Model of the iron hydrogenase active site covalently linked to a ruthenium photosensitizer: Synthesis and photophysical properties

Article abstract of DOI:10.1021/ic0303385

A model of the iron hydrogenase active site with the structure [μ-ADT)Fe₂(CO)₆] (ADT = azadithiolate (S-CH ₂-NR-CH₂-S), (2: R = 4-bromophenyl, 3: R = 4-iodophenyl)) has been assembled and covalently linked to a [Ru(terpy)₂] ²⁺ photosensitizer. This trinuclear complex 1 represents one synthetic step toward the realization of our concept of light-driven proton reduction. A rigid phenylacetylene tether has been incorporated as the linking unit in 1 in order to prolong the lifetime of the otherwise short-lived [Ru(terpy)₂]²⁺ excited state. The success of this strategy is demonstrated by comparison of the photophysical properties of 1 and of two related ruthenium complexes bearing acetylenic terpyridine ligands, with those of [Ru(terpy)₂]²⁺. IR and electrochemical studies reveal that the nitrogen heteroatom of the ADT bridge has a marked influence on the electronic properties of the [Fe₂(CO)₆] core. Using the Rehm-Weller equation, the driving force for an electron transfer from the photoexcited *[Ru(terpy)₂]²⁺ to the diiron site in 1 was calculated to be uphill by 0.59 eV. During the construction of the trinuclear complex 1, n-propylamine has been identified as a decarbonylation agent on the [(μ-ADT)Fe₂(CO)₆] portion of the supermolecule. Following this procedure, the first azadithiolate-bridged dinuclear iron complex coordinated by a phosphine ligand [(μ-ADT)Fe ₂(CO)₅PPh₃] (4, R = 4-bromophenyl) was synthesized.

Full text of DOI:10.1021/ic0303385

Inorg. Chem. 2004, 43, 4683−4692
Model of the Iron Hydrogenase Active Site Covalently Linked to a
Ruthenium Photosensitizer: Synthesis and Photophysical Properties
Sascha Ott,† Magnus Borgstro1m,‡ Mikael Kritikos,§ Reiner Lomoth,‡ Jonas Bergquist,|
,†
,†
Bjo1rn A° kermark,* Leif Hammarstro1m,‡ and Licheng Sun*
Departments of Organic Chemistry and Structural Chemistry, Arrhenius Laboratory, Stockholm
UniVersity, S-10691 Stockholm, Sweden, Department of Physical Chemistry, BMC, Uppsala
UniVersity, Box 579, S-75123 Uppsala, Sweden, and Department of Analytical Chemistry, BMC,
Uppsala UniVersity, Box 599, S-75124 Uppsala, Sweden
Received December 15, 2003
A model of the iron hydrogenase active site with the structure [(µ-ADT)Fe2(CO)6] (ADT ) azadithiolate (S−CH −
2
NR−CH −S), (2: R ) 4-bromophenyl, 3: R ) 4-iodophenyl)) has been assembled and covalently linked to a
2
[Ru(terpy)2]2+ photosensitizer. This trinuclear complex 1 represents one synthetic step toward the realization of our
concept of light-driven proton reduction. A rigid phenylacetylene tether has been incorporated as the linking unit in
1 in order to prolong the lifetime of the otherwise short-lived [Ru(terpy)2]2+ excited state. The success of this
strategy is demonstrated by comparison of the photophysical properties of 1 and of two related ruthenium complexes
bearing acetylenic terpyridine ligands, with those of [Ru(terpy)2]2+. IR and electrochemical studies reveal that the
nitrogen heteroatom of the ADT bridge has a marked influence on the electronic properties of the [Fe2(CO)6] core.
Using the Rehm−Weller equation, the driving force for an electron transfer from the photoexcited *[Ru(terpy)2]2+ to
the diiron site in 1 was calculated to be uphill by 0.59 eV. During the construction of the trinuclear complex 1,
n-propylamine has been identified as a decarbonylation agent on the [(µ-ADT)Fe2(CO)6] portion of the supermolecule.
Following this procedure, the first azadithiolate-bridged dinuclear iron complex coordinated by a phosphine ligand
[(µ-ADT)Fe2(CO)5PPh3] (4, R ) 4-bromophenyl) was synthesized.
Introduction
erentially the production of hydrogen.4,5 It is this remarkable
ability of the iron hydrogenases which has inspired chemists
of the bioinorganic community to synthesize close mimics
of the active site of the natural system in the search for active
hydrogen production catalysts.6-8 From crystallographic as
well as theoretical studies of the enzyme, the active site is
known to consist of two iron nuclei, which are in bonding
distance.9,10 They are linked by a dithiolate bridge, recently
suggested to possess the structure S-CH2-NH-CH2-S
(azadithiolate ) ADT).11,12 Apart from a cysteine-linked
Hydrogenases are a class of enzymes which catalyze the
metabolism of hydrogen in cyanobacteria and other micro-
organisms. They can thereby function either as sinks for
energy-rich electrons or provide the organisms with reducing
power from hydrogen oxidation.1-3 Hydrogenases are gener-
ally divided into two families, the nickel-iron (Ni-Fe)
hydrogenases and the iron (Fe-Fe) hydrogenases, reflecting
the different base metals present in the active site. Biologi-
cally, the two families differ from each other in that Ni-Fe
hydrogenases seem to be more involved in hydrogen oxida-
tion, whereas the iron hydrogenases tend to catalyze pref-
(4) Adams, M. W. W. Biochim. Biophys. Acta 1990, 1020, 115-145.
(5) Cammack, R. Nature 1999, 397, 214-215.
(6) Alper, J. Science 2003, 299, 1686-1687.
* Authors to whom correspondence should be addressed. E-mail:
licheng.sun@organ.su.se (L.S.); bjorn.akermark@organ.su.se (B.A.).
† Department of Organic Chemistry, Stockholm University.
‡ Department of Physical Chemistry, Uppsala University.
§ Department of Structural Chemistry, Stockholm University.
| Department of Analytical Chemistry, Uppsala University.
(1) Nicolet, Y.; Lemon, B. J.; Fontecilla-Camps, J. C.; Peters, J. W. Trends
Biochem. Sci. 2000, 25, 138-143.
(7) Darensbourg, M. Y.; Lyon, E. J.; Smee, J. J. Coord. Chem. ReV. 2000,
206-207, 533-561.
(8) Evans, D. J.; Pickett, C. J. Chem. Soc. ReV. 2003, 32, 268-275.
(9) Peters, J. W.; Lanzilotta, W. N.; Lemon, B. J.; Seefeldt, L. C. Science
1998, 282, 1853-1858.
(10) Nicolet, Y.; Piras, C.; Legrand, P.; Hatchikian, E. C.; Fontecilla-Camps,
J. C. Structure 1999, 7, 13-23.
(11) Nicolet, Y.; de Lacey, A. L.; Vernede, X.; Fernandez, V. M.;
Hatchikian, E. C.; Fontecilla-Camps, J. C. J. Am. Chem. Soc. 2001,
123, 1596-1602.
(2) Peters, J. W. Curr. Opin. Struct. Biol. 1999, 9, 670-676.
(3) Frey, M. ChemBioChem 2002, 3, 152-160.
10.1021/ic0303385 CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/02/2004
Inorganic Chemistry, Vol. 43, No. 15, 2004 4683

Ott et al.
[Fe4S4] cluster which is part of the electron-transfer chain
to and from the active site, carbonyl and cyanide ligands
occupy the remaining coordination sites around the iron
nuclei.
iron and ruthenium termini due to its linearity and rigidity.
