Terminal FeI-N2 and FeII...H-C interactions supported by tris(phosphino)silyl ligands

Article abstract of DOI:10.1002/anie.200701188

(Chemical Equation Presented) Taking hold: Monoanionic tris(phosphino)silyl ligands ([SiP^R₃] = [(2-R₂PC₆H ₄)₃Si]^_) stabilize unusual five-coordinate Fe^I complexes. The silane [SiP^Ph₃]H reacts with mesityliron(II) to give [(SiP^Ph₃)Fe^IIMes], which displays a C-H agostic interaction. Treatment with HCl and reduction affords the terminally bonded Fe^I-N₂ complex [(SiP ^Ph₃)FeN₂] (see scheme; structures show only donor atoms of [SiP^Ph₃]), which yields hydrazine under protolytic conditions.

Full text of DOI:10.1002/anie.200701188

Communications
DOI: 10.1002/anie.200701188
Dinitrogen Complexes
I
II
ꢀ
ꢀ
Terminal Fe N2 and Fe ···H C Interactions Supported by
Tris(phosphino)silyl Ligands**
Neal P. Mankad, Matthew T. Whited, and Jonas C. Peters*
Dinitrogen complexes of iron have been the target of
numerous synthetic studies in recent years.[1] An unusual
and intriguing formal oxidation state to explore in such
systems is iron(I), especially in the context of a Chatt-type[2]
of the apical neutral N-atom donor of a tetradentate NP3
ligand by an anionic Si-atom donor achieves the intended
goal. Herein we describe our efforts to prepare the mono-
anionic [SiPR3] ligands ([SiPR3] = [(2-R2PC6H4)3Si]ꢀ, R = Ph
I
IV
FeI/IV N2 fixation cycle (i.e. Fe N2 + 3H!Fe N + NH3;
and iPr) and to exploit them in the preparation of five-
ꢀ
ꢁ
I
I
FeIV N + 3H!Fe N2 + NH3). Yandulov and Schrockꢀs
recent work has provided impetus for such studies by
establishing the synthetic viability of a conceptually related
MoIII/VI N2 fixation cycle.[3] Both iron and molybdenum are
present in the FeMo cofactor of MoFe nitrogenases,[4] and a
coordinate [(SiPR3)Fe N2] complexes.
ꢁ
ꢀ
ꢀ
The tris(phosphino)silane precursors [SiPR3]H (R = Ph (1)
and iPr (2)) are prepared by ortho lithiation of 2-R2PC6H4Br
reagents with n-butyllithium and subsequent addition of
1/3 equivalent of trichlorosilane. Heating a benzene solution
pseudotetrahedral FeIV N species, a hypothetical intermedi-
of 1 and mesityliron(II) to 658Cresults in Si H activation and
ꢁ
ꢀ
ate of an FeI/IV N2 fixation mechanism, has recently been
characterized.[5] Both Hollandꢀs group and our own have
reported examples of three- and four-coordinate N2 adducts
of iron in the formal + 1 oxidation state, but in neither case
were terminally bonded N2 species identified or isolated;
extrusion of mesitylene, producing red [(SiPPh3)FeMes] (3,
Mes = 2,4,6-trimethylphenyl; Scheme 1).[8] Complex 3 has an
ꢀ
ꢀ
dinuclear end-on Fe NN Fe species were inevitably
obtained.[6] For the Schrock tris(amido)amine Mo N2 sys-
ꢀ
tems, it proved necessary to use a high degree of steric bulk to
prevent bimetallic adduct formation before the N2 fixation
chemistry of interest could be exposed. We have attempted to
retrofit the {(PhBPR3)Fe} scaffolds ([PhBPR3] = [PhBP-
(CH2PR2)3]ꢀ) with bulky substituents at the phosphorus
atoms to prevent dinuclear reaction pathways but have thus
I
ꢀ
far been unable to obtain terminally bonded Fe N2 adducts
using [PhBPR3] ligands.
Recently we began to turn our attention to new deriva-
tives of the classic Sacconi-type NP3 ligands (e.g.
N(CH2CH2PR2)3) to achieve access to a terminally bonded
I
Fe N2 synthon.[7] While {FeII(N2)(H)}+ species are readily
ꢀ
accessible using ligands of these types,[7] N2 adducts of iron(I)
have yet to be obtained. We find, however, that replacement
[*] N. P. Mankad, M. T. Whited, Prof. J. C. Peters
Division of Chemistry and Chemical Engineering
California Institute of Technology
Pasadena, California 91125 (USA)
Fax: (+1)626-577-4088
Scheme 1.
