Iron-mediated hydrazine reduction and the formation of iron-arylimide heterocubanes

Article abstract of DOI:10.1021/ic1021627

The reaction of Fe(N{SiMe₃}₂)₂ (1) with 1 equiv of arylthiol (ArSH) results in material of notional composition Fe(SAr)(N{SiMe₃}₂) (2), from which crystalline Fe ₂(μ-SAr)₂(N{SiMe₃}₂) ₂(THF)₂ (Ar = Mes) can be isolated from tetrahydrofuran (THF) solvent. Treatment of 2 with 0.5 equiv of 1,2-diarylhydrazine (Ar?NH-NHAr?, Ar? = Ph, p-Tol) yields ferric-imide-thiolate cubanes Fe ₄(μ₃-NAr?)₄(SAr)₄ (3). The site-differentiated, 1-electron reduced iron-imide cubane derivative [Fe(THF)₆][Fe₄(μ₃-N-p-Tol) ₄(SDMP)₃(N{SiMe₃}₂)]₂ ([Fe(THF)₆][4]₂; DMP = 2,6-dimethylphenyl) can be isolated by adjusting the reaction stoichiometry of 1/ArSH/Ar?NHNHAr? to 9:6:5. The isolated compounds were characterized by a combination of structural (X-ray diffraction), spectroscopic (NMR, UV-vis, Moessbauer, EPR), and magnetochemical methods. Reactions with a range of hydrazines reveal complex chemical behavior that includes not only N-N bond reduction for 1,2-di- and trisubstituted arylhydrazines, but also catalytic disproportionation for 1,2-diarylhydrazines, N-C bond cleavage for 1,2-diisopropylhydrazine, and no reaction for hindered and tetrasubstituted hydrazines.

Full text of DOI:10.1021/ic1021627

Inorg. Chem. 2011, 50, 1551–1562 1551
DOI: 10.1021/ic1021627
Iron-Mediated Hydrazine Reduction and the Formation of Iron-Arylimide
Heterocubanes
Michael J. Zdilla,† Atul K. Verma,‡ and Sonny C. Lee*
Department of Chemistry, Princeton University, Princeton, New Jersey 08544, United States and
Department of Chemistry, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1.
†Current address: Department of Chemistry, Temple University, Philadelphia, Pennsylvania 19122.
‡Current address: Avecia Biotechnology, Inc., Milford, Massachusetts 01757
Received October 27, 2010
The reaction of Fe(N{SiMe3}2)2 (1) with 1 equiv of arylthiol (ArSH) results in material of notional composition
Fe(SAr)(N{SiMe3}2) (2), from which crystalline Fe2(μ-SAr)2(N{SiMe3}2)2(THF)2 (Ar = Mes) can be isolated from
tetrahydrofuran (THF) solvent. Treatment of 2 with 0.5 equiv of 1,2-diarylhydrazine (Ar0NH-NHAr0, Ar0 = Ph, p-Tol) yields
ferric-imide-thiolate cubanes Fe4(μ3-NAr0)4(SAr)4 (3). The site-differentiated, 1-electron reduced iron-imide cubane
derivative [Fe(THF)6][Fe4(μ3-N-p-Tol)4(SDMP)3(N{SiMe3}2)]2 ([Fe(THF)6][4]2; DMP = 2,6-dimethylphenyl) can be
isolated by adjusting the reaction stoichiometry of 1/ArSH/Ar0NHNHAr0 to 9:6:5. The isolated compounds were
€
characterized by a combination of structural (X-ray diffraction), spectroscopic (NMR, UV-vis, Mossbauer, EPR), and
magnetochemical methods. Reactions with a range of hydrazines reveal complex chemical behavior that includes not only
N-N bond reduction for 1,2-di- and trisubstituted arylhydrazines, but also catalytic disproportionation for 1,2-diarylhy-
drazines, N-C bond cleavage for 1,2-diisopropylhydrazine, and no reaction for hindered and tetrasubstituted hydrazines.
Introduction
pathway isunclear. Recently, thefirst spectroscopic detection
of hydrazine interaction with the FeMo-cofactor has been
achieved in a mutant nitrogenase.6
The study of metal-mediated hydrazine reduction has long
been associated with efforts to establish the as-yet unknown
reaction chemistry of biological nitrogen fixation. The biolo-
gical reduction of dinitrogen to ammonia occurs at distinctive
iron-sulfur (Fe-S) biometalloclusters within nitrogenase
enzymes, of which the best characterized is the iron-
molybdenum cofactor (FeMo-cofactor) of molybdenum-
dependent nitrogenase.1 The transformation is reasonably
assumed to proceed through some type of metal-bound
hydrazinic fragment,2 and parent hydrazine (N2H4) is re-
leased upon acid or base-quenching of active, presteady-state
enzyme;3 N2H4 is also detected as a very minor side product
in vanadium-dependent nitrogenase.4 Parent hydrazine itself
can serve as a substrate for nitrogenase, although the reduction
efficiency is poor in wild-type enzyme5,6a,6c and the relation-
ship of this chemistry to the actual dinitrogen reduction
Synthetic investigations of metal-mediated hydrazine re-
duction are broad-ranging and span a number of transition
elements and a range of different ligand environments. With
respect to the metals present in the catalytic clusters of
nitrogenases, the most extensive studies have focused on
hydrazine reactivity in molybdenum complexes. The biolo-
gical connection stems from the presence of molybdenum in
the FeMo-cofactor, currently formulated with a [MoFe7S9X]
cluster core (X = C, N, or O), and it is predicated on the
supposition that molybdenum, as the unique heterometal in
an Fe-S environment, is the local substrate binding site
within the cluster.2
A diversity of molybdenum systems have been explored in
the context of hydrazine reduction, including mid-7,8 to high-
valent9 mononuclear complexes, dinuclear species,10 and
heterometallic, sulfide-containing clusters,11-13 and stoichio-
metric, disproportionative, and catalytic hydrazine reduc-
tions have been demonstrated. Metal-bound hydrazide
*To whom correspondence should be addressed. E-mail: sclee@
uwaterloo.ca.
(1) (a) Christiansen, J.; Dean, D. R.; Seefeldt, L. C. Annu. Rev. Plant
Physiol. Plant Mol. Biol. 2001, 52, 269. (b) Burgess, B. K.; Lowe, D. L. Chem.
Rev. 1996, 96, 2983. (c) Eady, R. R. Chem. Rev. 1996, 96, 3013.
(2) Seefeldt, L. C.; Hoffman, B. M.; Dean, D. R. Annu. Rev. Biochem.
2009, 78, 701.
(3) Thorneley, R. N. F.; Lowe, D. J. In Molybdenum Enzymes; Spiro, T. G.,
Ed.; Wiley-Interscience: New York, 1985; p 221.
(4) (a) Dilworth, M. J.; Eldridge, M. E.; Eady, R. R. Biochem. J. 1993,
289, 395. (b) Dilworth, M. J.; Eady, R. R. Biochem. J. 1991, 277, 465.
(5) Davis, L. C. Arch. Biochem. Biophys. 1980, 204, 270.
(6) (a) Barney, B. M.; Laryukhin, M.; Igarashi, R. Y.; Lee, H.-I.; Dos
Santos, P. C.; Yang, T.-C.; Hoffman, B. M.; Dean, D. R.; Seefeldt, L. C.
Biochemistry 2005, 44, 8030. (b) Barney, B. M.; Yang, T.-C.; Igarashi, R. Y.; Dos
Santos, P. C.; Laryukhin, M.; Lee, H.-I.; Hoffman, B. M.; Dean, D. R.; Seefeldt,
L. C. J. Am. Chem. Soc. 2005, 127, 14960. (c) Barney, B. M.; Igarashi, R. Y.; Dos
Santos, P. C.; Dean, D. R.; Seefeldt, L. C. J. Biol. Chem. 2004, 279, 53621.
r
2011 American Chemical Society
Published on Web 01/20/2011
pubs.acs.org/IC