In contrast to the large amount of research which has been
conducted on PDT-bridged diiron complexes, little work has
been reported on the azadithiolate (ADT) analogues [(µ-
ADT)Fe2(CO)6]25-27 and the elucidation of their photophysi-
cal and electrochemical properties. This is particularly
surprising since recent spectroscopic and theoretical studies
have shown that this arrangement may play a crucial role in
the natural system, providing a kinetically and thermody-
namically favored pathway for hydrogen production.11,12 We
therefore decided to include this structural feature into our
supramolecular assembly and 1 was selected as the target
complex.
The focus of the synthetic work has mainly been directed
toward ligand exchange reactions on the “parent” dithiolate-
bridged [(µ-PDT)Fe2(CO)6] (PDT ) propyldithiolate), a
strategy benefiting in particular from the synthetic ease with
which this complex is prepared.13 The nature of introduced
ligands ranges from isonitriles14 and cyanides15,16 to
thioethers17-19 and phosphines.20 In this context, Darensbourg
et al. have reported that [(µ-H)(µ-PDT)Fe2(CO)4(PMe3)2]+
is a catalyst for H2/D2 scrambling,20 whereas Rauchfuss et
al. demonstrated that [(µ-PDT)Fe2(CO)4PMe3(CN)]- serves
as a catalyst for electrochemical hydrogen evolution.21,22 We
have recently described the concept of light-driven proton
reduction, using a ruthenium-polypyridine complex as the
light-harvesting component and a model of the iron hydro-
genase as proton activation catalyst.23
Several potential problems have to be considered when
pursuing such a concept. First, it is important that the lifetime
of the sensitizer is sufficiently long to allow electron transfer
to the diiron moiety. Second, it is conceivable that an attached
iron unit could partly quench the excited state of the sensitizer
by energy transfer. This would decrease the lifetime of the
sensitizer and could seriously compete with electron transfer
events. Third, of course, the redox potential of the excited
state of the sensitizer should be sufficiently negative to permit
electron transfer to the iron complex. In an exploratory study
of ruthenium-sensitized hydrogen production, we chose
[Ru(terpy)2]2+as the photosensitizer due to its geometric
advantages over other Ru complexes with bidentate ligands.24
We anticipated that the short lifetime of the excited state of
this complex could be compensated at least partially by the
introduction of an acetylene substituent, which concomitantly
provides a synthetic handle for further elaboration to the
trinuclear complex. Furthermore, this acetylenic bridge offers
excellent control over the distance between the redox-active
The construction of 1 coincided with an unexpected ligand
substitution on the dinuclear iron site. n-Propylamine was
identified as the agent responsible for the decarbonylation
process. This new protocol could be utilized in the synthesis
of the first phosphine-substituted ADT-bridged diiron com-
plex 4. The photophysical and electrochemical properties of
the diiron complexes 2-6 were evaluated, and those of 1
were assessed in comparison with two novel reference
complexes 8a,b.
(12) Fan, H.-J.; Hall, M. B. J. Am. Chem. Soc. 2001, 123, 3828-3829.
(13) Winter, A.; Zsolnai, L.; Huttner, G. Z. Naturforsch. 1982, 37b, 1430-
1436.
(14) Lawrence, J. D.; Rauchfuss, T. B.; Wilson, S. R. Inorg. Chem. 2002,
41, 6193-6195.
(15) Lyon, E. J.; Georgakaki, I. P.; Reibenspies, J. H.; Darensbourg, M.
Y. Angew. Chem., Int. Ed. 1999, 38, 3178-3180.
(16) Lyon, E. J.; Georgakaki, I. P.; Reibenspies, J. H.; Darensbourg, M.
Y. J. Am. Chem. Soc. 2001, 123, 3268-3278.
Results and Discussion
Synthesis. The trinuclear complex 1 was expected to
become accessible by a Sonogashira cross coupling reaction28
between an advanced diironhexacarbonyl precursor, featuring
a suitable halogenated arene and 4′-ethynyl-2,2′:6′,2′′-terpyr-
idine. For this purpose, the bromo-equipped iron complex 2
and its iodo equivalent 3 were synthesized by applying a
recently published procedure for the preparation of some
related complexes.25 Thus, the lithium salt of diironhexa-
(17) Razavet, M.; Davies, S. C.; Hughes, D. L.; Pickett, C. J. Chem.
Commun. 2001, 847-848.
(18) Razavet, M.; Borg, S. J.; George, S. J.; Best, S. P.; Fairhurst, S. A.;
Pickett, C. J. Chem. Commun. 2002, 700-701.
(19) George, S. J.; Cui, Z.; Razavet, M.; Pickett, C. J. Chem. Eur. J. 2002,
8, 4037-4046.
(20) Zhao, X.; Georgakaki, I. P.; Miller, M. L.; Mejia-Rodriguez, R.;
Chiang, C.-Y.; Darensbourg, M. Y. Inorg. Chem. 2002, 41, 3917-
3928.
(21) Gloaguen, F.; Lawrence, J. D.; Rauchfuss, T. B. J. Am. Chem. Soc.
2001, 123, 9476-9477.
(22) Gloaguen, F.; Lawrence, J. D.; Rauchfuss, T. B.; Benard, M.; Rohmer,
M.-M. Inorg. Chem. 2002, 41, 6573-6582.
(25) Lawrence, J. D.; Li, H.; Rauchfuss, T. B. Chem. Commun. 2001,
1482-1483.
(26) Lawrence, J. D.; Li, H.; Rauchfuss, T. B.; Benard, M.; Rohmer, M.-
M. Angew. Chem., Int. Ed. 2001, 40, 1768-1771.
(27) Li, H.; Rauchfuss, T. B. J. Am. Chem. Soc. 2002, 124, 726-727.
(28) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975,
4467-4470.
(23) Ott, S.; Kritikos, M.; A° kermark, B.; Sun, L. Angew. Chem., Int. Ed.
2003, 3285-3288.
(24) Hammarstro¨m, L.; Barigelletti, F.; Flamigni, L.; Indelli, M. T.;
Armaroli, N.; Calogero, G.; Guardigli, M.; Sour, A.; Collin, J.-P.;
Sauvage, J.-P. J. Phys. Chem. A 1997, 101, 9061-9069.
4684 Inorganic Chemistry, Vol. 43, No. 15, 2004

Model of the Iron Hydrogenase ActiWe Site
Scheme 1
Scheme 2: Carbonyl-Triphenylphosphine Ligand Exchange Promoted
by (n-Pr)NH2
a (1) p-CH2O, CH2Cl2, 3 h, (2) SOCl2. b [(Li-S)2Fe2(CO)6], THF, - 78
°C, 10 min, 76% for 2, 76% for 3.
carbonyl disulfide29 was reacted with the respective N,N-
bis(chloromethyl)-4-haloaniline, which in turn was synthe-
sized from the p-haloaniline in two steps (Scheme 1).
Initial acetylene coupling experiments were conducted with
the bromo-substituted compound 2. However, stirring this
complex in the presence of trimethylsilyl acetylene, catalytic
amounts of copper iodide and [(PPh3)2PdCl2] (ca. 5 mol %)
in a mixture of toluene and n-propylamine at elevated
temperature did not afford the phenylacetylene 5. Instead, a
side-product was isolated in a yield of ca. 10%, in addition
to the recovered starting material. Whereas the low reactivity
of electron-rich bromoarene 2 in Sonogashira cross coupling
reactions may not be surprising,30 IR-studies of the product
revealed that the coordination sphere around the dimeric iron
site had been altered. A pattern typical for a monosubstituted
complex of type [Fe2(CO)5X] was observed in the carbonyl
region of the IR spectrum.31 Furthermore, the 1H NMR signal
for the methylene protons in the bridge, which appears as a
singlet in the starting material, had split into two doublets
(J ) 12.2 Hz), consistent with a decreased symmetry of the
system. The nature of the introduced ligand could unambigu-
ously be identified by single-crystal X-ray diffraction analysis
(Figure 4). It revealed that a triphenylphosphine ligand from
the palladium catalyst had entered the coordination sphere
of the dimeric iron site, thereby replacing one CO ligand.
