17- and 19-electron complexes [Fe(III)(η5-C5R5)(S2CNMe2)L](n+) (n = 1, 0): Electronic structure and substitution and redox chemistry. Formation of [Fe(IV)(η-C5R5)(dtc)2] and characterization of both 17e and 19e states of a transition-metal complex

Article abstract of DOI:10.1021/ja953603x

Oxidation of [FeCp*(η¹-dtc)(CO)₂], 1 (Cp* = η⁵-C₅Me₅, dtc = S₂CNMe₂), or [FeCp*(η²-dtc)(CO)], 2, using [Fe(III)Cp₂]⁺X^- (X^- = PF₆^- or BF₄^-, Cp = η⁵-C₅H₅) in THF cleanly gives [Fe(III)Cp*(η²-dtc)(CO)]⁺X^-, 2⁺X^-, as microcrystalline green, thermally stable, but substitution labile, salts. The substitution of CO in 2⁺PF₆^- by various solvents (CH₂Cl₂, THF, CH₃COCH₃, CH₃CN) (visible spectroscopy) follows pseudo-first-order kinetics but shows clearly the influence of the incoming solvent ligand on the substitution rate and, hence, is in good agreement with an associative mechanism. Displacement of the labile solvent ligand in these complexes by a phosphine results in the 17-electron (17e) cations [Fe(III)Cp*(η²-dtc)(L)]⁺PF₆^-, L = PPh₃ (7⁺PF₆^-) or η¹-dppe (8⁺PF₆^-). The same reaction in the presence of the anionic ligands CN^-, SCN^-, and Cl^- affords the corresponding neutral 17e Fe(III) complexes (respectively compounds 11, 13, and 14). All these 17e complexes were characterized by IR, ESR, and Mossbauer spectroscopies and elemental analysis. The cations were reduced to isostructural neutral Fe(II) complexes using 1 equiv of [Fe(I)Cp(C₆Me₆)] in THF or oxidized to the robust green 18e Fe(IV) complex [Fe(IV)(η⁵-C₅Me₅)(η²-S₂CNMe₂)₂]⁺PF₆^-, 9⁺PF₆^-, using Na⁺dtc^-·2H₂O. [Fe(IV)(η⁵-C₅(CH₂C₆H₅)₅)(η²-S₂CNMe₂)₂]⁺PF₆^-, 16⁺PF₆^-, was structurally characterized, and the dihapto mode of coordination of both dtc ligands was established. The 19e Fe(III) species 9 was shown to be an intermediate which further reduced H₂O. It could be alternatively synthesized by reduction of the 18e precursor 9⁺PF₆^- using 1 equiv of [Fe(I)Cp(C₆Me₆)] or by addition of anhydrous Na⁺dtc^- to 3⁺PF₆^- in MeCN at -40°C. The 19e complex 9 showed an ESR spectrum indicating an axial symmetry (two g values) in contrast with the ESR spectra of all the 17e species (2⁺-14) which show three g values characteristic of a rhombic distortion (for instance, the very close model 13). The Mossbauer doublet of 9 very slowly evolved to the new doublet of the thermally stable 17e complex 9'. In MeCN solution, the transformation of the blue complex 9 to the purple 17e complex 9' was much more rapid (above -40°C) as indicated by the rhombic spectrum of 9' in frozen solution and by low-temperature ¹³C NMR. In toluene, however, the 19e complex 9 showed a remarkable stability up to room temperature, which allowed recording of the ¹³C spectrum in d⁸-toluene. MO calculations have been performed on models for the 17e and 19e bis-dtc Fe(III) complexes. They suggest that the 17e species should have some significant sulfur spin density. The 19e species is found to have its odd electron occupying an antibonding metal-centered orbital. The cyclic voltammogram of 9⁺PF₆^- under continuous scanning for the monoelectronic reduction and the two monoelectronic reductions showed the decrease of the waves of 9⁺PF₆^- and the concomitant increase of those due to the partially decoordinated dtc complexes formed upon reduction. This permits an interpretation of the CV in terms of a triple-square scheme involving 9(+/0/-), 9'(+/0/-), and solvent (DMF) adducts in 18- and 19e states.

Full text of DOI:10.1021/ja953603x

J. Am. Chem. Soc. 1996, 118, 4133-4147
4133
17- and 19-Electron Complexes [FeIII(η5-C5R5)(S2CNMe2)L]n+
(n ) 1, 0): Electronic Structure and Substitution and Redox
Chemistry. Formation of [FeIV(η5-C5R5)(dtc)2] and
Characterization of both 17e and 19e States of a
Transition-Metal Complex
Marie-He´le`ne Delville-Desbois,*,† Stefan Mross,† Didier Astruc,*,† Jorge Linares,‡
Franc¸ois Varret,*,‡ Hassan Rabaaˆ,§,| Albert Le Beuze,§ Jean-Yves Saillard,*,§
Robert D. Culp, David A. Atwood, and Alan H. Cowley*,
Contribution from the Laboratoire de Chimie Organique et Organome´tallique, URA CNRS No.
35, UniVersite´ Bordeaux I, 351 Cours de la Libe´ration, 33405 Talence Ce´dex, France,
De´partement de Recherches Physiques, URA CNRS No. 71, UniVersite´ P. et M. Curie,
Tour 22, 4 Place Jussieu, 75252 Paris Ce´dex 05, France, Laboratoire de Chimie du Solide et
Inorganique Mole´culaire, URA CNRS No. 495, UniVersite´ de Rennes I, Campus de Beaulieu,
35042 Rennes Ce´dex, France, Laboratoire de Chimie The´orique, Faculte´ des Sciences, UniVersite´
Ibn Tofail, BP 133, Ke´nitra, Morocco, and Department of Chemistry and Biochemistry, The
UniVersity of Texas at Austin, Austin, Texas 78712-1167
ReceiVed October 27, 1995X
Abstract: Oxidation of [FeCp*(η1-dtc)(CO)2], 1 (Cp* ) η5-C5Me5, dtc ) S2CNMe2), or [FeCp*(η2-dtc)(CO)], 2,
using [FeIIICp2]+X- (X- ) PF6- or BF4-, Cp ) η5-C5H5) in THF cleanly gives [FeIIICp*(η2-dtc)(CO)]+X-, 2+X-,
as microcrystalline green, thermally stable, but substitution labile, salts. The substitution of CO in 2+PF6- by various
solvents (CH2Cl2, THF, CH3COCH3, CH3CN) (visible spectroscopy) follows pseudo-first-order kinetics but shows
clearly the influence of the incoming solvent ligand on the substitution rate and, hence, is in good agreement with
an associative mechanism. Displacement of the labile solvent ligand in these complexes by a phosphine results in
the 17-electron (17e) cations [FeIIICp*(η2-dtc)(L)]+PF6-, L ) PPh3 (7+PF6-) or η1-dppe (8+PF6-). The same reaction
in the presence of the anionic ligands CN-, SCN-, and Cl- affords the corresponding neutral 17e FeIII complexes
(respectively compounds 11, 13, and 14). All these 17e complexes were characterized by IR, ESR, and Mo¨ssbauer
spectroscopies and elemental analysis. The cations were reduced to isostructural neutral FeII complexes using 1
equiv of [FeICp(C6Me6)] in THF or oxidized to the robust green 18e FeIV complex [FeIV(η5-C5Me5)(η2-S2CNMe2)2]+-
PF6-, 9+PF6-, using Na+dtc-‚2H2O. [FeIV(η5-C5(CH2C6H5)5)(η2-S2CNMe2)2]+PF6-, 16+PF6-, was structurally
characterized, and the dihapto mode of coordination of both dtc ligands was established. The 19e FeIII species 9
was shown to be an intermediate which further reduced H2O. It could be alternatively synthesized by reduction of
-
-
the 18e precursor 9+PF6 using 1 equiv of [FeICp(C6Me6)] or by addition of anhydrous Na+dtc- to 3+PF6 in
MeCN at -40 °C. The 19e complex 9 showed an ESR spectrum indicating an axial symmetry (two g values) in
contrast with the ESR spectra of all the 17e species (2+-14) which show three g values characteristic of a rhombic
distortion (for instance, the very close model 13). The Mo¨ssbauer doublet of 9 very slowly evolved to the new
doublet of the thermally stable 17e complex 9′. In MeCN solution, the transformation of the blue complex 9 to the
purple 17e complex 9′ was much more rapid (above -40 °C) as indicated by the rhombic spectrum of 9′ in frozen
solution and by low-temperature 13C NMR. In toluene, however, the 19e complex 9 showed a remarkable stability
up to room temperature, which allowed recording of the 13C spectrum in d8-toluene. MO calculations have been
performed on models for the 17e and 19e bis-dtc FeIII complexes. They suggest that the 17e species should have
some significant sulfur spin density. The 19e species is found to have its odd electron occupying an antibonding
-
metal-centered orbital. The cyclic voltammogram of 9+PF6 under continuous scanning for the monoelectronic
reduction and the two monoelectronic reductions showed the decrease of the waves of 9+PF6- and the concomitant
increase of those due to the partially decoordinated dtc complexes formed upon reduction. This permits an
interpretation of the CV in terms of a triple-square scheme involving 9+/0/-, 9′+/0/-, and solvent (DMF) adducts in
18- and 19e states.
Introduction
the intermediacy of these species in redox catalysis and
electrocatalysis.3,4 Such types of catalysis involving electron
transfer (ET) are used in energy conversion5 and in multicatalytic
devices.6 In this context, the study of the 17e complexes
containing the chelating dithiocarbamate ligand seemed timely
(1) (a) Organometallic Radical Processes; Trogler, W. C., Ed.; Elsevi-
er: New York; J. Organomet. Chem. Lib. 1990, 22. (b) Paramagnetic
Species in ActiVation SelectiVity, Catalysis; Chanon, M., Juliard, M., Poite,
J. C., Eds.; Kluwer: Dordrecht, The Netherlands, 1988. (c) Astruc, D.
Electron Transfer and Radical Processes in Transition-Metal Chemistry;
VCH: New York, 1995.
The factors controlling the thermodynamic and kinetic
stability of 17e complexes1,2 are of current interest because of
† URA No. 35.
‡ URA No. 71. Present address: Laboratoire d’Optique et de Magne´-
tisme, URA No. 1531, Universite´ de Versailles, 45 Avenue des Etats-Unis,
78035 Versailles Cedex, France. Mo¨ssbauer studies.
§ URA No. 495. Theoretical studies.
| Universite´ Ibn Tofail.
The University of Texas at Austin. X-ray crystal structures.
X Abstract published in AdVance ACS Abstracts, April 1, 1996.
S0002-7863(95)03603-1 CCC: $12.00 © 1996 American Chemical Society

