Broadening the scope of ancillary phosphane-type ligands: A systematic comparison of PR3, PR2BH3-, and SiR3- and their chalcogen derivatives

Article abstract of DOI:10.1002/ejic.200601207

This work describes a systematic experimental and theoretical study of the properties of two series of isoelectronic and largely isosteric ligands, namely PPh₂Me, PPh₂BH₃^-, and SiPh ₂Me^- and SPtBu₃, SPrBu₂BH ₃^-, and SSitBu₃^-. In addition, we have also investigated the oxo derivatives OPPh₂BH₃ ^- and OSiPh₂Me^-. Based on X-ray crystal structure determinations (Fe-CO and C-O bond lengths) as well as NMR [e.g. δ(¹³CO)] and IR [ν(CO)] spectroscopic investigations of the corresponding [CpFe(CO)₂]⁺ complexes, we can conclude that, with respect to electron donor strength, phosphanyl borohydrides occupy an intermediate position between phosphanes (weakest donors) and silyl ligands (strongest donors). The same is true for the thio derivatives, although the differences are smaller. In the reaction with [CpFe(CO)₂] ⁺, the oxo derivative OPPh₂BH₃^- transfers a hydride ion rather than forming a stable [CpFe(CO) ₂(OPPh₂BH₃)] complex. The tendency to undergo hydride-transfer reactions was studied by density functional calculations for the series PtBu₂BH₃^-, OPtBu₂BH ₃, and SPtBu₂BH₃^-. The results reveal that OPtBu₂BH₃^- is the strongest and SPtBu₂BH₃^- the weakest hydride donor, in accordance with the experimental observations. Theoretical analysis indicates that the three derivatives PPh₂Me, PPh₂BH₃ ^_, and SiPh₂Me^- are truly isolobal species despite variations in their charge distributions. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejic.200601207

FULL PAPER
DOI: 10.1002/ejic.200601207
Broadening the Scope of Ancillary Phosphane-Type Ligands: A Systematic
–
–
Comparison of PR3, PR2BH3 , and SiR3 and Their Chalcogen Derivatives
Theresa I. Kückmann,[a] Franz Dornhaus,[a] Michael Bolte,[a] Hans-Wolfram Lerner,[a]
Max C. Holthausen,*[a] and Matthias Wagner*[a]
Keywords: Isoelectronic analogs / P ligands / Si ligands / Ligand design / Boron
This work describes a systematic experimental and theoreti- ences are smaller. In the reaction with [CpFe(CO)2]+, the oxo
–
cal study of the properties of two series of isoelectronic and
largely isosteric ligands, namely PPh2Me, PPh2BH3–, and
SiPh2Me– and SPtBu3, SPtBu2BH3–, and SSitBu3–. In addition,
derivative OPPh2BH3 transfers a hydride ion rather than
forming a stable [CpFe(CO)2(OPPh2BH3)] complex. The ten-
dency to undergo hydride-transfer reactions was studied by
density functional calculations for the series PtBu2BH3
–
–
we have also investigated the oxo derivatives OPPh2BH3
,
–
–
and OSiPh2Me–. Based on X-ray crystal structure determi-
nations (Fe–CO and C–O bond lengths) as well as NMR [e.g.
δ(13CO)] and IR [ν(CO)] spectroscopic investigations of the
corresponding [CpFe(CO)2]+ complexes, we can conclude
that, with respect to electron donor strength, phosphanyl
borohydrides occupy an intermediate position between phos-
phanes (weakest donors) and silyl ligands (strongest donors).
The same is true for the thio derivatives, although the differ-
OPtBu2BH3
,
and SPtBu2BH3
.
The results reveal that
OPtBu2BH3– is the strongest and SPtBu2BH3– the weakest hy-
dride donor, in accordance with the experimental observa-
tions. Theoretical analysis indicates that the three derivatives
PPh2Me, PPh2BH3–, and SiPh2Me– are truly isolobal species
despite variations in their charge distributions.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
broadening our focus to include negatively charged boron-
based P-donor ligands and their isoelectronic silyl analogs.
The purpose of this paper is to reveal the underlying rela-
tionship between two representatives of these ligand classes,
Introduction
The performance of a homogeneous catalyst depends
critically on the steric and electronic properties of its
ligands.[1,2] Intense research has therefore been devoted to
the development of donor molecules specifically designed
for the assembly of a particular catalyst with a well-defined
purpose. As a result, homogeneous catalysis with transi-
tion-metal complexes has enjoyed enormous success, and
catalyst research continues to gain momentum.[3] A wealth
of information has been gathered on numerous individual
ligand classes, which now must be compared in order to
establish their relationship to one another. The goal is to
create versatile toolboxes of ligands with smoothly varying
donor properties so that the most suitable derivative can
easily be selected for the rational de novo design of efficient
catalysts.