In an attempt to elucidate the role of the amine in this
ligand substitution, 2 and an excess of triphenylphosphine
were stirred in n-propylamine and in triethylamine in two
separate experiments. Whereas the reaction in the primary
amine resulted in quantitative ligand exchange after a few
hours, no such substitution could be detected in the tertiary
amine after 24 h (Scheme 2). From this experiment it appears
that the nucleophilic n-propylamine is responsible for a
decrease of the iron carbonyl bond stability and thereby
facilitates the carbonyl-to-phosphine ligand exchange. Whether
the initial attack proceeds on the iron center or on the
carbonyl ligand has not been determined in this study. Ligand
exchange reactions, facilitated by nucleophiles such as
halides or amines, are not unprecedented and have been
reported for Fe0 carbonyl species and other transition-metal
carbonyl clusters.32,33 Noteworthy is the fact that the mono-
substituted complex 4 is obtained exclusively, even at
prolonged reaction times and in the presence of a large excess
of triphenylphosphine. Two factors could account for this
phenomenon. The steric demand of the bulky triphenylphos-
phine ligand cannot be satisfied on both iron centers due to
the presence of the aryl group which resides directly over
one CO ligand. Furthermore, the increased electron density
at the substituted iron may be communicated to the adjacent
iron center, resulting in a decreased electrophilicity of the
monosubstituted complex.
To avoid ligand substitution, a tertiary amine was used in
the palladium-mediated cross coupling reaction for the
construction of the arylacetylene bond present in 1. Most
gratifyingly, we found that iodoarene 3, which is more
reactive than the bromo-correspondent 2, couples smoothly
to trimethylsilyl acetylene under these modified conditions
(Scheme 3). The acetylenic diiron complex 5 was isolated
in excellent yield and was characterized by standard tech-
niques, including X-ray analysis (Figure 3).
Subjecting 4′-ethynyl-2,2′:6′,2′′-terpyridine34,35 instead of
trimethylsilyl acetylene to otherwise identical coupling
conditions resulted in the formation of the terpyridine-
equipped diiron complex 6. Solutions of 6 in organic solvents
are unstable at room temperature and decompose within
minutes upon exposure to light, presumably due to light-
induced decarbonylation followed by intermolecular terpyr-
idine complexation. However, as a freshly prepared solution
in methanol which is kept in the dark, ligand 6 coordinates
readily to the (terpy)Ru fragment,36 giving rise to the desired
trinuclear complex 1 (Scheme 3).23
Over the past few years, the utilization of acetylene
moieties has become a popular strategy to prolong the
lifetime of the otherwise short-lived [Ru(terpy)2]2+ (τ < 1
ns) excited state. For example, ethynylterpyridine substitution
on one terpyridine ligand results in an increase of the excited-
state lifetime to 55 ns,37 whereas the introduction of an
ethynylpyrene prolongs the lifetime to 560 ns, although as a
mixture of metal- and pyrene-centered triplet excited states
(34) Potts, K. T.; Konwar, D. J. Org. Chem. 1991, 56, 4815-4816.
(35) Grosshenny, V.; Romero, F. M.; Ziessel, R. J. Org. Chem. 1997, 62,
1491-1500.
(29) Brandt, P. F.; Lesch, D. A.; Stafford, P. R.; Rauchfuss, T. B. Inorg.
Synth. 1997, 31, 112-116.
(30) Singh, R.; Just, G. J. Org. Chem. 1989, 54, 4453-4457.
(31) Ellgen, P. C.; Gerlach, J. N. Inorg. Chem. 1973, 12, 2526-2532.
(32) Angelici, R. J. Acc. Chem. Res. 1972, 5, 335-342.
(33) Lavigne, G. The Chemistry of Metal Cluster Complexes; Shriver, D.
F., Kaesz, H. D., Adams, R. D., Eds.; VCH: New York, 1990.
(36) Norrby, T.; Bo¨rje, A.; A° kermark, B.; Hammarstro¨m, L.; Alsins, J.;
Lashgari, K.; Norrestam, R.; Ma˚rtensson, J.; Stenhagen, G. Inorg.
Chem. 1997, 36, 5850-5858.
(37) Benniston, A. C.; Grosshenny, V.; Harriman, A.; Ziessel, R. Angew.
Chem., Int. Ed. Engl. 1994, 33, 1884-1885.
Inorganic Chemistry, Vol. 43, No. 15, 2004 4685

Ott et al.
Scheme 3
c [(PPh3)2PdCl2], CuI, Et3N, toluene, 40 °C, 3 h, 91% for 5, 57% for 6. d [(Terpy)RuCl2(DMSO)], MeOH, reflux, 4 h, 29%.
Scheme 4
in a quasi equilibrium.38 Despite this extensive utilization
of acetylene substituents, we were surprised to find that no
work has been reported on the most basic phenylethynyl-
substituted [Ru(terpy)2]2+. To have a reference system
comparable to 1, the 4-(N-piperidyl)phenylethynyl-terminated
[Ru(terpy)2]2+ complex 8a as well as its phenylethynyl-
analogue 8b were synthesized in procedures analogous to
that for the preparation of 1. Hence, iodo-4-(N-piperidyl)-
benzene39 was reacted with 4′-ethynyl-2,2′:6′,2′′-terpyridine35
to produce the acetylenic ligand 7a and phenyl acetylene
was reacted with 4′-[((trifluoromethyl)sulfonyl)oxy]-2,2′:
6′,2′′-terpyridine34,40 to form 7b. Subsequently, both ligands
were separately coordinated to the (terpy)Ru fragment in
refluxing ethanol, thereby forming 8a and 8b (Scheme 4).
e [(Terpy)RuCl2(DMSO)], EtOH, reflux, 16 h, 53% for 8a, 64% for 8b.
IR Spectroscopy. The IR spectra of complexes 1-3, 5,
and 6 (in solution of CH2Cl2) are literally identical in the
CO region and feature three signals at 2076, 2038, and 2001
cm-1. This result clearly indicates that there is only a
negligible electronic communication between the redox active
ruthenium and iron termini in the ground state. It might
therefore not be surprising that these values also correlate
with those reported for other azadithiolate-bridged diiron
complexes.26,27 It seems, however, that the nitrogen hetero-
atom in the ADT bridge slightly influences the electronic
properties of the carbonyl ligands. Compared to the IR
spectrum of [(µ-PDT)Fe2(CO)6] (νCO ) 2073, 2034, 1998
in CH2Cl2), the CO frequencies are shifted by 3-4 cm-1
toward higher wavenumbers. Reflecting the changed coor-
dination sphere around the iron cations, the IR spectrum of
the triphenylphosphine-substituted complex 4 exhibits mark-
edly different resonances in the carbonyl region at 2048,
(38) Harriman, A.; Hissler, M.; Khatyr, A.; Ziessel, R. Chem. Commun.
1999.
(39) Effenberger, F.; Agster, W.; Fischer, P.; Jogun, K. H.; Stezowski, J.
J.; Daltrozzo, E.; Kollmannsberger-von Nell, G. J. Org. Chem. 1983,
48, 4649-4658.
(40) Constable, E. C.; Ward, M. D. J. Chem. Soc., Dalton Trans. 1990,
1405-1409.