1552 Inorganic Chemistry, Vol. 50, No. 4, 2011
Zdilla et al.
(M=N-NH2) species are proposed as key intermediates in
dinitrogen reduction at mononuclear centers,14 and appro-
priate species have exhibited the relevant stepwise reactivity.7
From a bioinorganic standpoint, we note studies of the
catalytic chemical reduction of parent hydrazine by Fe-S
and M-Fe-S (M = Mo, V) cubane species that strongly
resemble, in whole or in part, known biological clusters;12
under the reported conditions, only labile, heterometal-sub-
stituted clusters were active in hydrazine reduction, suggest-
ing the need for a reactive heterometal site in this chemistry.15
With the discovery of vanadium-dependent nitrogenase,1c
vanadium has also been investigated in the same context and
with similar chemistry, although these studies are much more
limited in number and scope.16
fixation appears to be required in at least one class of
nitrogenase that contains no transition elements other
than iron.1c The focus on iron intensified following the
macromolecular crystallographic studies that yielded the first
molecular visualization of the entire FeMo-cofactor;17 the
structural data inspired mechanistic conjectures that specifi-
cally enlisted iron centers as the site of substrate reduction
chemistry in the heterometallic cofactor cluster.18 The recent
detection of nitrogenase intermediates in site-directed Mo-
dependent nitrogenase mutants has also implicated iron in
substrate binding based on the positions of the introduced
protein side chain substitutions.19
In comparison with the molybdenum chemistry, examples
of iron-mediated hydrazine reduction are fewer in number
and generally postdate the first nitrogenase protein structure
determinations. The earliest report to our knowledge is an
account, without molecular details, of the catalytic electro-
chemical reduction of N2H4 by [Fe4S4(SR)4]2- cubane
clusters.13 More recently, the reduction of parent hydrazine
has been found in β-diketiminate20 and polyphosphine21
complexes of Fe(II) and, catalytically, in a dinuclear cyclo-
pentadienyl/thiolate system.22 Reductive cleavage of the N-N
bond in organohydrazines and ligated organohydrazides has
also been observed in the last system, as well as in a few other
mono- and dinuclear complexes with cyclopentadienyl23 and
β-diketiminate24 ancillary ligands. These examples of Fe-
mediated hydrazine reduction, while few in number, never-
theless encompass a range of coordination environments and
reaction conditions. Although detailed mechanistic studies
are few,24c,25 the body of evidence suggests a number of
fundamentally different hydrazine reduction pathways de-
pending on reaction system.
Recent attention on this subject has expanded to include
iron-mediated chemistry. Iron-based biological nitrogen
(7) Mo(IV)-hydrazide(2-) complexes from protonated Mo(0)-dinitrogen
precursors: (a) Baumann, J. A.; Bossard, G. E.; George, T. A.; Howell, D. B.;
Koczon, L. M.; Lester, R. K.; Noddings, C. M. Inorg. Chem. 1985, 24, 3568.
(b) Gebreyes, K.; Zubieta, J.; George, T. A.; Koczon, L. M.; Tisdale, R. C. Inorg.
Chem. 1986, 25, 405. (c) Kaul, B. B.; Hayes, R. K.; George, T. A. J. Am. Chem.
Soc. 1990, 112, 2002. (d) Galindo, A.; Hills, A.; Hughes, D. L.; Richards, R. L.;
Hughes, M.; Mason, J. J. Chem. Soc., Dalton Trans. 1990, 283. (e) George,
T. A.; Rose, D. J.; Chang, Y.; Chen, Q.; Zubieta, J. Inorg. Chem. 1995, 34, 1295.
(8) (a) Hitchcock, P. B.; Hughes, D. L.; Maguire, M. J.; Marjani, K.;
Richards, R. L. J. Chem. Soc., Dalton Trans. 1997, 4747. (b) Watanabe, D.;
Gondo, S.; Seino, H.; Mizobe, Y. Organometallics 2007, 26, 4909.
(9) (a) Schrock, R. R.; Glassman, T. E.; Vale, M. G. J. Am. Chem. Soc.
1991, 113, 725. (b) Vale, M. G.; Schrock, R. R. Inorg. Chem. 1993, 32, 2767. (c)
Carrillo, D. C. R. Acad. Sci. II C 2000, 3, 175. (d) Yandulov, D. V.; Schrock,
R. R. J. Am. Chem. Soc. 2002, 124, 6252. (e) Yandulov, D. V.; Schrock, R. R.;
Rheingold, A. L.; Ceccarelli, C.; Davis, W. M. Inorg. Chem. 2003, 42, 796. (f)
Yandulov, D. V.; Schrock, R. R. Inorg. Chem. 2005, 44, 1103. (g) Vrubel, H.;
Verzenhassi, V. H. C.; Nakagaki, S.; Nunes, F. S. Inorg. Chem. Commun. 2008,
11, 1040.
(10) (a) Block, E.; Ofori-Okai, G.; Kang, H.; Zubieta, J. J. Am. Chem.
Soc. 1992, 114, 758. (b) Schollhammer, P.; Petillon, F. Y.; PoderGuillou, S.;
Saillard, J. Y.; Talarmin, J.; Muir, K. W. Chem. Commun. 1996, 2633. (c)
Schollhammer, P.; Guenin, E.; Petillon, F. Y.; Talarmin, J.; Muir, K. W.; Yufit,
D. S. Organometallics 1998, 17, 1922. (d) Petillon, F. Y.; Schollhammer, P.;
Talarmin, J.; Muir, K. W. Inorg. Chem. 1999, 38, 1954. (e) Le Grand, N.; Muir,
K. W.; Petillon, F. Y.; Pickett, C. J.; Schollhammer, P.; Talarmin, J. Chem.;Eur.
J. 2002, 8, 3115. (f) Schollhammer, P.; Didier, B.; Le Grand, N.; Petillon, F. Y.;
Talarmin, J.; Muir, K. W.; Teat, S. J. Eur. J. Inorg. Chem. 2002, 658.
(11) (a) Masumori, T.; Seino, H.; Mizobe, Y.; Hidai, M. Inorg. Chem.
2000, 39, 5002. (b) Seino, H.; Masumori, T.; Hidai, M.; Mizobe, Y. Organome-
tallics 2003, 22, 3424. (c) Takei, I.; Dohki, K.; Kobayashi, K.; Suzuki, T.; Hidai,
M. Inorg. Chem. 2005, 44, 3768.
In our own studies, we have assembled Fe(III)-arylimide-
thiolate cubane clusters, Fe4(NAr0)4(SAr)4, by reductive
scission of the N-N bond in 1,2-diarylhydrazines.26 With
respect to potentially relevant biomimetic chemistry, this
reaction system attracts our interest because of features in
common with biological Fe-S centers: (1) the iron sites are
weak-field, tetrahedral, and possess a partial sulfur-anion
environment; (2) the reaction chemistry couples substrate
reduction to biologically observed Fe(II)/Fe(III) redox
states; and (3) the product cluster structure maps onto the
archetypal biological Fe-S cubane geometry. Following an
initial communication,26 we present here a detailed account
(12) (a) Coucouvanis, D.; Mosier, P. E.; Demadis, K. D.; Patton, S.;
Malinak, S. M.; Kim, C. G.; Tyson, M. A. J. Am. Chem. Soc. 1993, 115,
12193. (b) Demadis, K. D.; Coucouvanis, D. Inorg. Chem. 1994, 33, 4195. (c)
Demadis, K. D.; Coucouvanis, D. Inorg. Chem. 1995, 34, 3658. (d) Demadis,
K. D.; Malinak, S. M.; Coucouvanis, D. Inorg. Chem. 1996, 35, 4038. (e)
Coucouvanis, D.; Demadis, K. D.; Malinak, S. M.; Mosier, P. E.; Tyson, M. A.;
Laughlin, L. J. J. Mol. Catal. A: Chem. 1996, 107, 123.
(17) See: Einsle, O.; Tezcan, F. A.; Andrade, S.; Schmid, B.; Yoshida, M.;
Howard, J. B.; Rees, D. C. Science 2002, 297, 1696, and references therein.
(18) Peters, J. W.; Szilagyi, R. K. Curr. Opin. Chem. Biol. 2006, 10, 101.
(19) (a) Hoffman, B. M.; Dean, D. R.; Seefeldt, L. C. Acc. Chem. Res.
2009, 42, 609. (b) Dos Santos, P. C.; Igarashi, R. Y.; Lee, H.-I.; Hoffman, B. M.;
Seefeldt, L. C.; Dean, D. R. Acc. Chem. Res. 2005, 38, 208.
(20) Yu, Y.; Brennessel, W. W.; Holland, P. L. Organometallics 2007, 26,
3217.
(21) (a) Crossland, J. L.; Zakharov, L. N.; Tyler, D. R. Inorg. Chem. 2007,
46, 10476. (b) Saouma, C. T.; M€uller, P.; Peters, J. C. J. Am. Chem. Soc. 2009,
131, 10358.
(22) Chen, Y.; Zhou, Y.; Chen, P.; Tao, Y.; Li, Y.; Qu, J. J. Am. Chem.
Soc. 2008, 130, 15250.
(13) Hozumi, Y.; Imasaka, Y.; Tanaka, K.; Tanaka, T. Chem. Lett. 1983,
897.
(14) (a) Pickett, C. J. J. Biol. Inorg. Chem. 1996, 1, 601. (b) Chatt, J.;
Dilworth, J. R.; Richards, R. L. Chem. Rev. 1978, 78, 589. (c) Schrock, R. R. Pure
Appl. Chem. 1997, 69, 2197.
(15) Coucouvanis, D. J. Biol. Inorg. Chem. 1996, 1, 594.
(16) (a) Rehder, D.; Woitha, C.; Priebsch, W.; Gailus, H. J. Chem. Soc.,
Chem. Commun. 1992, 364. (b) Lefloch, C.; Henderson, R. A.; Hughes, D. L.;
Richards, R. L. J. Chem. Soc., Chem. Commun. 1993, 175. (c) Gailus, H.;
Woitha, C.; Rehder, D. J. Chem. Soc., Dalton Trans. 1994, 3471. (d) Malinak,
S. M.; Demadis, K. D.; Coucouvanis, D. J. Am. Chem. Soc. 1995, 117, 3126. (e)
Ferguson, R.; Solari, E.; Floriani, C.; Osella, D.; Ravera, M.; Re, N.; Chiesi-Villa,
A.; Rizzoli, C. J. Am. Chem. Soc. 1997, 119, 10104. (f) Davies, S. C.; Hughes,
D. L.; Janas, Z.; Jerzykiewicz, L. B.; Richards, R. L.; Sanders, J. R.; Silverston,
J. E.; Sobota, P. Inorg. Chem. 2000, 39, 3485. (g) Davies, S. C.; Hughes, D. L.;
Janas, Z.; Jerzykiewicz, L. B.; Richards, R. L.; Sanders, J. R.; Silverston, J. E.;
Sobota, P. Inorg. Chem. 2000, 39, 3485. (h) Chu, W.-C.; Wu, C.-C.; Hsu, H.-F.
Inorg. Chem. 2006, 45, 3164.
(23) Ohki, Y.; Takikawa, Y.; Hatanaka, T.; Tatsumi, K. Organometallics.
2006, 25, 3111.
(24) (a) Smith, J. M.; Lachicotte, R. J.; Holland, P. L. J. Am. Chem. Soc.
2003, 125, 15752. (b) Vela, J.; Stoian, S.; Flaschenriem, C. J.; M€unck, E.; Holland,
P. L. J. Am. Chem. Soc. 2004, 126, 4522. (c) Sadique, A. R.; Gregory, E. A.;
Brennessel, W. W.; Holland, P. L. J. Am. Chem. Soc. 2007, 129, 8112.
(25) Yelle, R. B.; Crossland, J. L.; Szymczak, N. K.; Tyler, D. R. Inorg.
Chem. 2009, 48, 861.
(26) Verma, A. K.; Lee, S. C. J. Am. Chem. Soc. 1999, 121, 10838.