4134 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
DelVille-Desbois et al.
in view of the potential electron-reservoir property7 of the
delocalized iron-η2-dithiocarbamate (dtc) metallocyclic frag-
ment.8,9 It was hoped that this ligand would stabilize organoiron
radicals. With this system, we have disclosed the first example
of an electrocatalyzed reaction which can be initiated either by
an oxidant4b,10a,b or by a reductant:10c
[FeIIICp*(η2-dtc)(L)]x+ (x ) 0, 1). Their transformation into
the novel complexes [FeCp*(η2-dtc)2]n+ (FeIII for n ) 0, FeIV
for n ) 1) is shown.10d When n ) 0, one of the dtc ligands
can be monodentate (17e) or bidentate (19e), which leads to
the observation of an original transition-metal complex which
can have either 17 or 19 valence electrons (VE).
(e or hν8
+ CO
Results and Discussion
[FeCp*(η1-dtc)(CO)2]
2
[FeCp*(η -dtc)(CO)]
Synthesis of the 17e Complexes [FeIIICp*(η2-dtc)(CO)]+X-,
2
1
-
-
X- ) PF6 or BF4
. The reaction between 1 equiv of
(1)
-
- 14
[FeIIICp2]+X-, X- ) PF6 or BF4
,
in THF suspension and
1 equiv of either the monodentate complex [FeCp*(η1-dtc)-
(CO)2], 1, or the bidentate complex [FeCp*(η2-dtc)(CO)], 2,
cleanly gives a yellow-orange solution containing [FeCp2] and
an insoluble green powder. Attempts to solubilize and recrys-
tallize this green powder always led to a rapid color change
indicative of an evolution of this complex in solution. However,
elemental analyses were satisfactory for the 1e oxidation
Since the major problem in these ET chains is, as in organic
chemistry,11 the side reactions of the intermediate radicals,3c
we attempted to isolate, characterize, and study the chemistry
of such radicals.12 A number of 17e complexes are already
known,2 but no examples featuring the dtc ligand have been
reported before this study.12 The chemistry presented here is
rich and permits novel features of the 17e states to be elucidated.
The parent 18e complexes [FeCp(η1-dtc)(CO)2] and [FeCp(η2-
dtc)(CO)] (Cp ) C5H5) have been known for more than two
decades, but we found the C5Me5 (Cp*) series easier to study,9e
although still extremely labile in their 17e form. Complex 2
can also be obtained cleanly from 1 by visible-light photolysis13
(eq 1) whereas both oxidatively and reductively induced
electrocatalyzed chelations proceed in lower yields.10a-c In this
article, we explore the access, electronic structure, stability,
substitution lability, and redox processes of the 17e complexes
-
products 2+X- (X- ) PF6 or BF4-). The infrared spectra
recorded from KBr pellets showed the presence of terminal CO
stretches characteristic of the cationic nature of the complex
(νCO ) 2030 and 1985 cm-1); the presence of the dtc ligand in
its chelate form15 was suggested by the stretches corresponding
to CN and CS bonds (νCN ) 1550 cm-1, νCS ) 1015 cm-1).
The ESR spectra were consistent with an iron-centered radical
and an orthorhombic-axial transition in the solid-state sample
was observed at 200 K. The Mo¨ssbauer spectrum showed the
parameters of FeIII (IS ) 0.518 mm s-1 vs Fe; QS ) 0.686 mm
s-1) close to those obtained for the 17e FeIII complexes
[FeIIICp*(CO)(L)(CH3)]+PF6- or [FeIIICp*(CH3)(L)2]+PF6- (L
) phosphine).17,18 Additional evidence for the formation of
2+PF6- was provided by its monoelectronic reduction using [FeI-
Cp(C6Me6)] in THF solution at 20 °C (15 min), a reaction which
cleanly gave back the soluble complex 2 and a yellow precipitate
(2) For reviews, see Reference 1 and: (a) Baird, M. C. Chem. ReV. 1988,
88, 1217. (b) Connelly, N. G.; Geiger, W. E. AdV. Organomet. Chem. 1984,
23, 2. (c) Geiger, W. E.; Connelly, N. G. AdV. Organomet. Chem. 1985,
24, 87. (d) Stiegman, A. E.; Tyler, D. R. Comments Inorg. Chem. 1986, 5,
215. (e) Geoffroy, G. L.; Wrighton, M. S. “Organometallic Photochemistry;
Academic Press: New York, 1979. (f) Meyer, T. J.; Caspar, J. V. Chem.
ReV. 1985, 85, 187. (g) Brown, T. L. Ann. N.Y. Acad. Sci. 1980, 331, 80.
(h) Kobayashi, T.; Yasufuku, K.; Iwai, I.; Yesaka, H.; Noda, H.; Ohtani,
H. Coord. Chem. ReV. 1985, 64, 1.
(3) For reviews, see: (a) Kochi, J. K. J. Organomet. Chem. 1986, 300,
139. (b) Chanon, M. Acc. Chem. Res. 1987, 20, 214. (c) Astruc, D. Angew.
Chem., Int. Ed. Engl. 1988, 27, 643.
(4) Hersgberger, J. W.; Klinger, R. J.; Kochi, J. K. J. Chem. Soc., Chem.
Commun. 1982, 212. (b) Hersgberger, J. W.; Klinger, R. J.; Kochi, J. K.
J. Am. Chem. Soc. 1983, 105, 61. (c) Catheline, D.; Astruc, D. Coord.
Chem. ReV. 1982, 23, 41.
(5) See for instance: Lehn, J. M.; Sauvage, J. P.; Ziessel, R. NouV. J.
Chim. 1979, 3, 423.
-
of [FeIICp(C6Me6)]+PF6 (Scheme 1).
-
Although the system 2/2+PF6 is not fully chemically
reversible in THF solution because of the exchange of the CO
ligand with THF in the presence of [nBu4N+]BF4-, the
thermodynamic potential (E°) is easily accessible (+0.605 V
vs. SCE).10a,d As already shown,10d ferrocenium salts can
quantitatively oxidize 2 even if the thermodynamic potential
of the couple FeCp2/[FeCp2]+PF6- is only + 0.4 V vs. SCE.2b
If we assume that the solution is homogeneous, the redox
reaction (eq 2) is endergonic by 0.2 V (4.6 kcal/mol). However,
(6) Wrighton, M. S. Comments Inorg. Chem. 1985, 4, 269.
(7) Astruc, D. Acc. Chem. Res. 1986, 19, 377; Comments Inorg. Chem.
1987, 6, 61 (see also ref 1c, Chapter 6, and ref 3c).
(8) For reviews on dithiocarbamate complexes, see: (a) Fackler, J. P.
AdV. Chem. Ser. 1976, No. 150, 394. (b) Thorn, G. D.; Ludwig, R. A. The
Dithiocarbamates and Related Compounds; Elsevier: Amsterdam, 1962.
(c) Livingstone, S. E. Q. ReV. Chem. Soc. 1965, 19, 416. (d) Coucouvanis,
D. In Progress in Inorganic Chemistry; Lippard, S., Ed.; Wiley: New York,
1970; Vol. 11, p 233. (e) Coucouvanis, D. Prog. Inorg. Chem. 1979, 26,
301. (f) Willemse, J.; Cras, J. A.; Steggarda, J. J.; Keijzers, C. P. Struct.
Bonding (Berlin) 1976, 28, 83.
(9) For synthesis and studies of 18e [FeCp(dtc)] complexes, see: (a)
O’Connor, C.; Gilbert, J. D.; Wilkinson, G. J. Chem. Soc. A 1969, 84. (b)
Cotton, F. A.; McCleverty, J. A. Inorg. Chem. 1964, 5, 1398. (c) Roma´n,
E.; Catheline, D.; Astruc, D.; Batail, P.; Ouahab, L.; Varret, F. J. Chem.
Soc., Chem. Commun. 1982, 129. (d) Roma´n, E.; Astruc, D. Inorg. Chem.
1979, 18, 3284. (e) Catheline, D.; Roma´n, E.; Astruc, D. Inorg. Chem.
1984, 23, 4508. (f) Roma´n, E.; Catheline, D.; Astruc, D. J. Organomet.
Chem. 1982, 236, 229.
(10) (a) Amatore, C.; Verpeaux, J.-N.; Madonik, A. M.; Desbois, M.-
H.; Astruc, D. J. Chem. Soc., Chem. Commun. 1988, 200. (b) Verpeaux,
J.-N.; Desbois, M.-H.; Madonik, A. M.; Amatore, C.; Astruc, D. Organo-
metallics 1990, 9, 630. (c) Desbois, M.-H.; Astruc, D. J. Chem. Soc., Chem.
Commun. 1990, 943; 1995, 249. (d) Desbois, M.-H.; Astruc, D. Angew.
Chem., Int. Ed. Engl. 1989, 28, 460.
(11) (a) Bunnett, J. F. Acc. Chem. Res. 1978, 11, 413. (b) Save´ant, J.-
M. Acc. Chem. Res. 1980, 13, 323.
(12) See ref 10a,b.
-
2 + [FeCp2]+PF6- f 2+PF6
+
FeCp2 ∆G ) 4.6 kcal/mol (2)
-
-
in THF, neither [FeCp2]+PF6 nor 2+PF6 is soluble and the
displacement of the redox equation in favor of the formation
-
of 2+PF6 must be driven by its greater insolubility in this
solvent. Indeed, this complex is totally inert toward ligand
exchange in this relatively apolar solvent. Addition of a salt,
such as [nBu4N+]BF4-, or a cosolvent, such as CH2Cl2 (3/1),
(14) Catheline, D.; Astruc, D. J. Organomet. Chem. 1982, 226, C52.
(15) (a) Bonatti, F.; Hugo, R. J. Organomet. Chem. 1967, 10, 257. (b)
Malatesta, L.; Bonatti, F. Isocyanide Complexes of Metals; Wiley: New
York, 1969.
(16) Rajasekharan, M. V.; Giezynski, S.; Ammeter, J. H.; Ostwald, N.;
Michaud, P.; Hamon, J.-R.; Astruc, D. J. Am. Chem. Soc. 1982, 104, 2400.
(17) Morrow, J.; Catheline, D.; Desbois, M. H.; Manriquez, J. M.; Ruiz,
J.; Astruc, D. Organometallics 1987, 6, 2605.
(18) Greenwood, N. N.; Gibb, T. C. In Mo¨ssbauer Spectroscopy;
Chapmann and Hall: London, 1971; Chapter 8 and references cited therein.
See also this chapter for a few FeIV inorganic complexes.
(19) (a) Hamon, J.-R.; Astruc, D.; Michaud, P. J. Am. Chem. Soc. 1981,
103, 758. (b) Astruc, D.; Hamon, J.-R.; Lacoste, M.; Desbois, M.-H.;
Roma´n, E. Organomet. Synth. 1988, 4, 172.
(13) Desbois, M.-H.; Nunn, C. M.; Cowley, A. H.; Astruc, D. Organo-
metallics 1990, 9, 640.

[FeIII(η5-C5R5)(S2CNMe2)L]n+ Complexes
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4135
Scheme 1
Figure 1. Evolution of the yields of 2 and 2+PF6- during the electron-
-
transfer-chain catalysis between 1 and [FeCp2]+PF6
.
-
-
Back reduction of the 17e complexes 2+PF6 and 7+PF6
using Na/Hg gives very unsatisfactory results, due to the affinity
of Hg for sulfur, but the use of [FeICp(C6Me6)] is very
convenient and 1e reduction using this reagent is very clean. It
-
is noteworthy that replacement of the CO ligand in 2/2+PF6
increases the polarity of THF, solubilizes 2+PF6- in the medium,
and leads to 4+PF6- in which the last carbonyl group has been
replaced by a THF molecule. This was confirmed by electro-
chemical experiments which revealed that, when 1 and 2 were
oxidized electrochemically in THF on a preparative scale, they
gave a brown salt which was characterized as [FeIIICp*(η2-
dtc)(THF)]+BF4-, 4+BF4-, by comparison with an authentic
sample prepared chemically in the CH2Cl2/THF solvent mixture.
The oxidation of 1 by [FeCp2]+PF6-, which is supposed to
by PPh3 in 7/7+PF6- displaces the redox potential by almost 1
V. However, the same FeII (FeIII) complexes can be used for
1e reduction (oxidation) showing the generality of their ap-
plication in redox synthesis. The 18/19e system 9+PF6-/9 (Vide
infra) can be handled similarly using the FeI and FeIII sandwich
compounds for passing from one oxidation state to the other.
Substitution Lability of the 17e Complexes [FeIIICp*(η2-
-
dtc)(L)]+PF6 and Intermediacy of 19e States. Kinetic
Instability of 2+PF6- in Polar Solvents. The impossibility of
be even more endergonic (two carbonyl groups coordinated to
-
dissolving 2+PF6 in any solvent without rapid CO evolution
-
FeII), is also driven by the follow-up chemistry of 2+PF6
,
namely precipitation or ligand substitution, depending on
is indicative of the reactivity of this 17e complex. Dissolution
and color change from green to purple blue or orange (depending
on the solvent) are concomitant with the disappearance of the
CO stretches in the infrared spectra. Thus, the CO ligand is
working conditions. The thermodynamic potential of the 1/1+-
-
PF6 system is unknown since the electrochemical oxidation
is totally irreversible, even at scan rates as high as 5000 V s-1
,
-
very labile and the complex 2+PF6 is kinetically unstable in
which is indicative of the extremely short lifetime of the 17e
cation 1+.10a,20 However, the conversion of 1 to 2+X- proceeds
at Epa ) 1.25 V vs. SCE at 20 V s-1, and the E° value of the
closely related system 2/2+X- is known. If one assumes that
the replacement of CO by the free sulfur of the dtc ligand
displaces E° by about 1 V, E° (1/1+X-) should be approximately
+1.65 V vs. SCE. Thus, the oxidation of 1 to 1+X- by
[FeCp2]+X- is disfavored by 1.2 V (28 kcal/mol). It is driven
by the extremely fast chelation of 1+X-, by CO evolution, and
by the precipitation of 2+X-, since the yields of the crude
complex 2+X- are virtually quantitative (X- ) PF6- or BF4-).
Indeed, a study of the evolution of the respective yields of 2
and 2+PF6- in the ETC (electron transfer chain) transformation
solution despite the fact it can be stored at ambient temperature
for several months in the solid state. The rate of CO evolution
varies from one solvent to another, increasing with the
coordination ability of the incoming solvent (Scheme 2). This
CO substitution reaction leading to the FeIII centered-radicals
[FeIIICp*(η2-dtc)(L)]+PF6- was monitored by visible spectros-
copy in CH2Cl2 at 25 °C, in CH3COCH3 at 5 °C, and CH3CN
at 5 °C.10d The influence of the coordination ability of the
incoming ligand on the reaction is indicated by the straight lines
obtained by plotting ln(A - A∞) ) f(t), thus demonstrating
pseudo-first-order kinetics. The half-reaction times vary ap-
preciably with the nature of the incoming ligand (Figure 2).
-
of 1 into 2 as a function of the amount of catalyst [FeCp2]+PF6
The high rate of CO exchange with CH3CN (t1/2 ) 5 min at
shows (Figure 1) that this electrocatalytic system is rapidly
disrupted when the amount of the initiator is increased. The
side reaction is the precipitation of 2+PF6- in THF (the amount
-
5 °C) and CH3COCH3 (t1/2 ) 13.5 min at 5 °C) in 2+PF6
contrasts with its thermal stability. The pseudo-first-order
kinetics of these ligand exchange reactions and the large rate
dependence on the basicity of the incoming ligand are in
agreement with an associative mechanism involving 19e species
as intermediates or transition states.20 The associative mech-
anism for CO exchange with other ligands in 17e complexes
-
of precipitated 2+PF6 is linearly proportional to the added
quantities of the ferrocenium salt).
(20) (a) Fawcett, J. P.; Jackson, R. A.; Poe¨, A. J. J. Chem. Soc., Chem.
Commun. 1975, 733. (b) Fox, A.; Malito, J.; Poe¨, A. J. J. Chem. Soc.,
Chem. Commun. 1981, 1052. (c) Poe¨, A. J. Transition Met. Chem. 1982,
7, 65. (d) Herrinton, T. R.; Brown, T. L. J. Am. Chem. Soc. 1985, 107,
5788. (e) McCulleen, S. B.; Walker, H. W.; Brown, T. L. J. Am. Chem.
Soc. 1982, 104, 4007. (f) Shi, Q.-Z.; Richmond, T. G.; Trogler, W. C.;
Basolo, F. J. Am. Chem. Soc. 1982, 104, 4032; 1984, 106, 71. (g) Trogler,
W. C. Int. J. Chem. Kinet. 1987, 19, 1025. (h) Therien, M. J.; Ni, C.-L.;
Anson, F. C.; Osteryoung, J. G.; Trogler, W. C. J. Am. Chem. Soc. 1986,
108, 4037. (i) Zizelman, P. M.; Amatore, C.; Kochi, J. K. J. Am. Chem.
Soc. 1984, 106, 3771.
•
was first recognized and demonstrated by Poe¨20a-c for Re(CO)5 .
Subsequently this mechanism has been shown to be valid for
several other 17e complexes.20d-j It results from the very low
kinetic barrier for the 17 a 19e interconversion21 and from the
fact that 19e species have energies very close to those of 17e
species.22