4686 Inorganic Chemistry, Vol. 43, No. 15, 2004

Model of the Iron Hydrogenase ActiWe Site
Table 1: Selected Bond Lengths (Å) for 2, 4, and 5
2
4
5
Fe1-Fe2
Fe1-S1
Fe1-S2
Fe2-S1
Fe2-S2
Fe1-C1
Fe1-C2
Fe1-C3
Fe2-C4
Fe2-C5
Fe2-C6
S1-C7
2.494(1)
2.265(2)
2.270(2)
2.263(2)
2.283(2)
1.815(7)
1.816(7)
1.789(7)
1.802(7)
1.803(7)
1.804(7)
1.865(6)
1.858(6)
1.892(6)
2.5094(9)
2.2860(12)
2.2705(13)
2.2533(12)
2.2669(14)
2.2803(13)
1.761(5)
1.765(5)
1.779(5)
1.806(5)
1.775(5)
2.5035(8)
2.2569(10)
2.2603(11)
2.2660(10)
2.2516(11)
1.800(5)
1.796(5)
1.796(5)
1.797(5)
1.793(4)
1.796(5)
1.850(4)
1.864(4)
na
1.428(5)
1.424(5)
1.407(4)
1.432(6)
1.198(7)
1.824(5)
1.864(5)
1.858(5)
1.892(6)
S2-C8
Br1-C11/12
N1-C7
N1-C8
N1-C9
C12-C15
C15-C16
Si1-C16
na
na
na
na
na
na
Figure 1. ORTEP (ellipsoids at 30% probability level) view of C14H8-
BrNO6S2Fe2 (2).
Figure 3. ORTEP (ellipsoids at 30% probability level) view of C31H24-
BrNO5PS2Fe2 (4).
Fe-C and C-O distances and Fe-C-O angles are all
normal with one exception. The Fe-C-O angles for the
carbonyls closest to the amine functionality (i.e. trans to the
Fe-Fe bond) in 2 and 5 deviate significantly from the
expected 180° linear arrangements with values of 175.6(7)°
and 174.2(6)°, respectively. The nonbonding C‚‚‚N distance
between the amine and the closest carbonyl carbon atom in
2 (3.578(8) Å) is longer than the equivalent distance in 5
(3.506(5) Å). Moreover, the nitrogen atom deviates from the
plane defined by C7, C8, and C9 by 0.174(5) Å in 2 and by
0.147(3) Å in 5. This nonplanarity ruptures the potential
π-conjugation between the phenyl ring and the nitrogen
p-orbital. In 5, the acetylenic C-C triple bond and the Si-C
single bonds in the trimethylsilyl group adopt expected
values. Selected bond distances are summarized in Table 1.
Solid-State Structure of 4. The molecular structure of 4
is illustrated in Figure 4. In 4, the CO group located trans to
the Fe-Fe bond has been replaced by the PPh3 ligand. This
positioning of the phosphine group is consistent with its
bulkiness. The Fe-P bond distance (2.253(1) Å) is identical
to that reported for the related toluenedithiolate [(µ2-SC7H6S)-
Fe2(CO)5(PPh3)].41 In 4, the Fe-Fe single bond is somewhat
elongated (2.5094 (9) Å) compared to those in 2 and 5. As
found for 2 and 5, the CO group closest to the amine
functionality is significantly bent, 173.6(6)°. The carbonyl
to amine (i.e. C4‚‚‚N1) nonbonding distance is 3.518(7) Å.
Figure 2. ORTEP (ellipsoids at 30% probability level) view of C19H17-
SiNO6S2Fe2 (5).
1990, and 1935 cm-1. Analogous to the situation in the all-
carbonyl complexes, these values are shifted slightly toward
higher energy compared to those found for the PDT-bridged
[(µ-PDT)Fe2(CO)5(PPh3)] analogue.41 These slight differ-
ences indicate that the nitrogen in the bridge has a small
impact on the carbonyl ligands.
Solid-State Structures. Solid-State Structure of 2 and
5. The [(µ-S)2Fe2(CO)6] units of 2 and 5 are very similar
(Figures 1 and 2) and are to a large extent analogous to those
found in other [(µ-S)2Fe2(CO)6] structures reported.26 In the
central [(µ-SR)2Fe2(CO)6] unit the two Fe atoms and the two
S atoms form a butterfly conformation in which the metal
atoms are connected to each other through an Fe-Fe single
bond (2.494(1) and 2.5035(8) Å for 2 and 5, respectively).
In contrast to the Fe-Fe distances the S‚‚‚S contacts vary
slightly more, 3.075(2) Å in 2 and 3.058(2) Å in 5. In both
complexes the two Fe(CO)3 units are nearly eclipsed. The
(41) Hasan, M. M.; Hursthouse, M. B.; Kabir, S. E.; Malik, K. M. A.
Polyhedron 2001, 20, 97-101.
Inorganic Chemistry, Vol. 43, No. 15, 2004 4687

Ott et al.
5 suggesting that the oxidation of the diiron complexes may
be located on the tertiary aniline portion. Efforts to determine
the exact location of the electron hole in oxidized 2, 3, and
5 by spectroscopic techniques (ESR, UV/vis) were unsuc-
cessful due to the instability of the oxidized species. On the
cathodic scan, the all-carbonyl iron complexes 2, 3, and 5
typically feature an irreversible reduction around -1.55 (
0.02 V, whereas the electron-enriched complex 4 is reduced
at more negative potential at -1.90 V (recorded in CH2Cl2).
In comparison with [(µ-PDT)Fe2(CO)6] which is reduced at
-1.67 V vs Fc+/0 43
, the azadithiolate bridge seems to have
a considerable influence on the reduction potentials, render-
ing the ADT-bridged complexes 2, 3, and 5 more susceptible
for reduction by more than 100 mV. This feature makes [(µ-
ADT)Fe2(CO)6] complexes superior acceptors in electron-
transfer processes over the all-carbon bridged analogues.
Figure 4. Electronic absorption spectra of 2 (s s), 4 (‚‚‚‚‚‚‚), 6 (s), and
7a (s‚‚‚‚s) recorded in CH2Cl2.
The Ru(II)/Ru(III) couple in the [Ru(terpy)2]2+ complexes
1, 8a, and 8b is observed at 0.93 ( 0.02 V. Reduction of 1
at -1.49 V proceeds at a potential slightly less negative than
those usually observed for the reduction of the terpyridine
ligands and can thus be assigned to the reduction of the diiron
portion. The remaining reduction waves for 1 at -1.67 V
and -1.89 V and around -1.6 and -1.89 V for 8a and 8b
are typical for [Ru(terpy)2]2+ complexes and arise from the
reduction of the terpyridine ligands. The driving force for
an electron transfer from the photoexcited *[Ru(terpy)2]2+
to the diiron site in 1 can be calculated from these
electrochemical potentials in conjunction with the excited-
state energy (see emission section).
Table 2: Electrochemical Data for Complexes 1-5 and 8a,ba
compound
E
red/ox (RuII/III
)
Epa
E
pc (Fe2-site)
Epc (terpy)
2
3
na
na
na
0.54b
0.56b
0.34i
0.64i
0.65i
0.50i
na
-1.53b
-1.54b
-1.90i
-1.56i
-1.49i
na
na
na
na
4c
5
na
na
1
8a
8b
0.91
0.95
0.91
-1.67;d -1.89d
-1.60; -1.89
-1.57; -1.89
na
a All potentials were recorded for deaerated, 1 mM solutions in
acetonitrile containing Bu4NPF6 (0.1 M) as supporting electrolyte and are
given vs Fc+/0; scan rate 50 mV/s. Given are anodic (Epa) and cathodic
peak potentials Epc. b Irreversible. c Measured in CH2Cl2. d Detected by
differential pulse voltammetry.
Further, the N1 atom is displaced 0.196(4) Å from the plane
defined by C6, C7, and C8. These metrical data show that
the geometry of the azadithiolate functionality is rather
insensitive toward ligand substitution on the Fe atoms.
