACS and the active site of reduced Ni-SOD to advance our
understanding of how the amine versus amide coordination
environment influences the properties of the Ni ion. The
complex (NiII(HL2))- (3) is one of the few examples of a
synthetic mixed amine/amidate NiII species reported. Our
results demonstrate that the bisamidate complex (NiII(L1))2-
(2) is considerably more reactive toward O2 and susceptible
to trimer formation than the corresponding mixed amine/
amidate complex (NiII(HL2))- (3). SORCI calculations in
combination with S K-edge X-ray absorption spectroscopy
and low-temperature electronic absorption measurements
reveal that the additional amidate ligand in (NiII(L1))2- (2)
destabilizes the HOMO by ∼6.2 kcal mol-1 relative to (NiII-
(HL2))- (3). The consequence of this destabilization is mani-
fested in the nucleophilic activation of the doubly filled
HOMO, which makes (NiII(L1))2- (2) significantly more
reactive toward electrophiles. This explains why Ni-SOD
utilizes the NiN2S2 mixed amine/amidate coordination enviro-
nment; this coordination environment deactivates the nucleo-
philic S(π)-type HOMO toward attack (as demonstrated by
our model complexes) thereby allowing Ni-SOD to perform
Ni-centered oxidation and avoid thiolate oxygenation.
Acknowledgment. We acknowledge financial support
from the National Science Foundation (CHE-0348777).
We thank Dr. Atta Arif (University of Utah) for solving the
X-ray structure of 3 and Thomas Casey (Michigan State
University) for EPR measurements. Computational
resources provided by the Research Computing Grid at
the University of Nevada, Reno. J.S. thanks the NIH (P20
RR-016464) and the NSF (CHE-0844234) for support.
Supporting Information Available: Excited-state calculat-
ions of (NiII(L1))2- (2) and (NiII(HL2))- (3), computationally
€
derived structures, Mulliken population analysis, Lowdin popu-
lation analysis, results from the SORCI calculations, computa-
tionally derived coordinates, calculated EPR parameters, elec-
tronic absorption data from the low-temperature solid-state
samples, cyclic voltammograms, and X-ray structural data in
CIF format. This material is available free of charge via the
Internet at http://pubs.acs.org.