–
namely phosphanyl borohydrides PR2BH3 and triorganyl-
–
silanides SiR3 , and to compare them with the well estab-
lished phosphane ligands PR3.
A generic methyldiorganylphosphane A is shown in Fig-
ure 1. On moving to the phosphanyl borohydride B, the
phosphane methyl group is replaced by a BH3 moiety. This
leaves the central atom unchanged but results in a nega-
tively charged ligand. The substitution pattern is conserved
in the corresponding silyl compound C but the central
phosphorus atom is replaced by silicon, which again intro-
duces a negative charge. The same relationship holds true
for the chalcogen derivatives of these compounds (D–F;
Figure 1), which have been the subject of two recent stud-
ies.[11,12] We are interested in investigating the similarities as
well as the fundamental differences between these three re-
lated species. We commence with a brief review of selected
literature relevant to this pursuit.
Phosphanes are ubiquitous ligands in coordination
chemistry due to their σ-donating and π-accepting proper-
ties.[13,14] The degree to which these σ and π components
influence the overall coordination characteristics of a given
phosphane depends on the phosphorus-bound substituents.
The π-acceptor strength of PF3, for example, is comparable
to that of CO, which is often considered the archetypical π-
We have a long-standing interest in the development of
boron-based N-donor ligand systems [e.g. mono- and di-
topic bis- and tris(pyrazol-1-yl)borates] for applications
both in materials science and catalysis.[4–10] In order to ex-
tend the range of accessible complexes, we are presently
[a] Institut für Anorganische Chemie, Johann Wolfgang Goethe-
Universität Frankfurt am Main,
Max-von-Laue-Straße 7, 60438 Frankfurt, Germany
Fax: +49-69-798-29260
E-mail: Matthias.Wagner@chemie.uni-frankfurt.de
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
Eur. J. Inorg. Chem. 2007, 1989–2003
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1989

M. C. Holthausen, M. Wagner et al.
FULL PAPER
While not as common as the parent phosphanes, phos-
phane oxides also play an important role as ancillary li-
gands in coordination chemistry,[13] especially in the form
of bidentate bisphosphane monooxides [P(O)–P] or P(O)–
O and P(O)–N type ligands, which form part of efficient
catalytic systems for rhodium- and cobalt-catalyzed hydro-
formylation reactions.[31–33]
Transition metal siloxide complexes have also been inves-
tigated.[34–38] For example, rhodium siloxides are active in
catalytic hydrosilylation reactions,[34,39] and titanium, ni-
obium, and tantalum siloxide systems have been used to
investigate C–H activation[37] and olefin-to-alkylidene re-
arrangement.[38]
The corresponding chalcogen derivatives of phosphanyl
Figure 1. Isoelectronic phosphanes, phosphanyl borohydrides, and
silanides, and their chalcogen derivatives.
–
borohydrides, EPR2BH3 (E = O–Te), have only recently
been reported in the literature,[12] and their application as
ligands for transition metals has yet to be explored.
In order to shed more light on the relative importance of
donor atom and electronic structure for the ligand proper-
acceptor ligand. Trialkylphosphanes, on the other hand, are ties of phosphanes, phosphanyl borohydrides, and silanides,
primarily σ-donors and participate to only a small degree we have investigated two series of isoelectronic, isosteric
in π-backbonding.
compounds (Figure 2). In the case of the non-chalcogen
Transition metal complexes with silyl ligands, while not species, the diphenyl derivatives 1, 2a, and 3a were chosen
as widespread as phosphane complexes, have also received for detailed study. In order to still be able to evaluate the
–
significant attention in the literature.[15,16] Such studies have effect of increased steric demand, PtBu2BH3 (2b) was also
focused primarily on the reactivity of the Si–M bond, for investigated. For the chalcogen derivatives, both the phenyl
example in transition-metal-catalyzed hydrosilylation or C– and tert-butyl systems 4–8 were studied.