E-mail: jpeters@caltech.edu
[**] This work was supported by the NIH (GM 070757) and BP (MC2
program). N.P.M. was supported by an NSF Graduate Research
Fellowship, and M.T.W. by a Moore Foundation Fellowship. Larry
Henling and Dr. Angelo DiBilio are acknowledged for crystallo-
graphic assistance and EPR assistance, and Prof. John Bercaw
assisted with a Toepler pump analysis.
interesting solid-state structure (Figure 1) and is nominally
six-coordinate; the [SiPPh3] ligand and the mesityl ipso-carbon
atom together occupy five of the coordination sites, while the
sixth site (trans to the silyl donor) is occupied by an agostic
ꢀ
C H interaction with a mesityl ortho-methyl group. Two
molecules are found in the asymmetric unit, and both feature
very short distances from the iron center to the carbon atom
participating in the agostic interaction (2.564(4) and
2.569(4) Š).[9] It is plausible to attribute the structure of 3 to
the presence of strongly trans-influencing aryl and silyl donors
that do not favor a trans configuration.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author. CCDC-
638765 through CCDC-638770 and CCDC-644269 contain the
supplementary crystallographic data for this paper. These data can
be obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
5768
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 5768 –5771

Angewandte
Chemie
[(SiPPh3)FeCl] (4; Scheme 1). Though reaction between 2
and mesityliron(II) does not produce a tractable reaction
mixture, light orange [(SiPiPr3)FeCl] (5) can be accessed by
addition of CH3MgCl to
a mixture of 2 and FeCl2
(Scheme 1).[11] The crystal structures of 4 and 5 feature iron
centers with trigonal-bipyramidal coordination geometries
(Figure 3), and solution magnetic data indicate triplet ground
Figure 1. Left: Solid-state structure of [(SiPPh3)FeMes] (3) (only one
molecule from the asymmetric unit shown, agostic hydrogen atom
shown in calculated position). Right: Solid-state structure of
[(SiPPh3)FeN2] (6). Thermal ellipsoids are set at 50% probability;
hydrogen atoms not involved in the agostic interaction are omitted for
clarity.
Figure 3. Solid-state structures of [(SiPPh3)FeCl] (4; left) and
[(SiPiPr3)FeCl] (5; right). Thermal ellipsoids are set at 50% probability;
hydrogen atoms are omitted for clarity.
Solution NMR spectra of 3 are also interesting (Figure 2).
The room-temperature 1H NMR spectrum (C6D6) of diamag-
netic 3 features a sharp resonance at d = 1.84 ppm, corre-
states for both (meff = 2.9 mB (4) and 3.3 mB (5)). The cyclic
voltammogram (CV) of 4 in THF features two reversible
redox events, an FeIII/II couple at E0 ’ = ꢀ0.40 V and an FeII/I
couple at E0 ’ = ꢀ2.10 V (vs. Fc+/Fc). The corresponding
redox events for 5 are shifted to more negative potentials
by approximately 300 mV each owing to the more strongly
electron-donating phosphine substituents. The FeII/I couple in
5 is irreversible at 500-mVsꢀ1 scan rates, presumably because
of the enhanced lability of the chloride ligand in the more
electron-rich system.
Na/Hg reduction of 3 in THF produces the target terminal
iron(I) dinitrogen adduct [(SiPPh3)FeN2] (6; Scheme 1). The
closely related dinitrogen complex [(SiPiPr3)FeN2] (7) is
synthesized by Na(C10H8) reduction of 5 (Scheme 1). Solution
magnetic data indicate S = 1/2 spin states for these iron(I)
complexes 6 (meff = 1.8 mB) and 7 (meff = 2.2 mB). Accordingly,
the X-band EPR spectrum of 6 at 4 K (2-MeTHF glass)
features an intense rhombic signal with g1 = 2.013, g2 = 2.051,
g3 = 2.187 that is coupled to three P atoms (A1 = 58.0, A2 =
55.0, A3 = 5.8 G). The solid-state structures of these red-
orange complexes (Figure 1) confirm their identities as
terminally bonded N2 adducts of iron(I). Complex 6 has an
Figure 2. 1H NMR spectra of [(SiPPh3)FeMes] (3) at different temper-
atures (*=residual solvent peak).