Electronic Absorption Spectra. The UV/Vis spectra of
the dinuclear iron complexes 2-4 are dominated by a strong
absorption in the UV region around 270 nm. A second
intense absorption is visible in the near UV and shifts
substantially depending on the coordination sphere around
the iron nuclei. For example, it can be detected around 330
nm for 2 and 3 (graph not shown) up to 363 nm in 4 (Figure
5). This might be an indication for the involvement of metal
as well as ligand orbitals in this transition. A broad and
featureless absorption reaching up to 600 nm with an
extinction coefficient of maximal 1800 M-1 cm-1 is respon-
sible for the dark red color of these iron coordination
compounds. The piperidine-containing terpyridine 7a features
a strong absorption at 362 nm in addition to a band in the
far UV. The UV/Vis spectrum of ligand 6 exhibits three
intense absorptions at 278, 341, and 357 nm, thereby
suggesting that it can be regarded as a linear combination
of the spectra of 2 and that of the reference ligand 7a (Figure
4).
Electrochemistry. To assess a potential driving force for
an electron transfer from the photoexcited [Ru(terpy)2]2+
portion to the diiron site in 1, cyclic voltammograms were
recorded for all diiron and ruthenium complexes reported
herein (Table 2 and Supporting Information). This is of
particular interest since no electrochemical data of azadithi-
olate-bridged diiron complexes have been reported in the
literature to date. An irreversible oxidation of the iron
complexes 1-3 and 5 is observed at a potential around 0.60
( 0.05 V vs Fc+/0. Reflecting the donor capacity of the
triphenylphosphine ligand, complex 4 exhibits an irreversible
oxidation at a potential which is lower by 0.25 V compared
to those of the all-carbonyl analogues. A similar trend has
been observed by Poilblanc et al. during the introduction of
tertiary phosphines on [Fe2(µ-SR)2(CO)6]-type of com-
plexes.42 An interesting point, however, arises from an
inspection of the cyclic voltammogram of the piperidine-
containing [Ru(terpy)2]2+ complex 8a, which exhibits an
irreversible oxidation at 0.50 V. This oxidation, which
undoubtedly is centered on the tertiary aniline portion, is
suspiciously similar to that observed for complexes 1-3 and
The trinuclear ruthenium-diiron complex 1 and the
reference complexes 8a and 8b display electronic absorption
spectra typical for [Ru(terpy)2]2+ type of complexes (Figure
(43) The discrepancy of this value to the one reported in Razavet, M.;
Davies, S. C.; Hughes, D. L.; Barclay, J. E.; Evans, D. J.; Fairhurst,
S. A.; Liu, X.; Pickett, C. J. J. Chem. Soc., Dalton Trans. 2003, 586-
595 is a result of different reference systems. All potentials reported
(42) Mathieu, R.; Poilblanc, R.; Lemoine, P.; Gross, M. J. Organomet.
Chem. 1979, 165, 243-252.
herein are vs Fc+/0
.
4688 Inorganic Chemistry, Vol. 43, No. 15, 2004

Model of the Iron Hydrogenase ActiWe Site
Table 3: Luminescence Properties of Complexes 1, 8a, and 8b and
[Ru(terpy)2]2+ 51
295 K
77 K
complex
λ
ems/nm
φems
τ
ems/ns
6.5
λ
ems/nm
1
8a
8b
674
675
666
2 × 10-5
1.5 × 10-4
1.5 × 10-4
<10-5
657
658
650
598
23
13.5
[Ru(terpy)2]2+
0.250
piperidyl)phenylethynyl- 8a and the phenylethynyl-
substituted [Ru(terpy)2]2+ 8b, with those of unsubstituted
[Ru(terpy)2]2+.
The excited-state decay, as determined from the time-
resolved emission decay traces, gave lifetimes of 23 ns for
8a, 13.5 ns for 8b, and 6.5 ns for 1 (see Supporting
Information). A smaller lifetime component (c.a. 15% of the
total amplitude) with τ ≈ 2 ns was present in the samples of
all three complexes, and is attributed to a decomposition
product.50 Flash photolysis transient absorption experiments
with ca. 10 ns resolution show excited-state decays in
agreement with the emission data.
First, we note that the excited-state lifetimes in 1 and 8a,b
are much longer than that in the parent, unsubstituted
[Ru(terpy)2]2+ (τ ) 250 ps).51 This substantial improvement
can be explained by a delocalization of the MLCT excited
state over the conjugated phenylacethylene substituent. This
may decrease the rate of population of metal-centered states
that lead to rapid decay to the ground state, in analogy with
the results of previous studies.24,37,45,46
Second, the excited-state decay is significantly faster in 1
than in 8a and 8b. The quenching rate constant kq can be
estimated as kq ) 1/τ1 - 1/τ8a ) 1.1 × 108 s-1. Since the
electron transfer to the diiron moiety is strongly endothermic,
the quenching is not induced by electron transfer but rather
by dipole-dipole energy transfer. A calculation of the rate
constant using Fo¨rster’s equation52 with the orientation factor
κ set to 1 and a donor-acceptor distance of 8.5 Å gave a
value of a similar magnitude as the experimental one. The
exact location of the excited state in 1 is somewhat
ambiguous but a donor-acceptor distance of 8.5 Å is not
unreasonable, making Fo¨rster energy transfer a plausible
explanation for the observed quenching in 1. Energy transfer
quenching with a rate constant of 1.1 × 108 s-1 is modest
and will probably not seriously compete with electron transfer
in the next generation of Ru-Fe2 complexes.
Figure 5. Electronic absorption spectra of 1 (s), 8a (‚‚‚‚‚‚‚), and 8b (s
s), recorded in CH2Cl2.
5).44 The strongest absorption in the UV around 310 nm is
usually attributed to terpyridine-centered πfπ* transitions.
The longest wavelength absorption around 500 nm is caused
by metal-to-ligand charge-transfer transitions of the ruthe-
nium moiety. The band arising from this transition seems to
overlay the weak absorption from the diiron portion in 1.
Emission Data. The excited-state energy of complexes 1
and 8a,b were determined from the emission spectra recorded
in a solvent glass at 77 K (not shown). The peak with the
highest energy in the vibrational progression has been shown
to give a good estimate for the exited state energy.24,45,46 For
complexes 1, 8a, and 8b the excited-state energies were
calculated to be 1.89, 1.89, and 1.91 eV, respectively. As
an effect of the better electron accepting capacity of the
ligands in these complexes, these values are slightly lower
than that for [Ru(terpy)2]2+ (2.07 eV).47 The driving force
for electron transfer from the ruthenium to the diiron site
could be calculated with the Rehm-Weller equation.48 Using
the reduction potentials from the cyclic voltammetry data
and omitting the work term arising from Coulombic interac-
tions the driving force was calculated to be uphill with ∆G°
) 0.59 eV.49 The luminescence properties are summarized
in Table 3.
The acetylene linkage in 1 was incorporated to satisfy two
different requirements: to gain precise control over the
spatial separation between the redox active iron and ruthe-
nium termini and, more importantly, to increase the lifetime
of the [Ru(terpy)2]2+ photosensitizer. The success of this
strategy was determined by comparison of the excited-state
properties of 1 and two reference complexes, the 4-(N-
Conclusion
We have synthesized a model of the iron hydrogenase
active site [(µ-ADT)Fe2(CO)6] and have covalently linked
(44) Sauvage, J.-P.; Collin, J.-P.; Chambron, J.-C.; Guillerez, S.; Coudret,
C.; Balzani, V.; Barigelletti, F.; De Cola, L.; Flamigni, L. Chem. ReV.
1994, 94, 993-1019.
(50) The amplitude of the 2 ns component had increased to 50-75% of
the total amplitude when the experiments were repeated several months
later, with solutions freshly prepared from the powder sample that
had been stored in the freezer. This strongly suggests that the 2 ns
component is a decomposition product formed in the powder. Possibly,
the triple bond is oxidized by O2, which gives a complex with a lifetime
closer to that for the parent [Ru(terpy)2]2+. In addition, a minor
component (1-5% of the amplitude) with a lifetime of 150 ns was
present in all the samples, probably from an impurity.