Article
Inorganic Chemistry, Vol. 50, No. 4, 2011 1553
Scheme 1
obtained with very hindered chalcogenolates. We have
also explored the use of other arylthiols (Ar = 2,6-di-
methylphenyl, 2,4,6-diisopropylphenyl, 2-methylphenyl,
phenyl) in this chemistry. Although we have not pursued
their isolation at the heteroleptic ferrous amide/thiolate
stage, it seems likely that these other aryl substituents
exhibit similar chemistry based on equivalent reactions
between p-tolylthiol and 2,4,6-triisopropylbenzenesele-
nol and 1 and Mn(N{SiMe3}2)2, respectively, which yield
structural congeners of 2-THF.29,30
1
Dinuclear 2-THF exhibits a complex H NMR spec-
trum in C6D6 with a greater-than-expected number of
signals that suggest speciation. A characteristic, simpli-
fied spectrum can be obtained in THF-d8 solvent that is
identical for both pure 2-THF and for the protolysis
product 2 obtained from reaction in benzene solvent.
On the basis of this observation, we consider the product
2 from 1:1 MesSH/1 protolysis in benzene (the typical
initial in situ precursor state for the hydrazine reduction/
cluster assembly chemistry that follows) as a reactive ma-
terial of empirical formulation Fe(SAr)(N{SiMe3}2). We
attribute the isolation difficulties in the benzene reaction
system to the speciation behavior observed in C6D6,
which likely reflects oligomerization equilibria that
are suppressed in the presence of coordinating THF
solvent.
(b) Hydrazine Reduction and Cubane Formation. The
reaction of either 2-THF (1 equiv) or 2 (2 equiv, assum-
ing the notional Fe(SAr)(N{SiMe3}2) composition) with
1 equiv of 1,2-diarylhydrazine (Ar0NH-NHAr0) results
in an immediate darkening of solution color from red-
brown to maroon-black. After filtration to remove black,
benzene-insoluble material, concentration and addition
of a precipitating solvent yields the neutral cubane cluster
type Fe4(NAr0)4(SAr)4 (3) in black microcrystalline form.
Similar results are obtained using either benzene or THF
as the reaction solvent. Several variations of 3 have been
shown to form by 1H NMR spectroscopy (Ar = Mes, 2,
6-dimethylphenyl (DMP), 2,4,6-triisopropylphenyl, 2,4,6-tri-
phenylphenyl; Ar0 = Ph, p-Tol).26 The presence of 2,6-
substituents on the thiolate aryl group appears to be
required for the formation of stable cluster products;
reactivity is observed with Ar = Ph or 2-methylphenyl,
as evidenced by solution color changes to black, but no
signals associated with paramagnetic species were observed
by NMR spectroscopy. We speculate that less-hindered
arylthiolates favor oligomerization and/or speciation to
ill-defined products. We focus in this report on mesityl-
thiolate-ligated clusters with Ar0=Ph (3a) and p-Tol (3b).
The experimental stoichiometry fits the hypothetical
balanced cluster assembly reaction described by eq 1.
Clusters 3a and 3b are isolated in about 50% yield, with
azoarenes present as side products at significant levels.
The possible implications of azoarene formation in rela-
tion to eq 1, hydrazine reduction, and cluster assembly
mechanism are considered in a subsequent section.
of this synthetic chemistry as well as an investigation of the
scope of the hydrazine reduction.
Results and Discussion
Syntheses. The cluster synthesis chemistry is summar-
ized in Scheme 1.
(a) Ferrous Amide Thiolate Precursor. The starting
point for our syntheses is the ferrous amide precursor
Fe(N{SiMe3}2)2 (1).27 The bis(trimethylsilyl)amide liga-
tion in 1 provides both a strong latent base and a steric
barrier for selective protolysis chemistry, allowing the
introduction of a partial thiolate environment by treatment
with 1 equiv of arylthiol (ArSH). With very hindered
chalcogenols (e.g., 2,6-diphenylbenzenethiol), Fe2(μ-QR)2-
(N{SiMe3}2)2 (Q = O, S, Se) dimers with three-coordinate
iron centers are the known products of this reaction.28
In our investigation, the less hindered mesitylthiol (MesSH)
was employed. Protolysis gives an immediate solution
color change from transparent blue (in tetrahydrofuran
(THF)) or green (benzene) to dark red-brown. Reactions
in benzene did not yield well-defined material (2, vide
infra), whereas in THF reaction medium, crystalline
yellow Fe2(SMes)2(N{SiMe3}2)2(THF)2 (2-THF) is readily
isolated in about 50% yield. Complex 2-THF represents
a solvent adduct of the three-coordinate iron dimers
(27) (a) Andersen, R. A.; Faegri, K.; Green, J. C.; Haaland, A.; Lappert,
M. F.; Leung, W. P.; Rypdal, K. Inorg. Chem. 1988, 27, 1782. (b) Olmstead,
M. M.; Power, P. P.; Shoner, S. C. Inorg. Chem. 1991, 30, 2547.
(28) (a) Bartlett, R. A.; Ellison, J. J.; Power, P. P.; Shoner, S. C. Inorg.
Chem. 1991, 30, 2888. (b) Hauptmann, R.; Kliss, R.; Schneider, J.; Henkel, G. Z.
Anorg. Allg. Chem. 1998, 624, 1927. (c) Hauptmann, R.; Kliss, R.; Henkel, G.
Angew. Chem., Int. Ed. 1999, 38, 377. (d) Komuro, T.; Kawaguchi, H.; Tatsumi,
K. Inorg. Chem. 2002, 41, 5083. (e) Ohta, S.; Ohki, Y.; Ikagawa, Y.; Suizu, R.;
Tatsumi, K. J. Organomet. Chem. 2007, 692, 4792.
4FeðNfSiMe3g2Þ2 þ 4ArSH þ 2Ar0NHNHAr0
f Fe4ðNAr0Þ4ðSArÞ4 þ 8HNðSiMe3Þ2
ð1Þ
(29) Pryadun, R.; Holm, R. H. Inorg. Chem. 2008, 47, 3366.
(30) Bochmann, M.; Powell, A. K.; Song, X. Inorg. Chem. 1994, 33, 400.

1554 Inorganic Chemistry, Vol. 50, No. 4, 2011
Zdilla et al.
A second paramagnetic cluster type is evident as a
minor (<5%) but discernible component in the 1H NMR
spectra of crude reaction systems. The number and relative
intensities of the signals for this minority species suggested a
terminally heteroleptic, C3v-symmetric [Fe4(NAr0)4(SAr)3-
(N{SiMe3}2)]z cubane cluster. Reaction stoichiometries
and substituents (Ar=Mes, DMP; Ar0=Ph, p-Tol) were
adjusted to optimize the formation and isolation of this
new cluster, resulting ultimately in confirmation of the
cluster identity as the salt [Fe(THF)6][Fe4(N-p-Tol)4-
(SDMP)3(N{SiMe3}2)]2 ([Fe(THF)6][4]2). The cubane
oxidation state in 4 is reduced by one electron relative
to 3 based on the presence of the known [Fe(THF)6]2þ
dication31 in the crystal lattice. A balanced reaction can
be formulated via eq 2; application of this stoichiometry
does result in the production of 4 at significant levels,
although cluster type 3 remains the dominant product in
about 2:1 ratio by 1H NMR analysis.
Figure 1. Structure of Fe2(SMes)2(N{SiMe3}2)2(THF)2 (2-THF), with
thermal ellipsoids (50% probability level) and selected atom labels;
hydrogen atoms are not shown. Labels ending in “A” indicate atoms
related by crystallographic inversion symmetry.
9FeðNfSiMe3g2Þ2 þ 6ArSH þ 5Ar0NHNHAr0 þ
-
f Fe2þ þ 2½Fe4ðNAr0Þ4ðSArÞ3ðNfSiMe3g2Þꢀ
˚
Table 1. Selected Distances (A) and Angles (deg) for Fe2(SMes)2(N{SiMe3}2)2-
þ 16HNðSiMe3Þ2 þ Ar0NdNAr0
ð2Þ
(THF)2 (2-THF)
Fe1-S1
Fe1-S1A
Fe1-N1
Fe1-O1
2.4294(7)
2.3847(7)
1.935(2)
2.112(2)
3.534(1)
3.270(1)
S1-Fe1-S1A
N1-Fe1-S1
N1-Fe1-S1A
O1-Fe1-S1
O1-Fe1-S1A
N1-Fe1-O1
Fe1-S1-Fe1A
85.55(2)
119.31(7)
135.00(7)
106.11(6)
101.04(6)
106.02(9)
94.45(2)
Crystal Structures. Dinuclear 2-THF (Figure 1, Table 1)
is an example of the well-precedented thiolate-bridged
[Fe2(μ-SR)2] rhombic core geometry most commonly
associated with terminal thiolate32 or nitrosyl33 ligation.
A structural comparison of 2 against the close p-tolylthio-
late analogue Fe2(S-p-Tol)2(N{SiMe3}2)2(THF)2 has
been reported elsewhere,29 where it was observed that
the two core geometries differ in the nature of the rhombic
distortion: in 2-THF, the rhomb diagonal defined by the
Fe Fe separation is longer than the S S diagonal,
Fe1 Fe1A
3 3 3
S1 S1A
3 3 3
sulfur bridge, which opens the Fe-S-Fe angle as a
consequence.
Clusters 3a and 3b possess the well-recognized M4Q4L4
heterocubane geometry, in which Q is a μ3 core ligand
and L is a terminal ligand, and 4 is a site-differentiated
derivative thereof (Figure 2, Tables 2 and 3). For 3a and
3b, three separate crystal structures, comprising six in-
dependent cluster molecules, have been determined. Across
this set of observations, clusters either possess exact S4
symmetry or can be idealized to approximate this sym-
metry, as dictated in particular by the disposition of thiolate
substituents as shown in Figure 3. Deviations from
perfect S4 symmetry are associated primarily with canting
of the aryl substituents. In all structures, imide and
thiolate substituents associate in a pairwise stacking inter-
3 3 3
3 3 3
whereas the reverse is true for the tolyl derivative. This
structural difference was attributed to a steric effect of the
hindered mesityl substituent. We note here a further
difference in the pyramidalization at sulfur, which is greater
for the tolylthiolate dimer (—Fe-S-C = 105.32(11),
107.66(12)ꢀ for tolyl, 114.07(8), 124.74(9)ꢀ for mesityl).
This apparently induces a smaller Fe-S-Fe angle
(81.74(3)ꢀ for tolyl, 94.45(2)ꢀ for mesityl), resulting in
the reversed rhombic distortion compared to 2-THF. In
essence, the steric requirements of the ortho substituents
on the mesityl group force a more planar 3-coordinate
action of the aryl π faces; the closest C C contacts within
3 3 3
˚
the paired aryl groups occur at approximately 3.5 A,
consistent with packing at van der Waals radii for
aromatic carbons.36 The mixed terminal ligand set in 4
forces lower symmetry and disrupts the intramolecular
aryl stacking such that only two sets of thiolate/imide
substituents are paired.
(31) (a) Pohl, S.; Saak, W. Z. Naturforsch. B 1984, 39, 1236. (b) Saak, W.;
Pohl, S. Z. Anorg. Allg. Chem. 1987, 552, 186. (c) Bolte, M.; Lerner, H.-W.;
Scholz, S. Acta Crystallogr., Sect. E 2001, 57, m216.
(32) (a) Hagen, K. S.; Holm, R. H. Inorg. Chem. 1984, 23, 418. (b)
Hauptmann, R.; Lackmann, J.; Chen, C.; Henkel, G. Acta Crystallogr., Sect. C
1999, 55, 1084. (c) Hauptmann, R.; Lackmann, J.; Henkel, G. Z. Kristallogr.-
New Cryst. Struct. 1999, 214, 132. (d) Chen, C.; Hauptmann, R.; Lackmann, J.;
Henkel, G. Z. Kristallogr.-New Cryst. Struct. 1999, 214, 137. (e) Henkel, G.;
Chen, C. Inorg. Chem. 1993, 32, 1064. (f) Hammann, B.; Chen, C.; Florke, U.;
Hauptmann, R.; Bill, E.; Sinnecker, S.; Henkel, G. Angew. Chem., Int. Ed. 2006,
45, 8245.
The [Fe4(NR)4]q core occurs elsewhere as chloride-ligated
species [Fe4(NtBu)4Cl4]0/1-34 and [Fe4(NPh)4Cl4]2- 35
In
.
general, core geometries are largely equivalent between
these different cluster classes. In comparison with the
(33) (a) Harrop, T. C.; Song, D.; Lippard, S. J. J. Am. Chem. Soc. 2006,
128, 3528. (b) Tsou, C.-C.; Lu, T.-T.; Liaw, W.-F. J. Am. Chem. Soc. 2007, 129,
12626. (c) Chen, Y.-J.; Ku, W.-C.; Feng, L.-T.; Tsai, M.-L.; Hsieh, C.-H.; Hsu,
W.-H.; Liaw, W.-F.; Hung, C.-H.; Chen, Y.-J. J. Am. Chem. Soc. 2008, 130,
10929. (d) Thomas, J. T.; Robertson, J. H.; Cox, E. G. Acta Crystallogr. 1958, 11,
599. (e) Lu, T.-T.; Tsou, C.-C.; Huang, H.-W.; Hsu, I.-J.; Chen, J.-M.; Kuo, T.-S.;
Wang, Y.; Liaw, W.-F. Inorg. Chem. 2008, 47, 6040. (f) Harrop, T. C.; Song, D.;
Lippard, S. J. J. Inorg. Biochem. 2007, 101, 1730. (g) Wang, R.; Camacho-
Fernandez, M. A.; Xu, W.; Zhang, J.; Li, L. Dalton Trans. 2009, 777.
(34) (a) Verma, A. K.; Nazif, T. M.; Achim, C.; Lee, S. C. J. Am. Chem.
Soc. 2000, 122, 11013. (b) Link, H.; Decker, A.; Fenske, D. Z. Anorg. Allg.
Chem. 2000, 626, 1567.
(35) Duncan, J. S.; Nazif, T. M.; Verma, A. K.; Lee, S. C. Inorg. Chem.
2003, 43, 1211.
(36) (a) Bondi, A. J. Phys. Chem. 1964, 68, 441. (b) Rowland, R. S.; Taylor,
R. J. Phys. Chem. 1996, 100, 7384.