sponding to the mesityl para-methyl group, and two broad
resonances at d = 5.33 ppm and 0.58 ppm, corresponding to
the two ortho-methyl groups, which are inequivalent on the
ꢀ
NMR time scale because of restricted Fe Cipso rotation. The
1H NMR spectrum at ꢀ608C([D ]toluene) instead features
Fe N bond length of 1.819(2) Š and a short N N bond of
1.106(3) Š, consistent with the weak N2 activation implied by
ꢀ
ꢀ
8
three broad resonances for the ortho-methyl groups (d =
4.59 ppm, 1.10 ppm, and ꢀ0.24 ppm) in addition to the
sharp para-methyl singlet at d = 1.88 ppm, thus indicating
ꢀ
the relatively high-energy N N stretching vibrations
observed by IR spectroscopy (2041 cmꢀ1 for
6 and
ꢀ
that at low temperature even the exchange of C H bonds on
2008 cmꢀ1 for 7).[12] For comparison, tetrakis(phosphi-
ꢀ
the same methyl group is slower than the NMR time scale
ne)iron(0) dinitrogen complexes have N N vibrations rang-
ꢀ
owing to restricted C Crotation. Furthermore, the
ing from 1955 cmꢀ1 in [Fe(N2)(depe)2][13] to 2068 cmꢀ1 in
[Fe(N2)(dppe)2].[14] As expected, the N2 ligands in 6 and 7 are
in positions trans to the silyl donors. This structural feature is
unusual; transition-metal N2 complexes that also feature silyl
donors are rare,[15] especially where the ligands are trans-
disposed. Such an arrangement of ligands might be expected
to labilize the N2 ligand. Indeed, prolonged exposure of 6 to
vacuum results in loss of some of the N2-adduct species
31P{1H} NMR spectrum of 3 at ꢀ608Cresolves the inequiva-
lent phosphines into a doublet (2P) and a triplet (1P) with
2JPP = 11 Hz. While iron(II) species that exhibit C H agostic
ꢀ
interactions are known,[9] complexes of any transition metal in
ꢀ
which a C H agostic interaction is stable enough to freeze
ꢀ
rotation between the three C H positions of a methyl group
are quite rare.[10]
1
Protonolysis of ethereal solutions of 3 with HCl cleanly
produces mesitylene and paramagnetic, light orange
(determined by H NMR spectroscopy, Toepler pump anal-
ysis, and combustion analysis). The N2 ligand of 7 is similarly
Angew. Chem. Int. Ed. 2007, 46, 5768 –5771ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5769

Communications
labile. Hence, while each complex can be generated cleanly in
coordination of p-acidic ligands, whereas FeII accommodates
pure s donors, as in the thf-adduct complex 9.
solution and crystallographically characterized, neither pro-
vides satisfactory combustion analysis data upon rigorous
solvent removal.
Comparative examination of the solid-state structures of 6
and 7 provides insight as to why dinuclear complex formation
is observed for [{(PhBPiPr3)Fe}2(m-N2)] but not in the case of 6
and 7. Figure 4 shows space-filling models of the hypothetical
Previously characterized iron dinitrogen complexes have
been reported to release low yields (less than 15% per
Fe equivalent) of hydrazine and/or ammonia under strongly
protolytic conditions.[17] However, the addition of protic
reagents to either the previously reported diiron(I) systems
[{(PhBPiPr3)Fe}2(m-N2)] or the diiron(I) b-diketiminato com-
plex [LRFeNNFeLR][6a] did not lead to any detectable
production of either NH3 or N2H4. Therefore, our observation
of low yields of hydrazine upon addition of acids HX to 6
(17% per Fe equivalent for X = BF4; 7% per Fe equivalent
for X = Cl) represents a promising initial lead for the
{(SiPR3)Fe} systems. More interesting is the observation that
performing such protonations in the presence of CrX’2 as a
one-electron reductant increases yields of hydrazine signifi-
cantly (47% per Fe equivalent for X’ = Cl; 42% per Fe equi-
valent for X’ = Cp*; Cp* = C5Me5). The addition of similar
protic reagents to 8 rapidly and cleanly regenerates 6 with no
evidence for protonation at the N2 ligand. Analogous
conditions result in substantially lower yields of hydrazine
for 7 (9% per Fe equivalent) even in the presence of
[CrCp*2], presumably because the more reducing nature of
7 causes direct H+ reduction to H2 to vastly outcompete N2
reduction. Complex 7 is more basic than 6, and therefore
weaker acids that fail to react with 6 instead provide low
yields of hydrazine with 7 (e.g. 13% per Fe equivalent for
HX = [HNiPr2Et][BPh4]). One key to further advancing this
N2-reduction chemistry will be to control the delivery of
protons and electrons more carefully so that N2 reduction is
favored over H2 evolution.