(45) Caspar, J. V.; Meyer, T. J. Inorg. Chem. 1983, 22, 2444-2453.
(46) Treadway, J.; Loeb, B.; Lopez, R.; Anderson, P. A.; Keene, F. R.;
Meyer, T. J. Inorg. Chem. 1996, 35, 2242-2246.
(47) Maestri, M.; Armaroli, N.; Balzani, V.; Constable, E. C.; Cargill
Thompson, A. M. W. Inorg. Chem. 1995, 34, 2759-2767.
(48) Rehm, D.; Weller, A. Isr. J. Chem. 1970, 8, 259-271.
(49) The driving force for a second electron transfer from a regenerated
and photoexcited [Ru(terpy)2]2+ to an already reduced diiron complex
will be even more endothermic. However, protonation of the monore-
duced diiron site might compensate for this effect.
(51) Winkler, J. R.; Netzel, T. L.; Creutz, C.; Sutin, N. J. Am. Chem. Soc.
1987, 109, 2381-2392.
(52) Fo¨rster, T. Discuss. Faraday Soc. 1959, 7-17.
Inorganic Chemistry, Vol. 43, No. 15, 2004 4689

Ott et al.
Table 4. Selected Crystal Data for C14H8BrNO6Fe2 (2), C31H24BrNO5PS2Fe2 (4), and C19H17SiNO6S2Fe2 (5)
2
4
5
empirical formula
fw
C14H8BrNO6S2Fe2
541.94
C31H24BrNO5PS2Fe2
777.23
C19H17SiNO6S2Fe2
559.25
cryst syst
space group
a, Å
b, Å
c, Å
triclinic
triclinic
triclinic
P-1 (No. 2)
7.816(2)
11.046(3)
11.353(3)
74.81(3)
86.93(3)
80.04(3)
931.6(4)
2
1.932(1)
293
3.955
2152/236
0.915
0.0384, 0.0813
0.69, -0.68
P-1 (No. 2)
9.4590(17)
10.1888(19)
17.330(3)
73.47(2)
78.93(2)
88.07(2)
1571.0(5)
2
1.643(1)
293
2.420
4467/388
1.15
0.0522, 0.1365
1.05, -1.73
P-1 (No. 2)
7.6647(12)
9.6722(15)
16.918(3)
76.96(2)
81.12(2)
89.96(2)
1206.5(4)
2
1.539(1)
293
1.458
3425/283
1.25
0.0337, 0.0822
0.86, -1.27
R, °
â, °
γ, °
V, Å3
Z
F
calc, g cm-3
temperature, K
µ (MoKR), (mm-1
)
N(obs) (I > 2σ(I))/N(par)
goodness of fit
R1,a wR2, (I > 2σ(I))
∆Fmax, ∆Fmin (e/Å3)
a R values are defined as R1 ) ∑||Fo|-|Fc||/∑|Fo|, wR2 ) [∑[w(Fo -Fc )2]/∑[w(Fo )2]]1/2
.
2
2
2
it to a [Ru(terpy)2]2+ photosensitizer in a first attempt to
afford proton reduction by light. During the synthesis of 1,
n-propylamine has been identified as a decarbonylation agent
at the hexacarbonyldiiron core, and the new protocol has
been applied for the synthesis of a diiron complex 4
coordinated by a phosphine ligand. The IR spectra of all
novel [(µ-ADT)Fe2(CO)6] complexes were literally identical
in the CO region, pointing toward an electronic insulation
of the iron and ruthenium termini in the ground state. In
comparison with the IR spectrum of [(µ-PDT)Fe2(CO)6], the
redox-active nitrogen of the ADT bridge in 1-3, 5, and 6
seems to have a small impact on the CO frequencies.
Furthermore, the nitrogen heteroatom in the [(µ-ADT)Fe2-
(CO)6] complexes 1-3 and 5 has a marked effect on their
respective reduction potentials. The former complexes are
reduced at a potential more than 100 mV less negative than
the all-carbon bridged [(µ-PDT)Fe2(CO)6]. This finding
renders the ADT-bridged complexes superior to their PDT
analogues as acceptors in electron-transfer processes.
Compared to the unsubstituted [Ru(terpy)2]2+ which
exhibits an excited-state lifetime of 250 ps,51 introduction
of an acetylene tether resulted in increased lifetimes in 1,
8a, and 8b. In 1 the observed excited-state lifetime was 6.5
ns, substantially shorter than in 8a (23 ns) and 8b (13.5 ns).
Presumably, dipole-dipole energy transfer with a rate of 1.1
× 108 s-1 is responsible for the decrease of the excited-state
lifetime in 1. From the excited-state energy and the electro-
chemical data it becomes apparent that the driving force for
an electron transfer from the photogenerated *[Ru(terpy)2]2+
excited state to the diiron portion is uphill by 0.59 eV. We
anticipate, however, that protonation of the nitrogen in the
ADT bridge in 1 will shift the reduction potential toward
more positive values, thereby rendering this process ther-
modynamically more favorable. Alternatively, introduction
of different ligands on the diiron core, followed by proto-
nation of the iron-iron bond, could have a similar effect.
Finally, a photogenerated Ru(I) species, which is higher in
energy than the Ru(II) excited state, could provide sufficient
reducing power to drive an electron transfer to the biomimetic
model of the iron hydrogenase. Efforts in these directions
are currently in progress.
Experimental Section
A STOE-IPDS image plate diffractometer with a rotating-anode
using MoKR radiation (λ ) 0.71073 Å) was used for the single-
crystal data collection. Selected crystallographic data for compounds
2, 4, and 5 are given in Table 4. The intensities of the reflections
were integrated with the STOE software, and the numerical
absorption corrections were performed with the programs X-RED
and X-SHAPE.53 The structures were solved by direct methods and
refined by full matrix least-squares on F2 using the SHELX97 and
SHELXS97 software, respectively.54,55
Cyclic voltammograms were recorded using a Autolab Poten-
tiostat (Ecochimie Netherlands), controlled by GPES software
(version 4.8). A glassy carbon disk electrode (diameter 3 mm) was
used as the working electrode and was polished prior to each
experiment using an aqueous alumina powder slurry. A platinum
wire served as a counter and a nonaqueous Ag/Ag+ was used as
reference electrode. All potentials are given vs the Fc+/0 couple.
The oxidation potentials of the diiron complexes are anodic peak
potentials Epa, and the reduction potentials are given as cathodic
peak potentials Epc. Cyclic voltammograms were obtained for 1
mM solutions of the analyte in dry acetonitrile, containing 0.1 M
tetrabutylammonium hexafluorophosphate as supporting electrolyte
at a scan rate of 50 mV/s.
NMR spectra were measured on a Variant 300 MHz machine.
All electrospray mass spectra were acquired using a Bruker
Daltonics BioAPEX-94e superconducting 9.4 T Fourier transform
ion cyclotron resonance (FTICR) mass spectrometer (Bruker
Daltonics, Billerica, MA) in broadband mode. A home-built
apparatus controlled the direct infusion of sample. The sample was
delivered dissolved in 100% ACN using a helium gas container at
a pressure of 1.3 bar, pushing the sample through a 30 cm fused
silica capillary of inner diameter 20 µm. The flow rate was
approximately 100 nL/min. The ion source was coupled to an
(53) X-RED and X-SHAPE, STOE, Crystal Optimisation for Numerical
Absorption Corrections; Darmstadt, 1997.
(54) Sheldrick, G. M. Acta Crystallogr. 1994, A46, 467.
(55) Sheldrick, G. M. SHELX97. Computer Program for the Refinement
of Crystal Structures, Release 97-2; Sheldrick, G. M., Ed.; University
of Go¨ttingen: Go¨ttingen, Germany, 1997.