4136 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
DelVille-Desbois et al.
Scheme 2
The simplest way of obtaining the complex 7+PF6- involves
-
the reaction of 2+PF6 with 1 equiv of triphenylphosphine in
acetone or acetonitrile. It can be obtained as well starting from
1 or 2, ferrocenium hexafluorophosphate, and the desired
phosphorus ligand in acetone or dichloromethane. Whereas
7+PF6- is robust, the complex 8+PF6- is rather labile and must
be handled rapidly (see the characterization in Table 1). As in
the case of 2+PF6-, 7+PF6- can be reduced back to its neutral
form 7 in good yield using 1 equiv of [FeICp(C6Me6)]. The
oxidation of the latter is completed with 1 equiv of ferrocenium
hexafluorophosphate (Scheme 3).
Physical Properties of the 17e FeIII Complexes. The main
physical characterizations of the different complexes are given
in Table 1 (see Experimental Section for others). The ESR
parameters of all compounds show anisotropic g values char-
acteristic of low-spin FeIII complexes. The Mo¨ssbauer param-
eters isomer-shift (IS) and quadrupole splitting (QS) values, are
typical for Fe-dtc complexes.18,24
The details of the X-ray crystal structure studies of 11 are
summarized in Table 2, and significant bond distances and
angles are given in Tables 3 and 4, respectively. The Fe-S
distances are identical and close to those found in known
complexes.8,13,27 The S-C distances and the very short CN
bond length (1.28 Å) are indicative of the η2 coordination mode
of the dtc ligand and the large contribution of the resonance
Figure 2. Evolution of the variation of absorbance with time (ln(A -
A∞) ) f(t)) for the reaction of CO substitution in 2+PF6- by a solvent
molecule: ], in CH2Cl2 (25 °C); O, in CH3COCH3 (5 °C); b, in
CH3CN (5 °C). Semilogarithmic paper was used for the absorbance
scale.
Reactivity of 2+PF6- toward Neutral and Anionic Ligands.
All these experiments convergently show the influence of both
the ionic strength of the medium and the donor character of the
incoming ligand on the ligand exchange. These weak 2e donor
-
ligands in 2+PF6--6+PF6 could easily be displaced by
stronger ones; for instance, 2e phosphorus donors such as PPh3
-
or Ph2PCH2CH2PPh2 produce [FeIIICp*(η2-dtc)(PPh3)]+PF6
,
7+PF6-, and [FeIIICp*(η2-dtc)(η1-dppe)]+PF6-, 8+PF6-, re-
spectively, while the anionic ligands such as CN-, SCN-, and
Cl- generate [FeIIICp*(η2-dtc)(CN)], 11, [FeIIICp*(η2-dtc)-
(SCN)], 13, and [FeIIICp*(η2-dtc)(Cl)], 14 (Scheme 2).
(24) (a) Burger, K.; Korecz, L.; Mag, P.; Belluco, U.; busetto, L. Inorg.
Chim. Acta 1971, 5, 362. (b) Bancroft, G. N.; Butler, K. D.; Manzer, L.
E.; Shaver, A.; Ward, J. H. Can. J. Chem. 1974, 52, 782. (c) Long, G. J.;
Alway, D. G.; Barnett, K. W. Inorg. Chem. 1978, 17, 488.
(25) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; Wiley: New
York, 1980; Chapter 6.
(21) Philbin, C. E.; Granatir, C. A.; Tyler, D. R. Inorg. Chem. 1986, 25,
4806.
(22) (a) Astruc, D. Chem. ReV. 1988, 88, 1189. (b) Ruiz, J.; Lacoste,
M.; Astruc, D. J. Chem. Soc., Chem. Commun. 1989, 813. (c) Ruiz, J.;
Lacoste, M.; Astruc, D. J. Am. Chem. Soc. 1990, 112, 5471.
(23) Ko¨elle, U.; Gra¨tzel, M. Inorg. Chem. 1986, 25, 2689.
(26) Wenger, P. E. Analyse QuantitatiVe Mine´rale; Dunod: Paris, 1955;
p 163.

[FeIII(η5-C5R5)(S2CNMe2)L]n+ Complexes
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4137
Table 1. ESR and Mo¨ssbauer Data for Paramagnetic 17e (2+-8+ and 10+) and 19e (9) FeIII and Diamagnetic FeIV Complexes
ESRa
Mo¨ssbauer
g1
g2
g3
ISc
QSd
[FeCp(η1-dtc)(CO)2]
-0.148
0.027
0.210
0.524
1.688
1.827
1.940
2.188
[FeCp(η2-dtc)(CO)]
[FeCp*(η2-dtc)(CO)]
[FeCp*(η2-dtc)(PPh3)], 7
[FeCp*(η2-dtc)(THF)]+, 4+
[FeCp*(η2-dtc)((CH3)2CO)]+, 5+
[FeCp*(η2-dtc)(CH2Cl2)]+, 6+ b
[FeCp*(η2-dtc)(NCMe)]+, 3+
2.323
2.453
2.571
2.672
2.611b
2.476
2.666
2.636
2.237
2.050
2.035
2.135
2.007
2.188b
2.238
2.004
2.157
2.107
2.004
1.995
1.990
0.550
0.620
2.001b
1.960
1.941
1.975
2.049
[FeCp*(η2-dtc)(PPh3)]+, 7+
[FeCp*(η2-dtc*)(PPh3)]+, 10+
[FeCp*(η2-dtc)(η1-dppe)]+, 8+
[FeCp*(η2-dtc)(CO)]+, 2+
[FeCp*(η2-dtc)2]+, 9+
0.470
0.462
0.445
0.418
0.460
0.540
0.499
0.395
0.330e
0.659
0.631
0.665
0.686
0.220
0.940e
0.332
0.884
0.660e
[FeCp*(η2-dtc)2], 9
2.275
2.268
2.372
2.029
[FeCp*(η2-dtc)(η1-dtc)], 9′
[FeCp*(η2-dtc)(CN)], 11
2.126
2.268
2.035
1.973
[FeCp*(η2-dtc**)(CO)]+, 12+ f
[FeCp*(η2-dtc)(SCN)], 13
2.322
2.483
2.055
2.088
1.999
1.997
0.460
0.389e
0.850
0.850
0.910
0.580
0.550
0.660
0.587e
0.560g
1.180g
1.700h
1.040
0.600
[FeICp(η6-C6Me6)] (19e)41
2.063
2.000
1.864
[FeI(C6Me6)2]+ h (19e)16,41,42
2.086
2.144
2.352
2.453
2.430
1.996
2.0176
2.045
2.045
2.050
1.865
1.996
1.997
1.993
1.980
[FeI(η5-C6Me6H)(η6-C6Me6)] (19e)44,45
[FeIIICp*{(POMe)3}2(CH3)]+ (17e)43
[FeIIICp*(dppe)(CH3)]+ (17e)43
[FeICp(η1-dppe)(CO)] (17e)46
-
a Solid-state sample (10 K) except when specified. b Frozen solution (20 K). c mm s-1 vs Fe at RT (77 K). d mm s-1 (77 K). PF6 was used as
counteranion (dtc ) S2CNMe2; dtc* ) S2CNEt2). e Room temperature. f dtc** ) S2CN(CH2(C6H5))2.
Scheme 3
positive charge on the nitrogen atom and corresponds to a net
shift of electron density on the sulfur atoms which are now less
able to accept electron density from the iron.
-
Oxidation of [FeIIICp*(η2-dtc)(L)]+PF6 to [FeIVCp*(η2-
dtc)2]+PF6-. The substitution lability of the 2e ligand in the
complexes 2+PF6--8+PF6- can be extended to anionic ligands
as shown above, but depending on the ligand, the 19e
intermediates obtained can be very electron rich and act as
-
reducing agents. Thus, the complexes 2+PF6--8+PF6 react
with Na+dtc-‚2H2O (eq 3) to give the new complex [FeIVCp*-
(3)
form II to the structure with a considerable CN double-bond
character (confirmed by IR spectroscopy). In limiting form I,
a sulfur electron pair is delocalized within the metal-chelate
(η2-dtc)2]+PF6-, 9+PF6- (75% yield), Na+OH- (titrated by 0.01
N HCl), and H2 (detected by gas chromatography). The reaction
corresponds to the reduction of 1 mol of H2O by the complex
according to the following stoichiometry:
ring. On the other hand, in limiting form II, the nitrogen lone
pair has been donated to the C-N bond. This locates a formal
(27) (a) Asirvatham, V. S.; Yao, Z.; Klabunde, K. J. J. Am. Chem. Soc.
1994, 116, 5493 and references therein. (b) Collins, T. J.; Fox, B. G.; H,
Z. G.; Kostka, L.; Mu¨nck, E.; Rickard, C. E. F.; Wright, L. J. J. Am. Chem.
Soc. 1992, 114, 8724. (c) Pasek, E. A.; Straub, D. K. Inorg. Chem. 1972,
11, 259. (d) Sellmann, D.; Geck, M.; Knoch, F.; Ritter, G.; Dengler, J. J.
Am. Chem. Soc. 1991, 113, 3819. (e) Bakshi, E. N.; Delfs, C. D.; Murray,
K. S.; Peters, B.; Homborg, H. Inorg. Chem. 1988, 27, 4318. (f) Martin,
R. L.; Rohde, N. M.; Robertson, G. B.; Taylor, D. J. Am. Chem. Soc. 1974,
96, 3647.
[FeIIICp*(η2-dtc)(L)]+PF6- + Na+dtc- + H2O f
[FeIVCp*(η2-dtc)2]+PF6- + Na+OH- + 1/2H2 + L (4)
Complex 9+PF6- can also be formed upon contact of an organic