Si bond formation.[15,17] Recently, silyl derivatives have also
generated increasing interest as ancillary ligands.[18] Indeed,
it has been suggested that silyl ligands might enhance the
performance of soluble catalytic systems for hydrogenation
or hydroformylation reactions.[19] Silyl ligands are strongly
electron-releasing with comparatively poor π-accepting
properties and display a pronounced trans effect, which
makes them potentially complementary ligands to phos-
phanes.[18,20–22] It is clear, however, that the properties of
formally isoelectronic silyl and phosphane ligands differ sig-
nificantly and that this difference is too large to be consid-
ered a gradual transition.
Phosphanyl borohydrides PR2BH3– are intermediate spe-
cies in which the donor phosphorus atom remains as in
phosphanes, but which contain a negative charge, as do silyl
ligands. To date, these anionic phosphane analogues have
mainly been generated in situ and used as building blocks
Figure 2. Ligand numbering scheme for isoelectronic species 1–3
and 4–8.
for the preparation of chiral organophosphanes.[23–26] How-
ever, there is a growing number of phosphanyl borohydride
transition metal complexes in the literature. Manners et al.,
For the transition metal site, we chose the cyclopen-
for example, have exploited the regioselective insertion of tadienidoiron dicarbonyl fragment [CpFe(CO)2]+ (in future
the Pt(PEt3)2 fragment into the P–H bond of PPhBH3(H)2 referred to as Fp+) as a model system. This fragment is an
or PPh2BH3(H) in order to prepare the platinum phos- ideal candidate for our pursuit because it has a single vac-
phanyl borohydride complexes trans-[PtH(PPhRBH3)- ant coordination site (often occupied by halides or donor
(PEt3)2] (R = H, Ph).[27] Other examples of structurally solvent molecules in its isolable derivatives) and all of the
characterized phosphanyl borohydride complexes include ligands in question can be expected to coordinate to it in a
the compounds [(dppp)Pd(C6F5)(PPh2BH3)][28] and similar fashion, regardless of whether the result is a neutral
[(C5Me5)Fe(CO)2(PPh2BH3)].[29] In a 1996 paper, Fu and or cationic complex. In addition, its electronic properties
co-workers described diphenylphosphidoboratabenzene, can be related to the carbonyl IR stretching frequencies[40]
PPh2(BC5H5)–, as the anionic analog of triphenylphos- as well as to the 13C NMR shifts of the carbonyl[41] and
phane and investigated its coordination chemistry.[30]
cyclopentadienido[42] carbon atoms.
1990
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 1989–2003

Broadening the Scope of Ancillary Phosphane-Type Ligands
FULL PAPER
Terminal, unbridged phosphane complexes of the Fp+ solution than in the solid state. When the solid is stored
fragment are well known in the literature, with one of two under inert atmosphere and protected from light, the com-
synthetic routes generally being applied. The first involves pound is stable for several months. X-ray quality crystals
treating Fp+ complexes containing labile ligands such as thf can be obtained by gas-phase diffusion of diethyl ether into
or I– with phosphanes.[43–46] In the second, Fp2 is oxidized an acetonitrile solution of the complex.
in the presence of phosphane ligand, typically with ferro-
cenium or cobaltocenium ions.[47–49] Using a different ap-
proach, Malischhassucceededinisolating[(C5Me5)Fe(CO)2-
(PPh2Me)]I, which was prepared by reaction of the corre-
sponding diphenylphosphido complex with MeI and char-
acterized spectroscopically.[29]
To the best of our knowledge, no structurally charac-
terized monomeric FpSiR3 complexes in which the silyl li-
gand bears only alkyl and/or aryl substituents R have been
previously reported. Nevertheless, there are a number of si-
lyl complexes with known molecular structures where the
silicon atom is bound to electron-withdrawing functional
groups.[50–53] Complexes of Fp+ bearing silyl ligands with
solely aryl and/or alkyl substituents have, however, been
spectroscopically characterized and are typically obtained
by treating NaFp with a triorganylsilyl chloride.[54–57]
Scheme 1. Complexes of a phosphane and its isoelectronic ana-
logues, 11I–13.