Article
Inorganic Chemistry, Vol. 50, No. 4, 2011 1555
Figure 2. Structures of iron-arylimide cubane clusters Fe4(NPh)4(SMes)4 (3a 1.5C6H6, left), Fe4(N-p-Tol)4(SMes)4 (3b 0.5C6H6, tetragonal lattice,
3
3
center), and [Fe4(N-p-Tol)4(SDMB)3N(SiMe3)2]- ([Fe(THF)6][4]2 2THF, right), with thermal ellipsoids (50% probability level) and selected atom labels;
hydrogen atoms are not shown. Two independent, complete clusters occur in the asymmetric unit of 3a 1.5C6H6, of which one is shown above. In the
3
3
asymmetric unit of tetragonal 3b 0.5C6H6, three independent clusters exist, with two possessing crystallographic 4-fold rotoinversion symmetry and one
3
having 2-fold rotational symmetry; one of the S4-symmetric clusters is depicted above, with labels ending in “A”, “B”, or “C” indicating atoms related by
imposed symmetry. A second, monoclinic form of 3b is given as Supporting Information.
Table 2. Selected Distances (A) and Angles (deg) for Fe4(NPh)4(SMes)4 1.5C6H6 (3a 1.5C6H6), Fe4(N-p-Tol)4(SMes)4 0.5C6H6 (3b 0.5C6H6), and Fe4(N-p-Tol)4(SMes)4 (3b)a
˚
3
3
3
3
3a 1.5C6H6
3
3b 0.5C6H6 (tetragonal lattice)b
3b (monoclinic lattice)
a
3
nc meand of n
rangee
nc meand of n
rangee
nc meand of n
rangee
Fe-N
Fe-S
24
8
1.963(13)
2.214(9)
2.63(2)
2.904(12)
95.4(7)
121(3)
1.940(5)-1.985(5)
2.200(2)-2.225(2)
2.5928(12)-2.6662(11)
2.886(7)-2.925(7)
94.4(2)-97.1(2)
114.55(15)-127.6(2)
82.9(2)-85.2(2)
0.036-0.063
12
4
8
1.962(13)
2.215(4)
2.62(2)
2.91(2)
95.7(8)
121(3)
1.942(5)-1.983(5)
2.209(2)-2.218(2)
2.578(2)-2.6468(13)
2.875(9)-2.956(9)
94.6(2)-97.7(2)
115.3(2)-127.5(2)
82.2(2)-84.8(2)
0.028-0.067
12
4
6
1.966(8)
2.215(11)
2.63(3)
1.953(2)-1.976(2)
2.2044(9)-2.2275(9)
2.5712(6)-2.6445(6)
2.899(3)-2.958(3)
94.76(10)-98.29(10)
115.67(7)-124.62(8)
81.81(9)-84.51(9)
0.036-0.060
Fe Fe
N
12
12
24
24
24
3 3 3
N
8
6
2.92(2)
3 3 3
N-Fe-N
N-Fe-S
Fe-N-Fe
12
12
12
8
12
12
12
6
95.8(10)
121(3)
83.8(9)
84.3(6)
0.048(9)
84.0(7)
0.049(12)
Fe2N2 planarityf 12
0.052(11)
a Asymmetric unit contains two independent, complete clusters. b Asymmetric unit contains two independent quarter-clusters (crystallographic S4
symmetry) and a single half-cluster (crystallographic C2 symmetry). c Number of independent observations of the metric type. d Arithmetic mean, with
uncertainty representing the standard deviation from the mean. e Range of independent values. f rms displacement of the four rhomb atoms.
˚
Table 3. Selected Distances (A) and Angles (deg) for[Fe(THF)6][Fe4(N-p-Tol)4-
(SDMP)3(N{SiMe3}2)]2 2THF ([Fe(THF)6][4]2 2THF)
3
3
na
mean of nb
rangec
Fe-N
Fe-S
12
3
6
6
12
9
12
6
1
3
4
1.98(3)
2.248(8)
2.66(8)
2.93(4)
95(3)
121(6)
84(2)
0.05(4)
1.940(6)
122(9)
2.138(19)
1.927(5)-2.034(5)
2.242(2)-2.257(2)
2.5929(13)-2.8113(13)
2.853(7)-2.961(7)
89.8(2)-97.7(2)
113.0(2)-132.2(2)
82.2(2)-89.0(2)
0.019-0.102
Fe Fe
N
3 3 3
N
3 3 3
N-Fe-N
N-Fe-S
Fe-N-Fe
Fe2N2 planarityd
Fe-N(amide)
N(amide)-Fe-N
Fe-O(THF)
113.8(3)-132.2(2)
2.117(6)-2.159(4)
a Number of independent observations of the metric type. b Arith-
metic mean, with uncertainty representing the standard deviation from
the mean. c Range of independent values. d rms displacement of the four
rhomb atoms.
Figure 3. Projection along the pseudo-S4 axis of the C1-symmetric
structure of Fe4(NPh)4(SMes)4 (3a 1.5C6H6), illustrating the idealized
S4 cluster symmetry and the disposition of the aryl ligand substituents.
For clarity, top and bottom halves of the cluster are depicted in red and
blue, respectively.
3
NtBu clusters, the present arylimide clusters exhibit
˚
slightly longer Fe-N bonds (0.03 A) at the same all-ferric
oxidation state; this appears to be an intrinsic difference
associated with aryl versus alkyl substitution. In compar-
ing arylimide species, the reduced cluster possesses longer
The iron-imide (Fe-NR) cores of 3a, 3b, and 4 can also
be compared with the [Fe4S4] cores that constitute the
largest and most well-studied set of weak-field hetero-
cubanes.37 The core geometry of Fe4S4 clusters can be
Fe-N and Fe Fe distances, consistent with the repul-
sive effect of increased charge.
3 3 3