4138 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
DelVille-Desbois et al.
Table 5. Significant Interatomic Distances (Å) in Complex
Table 2. Crystallographic Data (Graphite-Monochromatized Mo
KR Radiation, λ ) 0.710 13 Å) for the X-ray Structure
- a
16+PF6
-
Determinations of 11 and 16+PF6
Fe-Cp(centroid)
Fe-S1
1.799(2)
2.329(2)
2.318(2)
2.334(2)
2.318(2)
2.124(4)
2.186(4)
2.221(4)
Fe-C10
Fe-C11
S1-C1
S2-C1
S3-C4
S4-C4
N1-C1
N2-C4
2.190(4)
2.117(4)
1.694(5)
1.693(5)
1.710(5)
1.685(5)
1.309(6)
1.302(6)
-
11
16+PF6
Fe-S2
formula
mol wt
C14H21N2S2Fe
337.30
C46H47F6N2S4PFe
956.92
Fe-S3
Fe-S4
cryst size, mm3
cryst syst
space group
cell dimens
a, Å
0.20 × 0.13 × 0.13 0.50 × 0.20 × 0.20
Fe-C7
triclinic
monoclinic
C2/c
Fe-C8
P1h
Fe-C9
8.637(2)a
9.087(2)
11.593(1)
94.95(3)
91.06(3)
113.67(3)
828.8(3)
2
18.639(3)
21.757(4)
24.283(5)
90
112.23(2)
90
9115(6)
8
1.395
a Numbers in parentheses are estimated standard deviations in the
least significant digits.
b, Å
c, Å
R, deg
â, deg
Table 6. Significant Intramolecular Angles (deg) in Complex
- a
16+PF6
S1-Fe-S2
S1-Fe-S3
S2-Fe-S4
S3-Fe-S4
C1-S1-Fe
C1-S2-Fe
C4-S3-Fe
C4-s4-Fe
71.95(6)
122.45(6)
119.34(6)
72.19(5)
90.1(2)
90.5(2)
89.4(2)
S1-C1-S2
S3-C4-S4
N1-C1-S1
N1-C1-S2
N2-C4-S3
N2-C4-S4
C2-N1-C3
C5-N2-C6
107.4(3)
107.6(3)
125.6(4)
127.0(4)
125.5(4)
126.8(4)
118.6(5)
116.7(5)
γ, deg
V, Å3
Z
1.352
11.50
d
calcd, g cm3
6.09
1275
1254
297(1)
7640
6895
297(1)
585
µ(Mo KR), cm-1
no. of data collcd
90.6(2)
no. of unique data used 175
a Numbers in parentheses are estimated standard deviations in the
least significant digits.
temp, K
3.5-50
3-50
no. of params varied
2θ range, deg
abs corr
max/min transm
ext coef, øb
R1
SHELXA
0.9179/0.0637
0.003(3)
0.0495
SHELXA
0.9552/0.4706
none
much stronger nucleophile than the neutral water molecule and
that Fe is not very oxophilic, the first step involves the attack
0.0385
0.0961
0.1262
-
of dtc on 2+PF6 (or 3+PF6--8+PF6-) which must be fast
when L is a weak ligand (2+PF6--6+PF6-) and slow with
bulky ligands such as PPh3 (7+PF6-) or dppe (8+PF6-). The
overall mechanism is shown in eqs 5-8 (PF6- as counteranion).
wR2
a Numbers in parentheses are estimated standard deviations in the
least significant digits. b F* ) F[1 + 0.001øF2λ3/sin(2τ)]-1/4
.
[FeCp*(η2-dtc)(L)]
2+-8+, 17e
+ + dtc f
2
1
Table 3. Significant Interatomic Distances (Å) in Complex 11a
[FeCp*(η -dtc)(η -dtc)(L)]
15, 19e
Fe-Cp(centroid)
Fe-S1
1.709(6)
2.285(4)
2.241(4)
2.10(2)
2.07(2)
2.126(12)
2.13(3)
Fe-C10
S1-C6
S2-C6
N1-C6
N1-C8
N1-C9
N2-C10
1.89(4)
1.687(13)
1.739(13)
1.28(2)
1.44(2)
1.49(2)
1.22(6)
(5)
Fe-S2
2
1
2
1
a
+ L
[FeCp*(η -dtc)(η -dtc)(L)] [FeCp*(η -dtc)(η -dtc)]
Fe-C1
Fe-C2
15, 19e
9′, 17e
(6)
Fe-C3
Fe-C4
2
2
1
Fe-C5
2.09(2)
f
(7)
[FeCp*(η -dtc)2]
[FeCp*(η -dtc)(η -dtc)]
9′, 17e
a Numbers in parentheses are estimated standard deviations in the
least significant digits.
9, 19e
2
+ H O f
[FeCp*(η -dtc)2]
9, 19e
2
Table 4. Significant Intramolecular Angles (deg) in Complex 11a
S1-Fe-S2
S1-Fe-C10
S2-Fe-C10
Fe-S1-C6
Fe-S2-C6
C6-N1-C8
76.4(2)
95.6(9)
97.0(6)
86.8(5)
87.0(5)
120.4(12)
C6-N1-C9
C8-N1-C9
Fe-C10-N2
S1-C6-N1
S2-C6-N1
S1-C6-S2
121.0(11)
118.6(12)
175.(3)
125.1(9)
125.2(9)
109.7(8)
[FeCp*(η2-dtc)2]+
9+, 18e
+ Na+OH- + 1/2H2 (8)
Considering the 17e a 19e equilibrium of eq 9 (Vide infra), it
is reasonable to assume that water reduction is achieved by the
more electron-rich 19e chelate form 9. Inner-sphere reduction
via protonation of 9 could be envisaged by comparison with
the known mechanism of reduction of water by the 19e complex
[CoCp2].23
a Numbers in parentheses are estimated standard deviations in the
least significant digits.
-
solution of the 17e complexes 2+PF6--8+PF6 with a slight
excess of water. Clearly, the second dtc ligand must come from
Crystal Structure of the FeIV Complex. The details of the
-
decomposition of another molecule of 2+PF6--8+PF6 and,
-
X-ray crystal structure studies of 16+PF6 are summarized in
as expected, the yields are lower (30-40%) than if the dtc ligand
is added. Workup of the reaction mixtures results in the
isolation of inorganic FeIII hydroxide as well as pentamethyl-
cyclopentadiene.
Table 2, and significant bond distances and angles are given in
Tables 5 and 6, respectively. Complex 16+PF6-was synthesized
The intermediacy of complex [FeIIICp*(dtc)2] in the mech-
-
-
anism of conversion of 2+PF6 to 9+PF6 (eq 4) is indicated
by the reduction of H2O. Water is present in the reaction as a
dihydrate in Na+dtc-‚2H2O and serves as an acceptor to trap
the 19e species 9 or 15. Since the conversion 15 a 9 at the
19e level is expected to be very fast as compared to water
reduction, we propose that reduction of water is effected by 9
and not by 15. Considering that the anionic dtc ligand is a
in order to obtain a crystal structure of the first C5R5FeIV

[FeIII(η5-C5R5)(S2CNMe2)L]n+ Complexes
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4139
Scheme 4
Figure 3. ORTEP view of the X-ray structure of the 17e complex 11.
-
Figure 5. On top ORTEP view of 16+PF6 showing the clockwise
directionality of the benzyl groups.
-
16+PF6 are very close to the values reported for other FeIV
complexes regardless of the spin configuration of the metal.27
These data confirm the important contribution of form II to the
electronic structure. The four Fe-S distances are identical
within experimental error and are short enough to be sure of
the electron density between the two atoms. Unlike in other
piano-stool Cp** complexes,39 all five benzyl groups are
directed away from the metal center and the benzylic carbons
are all slightly above the mean Cp plane. This is undoubtedly
due to the large steric effect of both dtc ligands. The orientation
of the phenyl groups is clockwise as illustrated in Figure 5.
The bond lengths for the Cp** ring are almost identical (ca.
1.42 Å) and close to the values reported for other Cp**
complexes.39 The Fe-Cp centroid distance is 1.799 Å. The
bond angles in the Cp** ring do not deviate from 108°, and
consequently, the ring is essentially planar even though it is
tilted by 3.5° from the normal plane.
-
Figure 4. ORTEP view of the X-ray structure of complex 16+PF6
.
organometallic complex (R * Me). Several different attempts
were made to obtain good-quality crystals of 9+PF6-. Even
though these attempts were apparently successful, these samples
underwent a slow degradation during data collection and only
allowed the determination of basic parameters (this complex
crystallizes in a monoclinic system: a ) 12.27, b ) 17.51, c
) 13.15 Å; â ) 90.6°; V ) 2826 Å3; Z ) 4; d ) 1.40 g/cm3).
However, use of the pentabenzylcyclopentadienyl ligand (Cp**)39
permitted the acquisition of a satisfactory X-ray crystal structure.
17e and 19e States of Complex [FeIIICp*(dtc)2]. Spectro-
scopic Data. The 19e complex [FeIIICp*(η2-dtc)2] can be
synthesized either by addition of anhydrous Na+dtc- to a 17e
-
solvent complex [FeIIICp*(η2-dtc)(solvent)]+ such as 3+PF6
at -45 °C in CH3CN solution or by one-electron reduction of
-
the 18e cation [FeIVCp*(η2-dtc)2]+PF6 using [FeICp(C6Me6)]
-
Examination of the different metrical parameters of 16+PF6
as reducing reagent in THF solution at -80 °C (Scheme 4).
Both reactions afford a blue solution whose ESR spectrum
at 20 K in the solid state exhibits axial symmetry with two g
values (g ) 2.275, g| ) 2.029). This spectrum (Figure 6) is
shows that it follows the same trends as those of 11 and other
FeIII or FeIV dithiocarbamate complexes (Figures 3 and 4).8,13,27
The Fe-S, S-C, and C-N bond distances for complex

4140 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
DelVille-Desbois et al.
Figure 6. ESR spectra of 9 and 9′ at 20 K (solid-state sample).
indicative of a metal-based radical and contrasts with the
rhombic distortion characterized by three g values which are
recorded for all the 17e complexes 2+-13. When the solvent
is removed from this blue solution at -35 °C, a blue micro-
crystalline powder can be isolated. Transfer to the Mo¨ssbauer
cell in the drylab leads to the observation of a Mo¨ssbauer
quadrupole doublet with IS ) 0.54 mm s-1 vs. Fe and QS )
0.94 mm s-1 at 77 K (Figure 7a and Table 1). These parameters
are markedly different from those of the 17e family. Whereas
9 has been shown to be thermally stable for up to 2 weeks in
the solid state, it decomposes in MeCN solution above -40 °C
to give a purple complex whose ESR spectrum exhibits the
characteristic three g values of the 17e family (see Table 1 and
Figure 6): g1 ) 2.268, g2 ) 2.126, g3 ) 2.035.
Figure 7. Iron-57 Mo¨ssbauer spectra of 9 recorded at 77 K showing
the transformation of complex 9 to 9′ with time: (a) fresh sample; (b)
after 6 days at 300 K; (c) after 26 days at 300 K.
The Mo¨ssbauer spectrum of 9 in the solid state at room
temperature also shows the disappearance of the doublet of the
19e complex and growth of a new doublet whose parameters
correspond to those of the 17e family (see Table 1 and Figure
7c). The 19e complex is more stable (a) in nonpolar solvents
and (b) when it is separated from the Na+PF6- salt which formed
during the synthesis. For instance, extraction of this complex
with toluene at -80 °C leads to a solution of the 19e form which
is stable for several hours at this temperature. The {1H}13C
NMR spectra of both the 17e and 19e forms were obtained.
The symmetrical 19e complex, 9, exhibits three peaks at 59.71,
30.37, and 11.22 in d8-toluene solution (298 K) and 68.41, 26.34,
and 9.93 ppm in CD3CN solution vs. SiMe4 (253 K, 273 K) for
Cquat (Cp*), NMe2, and C5Me5 respectively. For the 17e form,
both the mono- and bidentate dtc ligands were observed: 68.91
(Cquat (Cp*)), 38.52 and 31.20 (NMe2), and 9.93 (C5Me5) in
CD3CN (253 or 273 K). In addition, satisfactory elemental
analysis was obtained. It is striking, however, that the stability
of both the 17e and 19e species is much lower in a polar solvent
such as MeCN which is also a good ligand. In this solvent,
neither the 17e nor the 19e form is stable at room temperature,
which necessitates a more complete investigation of their
interconversion in this solvent. However, if it is assumed that
the 13C relaxation times are similar for both species, the 13C
NMR spectrum shows a proportion of 19e/17e complexes of
0.5 at 253 K, which is indicative of the following thermody-
namic equilibrium at this temperature:
(9)
The absence of signals for coordinated MeCN means that either
the 19e species [FeCp*(η2-dtc)(η1-dtc)(NCMe)] is not present
in a detectable amount or that its interconversion with the 17e
species is faster than the NMR time scale. The 19e form is