4690 Inorganic Chemistry, Vol. 43, No. 15, 2004

Model of the Iron Hydrogenase ActiWe Site
Analytica atmosphere vacuum interface (Analytica of Branford,
CT), and a potential difference of 2-4 kV was applied across a
distance of approximately 5 mm between the spraying needle and
the inlet capillary.
BrFe2NO6S2: C 31.03, H 1.48, N 2.59; found: C 31.21, H 1.62, N
1
2.56; H NMR (CDCl3): 7.39 (d, J ) 9.0 Hz, 2 H), 6.61 (d, J )
9.0 Hz, 2 H), 4.26 (s, 4 H, NCH2S); 13C NMR (CDCl3): 206.9,
143.8, 132.7, 117.4, 112.8, 49.7; IR(CH2Cl2): νCO 2077, 2038,
2000.
UV/vis absorption spectra were measured on a Varian Cary 50
Bio spectrophotometer. The solvent used in the emission studies
was acetonitrile of spectroscopic grade (Merck). The emission
spectra were recorded on a SPEX-Fluorolog II fluorimeter. Spectra
at 77 K were taken in dry butyronitrile solution in a coldfinger
Dewar. The infrared spectra were recorded on Perkin-Elmer
Spectrum 1. The time correlated single photon counting setup was
pumped with 100 kHZ pulses of 150 fs width generated in a
regenerative amplified Ti:Sapphire system from Coherent. The
wavelengths used for the experiments was 400 nm obtained from
doubling of the fundamental 800 nm light. The sample was
contained in a 1 × 1 cm quartz cuvette, and the emission light was
collected at magic angle conditions compared to the excitation light.
A 400 nm interference filter was used before the sample, and
different combinations of interference and cut off filters after the
sample were used to remove unwanted wavelengths. Emitted light
was collected by a water cooled Hammamatsu R38094-5 MCP-
PMT and resulted in a response function with a fwhm of 65 ps.
Flash photolysis experiments were performed with a frequency
tripled Q-switched Nd:YAG laser from Quantel. The out coming
light pumped an OPO generating <10 ns flashes tunable in the
range 410-660 nm. Analyzing light was provided by a pulsed 100
W Xe-lamp used in a spectrometer system from Applied Photo-
physics. Average energy of the laser pulses was 25 mJ.
[(µ-SCH2N(4-Iodophenyl)CH2S)Fe2(CO)6] (3). This compound
was synthesized in analogy to the preparation of 2, starting from
N,N-bis(chloromethyl)-4-iodoaniline (632 mg, 2 mmol). Red solid,
451 mg (0.76 mmol, 76%). Elemental analysis (%) calcd for C14H8-
Fe2NO6S2I: C 28.55, H 1.37, N 2.38; found: C 28.66, H 1.43, N
1
2.24; H NMR (CDCl3): 7.40 (d, J ) 9.0 Hz, 2 H), 6.63 (d, J )
9.0 Hz, 2 H), 4.29 (s, 4 H, NCH2S); 13C NMR (CDCl3): 206.8,
144.4, 138.6, 117.9, 82.3, 49.5; IR(CH2Cl2): νCO 2076, 2038, 2000.
[(µ-SCH2N(4-Bromophenyl)CH2S)Fe2(CO)5(PPh3)] (4). [(µ-
SCH2N(4-bromophenyl)CH2S)Fe2(CO)6] (40 mg, 0.075 mmol) was
dissolved in degassed n-propylamine (10 mL), and triphenylphos-
phine (65 mg, 0.25 mmol) was added. After 5 h of stirring at reflux,
TLC control indicated that all starting material had been consumed.
The red residue, obtained after removal of the solvent in vacuo,
was subjected to column chromatography (silica, 50% toluene in
hexane). Red solid, 43 mg (0.055 mmol, 74%). Elemental analysis
(%) calcd for C31H23BrFe2NO5PS2: C 47.92, H 2.99, N 1.80;
1
found: C 47.70, H 2.96, N 1.70; H NMR (CDCl3): 7.73 (m, 6
H), 7.45 (m, 9 H), 7.27 (d, J ) 7.8 Hz, 2 H), 6.43 (d, J ) 7.8 Hz,
2 H), 3.98 (d, J ) 12.2 Hz, 2 H), 2.95 (d, J ) 12.2 Hz, 2 H); 13C
NMR (CDCl3): 212.9, 212.7, 208.6, 145.3, 135.7, 135.1, 133.6,
133.5, 132.2, 130.4, 130.3, 128.7, 128.6, 117.1, 111.7, 47.0; IR-
(CH2Cl2): νCO 2048, 1990, 1935.
p-Formaldehyde, 4-iodoaniline, 4-bromoaniline, thionyl chloride,
[Pd(PPh3)2Cl2], CuI, and phenyl acetylene were purchased from
commercial suppliers and used as obtained. [(µ-S2)Fe2(CO)6],29 4′-
[((trifluoromethyl)sulfonyl)oxy]-2,2′:6′,2′′-terpyridine,34 4′-ethynyl-
2,2′:6′,2′′-terpyridine,35 iodo-4-(N-piperidyl)benzene,39 and [(terpy)-
RuCl2(DMSO)] were synthesized according to their respective
literature procedure. N,N-Di(chloromethyl)-4-haloaniline and (µ-
SCH2N(4-halophenyl)CH2S)[Fe(CO)3]2 were synthesized in analogy
to the published procedures.25 The synthesis and characterization
of 1 and 6 have been reported elsewhere.23
N,N-Bis(chloromethyl)-4-bromoaniline. p-Formaldehyde (780
mg, 25 mmol) was added to a solution of 4-bromoaniline (1.72 g,
10 mmol) in CH2Cl2 (12 mL). The resulting slurry was stirred at
room temperature for 3 h, before thionyl chloride (2.9 mL, 40
mmol) was added dropwise. After having been stirred for further
30 min, the solvents were removed in vacuo, and the remaining
solid was dried at 0 °C at high vacuum. The product was practically
pure by NMR analysis and was used without further purification.
1H NMR (CDCl3): 7.46 (d, J ) 9.0 Hz, 2 H), 7.09 (d, J ) 9.0 Hz,
2 H), 5.48 (s, 4 H, NCH2Cl).
N,N-Bis(chloromethyl)-4-iodoaniline. This compound was syn-
thesized in analogy to the preparation of N,N-bis(chloromethyl)-
4-bromoaniline, starting from 4-iodoaniline. 1H NMR (CDCl3): 7.64
(d, J ) Hz, 2 H), 6.98 (d, J ) Hz, 2 H), 5.47 (s, 4 H, NCH2Cl);
13C NMR (CDCl3): 143.0, 138.4, 118.2, 86.1, 65.4.
[(µ-SCH2N(4-Bromophenyl)CH2S)Fe2(CO)6] (2). Super hy-
dride (LiBEt3H, 1 M solution in THF, 2.06 mL, 2.06 mmol) was
added to a degassed solution of [(µ-S2)Fe2(CO)6] (344 mg, 1 mmol)
in THF (25 mL) at - 78 °C over 10 min. N,N-Bis(chloromethyl)-
4-bromoaniline (540 mg, 2 mmol) was added to the green solution,
causing an immediate change in color to red. After the reaction
mixture was allowed to warm to room temperature, the solvent was
removed in vacuo, and the resulting red solid was purified by
column chromatography (silica, 20% toluene in hexane). Red solid,
412 mg (0.76 mmol, 76%). Elemental analysis (%) calcd for C14H8-
[µ-SCH2N(4-Trimethylsilylethynylphenyl)CH2S][Fe(CO)3]2 (5).