Figure 4. Space-filling models of [{(PhBPiPr3)Fe}2(m-N2)] (left, one
{(PhBPiPr3)Fe} fragment omitted), [(SiPiPr3)FeN2] (7; center), and
[(SiPPh3)FeN2] (6; right).
{(PhBPiPr3)Fe N2} fragment, derived from the X-ray structure
ꢀ
of [{(PhBPiPr3)Fe}2(m-N2)] after stripping away one of the
{(PhBPiPr3)Fe} units, alongside its five-coordinate relative 7.
The N2 ligand of {(PhBPiPr3)Fe N2} extends well beyond the
ꢀ
protective pocket provided by the isopropyl substituents of
the phosphine donors, whereas even the b-N atom of the N2
ligand of 7 is nicely shrouded by the isopropyl substituents.
Energetically unfavorable interactions can be anticipated for
7 under the approach of another {(SiPiPr3)Fe} unit, thus
precluding formation of the dinuclear species. A similar
analysis holds for [(SiPPh3)FeN2] (6), where the terminal N2
ligand is buried even more deeply in the protective pocket
than in the case of 7 (Figure 4).
Received: March 17, 2007
Published online: June 28, 2007
Reversible one-electron reduction events at E0 ’ =
ꢀ1.93 V for 6 and E0 ’ = ꢀ2.72 V for 7 (vs. Fc+/Fc) are
observed by CV, along with irreversible oxidations at higher
potentials. Chemical reduction of 6 with an additional
equivalent of Na/Hg in the presence of [12]crown-4 affords
the dark purple formally iron(0) species assigned as
[(SiPPh3)FeN2][Na([12]c-4)2] (8) on the basis of a strong N2
Keywords: agostic interactions · dinitrogen · iron · phosphanes ·
Si ligands
.
[1] a) B. A. MacKay, M. D. Fryzuk, Chem. Rev. 2004, 104, 385;
b) J. C. Peters, M. P. Mehn, in Activation of Small Molecules
(Ed.: W. B. Tolman), Wiley-VCH, Weinheim, 2006, p. 81.
[2] J. Chatt, J. R. Dilworth, R. L. Richards, Chem. Rev. 1978, 78, 589.
[3] D. V. Yandulov, R. R. Schrock, Science 2003, 301, 76.
[4] a) L. C. Seefeldt, I. G. Dance, D. R. Dean, Biochemistry 2004, 43,
1401; b) O. Einsle, F. A. Tezcan, S. L. A. Andrade, B. Schmid, M.
Yoshida, J. B. Howard, D. C. Rees, Science 2002, 297, 1696.
[5] a) T. A. Betley, J. C. Peters, J. Am. Chem. Soc. 2004, 126, 6252;
b) M. P. Hendrich, W. Gunderson, R. K. Behan, M. T. Green,
M. P. Mehn, T. A. Betley, C. C. Lu, J. C. Peters, Proc. Natl. Acad.
Sci. USA 2006, 103, 17107.
1
vibration at 1967 cmꢀ1 in its IR spectrum, its diamagnetic H
and 31P{1H} (d = 84.3 ppm) NMR spectra, and combustion
analysis (Scheme 1). On the other hand, one-electron oxida-
tion of 6 with [FeCp2][BArF ] releases N2 and produces the
4
high-spin solvento species [(SiPPh3)Fe(thf)][BArF ] (9; ArF =
4
3,5-(F3C)2C6H4).[16] Note that anion 8 features a less labile N2
ligand than its neutral precursor 6 (see above) owing to
stronger backbonding from the more electron-rich Fe center,
and it therefore gives satisfactory combustion analysis data.