[FeIII(η5-C5R5)(S2CNMe2)L]n+ Complexes
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4141
Chart 1
stabilized by the entropy effect of the dtc ligand, although it is
a pseudoheptacoordinate complex. The 17e form is itself
stabilized by pseudohexacoordination. It seems that both the
17e and 19e states are relatively stable in nonpolar solvents and
in the solid state in contrast to the behavior in MeCN in the
-
presence of Na+PF6 salt. A reasonable explanation for this
observation is the salt-induced disproportionation of the radicals,
an effect already documented for other organometallic radicals.
The mechanism of the salt-induced disproportionation involves
double ion-pair exchange driving the dismutation electron
transfer reaction.22
Figure 8. Qualitative d-type MO diagram of a 17e CpML3 complex
(left) and 19e CpML4 complex (right).
so straightforward.31 The two isomers depicted in eq 9 belong
to the d5 CpML4 and CpML3 general types, respectively. The
qualitative d-type level ordering and occupation of such
complexes are shown schematically in Figure 8. A formally
hexacoordinated CpML3 system exhibits the usual splitting of
three nonbonding levels below two antibonding ones,32 with a
hole in the nonbonding block. On the other hand, the formally
heptacoordinated CpML4 species presents the opposite splitting
of two below three,33 with one single electron located in the
antibonding block. For each system, it is of course not possible
to predict a priori which is the singly occupied level among the
group of three nonbonding (17e) or antibonding (19e) levels.
In addition, more complicated situations can arise from the
peculiar nature of the ligands. For example, the 19e situation
shown in Figure 8 can be different if one of the ligands possesses
a low-lying antibonding orbital which is capable of full or partial
Theoretical Study. In order to provide a better insight into
the electronic factors governing the equilibrium represented in
eq 9, we have carried out MS-SCF-XR calculations on the
17e and 19e models [CpFe(η1-S2CNH2)(η2-S2CNH2)] (A) and
[CpFe(η2-S2CNH2)2] (B). This type of calculation provides
reliable qualitative results. Thus, previous calculations on 19e
iron complexes28 have shown that they correctly reproduce the
level ordering and occupation, as well as the energy gaps, which
is not always the case for this type of electron-rich complex
when modeled with simple extended-Hu¨ckel MO calculations.29
The choices for the geometries of A and B (Chart 1), were based
on the experimental structures of related 18e and 17e compounds
(see the Experimental Section for the geometrical and compu-
tational details).
delocalization of the odd electron away from the metal.28a 30
,
This could, in fact, be the case for 9 with the π* LUMO of the
dtc ligand acting as an electron sink. Such a situation would
correspond to a formally 18e FeIV metal with a reduced ligand
system. On the other hand, the 17e MO diagram of Figure 8
can also be different if a ligand possesses some high-lying
nonbonding lone pairs which can pass above the nonbonding d
block. In such a case, the odd electron could be located in a
ligand-based rather than in a metallic orbital. This might be
the case of the 17e species reported here, owing to the rather
high energy of a sulfur lone pair. Such a situation would
formally correspond to an 18e FeII system with an electron
deficiency on the ligand.
The computed energies and charge distributions for the
highest occupied and lowest unoccupied levels of A are listed
(31) Lin, Z.; Hall, M. B. Inorg. Chem. 1992, 31, 2791.
(32) Albright, T. A.; Burdett, J. K.; Whangbo, M. H. Orbital Interactions
In Chemistry; Wiley-Interscience: New York, 1985.
(33) Kubacek, P.; Hoffmann, R.; Halvas, Z. Organometallics 1982, 1,
180.
(34) At the suggestion of a reviewer, we have explored the viability of
this hypothesis experimentally. The reaction of n-Bu3SnH with the purple
17-electron complex 9′ was carried out at -30 °C in THF solution and
yielded a mixture of Cp*H, Me2NC(S)H, and (Bu3Sn)2S. The most plausible
origin of Cp*H is that it arises from the decomposition of [FeCp*(dtc)H]
which, in turn, is formed via hydrogen atom abstraction from n-Bu3SnH
by the radical iron center. On the other hand, Me2NC(S)H and (n-Bu3-
Sn)2S are probably formed by attack of n-Bu3SnH or n-Bu3Sn at the S-C
bond of the dtc ligand. By means of independent experiments, it was
established that these compounds are not formed by the reaction of n-Bu3-
SnH with diamagnetic sources of dithiobarbamate. We therefore conclude
that the observed reactivity pattern is in accord with the theoretical and
spectroscopic arguments and that collectively these results imply that the
spin density is delocalized on both the iron and the sulfur atoms. However,
we also know that organometallic radicals can react with the part of the
molecule that possesses the lowest ground-state spin density.22a
The problem of the relative stabilities of the 17e/19e states
is directly related to the electronic structure of their parent 18e
species. The MO diagram of a stable 18e complex generally
exhibits a significant gap which separates its nonbonding (or
bonding) HOMO from its antibonding LUMO. This gap confers
stability on the complex. This stability is reduced if one
supplementary electron is added, since this electron is forced
to occupy an antibonding level. A simple way to cancel the
antibonding nature of the singly occupied HOMO of this 19e
system is via ligand dissociation. If one ligand is lost, an
antibonding level becomes nonbonding, and the situation
becomes that of a stable 18e system which has lost one electron.
Apparently, this 17e state appears more favorable than the 19e
one. However, it should be recognized that the effect of ligand
loss is not only the stabilization of the odd electron but also
the destabilization of two M-L σ-bonding electrons which
become the nonbonding pair of the dissociated ligand. Con-
sequently, the relative stability of the 17e/19e states is the result
of a delicate balance between bonding and antibonding effects.
This situation is somewhat related to the 3e bond problem,30
although it has been shown that such a relationship is not always
(28) (a) Lacoste, M.; Rabaaˆ, H.; Astruc, D.; Ardoin, N.; Varret, F.;
Saillard, J.-Y.; Le Beuze, A. J. Am. Chem. Soc. 1990, 112, 9548. (b)
Lacoste, M.; Rabaaˆ, H.; Astruc, D.; Ardoin, N.; Pre´cigoux, C.; Saillard,
J.-Y.; Le Beuze, A.; Courseille, C. Organometallics 1989, 8, 2233. (c)
Rabaaˆ, H.; Chraibi, M.; Saillard, J.-Y.; Le Beuze, A.; Astruc, D. J. Mol.
Struct. (Theochem.) 1991, 251, 57.
(29) As a matter of fact, standard extended-Hu¨ckel calculations on B
lead to a wrong ground-state electronic configuration, with the odd electron
located in a π*(dtc) combination (unpublished results).
(30) Tyler, D. R. Acc. Chem. Res. 1991, 24, 325.

4142 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
DelVille-Desbois et al.
Table 7. Energy and Charge Distribution of the Highest Occupied
and Lowest Unoccupied One-Electron Levels of
[CpFe(η1-S2CNH2)(η2-S2CNH2)] (A)
charge distribn (%)
level occ E (eV) Fe Cp η1-dtca η2-dtca [I + O]b
18a′′
26a′
25a′
17a′′
24a′
16a′′
15a′′
14a′′
23a′
13a′′
22a′
12a′′
0
0
0
0
1
2
2
2
2
2
2
2
-3.08
-4.50
-5.03
-5.02
-6.70
-6.88
-7.24
-7.31
-7.34
-7.89
-8.35
-8.55
01
02
53
54
17
40
60
09
55
07
25
15
00
00
14
14
01
03
08
02
01
02
06
07
56
01
12
00
44
28
07
48
20
22
25
38
14
66
06
18
13
10
13
15
07
39
22
11
29
31
15
15
25
19
12
26
17
30
22
29
a dtc ) S2CNH2. b I ) intersphere region (always < 2%); O )
outer-sphere extramolecular region.
Figure 10. Plots of some frontier orbitals of A (top) and B (bottom):
24a′, in the xz plane; 16a′′ in a plane parallel to xz and situated 0.5 au
away from Fe; 23a′, 18a′′, and 19a′′, in Fe(η2-dtc) plane.
Table 8. Energy and Charge Distribution of the Highest Occupied
and Lowest Unoccupied One-Electron Levels of
[CpFe(η2-S2CNH2)2] (B)
charge distribn (%)
level
occ
E (eV)
Fe
Cp
dtca
[I + O]b
20a′′
24a′
19a′′
23a′
18a′′
22a′
17a′′
16a′′
21a′
20a′
19a′
15a′′
18a′
0
0
0
0
1
2
2
2
2
2
2
2
2
-3.09
-3.36
-5.21
-5.30
-5.31
-7.05
-7.27
-7.66
-7.91
-8.04
-8.36
-8.92
-8.98
00
00
40
51
51
65
02
00
88
04
06
24
06
01
10
03
18
20
02
00
06
02
04
17
01
20
75
50
41
17
17
21
71
68
02
61
49
50
44
24
40
16
15
15
12
27
26
08
31
28
25
30
a dtc ) S2CNH2. b I ) intersphere region (always < 2%); O )
outer-sphere extramolecular region.
modify this result. In particular, a variation of the Fe-S-C
angle associated with the η1-dtc ligand, from the assumed value
of 120 to 113°, leads to the quasi-degeneracy of both configura-
tions and results in a larger metallic character for the 16a′′ level
(44%). On the other hand, an increase of this bond angle tends
to reinforce slightly the preference for the (24a′)2(16a′′)1 ground
state. From these results, it can be concluded that some
significant spin delocalization occurs on the ligands.
Figure 9. MO diagram of 17e and 19e models [CpFe(η1-S2CNH2)(η2-
S2CNH2)] (A) and [CpFe(η2-S2CNH2)2] (B). They have been rescaled
by placing their essencially nonbonding x2-y2 levels at the same energy.
The energies and charge distributions for the highest occupied
and lowest unoccupied levels of B are listed in Table 8. The
MO diagram for B is shown on the right-hand side of Figure 9.
As in the case of A, the nonbonding fully occupied metal and
sulfur blocks are overlapping. On the other hand, the three
antibonding d-type levels (18a′′, 23a′, and 19a′′, plotted in Figure
10) are fairly isolated (by 1.95 eV) with respect to the two lowest
levels derived from the π* dtc orbitals (24a′ and 20a′′), thus
indicating that the odd electron is forced to occupy a metallic
orbital. In fact, the (18a′′)1 and (23a′)1 configurations are
computed to be quasi-degenerate.31 A way of stabilizing a π*
dtc orbital, and perhaps to render it occupied in B, is to rotate
the NR2 group by 90° around the CN axis. Since this orbital is
somewhat C-N antibonding in the planar form, this antibonding
character is lost when the CS2 and NR2 planes are perpendicular.
The calculations were performed with the NH2 groups of B
oriented perpendicular to their corresponding CS2 groups, thus
maintaining the overall Cs symmetry. As a consequence, the
in Table 7. The MO diagram for A is shown on the left-hand
side of Figure 9. Although the expected three-below-two
splitting of the d-type orbitals is observed, the nonbonding metal
and sulfur blocks are overlapping in such a way that some
mixing between them occurs. The HOMO 24a′ is a ligand
orbital which is preponderantly located on the uncoordinated
sulfur atom (41%). The next occupied MO (16a′′) is more
metal-centered, but it also possesses some sulfur character (14%
on each sulfur of the monodentate dtc). The 24a′ and 16a′′
orbitals are plotted in Figure 10. They are very close in energy.
The ground-state configuration is computed to be (16a′′)2(24a′)1,
0.30 eV more stable than the (24a′)2(16a′′)1 configuration. Spin-
polarized calculations lead to a similar value (0.35 eV), with a
ground-state spin density of 60% on the uncoordinated sulfur
atom. At level of accuracy of the present calculations, this
energy difference lies at the limit of significance. We have
checked that reasonable structural changes do not significantly

[FeIII(η5-C5R5)(S2CNMe2)L]n+ Complexes
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4143
Scheme 5
electrochemical time scale. It is also suspected that a solvent
molecule rapidly (and reversibly) binds to the 17e metal center
to give the 19e state 10. Thus, we suggest that the two new
waves are due to the 17e monodentate form 9′ and the solvent
adduct 9*. An additional observation that supports the postulate
of a solvent adduct is the fact that anodic oxidation of the 17e
species 9′ gives a very reactive 16e species 9′+. The two ways
that 9′+ can recover the favorable 18e state are (i) by chelating
the monodentate dtc ligand, which gives 9+, and (ii) by binding
a solvent molecule to give 9*+. The cationic solvent complex
rapidly gives 9+ and the solvent molecule, but fast cathodic
reduction may compete, which leads to the observation of the
assigned R2/O3 wave of small intensity. Since the solvent
adduct is more electron rich than 9+/0 while 9′+/0 is less electron
rich than 9+/0, the wave R0/O1 positive of 9+/0 is attributed to
9′+/0, whereas R2/O3 negative of 9+/0 corresponds to that of
Figure 11. Cyclic voltammogram under continuous scan of 2.22 ×
-
n
10-5 M 2+PF6 in DMF solution (0.1 M Bu4NBF4, at Pt electrode,
scan rate ) 0.6 V/s, 20 °C): (a) switching potential beyond the second
electron transfer; (b) switching potential between both electron transfers.
the solvent adduct 9*+/0
.
π* dtc levels were found to be lower in energy but still of too
high an energy to be occupied (0.66 eV above the antibonding
d-block). On the basis of this result, one can conclude that B
is a genuine FeIII 19e complex, in full agreement with the
experimental data.
Although these conclusions can be inferred only from the
continuous scan of the first zone (left side, Figure 11b), they
are well confirmed by the continuous scan of the overall CV
shown in Figure 11a. Indeed, in passing from the 19e to the
20e state 9-, the dtc partial decoordination becomes very
exergonic and significantly more rapid (Vide infra). As
expected, the decrease of the wave R1/O2 of 9+/0 is now much
more dramatic, and the new waves R0/O1 and R2/O3 exhibit a
much higher intensity. It is therefore concluded that the partial
decoordination of one dtc ligand is slow but observable at the
neutral level 9 on the electrochemical time. On the basis of
experiments carried out at a variety of scan rates between 0.02
and 0.8 V s-1, it is estimated that the first-order rate constant
for the decoordination of 9 (19e) f 9′ (17e), k, is approximately
5 × 10-2 s-1 at 25 °C. This same process is fast at the anionic
20e level 9-. The observations of the second zone are fully
consistent with these conclusions inferred from the first zone.
The cathodic wave 9 f 9- is in fact almost totally chemically
irreversible. Only a weak shoulder can be observed for the
anodic oxidation 9- f 9 (O5). A very large wave is apparent
on the anodic side (O4) and corresponds to the oxidation of the
decoordinated 18e form 9′-. Also note that the cathodic wave
R3 has a significantly reduced intensity. This observation is
explained by the fast chelation of 9′ in the course of the full
CV scanning which involves oxidation to 9′+ (Vide supra). Thus,
the overall mechanism for the double monoelectronic process
involving partial decoordination of one dtc ligand and subse-
quent coordination of a solvent molecule to the unsaturated metal
center is consistent with a triple-square mechanism (Scheme
5).
-
Electrochemical Data. The voltammogram for 9+PF6
provides additional information concerning both the neutral
radical 9 and the anion 9- generated by cathodic reduction, with
particuler reference to the decoordination of a dtc ligand for
these species on the electrochemical time scale. Although the
CV looks somewhat complex at first sight, it is possible to obtain
substantial information by continuous cycling. Two regions can
be clearly distinguished: the left-hand side corresponds to redox
interconversions between the cationic and the neutral species
(derived from 9+/0), and the right-hand side relates to redox
interconversions between the neutral and the anionic species
(derived from 90/-).
Figure 11a shows the entire CV, while Figure 11b represents
the continuous scan of the left zone only (9+/0) with a switching
potential situated before the second reduction wave. The
comparison of the two figures emphasizes the observation of
the decoordination of one dtc ligand. In Figure 11b, the
Nernstian wave corresponding to 9+/0 is clearly apparent (R1/
O2). Upon continuous scan, the intensities of both the cathodic
and anodic sides of the wave decrease, thus indicating that the
couple 9+/0 slowly disappears from the proximity of the
electrode. Concomitant with the disappearance of 9+/0, two
waves increase upon continuous scanning: R0/O1 at a less
negative potential than 9+/0 and a weak shoulder R2 negative
of 9+/0. It is known from other spectroscopic methods and from
the absence of CV waves due to dtc (Vide supra) that no
decomposition occurs on such a rapid time scale. The only
rearrangement recorded is the decoordination of one dtc ligand
from 9 to give the 17e radical 9′, which is stable on the
Conclusions
(1) The extremely sensitive 17e complex [FeIIICp*(dtc)-
CO]+PF6-, 2+PF6-, previously reported in electrocatalytic