[(PPh3)2PdCl2] (7 mg, 0.01 mmol) and CuI (2 mg, 0.01 mmol) were
added successively to a degassed solution of iodoarene 3 (59 mg,
0.1 mmol) and trimethylsilylacetylene (20 mg, 0.2 mmol) in
triethylamine (8 mL) at 40 °C. After 90 min of stirring at this
temperature, the mixture was concentrated in vacuo, and the black
residue was subjected to column chromatography (silica, 40%
toluene in hexanes). The title compound was afforded as a red solid,
51 mg (0.091 mmol, 91%). Single crystals suitable for X-ray
analysis were grown from concentrated pentane solutions at -20
°C. Elemental analysis (%) calcd for C19H17Fe2NO6S2Si: C 40.81,
H 3.06, N 2.50; found: C 40.96, H 3.00, N 2.41; 1H NMR
(CDCl3): 7.57 (d, J ) 9.0 Hz, 2 H), 6.50 (d, J ) 9.0 Hz, 2 H),
4.26 (s, 4 H, NCH2S); 13C NMR (CDCl3): 206.8, 144.4, 133.8,
115.1, 114.6, 105.0, 93.0, 49.5, 0.1; IR(CH2Cl2): νCdC 2153, νCO
2076, 2038, 2001.
4′-(4-N-Piperidylphenylethynyl)-2,2′:6′,2′′-terpyridine (7a).
Copper iodide (2 mg, 0.01 mmol) was added to a degassed solution
of 4-N-piperidyliodobenzene (38 mg, 0.13 mmol), 4′-ethynyl-2,2′:
6′,2′′-terpyridine (34 mg, 0.13 mmol) and [Pd(PPh3)2Cl2] (7 mg,
0.01 mmol) in toluene, and Et3N (1:1, 10 mL) at 70 °C. After having
been stirred at this temperature for 3 h, the solvents were removed
in vacuo, and the resulting black solid was subjected to column
chromatography (silica, 2% acetone in CH2Cl2). Tan white solid,
33 mg (0.08 mmol, 61%). 1H NMR (CDCl3): 8.69 (m, 2 H), 8.59
(m, 2 H), 8.51 (s, 2 H), 7.84 (m, 2 H), 7.42 (d, J ) 8.7 Hz, 2 H),
7.31 (m, 2 H), 6.86 (d, J ) 8.7 Hz, 2 H), 3.24 (m, 4 H), 1.65 (m,
6 H); 13C NMR (CDCl3): 155.9, 155.3, 151.8, 149.1, 136.8, 134.2,
133.2, 123.9, 122.5, 121.2, 115.0, 111.2, 95.3, 86.2, 49.4, 25.4,
24.3.
4′-(Phenylethynyl)-2,2′:6′,2′′-terpyridine (7b). This compound
was synthesized in a procedure analogous to that for the preparation
of ligand 7a, starting from phenyl acetylene (61 mg, 0.6 mmol)
and 4′-[((trifluoromethyl)sulfonyl)oxy]-2,2′:6′,2′′-terpyridine (114
1
mg, 0.3 mmol). White solid, 83 mg (0.25 mmol, 83%). H NMR
Inorganic Chemistry, Vol. 43, No. 15, 2004 4691

Ott et al.
(CDCl3): 8.69 (m, 2 H), 8.58 (m, 2 H), 8.55 (s, 2 H), 7.82 (m, 2
H), 7.56 (m, 2 H, Ph), 7.36 (m, 3 H, Ph), 7.30 (m, 2 H); 13C NMR
(CDCl3): 155.6, 155.4, 149.1, 136.8, 133.3, 131.9, 129.0, 128.4,
123.9, 122.8, 122.4, 121.2, 93.7, 87.5.
[(Terpy)Ru(4′-(Phenylethynyl)-2:2′,6′:2′′-terpyridine)] (PF6)2
(8b). Synthesized in analogy to 8a, starting from ligand 7b (77
mg, 0.23 mmol) and [(terpy)RuCl2‚DMSO] (111 mg, 0.23 mmol).
Red solid, 141 mg (0.15 mmol, 64%). Elemental analysis (%) calcd
for C38H26F12N6P2Ru: C 47.66, H 2.74, N 8.77; found: C 47.42,
[(Terpy)Ru(4′-(4-N-Piperidylphenylethynyl)-2:2′,6′:2′′-terpyr-
idine)] (PF6)2 (8a). [(Terpy)RuCl2‚DMSO] (20 mg, 0.04 mmol)
was added to a degassed solution of ligand 7a (17 mg, 0.04 mmol)
in ethanol (20 mL). After stirring at reflux for 16 h, the solvent
was removed in vacuo, and the resulting red solid was subjected
to flash chromatography (silica, CH3CN:H2O:saturated aqueous
KNO3 ) 90:9:1). The fractions containing product were concen-
trated in vacuo, and the remaining red solid was dissolved in water
(5 mL) to which a saturated aqueous solution of NH4PF6 (10 mL)
was added. After extraction with CH2Cl2) (3*15 mL) the combined
organic layers were washed with a saturated aqueous solution of
NH4PF6 (2*15 mL) and dried over Na2SO4. Filtration and removal
of the solvent in vacuo was followed by recrystallization from
CH2Cl2/Et2O to afford the title compound as a red solid, 22 mg
(0.021 mmol, 53%). Elemental analysis (%) calcd for C43H35F12N7P2-
Ru: C 49.59, H 3.39, N 9.41; found: C 49.41, H 3.62, N 9.24; 1H
NMR (CD3CN): 8.78 (s, 2 H), 8.73 (d, J ) 8.4 Hz, 2 H), 8.47 (m,
4 H), 8.40 (t, J ) 8.4 Hz, 1 H), 7.91 (m, 4 H), 7.59 (d, J ) 9.0 Hz,
2 H), 7.39 (m, 2 H), 7.32 (m, 2 H), 7.15 (m, 4 H), 7.03 (d, J ) 9.0
Hz, 2 H), 3.37 (m, 4 H), 1.67 (m, 6 H); 13C NMR (CD3CN): 158.9,
158.7, 156.2, 156.1, 153.6, 153.5, 139.1 (2 H), 136.9, 134.5, 132.1,
128.5, 128.4, 125.6, 125.4 (2 H), 124.7, 115.6, 109.6, 100.4, 86.3,
49.4, 26.2, 25.1; FTICR ESI-MS: Ru-isotopic pattern centered at
m/z ) 896.19 (100%) [M - PF6-]+, calcd for [M - PF6-]+: m/z
) 896.16.
1
H 2.90, N 8.72; H NMR (CD3CN): 8.90 (s, 2 H), 8.78 (d, J )
8.1 Hz, 2 H), 8.53 (m, 4 H), 8.42 (t, J ) 8.1 Hz, 1 H), 7.92 (m, 4
H), 7.79 (m, 2 H, Ph), 7.56 (m, 3 H, Ph), 7.38 (m, 4 H), 7.17 (m,
4 H); 13C NMR (CD3CN): 158.9, 158.6, 156.4, 156.2, 153.6, 153.5,
139.2 (2 H), 137.1, 133.1, 131.3, 131.0, 130.1, 128.7, 128.5, 126.2,
125.6, 125.5, 124.8, 122.3, 97.7, 87.2; FTICR ESI-MS: Ru-isotopic
pattern centered at m/z ) 813.11 (100%) [M - PF6-]+, calcd for
[M - PF6-]+: m/z ) 813.09.
Acknowledgment. Financial support for this work was
provided by the Swedish Energy Agency, the Knut and Alice
Wallenberg Foundation, and the Swedish Research Council
(VR). J. B. acknowledges the financial support by grants
from the Knut and Alice Wallenberg Foundation and the
Swedish Research Council (621-2002-5261, 629-2002-6821).
L.H. acknowledges a research fellow position from the Royal
Swedish Academy of Science.
Supporting Information Available: A cyclic voltammogram
typical for [(µ-ADT)Fe2(CO)6] type complexes and emission decay
traces for complexes 1 and 8b. This material is available free of
charge via the Internet at http://pubs.acs.org.
IC0303385
4692 Inorganic Chemistry, Vol. 43, No. 15, 2004