The N2 ligand in 6 can be displaced by CO (1 atm) to
provide [(SiPPh3)Fe(CO)] (10; n˜(CO) = 1881 cmꢀ1). The CO
stretching frequency for 10 is virtually identical to that for the
structurally related but cationic FeI carbonyl complex [{N(2-
diisopropylphosphino-4-methylphenyl)3}Fe(CO)][BPh4],
where isopropyl rather than phenyl substituents decorate the
phosphine donors.[7a] No reaction, however, is observed
between 6 and excess PMe3 or NH3. The FeI site trans to the
silyl ligand therefore appears preferentially disposed to
[6] a) J. M. Smith, A. R. Sadique, T. R. Cundari, K. R. Rodgers, G.
Lukat-Rodgers, R. J. Lachicotte, C. J. Flaschenriem, J. Vela, P. L.
Holland, J. Am. Chem. Soc. 2006, 128, 756; b) T. A. Betley, J. C.
Peters, J. Am. Chem. Soc. 2003, 125, 10782.
[7] a) C. E. MacBeth, S. B. Harkins, J. C. Peters, Can. J. Chem. 2005,
83, 332; b) P. Stoppioni, F. Mani, L. Sacconi, Inorg. Chim. Acta
1974, 11, 227.
[8] Rh complexes of the related ligand [(R3PCH2CH2)3Si]ꢀ have
been reported: F. L. Joslin, S. R. Stobart, J. Chem. Soc. Chem.
Commun. 1989, 504.
5770
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 5768 –5771

Angewandte
Chemie
[9] Similar Fe···Cdistances have been observed in other systems for
[12] In the X-ray structure of 7, a Cl ligand is present in place of the
ꢀ
ꢀ
N2 ligand with approximately 3% occupancy. As a result, Fe N
which C H agostic interactions have been assigned. Recent
ꢀ
and N N bond lengths are not discussed here.
examples: a) S. W. Kohl, F. W. Heinemann, M. Hummert, W.
Bauer, A. Grohmann, Chem. Eur. J. 2006, 12, 4313; b) N. A.
Eckert, J. M. Smith, R. J. Lachicotte, P. L. Holland, Inorg. Chem.
2004, 43, 3306; c) K. Rachlewicz, S.-L. Wang, C.-H. Peng, C.-H.
Hung, L. Latos-Grazynski, Inorg. Chem. 2003, 42, 7348.
[13] S. Komiya, M. Akita, A. Yoza, N. Kasuga, A. Fukuoka, Y. Kai, J.
Chem. Soc. Chem. Commun. 1993, 787.
[14] R. A. Cable, M. Green, R. E. Mackenzie, P. L. Timms, T. W.
Turney, J. Chem. Soc. Chem. Commun. 1976, 270.
[15] a) R. J. Trovitch, E. Lobkovsky, P. J. Chirik, Inorg. Chem. 2006,
45, 7252; b) V. K. Dioumaev, K. Plossl, P. J. Carroll, D. H. Berry,
J. Am. Chem. Soc. 1999, 121, 8391; c) D. G. Gusev, F. G.
Fontaine, A. J. Lough, D. Zargarian, Angew. Chem. 2003, 115,
226; Angew. Chem. Int. Ed. 2003, 42, 216.
[10] Selected examples: a) M. D. Leatherman, S. A. Svejda, L. K.
Johnson, M. Brookhart, J. Am. Chem. Soc. 2003, 125, 3068;
b) F. M. Conroy-Lewis, L. Mole, A. D. Redhouse, S. A. Litser,
J. L. Spencer, J. Chem. Soc. Chem. Commun. 1991, 1601.
[11] Precomplexation of 2 with FeCl2 is observed upon mixing, and
the [L2FeCl2] (L2 = k2-[SiPiPr3]H) intermediate has been charac-
terized by X-ray crystallography (See Supporting Information).
Though no reaction is observed between 1 and FeCl2, adding
CH3MgCl to the mixture does produce 4 in modest yields.
[16] The related complex [(SiPPh3)Fe(thf)]BPh4 has been identified
by X-ray crystallography. See the Supporting Information.
[17] a) J. D. Gilbertson, N. K. Szymczak, D. R. Tyler, J. Am. Chem.
Soc. 2005, 127, 10184; b) T. A. George, D. J. Rose, Y. Chang, Q.
Chen, J. Zubieta, Inorg. Chem. 1995, 34, 1295; c) G. J. Leigh,
Acc. Chem. Res. 1992, 25, 177.
Angew. Chem. Int. Ed. 2007, 46, 5768 –5771ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5771