4144 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
DelVille-Desbois et al.
studies, has been synthesized from a neutral precursor [FeII-
calibrated with polystyrene. Samples were examined in solution (0.1
mm cells with NaCl windows), between NaCl disks in Nujol, or in
KBr pellets. 1H NMR spectra were recorded with a Brucker AC 200
(200 MHz) spectrometer. 13C NMR spectra were obtained in the pulsed
FT mode at 50.327 MHz with a Brucker AC 200 spectrometer. All
chemical shifts are reported in parts per million (δ, ppm) with reference
to the solvent or Me4Si. Electronic spectra (UV and visible) were
recorded at 5 and 25 °C with a Cary 219 spectrophotometer with 10 or
1 mm quartz cells. ESR spectra were recorded with a Brucker ER
200 ttX band spectrometer. Cyclic voltammetry data were recorded
with a PAR 273 potentiostat galvanostat. Care was taken in the CV
experiments to minimize the effects of solution resistance on the
measurements of peak potentials (the use of positive feedback iR
compensation and dilute solutions (≈10-3 mol/L) maintained currents
between 1 and 10 µA).25 The additional redox couple ferrocene/
ferrocenium was used when possible as a control for iR compensation;
∆Ep was 60 mV throughout the experiments. Thermodynamic
potentials were recorded with reference to an aqueous SCE in THF
Cp*(η2-dtc)CO] or [FeIICp*(η1-dtc)(CO)2] by one-electron
-
oxidation using [FeIIICp2]+PF6
. This complex has been
characterized by elemental analysis, ESR (rhombic distortion)
and Mo¨ssbauer spectroscopies, and by its one-electron reduction
to the isostructural FeII complex 2 using the electron-reservoir
complex [FeICp(C6Me6)].
(2) In 2+PF6-, the CO ligand is extremely labile and is
exchanged rapidly by a solvent molecule (MeCN, CH3COCH3,
CH2Cl2) at low temperature to give isoelectronic 17e complexes
in which exchange of the solvent by another ligand is facile.
Several stable 17e complexes have been synthesized in this way
or via ferrocenium oxidation of 1 or 2 and characterized using
various spectroscopic techniques and by X-ray crystallography.
The 17/18e complexes are new redox couples for which both
forms can be isolated using [FeIIICp2]+PF6- and [FeICp(C6Me6)]
as exclusive synthetic redox reagents.
n
(0.1 M Bu4NBF4). When necessary, the reference electrode was an
(3) Kinetic studies of ligand exchange in the very substitution
labile 17e complexes indicate that 19e intermediates or transition
states are involved in associative mechanisms. With dtc as the
incoming ligand, 19e species can be isolated and characterized.
This reaction ends up by one-electron oxidation of this 19e
Ag quasi-reference electrode (QRE). The silver wire was pretreated
by immersion in 10 M HNO3 for 5 min before use. The counter
electrode was platinum. The working electrode was a polished platinum
disk embedded in glass (7.85 × 10-3 cm2) treated first with 0.1 N HNO3
solution and then with a saturated ammonium iron hexahydrate sulfate
solution. The QRE potential was calibrated by adding the reference
couple (FcH/FcH+; FcH ) ferrocene). Thermodynamic potentials were
recorded with reference to SCE. Mo¨ssbauer spectra were recorded with
a 25 mCi 57Co source on Rh, using a symmetric triangular sweep mode.
Elemental analyses were performed by the Centre of Microanalyses of
the CNRS at Lyon-Villeurbanne, France.
-
complex 9 by H2O to give the 18e cation 9+PF6 and H2. A
novel series of complexes [FeCp*(η2-dtc)2n+, n ) 0 and 1
-
(9+PF6 and 9), is accessible via this route.
-
(4) The cation 9+PF6 constitutes the first example of an
FeIVCp* complex. This is one member of a series of robust
compounds which has been characterized by various spectro-
scopic methods including Mo¨ssbauer spectroscopy. The C5(CH2-
Ph)5 derivative has been characterized by X-ray crystallography.
(5) The neutral radical [FeIIICp*(η2-dtc)2], 9, represents the
first example of a 19e complex which has been generated either
by addition of dtc- to a 17e complex or by one-electron
reduction of the 18e cationic precursor 9+ (using [FeICp(C6-
Me6)]).
Kinetic Studies. The rate of substitution was determined by
monitoring the changes in the visible spectra at an appropriate
wavelength as a function of time. Isosbestic points were observed (see
text). Solutions were prepared at low temperature when required. An
adapted system to regulate the temperature was used. Reactions were
carried out under pseudo-first-order conditions with respect to the
solvent. Pseudo-first-order constants and half-life times were obtained
from the slope of a plot of ln(A - A∞) vs. time by least-squares analysis.
(6) Both the 17e and 19e states of the radical 9 have been
characterized by ESR, Mo¨ssbauer, and NMR spectroscopies and
cyclic voltammetry. This is the first example of the concomitant
observation of both the 17 and 19e states. MS-SCF-XR
calculations on model compounds for the 17e (A) and the 19e
(B) states of 9 indicate that the two highest occupied levels of
A are nonbonding with some significant sulfur admixture. On
the other hand, the electronic structure of form B corresponds
to that of a typical authentic 19e system, in which the odd
electron occupies an antibonding level of dominant metallic
character, with no participation of the π* dtc orbital.
(7) A triple-square mechanism involving the partial decoor-
dination of the dtc ligand in 90/- and the coordination of a
solvent molecule to the monodentate dtc complexes 9*+/0 takes
into account the different waves observed by continuous
scanning of either the first one-electron reduction zone or the
two monoelectronic zones. The 20e species 9- decoordinates
much more rapidly than the neutral form 9 but is still observable
on the electrochemical time scale.
X-ray Analyses. A summary of the crystal data collection param-
eters and structure refinement is given in Table 2. Suitable crystals of
-
11 and 16PF6 were mounted inside thin-walled glass capillaries and
sealed under argon and placed at 297(1) K on an Enraf-Nonius CAD-4
diffractometer using graphite-monochromated Mo KR radiation (λ )
0.710 73 Å) in the θ-2θ scan mode. The structures were solved using
direct methods and refined using full-matrix least-squares on F2 with
the SHELXTL PC 5.0 software package (Sheldrick, G. M. Siemens
Analytical Instruments, 1994). All non-hydrogen atoms were refined
with anisotropic thermal parameters. The methyl hydrogens in 16+PF6-
were placed in idealized positions after using a difference electron
density synthesis to set the initial torsion angle and refined as a riding-
rotating model with general isotropic thermal parameters. All other
hydrogen atoms were placed in idealized positions and refined as a
riding model with general isotropic thermal parameters. The PF6- anion
is disordered, and the positions for four of the fluorine atoms were
refined with unrestrained anisotropic displacement parameters over two
sites with occupation factors of 0.725(18) and 0.275(18). In 11, the
carbon and nitrogen atoms were subjected to “rigid bond” restraints;
i.e., the components of the anisotropic parameters in the direction of a
bond between two carbon or nitrogen atoms are restrained to be equal
within an effective standard deviation of 0.01. This proved to be
necessary in order to keep the anisotropic refinement stable.
Experimental Section
General Data. Reagent-grade tetrahydrofuran (THF), diethyl ether,
and pentane were predried over Na foil and distilled from sodium-
benzophenone ketyl under argon immediately prior to use. Acetonitrile
(CH3CN) was stirred under argon overnight over phosphorus pentoxide,
distilled from sodium carbonate, and stored under argon. Methylene
chloride (CH2Cl2) was distilled from calcium hydride just before use.
All other chemicals were used as received. All manipulations were
carried out using Schlenk techniques or in a nitrogen-filled Vacuum
Atmospheres drylab. Infrared spectra were recorded with a Perkin-
Elmer 1420 ratio recording infrared spectrophotometer which was
SCF-MS-Xr Calculations. The standard version of the (spin
restricted) density functional SCF-MS-XR method35 was used and
applied to the 17e and 19e models [CpFe(η1-S2CNH2)(η2-S2CNH2)]
(A) and [CpFe(η2-S2CNH2)2] (B). Some spin-polarized calculations
were also carried out on B.36 The atomic radii of the muffin-tin spheres
and the exchange scaling parameters R were taken from the tabulation
(35) Johnson, K. H. AdV. Quantum Chem. 1973, 7, 143.
(36) (a) Slater, J. C.; Johnson, K. H. Phys. ReV. B 1972, 5, 844. (b)
Slater, J. C.; Wod, J. H. Int. J. Quantum Chem. 1971, 45, 3.