1556 Inorganic Chemistry, Vol. 50, No. 4, 2011
Zdilla et al.
Table 4. Spectroscopic Data
c
electronic absorption:a λ, nm
(εM, L mol-1 cm-1
Mossbauer: δ (|ΔEQ|),
mm/s
€
)
1H NMR:b δ, ppm
3
3
Fe2(SMes)2(N{SiMe3}2)2-
(THF)2 (2-THF)
255 (21,800), 285 (13,900)
THF-d8: 15.7, 22.1, 23.6, 27.9, 33.0 (all v br); CD3CN:d
1.8 (THF), 3.6 (THF), 15.4 (br, 18H, SiMe3),
30.1 (br, 6H, o-Me), 34.3 (br, 2H, m-H),
0.87 (1.84)
39.4 (br, 3H, p-Me)
Fe4(NPh)4(SMes)4 (3a)
-0.99 (br, 4H, Ph-p-H), 0.65 (br, 8H, Ph-o-H),
5.01 (12H, Mes-p-Me), 5.49 (br, 24H, Mes-o-Me),
11.77 (8H, Mes-m-H), 13.51 (8H, Ph-m-H)
1.8 (br, 8H, Tol-o-H), 4.47 (12H, Mes-p-Me),
5.40 (br, 24H, Mes-o-Me), 7.71 (12H, Tol-p-Me),
11.623 (8H, Mes-m-H), 12.91 (8H, Tol-m-H)
-2.79 (br, 6H, Tol-o-H), 0.0 (br, 2H, Tol-o-H),
1.66 (br, 24H, Fe-THF-β-H), 1.96 (18H, SiMe3),
4.15 (br, 24H, Fe-THF-R-H), 5.27 (3H, DMP-p-H),
5.40 (18H, DMP-o-Me), 9.44 (3H, Tol-p-Me),
11.77 (6H, DMP-m-H), 12.75 (9H, Tol-p-Me),
13.24 (2H, Tol-m-H), 14.88 (6H, Tol-m-H)
Fe4(N-p-Tol)4(SMes)4 (3b)
240 (62,800), 455 (36,400)
235 (128,000), 326 (63,900)e
0.35 (0.70)
[Fe(THF)6][Fe4(N-p-Tol)4-
(SDMP)3(N{SiMe3}2)]2
([Fe(THF)6][4]2)
a THF solution, ca. 22 ꢀC. b THF-d8 and CD3CN for 2-THF, C6D6 solution for all others, ca. 22 ꢀC. c Isomer shifts measured at 4.2 K, referenced to Fe
metal at room temperature; errors in fit estimated at 0.02 mm/s. d The spectrum of crystalline 2-THF in CD3CN also shows an additional feature, an
upfield shoulder to the SiMe3 resonance at ca. 15% relative intensity; this signal appears to correspond to 1 in CD3CN and can be eliminated by titration
with additional MesSH. e Molar absorption coefficients reported per mole of cluster 4 to facilitate comparison with values for 3b.
viewed as a polyhedron constructed from concentric
interpenetrating Fe4 and S4 tetrahedra, with distinctly
nonplanar Fe2S2 rhomb faces; S-Fe-S angles approach
tetrahedral angles, but Fe-S-Fe angles are markedly
acute (ca. 73ꢀ, on average). By comparison, the imide
cubanes are more compact because of shorter Fe-Q bond
lengths and also more cubic, with flatter Fe2N2 rhombs
and rhomb interior angles that more closely approach
derivatization. Cluster paramagnetism generates isotro-
pically shifted signals for the aryl substituents that alternate
in displaced direction with alternant hydrocarbon posi-
tion (downfield: m-H, o-Me, p-Me; upfield: o-H, p-H),
implying a predominant spin polarization (delocalization)
mechanism.38
Solution electronic absorption spectra of the Fe-NR
cubane clusters display intense charge transfer bands in
the near-UV-to-visible regions; comparative UV-vis spectra
(including that of 2) are presented in Figure 5. The all-
ferric cubanes 3 show much more extensive absorption
into the visible range of the spectrum in comparison with
the one-electron reduced and ligand substituted cluster 4,
resulting in visually apparent color differences of maroon
(3a, 3b) versus brown (4). This is consistent with LMCT
behavior upon replacement of a thiolate by an amide
donor and upon reduction of core oxidation state, both of
which destabilize further core reduction and shift LMCT
transitions to higher energy.39 Redox properties, how-
ever, could not be quantified directly; cyclic voltammetry
revealed only broad, irreversible processes for 3a, and no
measurements were made on side product 4.
90ꢀ. Fe Fe separations are slightly shorter in the imide
3 3 3
˚
system by about 0.1 A. Assuming rigid Fe-Q bond
lengths and factoring the difference in lengths, the struc-
tural observations suggest that nonbonded, repulsive
interactions within the Fe2Q2 rhombs account for the
angular differences. For the larger sulfide donor, a longer
Q
Q separation is enforced, resulting in sharper Fe-
3 3 3
Q-Fe angles. The smaller nitrogen donors allow closer
Q
while repulsive Fe Fe interactions prevent more open
Q contacts, therefore more open Fe-Q-Fe angles,
3 3 3
3 3 3
Q-Fe-Q angles.
˚
At a finer level (deviations of ca. 0.03 A for Fe-S
distances), the cores of Fe4S4 clusters also exhibit slight,
regular distortions away from idealized Td symmetry,
usually in the form of a D2d distortion.37 The Fe4(NR)4
cores show evidence of similar deformations, although
the deviations are less internally consistent than those
observed in the Fe-S cubanes. We note that the promi-
nent interligand stacking interactions in the present sys-
tem may affect core metrics, precluding the identification
of subtle intrinsic core geometry preferences.
In the solid state, zero-field 57Fe Mossbauer measure-
€
ment of polycrystalline 3b reveals a single quadrupole
doublet with isomer shift and quadrupole splitting con-
sistent with high-spin Fe(III) and quite similar to those of
the only other all-ferric Fe-NR cubane, Fe4(NtBu)4Cl4,
34
characterized by Mossbauer spectroscopy. Magnetic
€
susceptibility measurements on 3b reveal temperature-
dependent susceptibility that decreases to a minimum
with decreasing temperature (1.1 μB at 55 K) before rising
at still lower temperatures (Figure 6). This behavior is
well-known in Fe-S cubane systems and indicates a net
diamagnetic cluster ground state in the presence of trace
paramagnetic contamination.39,40 Magnetically active,
Physical Properties of Cubane Clusters. Spectroscopic
data are compiled in Table 4.
In solution, cubane cluster types 3 and 4 are readily
identified by their characteristic 1H NMR spectra (Figure 4).
All derivatives investigated show spectral patterns con-
sistent with freely rotating ligand substituents and highest
averaged point symmetry (Td for 3, C3v for 4), and reso-
nances were completely assigned through a combination
1
(38) (a) Horrocks, W. D., Jr. Analysis of Isotropic Shifts. In NMR of
Paramagnetic Molecules: Principles and Applications; LaMar, G. N., Horrocks,
W. D., Jr., Holm, R. H., Eds.; Academic Press: New York, 1973; Chapter 4. (b)
Holm, R. H.; Phillips, W. D.; Averill, B. A.; Mayerle, J. J.; Herskovitz, T. J. Am.
Chem. Soc. 1974, 96, 2109.
of relative peak areas, H COSY, NOEDS, and ligand
(37) (a) Rao, P. V.; Holm, R. H. Chem. Rev. 2004, 104, 527. (b) Berg, J. M.;
Holm, R. H. In Iron-Sulfur Proteins; Spiro, T. G., Ed.; Wiley: New York, 1982;
Chapter 1.
(39) Sharp, C. R.; Duncan, J. S.; Lee, S. C. Inorg. Chem. 2010, 49, 6697.

Article
Inorganic Chemistry, Vol. 50, No. 4, 2011 1557
Figure 4. 1H NMR spectra (C6D6, 500 MHz, 295 K) of Fe4(N-p-Tol)4(SMes)4 (3b, top), and [Fe(THF)6][Fe4(N-p-Tol)4(SDMB)3N(SiMe3)2]2
([Fe(THF)6][4]2, bottom). Resonances are labeled with superscripts indicating ligand type (N=imide, S=thiolate) and x=trace diamagnetic con-
taminants (solvent or silylmethyl-containing species).
Figure 6. Temperature dependence plot of the molar magnetic suscept-
Figure 5. UV-visible absorption spectra of Fe2(SMes)2(N{SiMe3}2)2-
(THF)2 (2-THF), Fe4(N-p-Tol)4(SMes)4 (3b), and [Fe(THF)6][Fe4(N-p-
ibility of polycrystalline Fe4(N-p-Tol)4(SMes)4 (3b).
Tol)4(SDMB)3N(SiMe3)2]2 ([Fe(THF)6][4]2) in THF. The spectrum of 2-
THF (1.0 mM) cannot be associated definitively with the solid-state
structure because of possible speciation in solution. The molar absorptiv-
ity of [Fe(THF)6][4]2 is normalized per cubane cluster.
but spectroscopically silent, impurities are persistent in
this system and must be minimized by careful recrystalli-
zation; indeed, our earlier magnetic measurements and
spin state assignment for cluster 3a (originally reported
as a S=2 ground state) are probably in error because
of magnetic contaminants.26 For cluster 4, zero-field
€
Mossbauer measurement yielded broad absorptions that
could not be reliably fitted, and magnetochemical studies
were not attempted. The electron paramagnetic resonance
(EPR) spectrum of 4a at 12 K in frozen toluene glass
(Figure 7) shows an axial signal pattern with g=2.034,
2.16(4) that is consistent with a S=1/2 ground state.
N-N Bond Reduction: Mechanism and Scope. Given
the reaction stoichiometry proposed in eq 1 and the dinuclear
solid-state structures of precursors 127b and 2-THF, it is
Figure 7. Perpendicular mode EPR spectrum of [Fe(THF)6][Fe4(N-p-
Tol)4(SDMB)3N(SiMe3)2]2 ([Fe(THF)6][4]2) (12 K, frozen toluene glass),
with axial signal features indicated.
tempting to propose that cluster assembly proceeds via
the simple three-step sequence outlined in Scheme 2. This
mechanism assumes tetrahedral, 4-coordinated Fe(II/III)
environments and the intermediacy of bridged dinuclear
species. The reaction begins with protolysis of the latent
amide base in 2 and concomitant formation of a μ:η2,η2
hydrazide(2-)-bridged dimer. Reductive scission of the
(40) (a) Laskowski, E. J.; Frankel, R. B.; Gillum, W. O.; Papaefthymiaou,
G. C.; Renaud, J.; Ibers, J. A.; Holm, R. H. J. Am. Chem. Soc. 1978, 100,
5322. (b) Herskovitz, T.; Averill, B. A.; Holm, R. H.; Ibers, J. A.; Phillips, W. D.;
Weiher, J. F. Proc. Natl. Acad. Sci. U.S.A. 1972, 69, 2437. (c) Cleland, W. E.;
Holtman, D. A.; Sabat, M.; Ibers, J. A.; DeFotis, G. C.; Averill, B. A. J. Am.
Chem. Soc. 1983, 105, 6021.