[FeIII(η5-C5R5)(S2CNMe2)L]n+ Complexes
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4145
-
of Schwartz37 for heavy elements and from Slater38 for H. Those
relative to the extra molecular (outer-sphere) and the inner-sphere
regions are weighted-average values of the atomic ones. The muffin-
tin atomic spheres were at first adjusted assuming tangent spheres. In
order to provide a better description of the ring systems, the carbon
spheres were subsequently enlarged by 25% and an additional empty
sphere was located in the center of each ring.38 The maximum l values
in the partial wave expansion included in the calculations were l ) 2
for Fe and outer spheres, l ) 1 for S, N, and C spheres, and l ) 0 for
H spheres. SCF calculations were converged to better than (0.0001
Ry on each level. The geometries considered in the calculations were
derived from the experimental structures of related 18e and 17e
complexes: [FeCp*(η1-S2CNMe2)(CO)2], 1,13 [FeCp*(η2-S2CNMe2)-
(PPh3)], 7,13 [FeCp*(η2-S2CNMe2)(CN)], 11, and [Cp**Fe(η2-S2-
CNMe2)2]+, 16+PF6- (Cp** ) η5-C5(CH2Ph)5. Both compounds were
idealized to Cs symmetry (pseudo-C2V in the case of B). The real
symmetry plane of B is the one connecting the two dtc ligands. Model
A was found sterically hindered. Only the conformation shown in Chart
1, with the uncoordinated S-atom lying “down” was found reasonable
on steric grounds, in agreement with the X-ray structure of Cp*Fe(η1-
S2CNMe2)(CO)2.13 Unless specified in the text, the major bond
distances (Å) and angles (deg) are those listed below. Model A: Fe-C
) 2.11, Fe-S ) 2.28, (η2-S)-C ) 1.72, (η1-S)-C ) 1.76, C(η2-
dtc)-N ) 1.32, C(η1-dtc)-N ) 1.34, C-S ) 1.65; S-Fe-S ) 76,
S-C-S ) 110, Fe-(η1-S)-C ) 120, (S-C-S) ) 125 (η1-dtc) or
115 (η1-dtc). Model B: Fe-C ) 2.17, Fe-S ) 2.32, S-C ) 1.72,
C-N ) 1.32; S-Fe-S ) 72, S-C-S ) 106, Fe-S-C ) 91. In
both models the C-C, C-H, and N-H bond distances were (Å) 1.42,
1.09, and 1.01, respectively.
Complex 2+ could also be isolated as a BF4 salt, using the same
-
procedure with [FeCp2]+BF4 (86% yield). IR (KBr pellet; ν, cm-1):
2025, 1985 (CO); 1550 (CN); 1020 (CS). ESR (solid state sample, 12
K): g1 ) 2.512, g2 ) 2.075, g3 ) 1.986. Anal. Calcd for C14H21-
FeNOBS2F4: C, 39.46; H, 4.96; N, 3.28. Found: C, 39.75; H, 4.85;
Fe, 12.11; N, 2.92.
-
[FeCp*(η2-dtc)(CH3CN)]+PF6-, 3+PF6
. A 0.1 g sample of
-
complex 2+PF6 (0.206 mmol) was introduced under argon into a
Schlenk tube. A 10 mL volume of degassed CH3CN was added, and
the mixture was stirred under argon for 10 min at room temperature.
The green suspension rapidly gave a bright purple solution. It was
then filtered into another Schlenk tube. CH3CN was then concentrated,
and the complex was precipitated by addition of dry ether. The solid
was then filtered out under argon on a frit and washed with 3 × 10
-
mL portions of dry pentane resulting in 89 mg (87% yield) of 3+PF6
.
Mo¨ssbauer data (mm/s vs. Fe at RT, 77 K): IS 0.550, QS 0.620. IR
(KBr pellet; ν, cm-1): 1535 (CN); 1035 (CS). ESR (CH3CN, 20 K):
g1 ) 2.611, g2 ) 2.184, g3 ) 2.001; (solid state sample, 12 K) g1 )
2.672, g2 ) 2.007. Visible (CH3CN; λ, nm (ꢀ, L mol-1 cm-1), RT):
514 (1820). Cyclic voltammetry (CH3CN, 0.1 M nBu4NBF4, RT, V )
0.4 V s-1, Pt; E° (V vs. SCE)): 0.570 (∆E ) 60 mV; ia/ic ) 0.62);
+0.835 (∆E ) 120 mV; ic/ia ) 0.81; R ) 0.70; D ) 2.30 × 10-5 cm2
s-1; ks ) 1.3 × 10-2 ((0.1) cm s-1). Anal. Calcd for C15H24FeN2-
PS2F6: C, 36.23; H, 4.86; N, 5.63. Found: C, 36.48; H, 4.63; N, 5.08.
[FeCp*(η2-dtc)(THF)]+PF6-, 4+PF6-. A 0.1 g sample of complex
-
2+PF6 (0.206 mmol) was stirred under argon for 3 h, at room
temperature, in CH2Cl2/THF (3/1) solution. The mixture slowly
assumed the orange color of a clear solution. It was then filtered under
argon into another Schlenk tube. The solvent was evaporated, and an
orange-brown solid began to precipitate. The precipitation was
completed by the addition of dry ether. The resulting solid was then
filtered out under argon on a frit and washed with 3 × 10 mL portions
of dry pentane (95 mg; 87% yield). IR (THF; ν, cm-1): 1510 (CN);
1020 (CS). ESR (THF, 10 K): g1 ) 2.323, g2 ) 2.050, g3 ) 2.004;
(solid state sample, 8 K) g1 ) 2.313, g2 ) 2.037, g3 ) 1.990. Cyclic
voltammetry (THF, 0.1 M n-Bu4NBF4, RT, V ) 0.4 V/s, Pt; E° (V vs.
SCE)): -0.248 (∆E ) 75 mV; ia/ic ) 0.89; D ) 1.07 × 10-5 cm2/s);
-0.765 (∆E ) 80 mV; ia/ic ) 0.88; D ) 1.46 × 10-5 cm2/s).
-
Preparation. [FeCp*(η2-dtc)(CO)]+PF6-, 2+PF6
. A 0.764 g
sample of complex 1 (2.082 mmol) in 10 mL of THF was stirred with
0.689 g (2.082 mmol) of [FeCp2]+PF6- for 15 min, at room temperature,
under argon, in the dark. The red solution immediately turned orange
with formation of a green precipitate. This precipitate was filtered out
in air, washed with 3 × 20 mL portions of dry ether, and dried in
Vacuo (0.925 g; 95% yield). Ferrocene, 0.360 g (93% yield), was also
recovered. The green complex could not be crystallized because of its
decomposition in solution (Vide infra for more details). Careful
treatment and washings nevertheless provided a satisfactory sample
for elemental analysis. Mo¨ssbauer data (mm/s vs. Fe at RT (room
temperature), 77 K): IS 0.518, QS 0.686. IR (KBr pellet; ν, cm-1):
2030, 1985 (CO); 1550 (CN); 1015 (CS). ESR (solid state sample, 10
K): g1 ) 2.239, g2 ) 2.107, g3 ) 2.049; an evolution of the spectrum
with the temperature was observed with an orthorhombic-axial
[FeCp*(η2-dtc)(L)]+PF6 [L ) CH3COCH3 (5+PF6-), CH2Cl2
-
(6+PF6-)]. These complexes were formed by dissolving complex
-
2+PF6 in the appropriate solvent giving orange solutions.
5+PF6-: IR (CH3COCH3; ν, cm-1, RT); 1510 (CN); 1015 (CS). ESR
(CH3COCH3, 20 K): g1 ) 2.453, g2 ) 2.035, g3 ) 1.995. Visible
(CH3COCH3; λ, nm (ꢀ, L mol-1 cm-1), RT): 460 (1853).
6+PF6-: IR (CH2Cl2; ν, cm-1, RT): 1515 (CN); 1020 (CS). ESR
(CH2Cl2, 20 K): g1 ) 2.571, g2 ) 2.135, g3 ) 1.990. Visible (CH2-
Cl2; λ, nm (ꢀ, L mol-1 cm-1), RT): 446 (1792).
transition at 200 K giving two new g values g1 ) 2.237 and g2,3
)
2.071. Visible (CH3CN; λ, nm (ꢀ, L mol-1 cm-1); 5 °C): 584 (811);
452 (2450). Anal. Calcd for C14H21FeNOPS2F6: C, 34.72; H, 4.37;
Fe, 11.53; N, 2.89. Found: C, 34.19; H, 4.36; Fe, 12.11; N, 2.77.
[FeCp*(η2-dtc)(PPh3)]+PF6-, 7+PF6-. A 0.3 g sample of complex
-
2+PF6 (0.620 mmol) was stirred in 10 mL of either CH3COCH3 or
(37) (a) Schwartz, K. Phys. ReV. B 1972, 5, 2466; (b) Schwartz, K.
Theoret. Chim. Acta 1974, 34, 225.
(38) Slater, J. C. Int. J. Quantum Chem. 1973, 47, 533.
(39) (a) Roch, N.; Klemperer, W. G.; Johnson, K. H. Chem. ReV. Lett.
1973, 23, 149. (b) Weber, J.; Geoffroy, M.; Pe´nigault, E. J. Am. Chem.
Soc. 1978, 100, 3508. (c) Weber, J.; Goursot, A.; Pe´nigault, E.; Ammeter,
J. H. J. Am. Chem. Soc. 1982, 104, 1491.
(40) (a) Chambers, J. W.; Baskar, A. J.; Bott, S. G.; Atwood, J. L.;
Rausch, M. D. Organometallics 1986, 5, 1641. (b) Rausch, M. D.; Tsai,
W. M.; Chambers, J. W.; Rogers, R. D.; Alt, H. G. Organometallics 1989,
8, 816.
(41) Hamon, J.-R.; Astruc, D.; Michaud, P. J. Am. Chem. Soc. 1981,
103, 758.
(42) Brintzinger, H. H.; Palmer, G.; Sands, R. H. J. Am. Chem. Soc.
1966, 88, 623.
(43) Morrow, J. R.; Astruc, D. Bull. Soc. Chim. Fr. 1992, 129, 319.
(44) Michaud, P.; Astruc, D.; Ammeter, J. H. J. Am. Chem. Soc. 1982,
104, 3755.
(45) Astruc, D.; Mandon, D.; Madonik, A. M.; Michaud, P.; Ardoin,
N.; Varret, F. Organometallics 1990, 9, 2155.
CH3CN with 0.162 g (0.620 mmol) of triphenylphosphine for 30 min,
under argon, at room temperature. The resulting purple solution was
then filtered through a Celite column (1 cm height) and washed with
CH3CN. Addition of absolute alcohol was followed by a slow
recrystallization by evaporation of CH3CN at ambient temperature.
Purple microcrystalline powder 7+PF6- was isolated in 86% yield (351
mg). Mo¨ssbauer data (mm/s vs. Fe at RT, 77 K): IS 0.470, QS 0.659.
IR (KBr pellet; ν, cm-1): 1530 (CN); 1015 (CS). ESR (solid state
sample, 10 K): g1 ) 2.478, g2 ) 2.238, g3 ) 1.960; an evolution of
the spectrum with the temperature was observed with an orthorhombic-
axial transition at 100 K giving two new g values of g1 ) 2.467 and
n
g2,3 ) 2.023. Cyclic voltammetry (CH3CN, 0.1 M Bu4NBF4, RT, V
) 0.4 V/s, Pt; E° (V vs. SCE)): -0.52 (∆E ) 60 mV; ia/ic ) 1.0; D
) 2.37 × 10-5 cm2/s). Anal. Calcd for C31H36F6FeP2NS2: C, 51.82;
H, 5.05; Fe, 7.71; N, 1.95; P, 8.62. Found: C, 52.13; H, 5.15; Fe,
-
7.71; N, 1.93; P, 8.81. Complex 7+PF6 was also be obtained by
stoichiometric oxidation of complex 1 in the presence of 1 equiv of
-
PPh3 using 1 equiv of [FeCp2]+PF6 as oxidant.10a
(46) Lapinte, C.; Catheline, D.; Astruc, D. Organometallics 1984, 3, 817.
(47) For ESR data and studies of inorganic and organometallic com-
plexes, see ref 1c, Chapter 4, and the following: (a) Ammeter, J. H. J.
Magn. Reson. 1978, 30, 299. (b) Connelly, N. G.; Geiger, W. E.; AdV.
Organomet. Chem. 1984, 83, 1. (c) Rieger, P. H. In ref 1a, p 270. (d)
Baird, M. C. Chem. ReV. 1988, 88, 1217. (e) Kaim, W. Coord. Chem.
ReV. 1987, 76, 187.
The same procedure was also used to obtain an 80% yield of [FeCp*-
(η2-S2CNEt2)(PPh3)]+PF6-, 10+PF6-, starting from [FeCp*(η2-S2-
CNEt2)(CO)]+PF6- 10e and PPh3 or [FeCp*(η1-S2CNEt2)(CO)2], 1 equiv
of [FeCp2]+PF6-, and PPh3. Mo¨ssbauer data (mm/s vs. Fe at RT, 77
K): IS 0.462, QS 0.631. IR (KBr pellet; ν, cm-1): 1500 (CN); 1015