1558 Inorganic Chemistry, Vol. 50, No. 4, 2011
Zdilla et al.
Scheme 2a
based on test reactions of 2 (2 equiv) with various
hydrazines (1 equiv) in benzene, with reaction assays
summarized in Table 5.
(a) Mass Balance and Azoarene Formation. The for-
mation of significant quantities of azoarenes in the synth-
esis of 3 requires explanation, particularly as the reaction
stoichiometry proposed in eq 1 does not produce azoar-
ene. Hydrazine disproportionation (eq 3) is one potential
source of diazenes; this reaction displays no net metal-
based redox changes and appears to be the primary
mechanism of N-N bond reduction in several iron-
mediated reaction systems.21a,24
2PhNHNHPh f 2PhNH2 þ PhNdNPh
ð3Þ
To probe the relationship of azoarene formation to N-N
bond reduction in the present system, we endeavored to
quantitate the mass balance following cluster assembly.
NMR spectroscopic analysis of the crude reaction mix-
tures was found to be inadequate for this purpose
because of ill-defined, magnetically active contaminants
that broadened resonances and precluded accurate signal
integration. Instead, organic products were identified and
quantified by gas chromatographic analysis following
acid digestion to release metal-bound moieties and deri-
vatization (benzoylation42) to detectable analytes.
Analyses of the reactions with 1,2-Ar0NHNHAr0 give
47 and 55% Ar0NHBz (benzoylated arylamine) for Ar0 =
Ph and p-Tol, respectively, and 18% Ar0NdNAr0 (for
either aryl group) based on initial hydrazine nitrogen
content, with no other arylnitrogen species detected.
The total recovery of diarylhydrazine-derived species is
about 70%, which is similar to recoveries from control
assays under equivalent conditions (ca. 80% maximum,
see Experimental Section); we therefore believe these results
accurately reflect the distribution of arylnitrogen-containing
species in the crude reaction systems. These assays indicate
that about 75% of the hydrazine is reduced and 25% is
oxidized. This immediately excludes disproportionation
as the primary pathway for N-N bond reduction, inasmuch
as eq 3 can only explain one-third of the reduced nitrogen
content based on azoarene yields.
Nonetheless, we believe that disproportionation does
contribute to the product distribution. The product het-
erocubane clusters 3a and 3b, or unidentified species derived
therefrom, catalyze the disproportionation of 1,2-diaryl-
hydrazines to arylamines and azoarenes. Disproportio-
nation is evident and effective at low cluster loadings
(0.1 mol % vs hydrazine), and the reaction is complete
within minutes at cluster and hydrazine concentrations
similar to actual preparative conditions. Azoarenes them-
selves are unreactive with the species in this system. We
have observed facile, catalytic diarylhydrazine dispropor-
tionation in an unrelated ferrous thiolate system;43 these
findings, along with other documented roles in Fe-mediated
hydrazine reduction chemistry,21a,24 suggest that dispro-
portionation is a common reactivity mode. If we attribute
all azoarene formation to this catalytic process, thereby
accounting for one-third of the reduced nitrogen con-
tent, and we assume eq 1 as the operative pathway that
a Note: The 3-coordinate iron centers are notional; solvent coordination or
bridging interactions to give 4-coordinate iron are likely.
Table 5. Reactivity of 2 (2 equiv) with N-N/NdN Substrates (1 equiv)
substrate
productsa (% yield)
PhHN-NHPh
PhNHBz (47), PhNdNPh (18)
(p-Tol)HN-NH(p-Tol) (p-Tol)NHBz (55), (p-Tol)NdN(p-Tol) (18)
PhHN-N(Me)Ph
Ph(Me)N-N(Me)Ph
iPrHN-NHiPr
tBuHN-NHtBu
PhNdNPh
3 and other clustersb
no reaction
iPrNHBz (13), iPr(Bz)N-NHBz (38)
no reaction
no reaction
H2N-NH2
H2NBz (5)
a Characterized by GC assay following acidification/extraction/ben-
zoylation protocol unless otherwise indicated. All peaks in the chroma-
tograms were identified, including thiolate-derived species (MesSH,
MesSSMes, MesSBz), BzOH, and BzOBz; only substrate-derived pro-
ducts are reported. b Identified by NMR assay of crude reaction mixture.
N-N bond coupled with simultaneous 2-electron oxida-
tion of the two ferrous centers gives a diferric diimide
dimer as the next step. Finally, fragment condensation of
two diimide dimers yields the ultimate cubane product.
The feasibility of the final step has been demonstrated and
is discussed elsewhere as a separate report;41 we address
here the chemistry of N-N bond reduction. We do not
exclude the possibility of alternate, non-concerted, one-
electron mechanisms in this chemistry; Scheme 2 merely
offers a simple starting hypothesis for consideration.
The coupling of hydrazine N-N bond reduction with
oxidation of Fe(II) to Fe(III) in dinuclear complexes per
Scheme 2 has been invoked in a dinuclear cyclopentadie-
nyl-ligated system23 and, very recently, observed as a
direct transformation in a triphosphinoborate-ligated
complex.21b In the present system, mechanistic investiga-
tion is hindered by factors intrinsic to weak-field cluster
self-assembly chemistry. The weak-field, 4-coordinate
iron sites are exchange-labile and undergo substantial,
rapid reorganization over the course of product formation.
Solution spectroscopic compound identification is chal-
lenging without prior physical isolation and structural
characterization, and paramagnetic line broadening in-
creases the difficulty in detecting low-concentration spe-
cies. In general, the direct tracking of a reaction sequence
is not possible under these conditions. Operating within
these constraints, we offer the following observations
(42) Barreira, E. S.; Parente, J. P.; Wilson de Alencar, J.. J. Chromatogr.
1987, 398, 381.
(43) Zdilla, M. J.; Verma, A. K.; Lee, S. C. Inorg. Chem. 2008, 47, 11382.
(41) Zdilla, M. J.; Lee, S. C., manuscript in preparation.

Article
Inorganic Chemistry, Vol. 50, No. 4, 2011 1559
generates the remainder of the reduced nitrogen, we
expect a 50% maximum yield of cubane 3, which is
comparable to the actual crude isolated yields.
Parent and alkyl-substituted hydrazines were also tested
for reactivity. Not all substrates are reactive, but those
that are give black product mixtures visually similar to the
arylhydrazine reaction systems. The dark solids isolated
from these reactions, however, proved difficult to char-
acterize, with poor solubility in benzene and no diagnos-
tic NMR signals in THF-d8 solution.
Parent hydrazine reacts to form a black reaction mixture
within 5 min. Analysis of the crude reaction mixture by
either acid digestion/benzoylation/GC or spectroscopic
indophenol assay44 confirm the presence of ammonia in
low yield (5% based on total N2H4 and [N(SiMe3)2]-
nitrogen content). This recovery is too low to interpret,
especially as silylamide hydrolysis provides an additional
source of reduced nitrogen. It is possible that ammonia is
lost during our workup protocol, that reduced nitrogen is
trapped as insoluble iron-containing precipitate, or that
N2H4 functions as a reductant in this system, with release
of N2 and formation of reduced iron.
1,2-Diisopropylhydrazine reacts at diminished rates
relative to the reactions with 1,2-diarylhydrazines (darken-
ing to opacity in ca. 45 min vs immediate). Assay of the
crude reaction mixture by benzoylation/GC analysis
reveals that 13% of the nitrogen originates from N-N
bond cleavage, while 38% of the nitrogen content appears
in the form of iPrBzNNHBz; azo-isopropane, if present,
was not detected because of its low boiling point. The
recovery of unreduced hydrazine with a missing isopropyl
substituent indicates the existence of an alternate reaction
path. Although the fate of the alkyl group is uncertain,
one possibility is hydrogen elimination at the R-position
to form a hydrazone, which hydrolyzes upon acidic aqueous
workup (eq 5) to give, after derivatization, the final analyte.
Similar behavior has been observed in other Fe-NR cluster
systems and has precluded preparation of iron-alkylimide
clusters with hydrogens at positions R to nitrogen.45
The other product formed in this system is a black,
benzene-insoluble powder. 1H NMR analysis of this
material in THF-d8 reveals broad, weak signals indicative
of paramagnetic material but is otherwise uninforma-
tive. Acid digestion/benzoylation/GC assay of the Ar0 =
Ph reaction shows the presence of mesitylthiolate-
derived analytes (MesSH, MesSBz, (MesS)2) at 14% total
yield, as well as low levels of PhNHBz (3%), indicating
that the material is predominantly iron-thiolate in com-
position.
(b) [NAr] Crossover. A crossover experiment involving
the reaction of a 1:1 mixture of 1,2-diphenylhydrazineand
1,2-di-p-tolylhydrazine (at 1 equiv of total hydrazine con-
tent) was conducted to obtain evidence of the concerted
dinuclear mechanism proposed in Scheme 2. The detec-
tion of a mixed imide Fe4(NPh)n(N-p-Tol)n-4(SMes)4 cu-
bane distribution favoring n=0, 2, and 4 congeners would
support this mechanism. Instead, extensive core mixing is
observed, with essentially statistical distributions of all
possible cluster core compositions. This result is incon-
clusive, however, as the production of arylamine via
disproportionation (eq 3) gives rise to a side reaction,
specifically the rapid exchange of core imides with free
amines (eq 4). The details of this core exchange chemistry
and full spectroscopic assignments are described in a
separate report.41 We note that azoarenes obtained in
this crossover experiment, as well as in catalytic dispro-
portionation assays involving mixed hydrazines, are almost
exclusively (90-95%) symmetric by GC/MS analysis,
strongly suggesting that azoarene does not arise from
separated ArN fragments; the low levels of mixed azoar-
ene product probably originate from the oxidation of
reduced arylnitrogen species during the aerobic workup
prior to GC analysis.
i
iPrHNNdCMe2 þ H2O f PrNHNH2 þ Me2CdO
Fe4ðNAr0Þ4ðSArÞ4 þ ArꢁNH2
ð5Þ
f Fe4ðNArꢁÞnðNAr0Þ4 - nðSArÞ4
ðn ¼ 0- 4Þ
ð4Þ
Hydrazine sterics also appear to influence the reactiv-
ity. Thus, 1,2-diethylhydrazine reacts in a qualitatively
similar fashion to 1,2-diisopropylhydrazine, although at a
faster rate (ca. 30 s); product analysis was not attempted
on this system because of the volatility of the ethylamine
product prior to derivatization. 1,2-Di-t-butylhydrazine,
by contrast, shows no color change when mixed with 2,
although a small quantity of light-colored precipitate is
observed. 1H NMR analysis suggests the formation of a
weak adduct in C6D6 based on the appearance of new,
sharp, isotropically shifted resonances; addition of THF,
however, returns crystalline 2 in quantitative yield. The
absence of reactivity for this hydrazine likely arises from
the congested steric environment about the N-N bond,
which may hinder adoption of ligand geometries neces-
sary for N-N bond scission.
(c) Scope of N-N Bond Reduction. Other hydrazines
were surveyed to delineate the extent of N-N reduction
chemistry. Tetrasubstituted PhMeN-NMePh is unreactive,
while trisubstituted PhMeN-NHPh leads to a dark solu-
tion that reveals, by 1H NMR analysis, a complex array of
isotropically shifted resonances characteristic of para-
magnetic cluster species, including the signal sets for 3a
and the (NPh,SMes)-derivative of 4. These results suggest
that N-H deprotonation is necessary for N-N bond
reduction in this system and that a single deprotonation
to hydrazide(1-) is sufficient to achieve this. Perhaps the
formation of a dinuclear N-bridging interaction facil-
itates the reduction of the N-N bond by organizing a
2-electron reduction from 1-electron Fe(II) reductants;
deprotonation of the substrate to the nitrogen anion form
would favor N-bridging. No azoarene is detected in the
trisubstituted hydrazine system, consistent with the con-
clusions from the mass balance and crossover experiments
that argue against a direct role for hydrazine dispropor-
tionation in the formation of the cubane clusters.
Summary and Implications
Ferrous amide thiolate species can react with 1,2-diarylhy-
drazines to form iron-arylimide-thiolate cubanes. The imide
(44) Chaney, A. L.; Marbach, E. P. Clin. Chem. 1962, 8, 130.
(45) Duncan, J. S.; Zdilla, M. J.; Lee, S. C. Inorg. Chem. 2007, 46, 1071.

1560 Inorganic Chemistry, Vol. 50, No. 4, 2011
Zdilla et al.
ligands arise from the 2-electron reduction of the hydrazine
N-N bond, through a transformation that appears to be
coupled to the 1-electron oxidation of Fe(II) to Fe(III).
Although reactivity is evident with a number of different
thiolate ligands and hydrazine substrates, well-defined iron-
containing products (cubane clusters, exclusively) are found
only from the specific combination of sterically hindered
arylthiolate ligands and 1,2-diarylhydrazines. Analysis of the
organic products reveals behavior that is variable and sub-
strate-dependent, with catalytic hydrazine disproportiona-
tion and organohydrazine N-C bond cleavage also observed
under appropriate conditions. These results establish that
simple N-N bond reduction is but one of several possible
reactivity modes available in the reaction of simple iron
precursors with hydrazines.
Observations derived from synthetic systems can only be
extrapolated to N-N bond reduction in nitrogenases with
caution. At present, the enzyme systems provide limited
information on reactive states, synthetic complexes mimic the
FeMo-cofactor environment incompletely at best, and iron-
hydrazine reaction chemistry is demonstrably complex. Un-
der these circumstances, synthetic results only outline chemi-
cal possibilities. In previous studies of synthetic weak-field
iron-sulfur environments, parent and alkyl-substituted hy-
drazines were found inert to reaction chemistry beyond simple
N-coordination using low-coordinate ferrous thiolate42 or
[Fe4S4(SR)4]2-12 species under aprotic or reducing/weakly
protonating conditions; electrochemical reductions in protic
solvents, however, have been noted.13 For arylhydrazines,
N-N bond cleavage from disproportionation or rearrange-
ment reactions has been documented with ferrous thiolate42
and diferrous sulfide-bridged β-diketiminate24b complexes in
aprotic solvents. The present study expands the known
chemistry, with N-N bond reduction proven in both aryl-
and alkylhydrazines; activation via hydrazine deprotonation
appears important in achieving this reactivity. Speculatively,
these observations seem to fit with the limited activity of the
nitrogenase enzymes for exogenous hydrazine substrate,
although access to the reaction pocket and selective delivery
of protons may be the limiting factors in the enzymatic
reduction.5,6 Partially reduced, anionic dinitrogen intermedi-
ates, that is, ligated hydrazides, are expected to undergo
N-N cleavage chemistry readily in weak-field ferrous envir-
onments.
with CH2Cl2, preliminary drying over KOH, and final distilla-
tion from CaH2. Fe(N{SiMe3}2)2 (1) was synthesized by the
procedure of Anderson et al,27a using FeCl2 in place of FeBr2-
(THF)2.50 All other chemicals were obtained from commercial
sources or prepared as described below.
Fe2(μ-SMes)2(N{SiMe3}2)2(THF)2 (2-THF). A clear solution
of MesSH (0.304 g, 2.00 mmol) in THF (10 mL) was added to a
stirred blue solution of 1 (0.753 g, 2.00 mmol) in THF (10 mL) to
afford a color change from blue to red-brown. The reaction
mixture was stirred 30 min and concentrated in vacuo to 10 mL,
at which point yellow microcrystals were observed on the sides
of the flask. To this mixture, 10 mL of n-pentane was added,
crystallizing more material. The mixture was held at -30 ꢀC for
20 min, then filtered to give 0.396 g of yellow microcrystalline 2
(0.450 mmol, 45%). Anal. Calcd for C38H74O2N2S2Si4Fe2: C,
51.91; H, 8.48; N, 3.19. Found: C, 51.54; H, 8.52; N, 3.14.
Fe4(μ3-NPh)4(SMes)4 (3a). MeSH (0.405 g, 2.66 mmol) in
10 mL of benzene was added dropwise over the course of 5 min
to a stirred green solution of 1 (0.753 g, 2.0 mmol) in 10 mL of
benzene to afford a brown solution with a brown precipitate,
which dissolved upon further stirring for 10 min. To this brown
solution, N,N0-diphenylhydrazine (0.245 g, 1.33 mmol) in 10 mL
of benzene was added, resulting in an immediate color change to
black. The reaction was stirred for 20 h, filtered to remove
benzene-insoluble black powder, concentrated in vacuo to 5 mL,
and diluted with 30 mL of hexamethyldisiloxane to give black
microcrystalline material. The mixture was chilled for 20 min
at -30 ꢀC and filtered to isolate 0.193 g of black microcrystals.
This first crop is contaminated by small amounts of NMR-silent,
paramagnetic impurities based on NMR, magnetochemical, and
elemental analyses. Product of improved purity was obtained
from the filtrate following the initial crop by evaporation in
vacuo to a total solution volume of 1-3 mL and storage for 1-2
days at -30 ꢀC. The black crystals that result were isolated and
dried (0.180 g, 0.151 mmol, 22.7%; total crude yield, including
first crop: 47%). Anal. Calcd for C60H64N4S4Fe4: C, 60.42; H,
5.41; N, 4.70. Found: C, 59.86; H, 4.93; N, 4.50.
Fe4(μ3-N-p-Tol)4(SMes)4 (3b). A procedure analogous to
the preparation of 3a was followed using MesSH (0.304 g, 2.00
mmol, in 10 mL of benzene), 1 (0.753 g, 2.00 mmol, in 10 mL of
benzene), and N,N0-di-p-tolylhydrazine (0.213 g, 1.00 mmol in
10 mL of benzene). After the first filtration to remove benzene-
insoluble black powder, the filtrate was concentrated in vacuo to
3 mL and treated with 15 mL of pentane to give black micro-
crystals. The mixture was held at -30 ꢀC for 20 min, then filtered
to yield 0.256 g of black microcrystalline 3b (0.199 mmol, 40%
crude yield). Analytically pure samples were obtained by re-
crystallization of the initial microcrystalline product via
n-pentane/benzene vapor diffusion at room temperature as fol-
lows: 50 mg of crude 3b was dissolved in 0.4 mL of benzene, into
which 2 mL of n-pentane was diffused; after 10 d, the super-
natant was decanted, and the crystals rinsed with cold n-pentane
to give 27 mg of pure crystalline 3b (54% recovery to give 22%
Experimental Section
General Considerations. All manipulations were conducted at
room temperature under a pure dinitrogen atmosphere unless
overall yield). Anal. Calcd for C64H72N4S4Fe4 0.5C6H6; C,
otherwise indicated. Previously described anaerobic synthetic
€
3
1
62.48; H, 5.87; N, 4.35. Found: C, 62.41; H, 5.94; N, 4.32.
[Fe(THF)6][Fe4(μ3-N-p-Tol)4(SDMB)3N(SiMe3)2]2 ([Fe(THF)6]-
[4]2). A procedure analogous to the preparation of 3a was
followed using 2,6-dimethylbenzenethiol (DMBSH; 0.185 g,
1.34 mmol, in 10 mL of benzene), 1 (0.753 g, 2.00 mmol, in
10 mL of benzene), and N,N0-di-p-tolylhydrazine (0.236 g,
1.11 mmol in 10 mL of benzene). The black reaction mixture was
stirred for 20 h, then concentrated in vacuo to 5 mL, diluted with
15 mL of n-pentane, chilled for 20 min at -30 ꢀC, and filtered to
give 0.060 g of black powder, which contained 4 contaminated
with substantial amounts of SDMB-ligated 3. The filtrate was
and characterization ( H NMR, UV-vis, Mossbauer, EPR
spectroscopies; magnetic susceptibility measurement; elemental
analysis) protocols35,45 were used in the present study. 1,2-Diiso-
propylhydrazine,46 isopropylhydrazine hydrochloride,46 1,2-di-
t-butylhydrazine,47 1,2-di-p-tolylhydrazine,48 1-methyl-1,2-
diphenylhydrazine,49 1-methyl-1,2-di-p-tolylhydrazine,49 and
1,2-dimethyl-1,2-diphenylhydrazine49 were prepared by litera-
ture methods. 1,2-Diethylhydrazine was obtained from the
commercially available dihydrochloride salt (Aldrich) by treat-
ment with excess 40% aqueous KOH, followed by extraction
(46) Ghali, N. I.; Venton, D. L.; Hung, S. C.; Le Breton, G. C. J. Org.
Chem. 1981, 46, 5413.
(50) The presence of THF in the reaction system can lead to THF
contamination in the final distilled product; details are provided as Support-
ing Information.
(47) Stowell, J. C. J. Org. Chem. 1967, 32, 2360.
(48) Carlin, R. B.; Wich, G. S. J. Am. Chem. Soc. 1958, 80, 4023.
(49) Katritzky, A. R.; Wu, J.; Verin, S. V. Synthesis 1995, 651.