4146 J. Am. Chem. Soc., Vol. 118, No. 17, 1996
DelVille-Desbois et al.
-
(CS). ESR (solid state sample, 10 K): g1 ) 2.666, g2 ) 2.004, g3 )
1.941 with an orthorhombic-axial transition at 95 K (g1 ) 2.663; g2,3
) 1.977). Anal. Calcd for C33H40FeP2NS2: C, 53.08; H, 5.39; Fe,
7.47; N, 1.87; S, 8.58. Found: C, 53.03; H, 5.25; Fe, 7.36; N, 1.91;
Complex 9+PF6 was also be obtained by slow evolution of a
solution of either complex 2+PF6-, 7+PF6-, or 8+PF6- in either CH2-
Cl2, CH3COCH3, or CH3CN solution in the presence of an excess of
water. Neither the action of O2, prolonged visible photolysis, nor slight
warming gave the desired product. The reaction times were dependent
on the solvent but generally took between 12 and 24 h. Starting from
2+PF6-, the following yields were obtained: 32% (CH2Cl2), 37% (CH3-
COCH3), and 30% (CH3CN). A fraction insoluble in organic solvents
consisted of hydrated iron oxides FeIII(OxHy) ((0,0) < (x,y) < (3,3)) as
shown by IR spectroscopy (KBr pellet; ν, cm-1; RT: 3300 (OH); 1100
(M-OH)) and qualitative tests to determine the oxidation state of iron.26
ESR spectroscopy revealed an absorption centered at g ) 2.0728 (∆H
) 1543 G). C5Me5H was also recovered in the organic layer and
identified by comparison with an authentic sample.
S, 8.57.
-
[FeCp*(η2-dtc)(η1-dppe)]+PF6-, 8+PF6
. A 0.1 g sample of
-
complex 2+PF6 (0.206 mmol) in 10 mL of CH3CN was stirred with
0.082 g (0.206 mmol) of dppe ((diphenylphosphino)ethane) for 30 min,
under argon, at room temperature. The resulting purple solution was
then rapidly filtered through a Celite column (1 cm height), concen-
trated, and recrystallized after the addition of absolute alcohol to give
70 mg (40% yield) of complex 8+PF6-. Mo¨ssbauer data (mm/s vs. Fe
at RT, 77 K): IS 0.445, QS 0.665. IR (KBr pellet; ν, cm-1): 1530
(CN); 1015 (CS). ESR (solid state sample, 10 K): g1 ) 2.636, g2 )
2157, g3 ) 1.975 with an orthorhombic-axial transition at 110 K (g1
) 2.635; g2,3 ) 2.066). Anal. Calcd for C39H45FeNP3S2: C, 54.80;
H, 5.30; N, 1.65; Fe, 6.13. Found: C, 54.49; H, 5.27; N, 1.61; Fe,
6.00.
[FeCp*(η2-dtc)2], 9. The following two procedures were used:
Method A. The same procedure as that described for in 7 was used
for the reduction of a 0.200 g sample of [FeCp*(η2-dtc)2]+PF6-, 9+PF6-
,
by 1 equiv of [FeICp(C6Me6)] at -80 °C in THF solution. Extraction
with toluene at -30 °C gave a blue-ink solution already observed during
the synthesis of 9+PF6-. Further crystallization gave 0.048 g (31%)
of the blue complex 9. Mo¨ssbauer data (mm/s vs. Fe at RT, 77 K):
see Table 1. IR (KBr pellet; ν, cm-1): 1520 (CN); 1015 (CS). ESR
(solid state sample, 12 K): g1 ) 2.268, g2,3 ) 2.029. Anal. Calcd for
C16H27FeN2S4: C, 44.54; H, 6.31. Found: C, 45.09; H, 6.63. Similar
treatment of the residual solid gave 0.063 g (85%) of [FeCp-
[FeCp*(η2-dtc)(PPh3)], 7.13 A 0.200 g sample of the cationic
precursor 7+PF6- (0.278 mmol) in THF solution was stirred with 0.079
g of [FeICp(C6Me6)] (0.278 mmol) for 15 min, under argon. The green
solution of the FeI complex immediately turned deep red with formation
of a yellow precipitate of [FeCp(C6Me6)]+PF6-. The THF was then
removed in Vacuo and the crude product washed 3 times with degassed
pentane. Filtration into another Schlenk tube and slow evaporation of
the solvent gave 0.130 g (81%) of red-brown microcrystals; 0.093 g
-
(C6Me6)]+PF6 after elimination of black impurities.
-
(79%) of [FeCp(C6Me6)]+PF6 was also recovered after filtration
Method B. A 0.074 g (4.132 mmol) sample of Na(dtc)‚2H2O was
dried in Vacuo at 80 °C during 48 h, and 0.200 g (4.132 mmol) of
complex 2+PF6- was dissolved in CH3CN. The Na(dtc) salt was then
dissolved in the same solvent and added dropwise to the FeIII solution
at -40 °C. The bright-purple color of the acetonitrile complex
immediately turned blue, and microcrystals formed in the medium. The
solvent was then removed by filtration to eliminate the NaPF6 salt.
Further crystallization from CHCl3 was carried out and gave 0.134 g
(75%) of the desired complex 9. Different attempts at crystallization
failed to yield good-quality crystals for X-ray structure determination.
Study of the Decomposition of Complex [FeCp*(η2-dtc)(CO)]+-
PF6-. In various solvents (CH3CN, 5 °C; CH3COCH3, 5 °C; CH2Cl2,
25 °C), 10-4 mol/L solutions of the complex were prepared under argon
at the desired temperature and immediately examined and monitored
by UV visible spectroscopy. As explained in the General Data section
the plots ln(A - A∞) vs. time were obtained and gave pseudo-first-
order kinetics with half-life times of 5 min (CH3CN, 5 °C), 13.5 min
(CH3COCH3, 5 °C), and 57 min (CH2Cl2, 25 °C) without any shaking
through an alumina column and recrystallization. 1H NMR (C6D6, 20
°C): δ 7.00 (m, C6H5-, 15H); δ 2.27 (s, N(CH3)2, 6H); δ 1.54 (s,
CH3, 15H). 13C NMR (C6D6, 20 °C): δ 201.11 (CN); δ 138.61, 135.62,
135.27, 130.15 (C6H5-); δ 79.80 (C5(CH3)5); δ 37.10 (N(CH3)2); δ
10.30 (C5(CH3)5). Mo¨ssbauer data (mm/s vs. Fe at RT, 77 K): IS
0.520, QS 2.190. IR (KBr pellet; ν, cm-1): 1480 (CN); 1000 (CS).
Anal. Calcd for C31H36FeNPS2: C, 64.91; H, 6.33; Fe, 9.74; N, 2.44;
P, 5.40. Found: C, 64.79; H, 6.22; Fe, 9.21; N, 2.44; P, 5.53.
Oxidation of Complex 2 by 1 equiv of [FeCp2]+PF6-. A 0.062 g
sample of the chelate complex 2, [FeCp*(η2-dtc)(CO)] (0.183 mmol),
-
in THF solution was stirred under argon with 0.060 g of [FeCp2]+PF6
(0.183 mmol) for 30 min. The red-brown solution rapidly turned orange
with formation of a green precipitate. The same workup as that
-
described for 1 afforded 0.073 g (83% yield) of complex 2+PF6 and
0.029 g (85%) of ferrocene.
Reduction of 2+PF6- by 1 equiv of [FeICp(C6Me6)]. An identical
procedure to that described for 7 was used for the reduction of a 0.239
g sample of complex 2+PF6-. Recrystallization afforded 0.136 g (81%)
or stirring.
-
-
of complex 2 and 0.186 g (88.6%) of [FeCp(C6Me6)+PF6
.
Both
[FeCp*(η2
-dtc)(CN)], 11. A 0.3 g sample of 2+PF6 (0.620
complexes were identified by comparison with authentic samples.9
[FeCp*(η2-dtc)2]+PF6-, 9+PF6-. A 0.3 g sample of 2+PF6- (0.620
mmol) in 15 mL of oxygen-free CH2Cl2 (CH3COCH3 or CH3CN could
also be used) was stirred under argon, at 20 °C, with 0.111 g of Nadtc‚-
2H2O (0.620 mmol) for 1 h. The solution initially assumed the orange
color of complex 6+PF6-, followed by a blue-ink intermediate color
which finally disappeared and gave a bright breen solution. This
solution was washed three times with distilled water, dried over Na2-
mmol) in 15 mL of oxygen-free CH3CN (CH3COCH3 or CH2Cl2 could
also be used) was stirred under argon, at 20 °C, for 2 h. A 30 mg
amount of NaCN (0.620 mmol) was added and left to react for 12 h.
-
The solution initially assumed the purple color of complex 3+PF6
,
followed by the red-burgundy color characteristic of the desired
complex. The solvent was evaporated in Vacuo; the crude product was
washed with pentane and ether, respectively, and then extracted with
freshly distilled and degassed CHCl3. Filtration into another Schlenk
tube and concentration of the solvent gave 0.170 g (81%) of a purple
microcrystalline powder. Mo¨ssbauer data (mm/s vs. Fe at RT): 77 K,
IS 0.395, QS 0.884; RT, IS 0.330, QS 0.660. ESR (solid state sample,
-
SO4, and filtered, and complex 9+PF6 was recrystallized at -30 °C
overnight (0.254 g; 71%). The aqueous layer containing NaOH was
titrated with a 0.01 N solution of HCl according to the stoichiometry
of the reaction (69%). The evolution of dihydrogen was monitored
by gas chromatography as the reaction proceeded. 1H NMR (CD3CN,
20 °C): δ 3.228 (s, CH3, 12H); δ 1.279 (s, CH3, 15H). 13C NMR: δ
200.56 (CN); δ 106.04 (C5(CH3)5); δ 38.42 (N(CH3)2); δ 9.75 (C5-
(CH3)5). Mo¨ssbauer data (mm/s vs. Fe at RT, 77 K): IS 0.460, QS
0.220. IR (KBr pellet; ν, cm-1, RT): 1540 (CN); 1005 (CS). Cyclic
4 K): g1 ) 2.372, g2 ) 2.268, g3 ) 1.973. IR (mineral oil; ν, cm-1
,
RT): 2080 (CtN); 1510 (CN); 1015 (CS). Cyclic voltammetry (THF,
0.1 M nBu4NBF4, RT, V ) 0.4 V/s, Pt E° (V vs. SCE)): -0.842 (∆E°
) 77 mV; ia/ic ) 0.86; D ) 1.1 × 10-5 cm2/s; Epa ) +0.755 V; ic/ia
) 0). Anal. Calcd for C14H21FeN2S2: C, 49.85; H, 6.27; N, 8.30; S,
19.01. Found: C, 49.67; H, 6.22; N, 3.17; S, 19.19.
[FeCp*(η2-dtc)(SCN)], 13. A 0.3 g sample of 2+PF6- (0.620 mmol)
in 15 mL of oxygen-free CH3CN (CH33COCH3 or CH2Cl2 could also
be used) was stirred under argon, at 20 °C, for 2 h. A 46 mg amount
of NH4SCN (0.620 mmol) was added, and the mixture was left to react
for 12 h. The solution initially assumed the purple color of complex
3+PF6-, followed by the winy purple color of complex 13 in solution
and accompanied by formation of a purple precipitate. The solvent
was evaporated in Vacuo; the crude product was washed with pentane
and ether, respectively, and then extracted with freshly distilled and
n
voltammetry (DMF, 0.1 M Bu4NBF4, RT, V ) 400 mV/s, Pt; E° (V
vs. SCE)): -0.248 (∆E° ) 60 mV; ia/ic ) 0.9; D ) 1.07 × 10-5 cm2/
s; -0.765 (∆E° ) 70 mV; ia/ic ) 0.88; D ) 1.46 × 10-5 cm2/s).
Crystals suitable for X-ray structure determination were obtained, but
slow decomposition during data collection permitted only the deter-
mination of lattice parameters. This complex crystallizes in a mono-
clinic system: a ) 12.27, b ) 17.51, c ) 13.15 Å; â ) 90.6°; V )
2826 Å3; Z ) 4; d ) 1.40. Anal. Calcd for C16H27FeN2PF6S4: C,
33.33; H, 4.72; N, 4.85. Found: C, 33.59; H, 4.71; N, 4.65.

[FeIII(η5-C5R5)(S2CNMe2)L]n+ Complexes
J. Am. Chem. Soc., Vol. 118, No. 17, 1996 4147
degassed CHCl3. Filtration into another Schlenk tube and concentration
of the solution gave 0.183 g (80%) of a black-purple microcrystalline
powder. Mo¨ssbauer data (mm/s vs. Fe): 77 K, IS 0.460, QS 0.660;
RT, IS 0.389; QS 0.587. IR (mineral oil; ν, cm-1, RT): 2060 (SCN);
1520 (CN); 1015 (CS). ESR (solid-state sample, 3.5 K): g1 ) 2.483,
g2 ) 2.088, g3 ) 1.997. Cyclic voltammetry (THF, 0.1 M nBu4NBF4,
RT, V ) 0.4 V/s, Pt, E° (V vs. SCE)): 0.725 (∆E° ) 70 mV; ic/ia )
0.95; D ) 1.34 × 10-5 cm2/s); -0.740 (∆E° ) 90 mV; ia/ic ) 0.80;
1.00; D ) 1.32 × 10-5 cm2/s); -0.696 (∆E° ) 82 mV; ia/ic ) 0.85;
D ) 5.37 × 10-5 cm2/s). Anal. Calcd for C13H21FeNS2Cl: C, 45.03;
H, 6.10; N, 4.04; S, 18.49; Cl, 10.22. Found: C, 44.79; H, 6.07; N,
4.23; S, 18.49; Cl, 10.16.
Acknowledgment. We thank Dr. J. M. Grene`che (Universite´
du Mans) for kindly recording Mo¨ssbauer spectra, Dr. J. M.
Dance, E. Marquestaut, and B. Barbe (Universite´ Bordeaux I)
for ESR and NMR assistance, and the Institut Universitaire de
France, the CNRS, the Universities Bordeaux I, Paris VI, and
Rennes I, the Re´gion Aquitaine, and the U.S. National Science
Foundation (NSF) for financial support. R.D.C. thanks the NSF
for a Graduate Research Fellowship.
n
D ) 1.38 × 10-5 cm2/s). Cyclic voltammetry (CH2Cl2, 0.1 M Bu4-
NBF4, RT, V ) 0.4 V/s, Pt; E° (V vs. SCE)): 0.715 (∆E° ) 70 mV;
ic/ia ) 0.85; D ) 1.13 × 10-5 cm2/s; Epc ) -0.940, ia/ic ) 0). Anal.
Calcd for C14H21FeN2S3: C, 45.52; H, 5.73; N, 7.58. Found: C, 45.34;
H, 5.72; N, 7.64.
[FeCp*(η2-dtc)(Cl)], 14. A 0.2 g sample of 2+PF6- (0.410 mmol)
in 10 mL of oxygen-free CH3CN was stirred under argon, at 20 °C,
for 2 h, and 1 equiv of dry NMe4Cl (0.410 mmol) in CH3CN solution
was then added dropwise at -45 °C over a period of 15 min. The
solution was slowly warmed to 20 °C and stirred for one more hour at
this temperature. The color of the solution changed from bright-purple
to blue-ink. The solvent was evaporated in vacuo; the crude product
was washed with pentane and ether, respectively, and then extracted
with freshly distilled and degassed CHCl3. Filtration into another
Schlenk tube and concentration of the solvent gave 0.115 g (80%) of
black pyramidal crystals. IR (mineral oil; ν, cm-1, RT): 2060 (SCN);
1520 (CN); 1015 (CS). ESR (solid-state sample, 2.7 K): g1 ) 2.277,
Supporting Information Available: Text describing X-ray
procedures and Tables of X-ray parameters, positional and
thermal parameters, bond distances and angles, least-squares
planes, and anisotropic thermal parameters of the non-hydrogen
-
atoms for 11 and 16+PF6 (23 pages). This material is
contained in many libraries on microfiche, immediately follows
this article in the microfilm version of the journal, can be ordered
from ACS, and can be downloaded from the Internet; see any
current masthead page for ordering information and Internet
access instructions.
g2 ) 2.169, g3 ) 2.061; with a rhombic transition at 120 K, g1,2
2.191, g3 ) 2.065. Cyclic voltammetry (THF, 0.1 M Bu4NBF4, RT,
V ) 0.4 V/s, Pt; E° (V vs. SCE)): +0.562 (∆E° ) 65 mV; ic/ia )
)
n
JA953603X