Article
Inorganic Chemistry, Vol. 50, No. 4, 2011 1561
dried in vacuo to give 0.279 g of black powder. This powder was
recrystallized by vapor diffusion of n-pentane/THF at -30 ꢀC.
After 60 h, black crystals were isolated, washed with cold (-30 ꢀC)
n-pentane, and dried in vacuo to give 0.070 g of 4 (0.02 mmol,
temperature program set to an initial temperature of 70 ꢀC,
ramping at 10 ꢀC/min for 8 min, holding at 150 ꢀC for 2 min,
ramping at 20 ꢀC/min for 5 min, and holding at a final
temperature of 250 ꢀC for 2 min.
10%). Anal. Calcd for C140H194N10O6S6Si4Fe9 2C4H8O: C, 58.00;
3
(a) GC Assay for Fe-Mediated Hydrazine Reactivity. The
organic products of the reaction between 2 and various hydra-
zines were separated and analyzed as follows. Under anaerobic
conditions, MesSH (0.021 g, 0.20 mmol) in 2 mL of benzene
was added dropwise to a stirred green solution of 1 (0.076 g,
0.20 mmol) in 2 mL of benzene, affording a color change to
brown. After stirring for 5 min, 0.10 mmol of hydrazine (neat for
diisopropylhydrazine, dissolved in 1 mL of benzene for dia-
rylhydrazines) was added, darkening the solution color to black
(immediate for aryl hydrazines, over the course of 45 min for
diisopropylhydrazine). After 30 h, the reaction vessel was
opened to air, and 5 mL of 6 M HCl added to liberate ligands
from iron by protonolysis. The bilayer was mixed by vigorous
stirring for 2 h, then partitioned.
The benzene layer, which contained neutral organic species
(e.g., MesSH, (MesS)2, azo compounds), was rinsed with 1 mL
of 6 M HCl, partitioned, and dried over anhydrous Na2SO4.
After filtration, a known quantity of internal integration standard
(benzophenone or anthracene, typically 30 mg) was added, and
the solution contents analyzed by GC.
H, 6.91; N, 4.57. Found: C, 58.42; H, 6.56; N, 4.08.
Other [Fe4(μ3-NAr0)4(SAr)3(N{SiMe3}2)]- Cubanes. In our
efforts to identify the cluster type represented by 4, the follow-
ing derivatives were also prepared: [Fe4(μ3-NPh)4(SMes)3(N-
{SiMe3}2)]-, [Fe4(μ3-N-p-Tol)4(SMes)3(N{SiMe3}2)]-, [Fe4(μ3-
NPh)4(SDMB)3(N{SiMe3}2)]-. These derivatives were synthe-
sized by appropriate combination of thiol and 1,2-diarylhydra-
zine per the synthesis of 4, although the isolation of highly
crystalline products in these cases proved problematic. By
comparison with the spectrum of 4, 1H NMR data unambigu-
ously identify the cluster type in the derivatives and are provided
for reference as Supporting Information.
Gas Chromatographic Analysis. GC data were obtained using
an Agilent 6850 Series GC system equipped with a flame
ionization detector and a J&W Scientific HP-1 capillary column
(dimethylpolysiloxane, 30 m length ꢂ0.32 mm i.d. ꢂ 25 μm
film). The injection chamber was held at 260 ꢀC, He carrier gas
delivered at 27.71 PSI at a flow rate of 4.5 mL/min, and the oven
The acidic aqueous fractions contained primarily iron, am-
monium, and hydrazonium salts, with trace amounts of MesSH.
These fractions were combined, chilled in an ice bath, and
treated with Na2EDTA to sequester iron, resulting in a white
precipitate. The solution was made basic to litmus by dropwise
addition of 5 M NaOH with mixing, resulting in dissolution of
the white precipitate. To allow analysis of low-boiling compo-
nents and assist in extractive separation of reaction products,
the organic species were derivatized by benzoylation,42 via the
addition of 0.25 g NaHCO3, followed by 0.1 mL of benzoyl
chloride; for the di-p-tolylhydrazine system, 1 mL of CH2Cl2
was also added to dissolve precipitated p-toluidine. The reaction
vessel was then sealed, stirred vigorously for 2 h, and extracted
with 3 ꢂ 4 mL portions of CH2Cl2 to isolate the benzoylated
products. The organic fractions were combined and dried for 5
min over anhydrous Na2SO4. After filtration, a known quantity
of internal integration standard (benzophenone or anthracene,
typically ca. 5 mg) was added, and the solution contents
analyzed by GC. Retention times were compared with authentic
reference compounds as prepared below, and yields were deter-
mined by comparison of peak integrals with internal standards.
Table 6. Extraction and Derivatization Control Assays
additional
components
products
(% yield)
substrate
iPrNH2
iPrNHBz (82)
iPrNHBz (40)
iPrNHBz (70)
iPrNHBz (75)
iPrNH2
iPrNH2
iPrNH2
FeCl2
FeCl3
FeCl2 þ
Na2EDTA
FeCl2 þ
Na2EDTA
iPrHN-NHiPr
iPr(Bz)N-NHiPr (82),
iPr(Bz)N-N(Bz)
iPr (2), iPrNHBz (4)
iPr(Bz)N-NHiPr (82),
iPr(Bz)N-N(Bz)
iPr (2), iPrNHBz (5)
PhNdNPh (86)
iPrHN-NHiPr
FeCl3 þ
Na2EDTA
PhNdNPh
PhNdNPh
PhNdNPh
HN(SiMe3)2
FeCl2
FeCl3
PhNdNPh (71)
PhNdNPh (87)
FeCl2 or FeCl3 þ
BzNH2 (1-33)a
Na2EDTA
a Recoveries of benzamide via the benzoylation protocol varied but
were typically less than 5%.
Table 7. Crystallographic Data for Fe2(SMes)2(N{SiMe3}2)2(THF)2 (2-THF), Fe4(NPh)4(SMes)4 1.5C6H6 (3a 1.5C6H6), Fe4(N-p-Tol)4(SMes)4 0.5C6H6 (3b 0.5C6H6),
3
3
3
3
Fe4(N-p-Tol)4(SMes)4 (3b), and [Fe(THF)6][Fe4(N-p-Tol)4(SDMP)3(N{SiMe3}2)]2 2THF ([Fe(THF)6][4]2 2THF)a
3
3
2-THF
3a 1.5(C6H6)
3
3b 0.5C6H6
3
3b
[Fe(THF)6][4]2 2THF
3
formula
fw
space group
C
879.17
P1 (no. 2)
1
10.4231(3)
10.9167(5)
12.8489(5)
66.488(2)
69.768(2)
85.210(2)
1255.40(8)
1.163
27.35
99.9
0.786
4.14 (10.33)
1.026
38H74Fe2N2O2S2Si4
C69H73Fe4N4S4
1309.96
P1 (no. 2)
4
15.2729(7)
18.6174(7)
23.6395(10)
105.992(2)
96.730(2)
93.088(2)
6390.9(5)
1.361
23.00
99.8
1.064
5.91 (9.27)
1.022
C67H75Fe4N4S4
1287.95
P4n2 (no. 118)
8
17.1488(8)
17.1488(8)
43.475(2)
90
C64H72Fe4N4S4
1248.90
P21/c (no. 14)
4
20.7971(5)
11.4905(3)
24.9620(6)
90
91.5564(11)
90
5962.9(3)
1.391
27.52
C148H210Fe9N10O8S6Si4
3064.63
C2/c (no. 15)
4
25.5272(5)
21.9314(5)
30.2981(6)
90
113.8988(10)
90
15508.0(6)
1.313
22.98
Z
a, A
˚
˚
b, A
˚
c, A
R, deg
β, deg
γ, deg
90
90
3
˚
V, (A )
F
12785.2(10)
1.338
22.50
99.9
1.063
calc, g/cm3
θ
max, deg
total data,b %
μ, mm-1
99.7
1.137
4.64 (9.46)
1.021
100.0
0.984
6.86 (14.66)
1.033
R1 (wR2),c %
Sd
4.14 (8.72)
1.020
a
˚
Data collected at T = 170(2) (for 3) or 200(2) (all others) K using Φ and ω scans with graphite-monochromatized Mo KR radiation (λ= 0.71073 A).
P
P
P
b Percent completeness of (unique) data collection within the θmax limit. c Calculated for I > 2σ(I): R1 = ||Fo| - |Fc||/ |Fo|, wR2 = { w(Fo2 - Fc2)2/
P
P
w(Fo2)2}1/2
.
d S = goodness of fit = { [w(Fo2 - Fc2)2]/(n-p)}1/2, where n is the number of reflections and p is the number of parameters refined.

1562 Inorganic Chemistry, Vol. 50, No. 4, 2011
Zdilla et al.
Important control analyses are summarized in Table 6.
(c) Reference Compounds. Benzoylated reference com-
(b) Control Assays. The efficiency and recovery limits of the
i
i
benzoylation procedure were tested using PrNH2, PrNHN-
HiPr, and HN(SiMe3)2. A known amount of neat amine or
hydrazine (typically 10 μL) was dissolved in aqueous 6 M HCl
with stirring. The solution was chilled in an ice bath and made
basic to litmus by addition of 5 M NaOH. The solution was
treated with NaHCO3 (0.25 g), followed by benzoyl chloride (1.1
equiv per nitrogen atom), then stirred vigorously for 1-2 h. The
resulting mixture was extracted into CH2Cl2, dried over anhy-
drous Na2SO4, and filtered. After addition of a known quantity
of internal integration standard (benzophenone), the solution
was analyzed by GC to determine the recovery of amine or
hydrazine as benzoylated species. The effect of iron on this
protocol was assessed by the introduction of FeCl2 or FeCl3
(typically 20 mg) prior to the addition of amine or hydrazine.
The presence of the iron chlorides resulted in the formation of
gelatinous solids after basification with NaOH (green-colored
from FeCl2, perhaps Fe(OH)2, and rust-colored from FeCl3,
perhaps FeO(OH)); FeCl2, in particular, hindered the quantita-
tion of organonitrogen products by GC. The addition of
Na2EDTA chelating agent prior to basification was successful
in solubilizing the metal ions and improving the recoveries of
benzoylated products. Limited hydrazine reduction (ca. 5%)
was observed in the presence of iron salts.
i
i
pounds PrNHBz, PhNHBz, p-TolNHBz, Pr(Bz)N-NHBz,
i
iPr(Bz)N-NHiPr, Pr(Bz)N-N(Bz)iPr were prepared by mod-
ification of the derivatization protocol described in parts
(a) and (b). Compound identities were verified by GC/MS,
1H NMR spectroscopy, and melting point comparison with
literature values. Synthesis and characterization details are
provided as Supporting Information.
X-ray Crystallography. Single crystals suitable for X-ray
diffraction analysis were obtained from the following condi-
tions: 2 as yellow-orange blocks, from THF/n-pentane solution
at -30 ꢀC; 3a and 3b as black plates, from C6H6/n-pentane
vapor diffusion at 25 ꢀC; 4 as black plates, from THF/n-pentane
vapor diffusion at 25 ꢀC.
General crystallographic procedures are detailed elsewhere.51
ψ-Scan absorption corrections were applied using the Siemens
SHELXTL software suite,52 and multiscan absorption correc-
tions were applied using PLATON.53 Appropriate disorder
models and restraints were employed as needed. Essential
crystallographic data for the compounds in this work are
summarized in Table 7, with specific details for individual
structure determinations available as Supporting Information.
The recovery of azobenzene was determined by the following
procedure. Azobenzene (typically 20 mg) was dissolved in 2 mL
of C6H6 and agitated with 3 mL of 6 M HCl for 1-2 h. The
mixture was then extracted into CH2Cl2, partitioned, and dried
over anhydrous Na2SO4. After filtration, a known amount of
internal integration standard (anthracene) was added and the
solution analyzed by GC. The addition of either FeCl2 or FeCl3
(10 mg) to the 6 M HCl solution before the addition of the
azobenzene solution was found to have no effect on the GC analysis.
Acknowledgment. This research was supported by the
Arnold and Mabel Beckman Foundation (Beckman Young
Investigator Award), the National Science Foundation (U.S.A.,
CAREER CHE-9984645), and the Natural Sciences and En-
gineering Research Council (Canada). We thank Prof. R. Cava
and Dr. T. Klimczuk (Princeton) for magnetochemical measure-
€
ments, Prof. R. H. Holm (Harvard) for access to a Mossbauer
spectrometer, and Lay Ling Tan for experimental assistance.
Supporting Information Available: Crystallographic data;
syntheses and characterization data for reference compounds.
This material is available free of charge via the Internet at http://
pubs.acs.org.
(51) Kayal, A.; Ducruet, A. F.; Lee, S. C. Inorg. Chem. 2000, 39, 3696.
(52) Sheldrick, G. M. SHELXTL, Version 5.04; Siemens Analytical X-ray
Instruments: Madison, WI, 1996.
(53) Spek, A. L. ActaCrystallogr., Sect. A: Found. Crystallogr. 1990, 46, C34.