The σ-donor properties of the phosphanyl borohydride
2a (Figure 2) towards selected main group Lewis acids have
already been explored.[58] It was found that 2a forms
stronger Lewis acid–base adducts than the related phos-
Preparation of the phosphanyl borohydride homologue
phane PPh2Me (1). Transition metal complexes of phos- 12a in an analogous reaction using BH3·thf instead of MeI
phanyl borohydrides are rare in the literature, with only a has been reported.[59] However, the general utility of phos-
handful of structurally characterized complexes {e.g. phanyl borohydrides as ligands can best be achieved if it is
[FpPPh2(BC5H5)] and [(C5Me5)Fe(CO)2(PPh2BH3)]}.[27–30] possible to form metal complexes by introducing the pre-
The complex [FpPPh2BH3] was prepared from FpPPh2 and formed ligands, rather than assembling them with the phos-
BH3·thf and was characterized spectroscopically.[59]
phorus atom connected to a metal fragment. With this goal
The few examples of Fp-phosphane chalcogenide com- in mind, FpI was treated with KPPh2BH3 (K2a) in thf,
plexes with trialkyl- and triarylphosphane chalcogenides which resulted in nearly quantitative formation of the de-
EPR3 (R = iPr, tBu, Ph; E = O, S, Se) were prepared by sired compound 12a. It also proved possible to synthesize
reaction with [Fp(thf)]BF4 and characterized spectroscopi- [FpPtBu2BH3] (12b) by the same method. Both complexes
cally[60] and, in part, by cyclic voltammetry.[61] The same were isolated as X-ray quality crystals by gas-phase dif-
synthetic route can be followed for silyl thiolates, and a fusion of pentane into a concentrated toluene solution of
small number of complexes of the type FpSSiR3 have been 12a or 12b. The yellow solids can be handled in air for short
isolated.[62] The coordination of chalcogenated phosphanyl periods of time but decompose upon longer exposure.
borohydrides to the Fp+ fragment has not been described
previously.
Considering the similarities between silanides and phos-
phanyl borohydrides, it would be desirable to apply the
same synthetic route used for complexes of the latter to the
preparation of complexes of the former. The sodium silan-
ides NaSiPh2Me[64] (Na3a) and NaSitBu2Me[65] (Na3b) are
both known in the literature, although the former has been
characterized only by boiling point determination of the
Results and Discussion
Synthesis
In order to conduct a detailed comparison of transition corresponding silane after hydrolysis. The di-tert-butyl spe-
metal complexes of the non-chalcogen ligands, compounds cies Na3b is well documented, but the synthesis of the do-
[FpPPh2Me]I (11I), [FpPPh2BH3] (12a), [FpPtBu2BH3] nor-free compound as described by Sekiguchi, which in-
(12b), and [FpSiPh2Me] (13) were synthesized (Scheme 1). volves reduction of the bromosilane with elemental sodium
Compound 11I was prepared from FpI (9), LiPPh2, and in boiling heptane, results in a significant amount of disil-
MeI in thf. Reports of the synthesis of this complex directly ane impurity.[65] Another common synthesis of sterically
from FpI and PPh2Me (1) describe impurities due to the hindered sodium silanides, in this case donor-supported, in-
CO-substitution product [CpFe(CO)(PPh2Me)I].[63] In con- volves the reduction of the corresponding bromosilanes
trast, the procedure followed here produces the product as with sodium metal in thf.[66] Biphenyl can be used as a re-
a yellow solid in high purity and good yield. Complex 11I dox catalyst to allow lower reaction temperatures. In an ef-
is, however, sensitive to light, and photochemical decompo- fort to avoid unnecessary and unwanted by-products, this
sition can be observed, which proceeds more quickly in method was used to prepare (thf)xNaSitBu2Me. The NMR
Eur. J. Inorg. Chem. 2007, 1989–2003
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1991

M. C. Holthausen, M. Wagner et al.
FULL PAPER
chemical shifts (1H, 13C, 29Si) of the sodium silanide pro-
The reaction of KSPPh2BH3 (K6a) with [Fp(thf)]PF6
duced by this route were in good agreement with the pub- (14PF6) was more successful. Thus, when the reactants were
lished values. Slight variations can be attributed to the fact combined in CH2Cl2 at –78 °C and warmed to room tem-
that the procedure reported here results in a donor-sup- perature overnight, a colorless precipitate (KPF6) formed.
ported silanide, while the literature describes the synthesis Changing the solvent to toluene with subsequent filtration
of the donor-free compound. Unfortunately, this synthetic through a plug of silica gel on a frit allowed the product
method led to unexpected side products. In an attempt to complex [FpSPPh2BH3] (16a, Scheme 2) to be isolated as
grow crystals of the silanide, dark green blocks of the si- an orange oil. Although the NMR and IR data clearly indi-
lylated biphenyl radical [(18-crown-6)(thf)2Na][4,4Ј-(Si- cate the formation of the complex, attempts to obtain single
tBu2Me)2(C6H4)2] were isolated.[67] In light of this result crystals of 16a failed.
and the poor characterization of NaSiPh2Me in the litera-
ture, no further attempts to synthesize [FpSiR2Me] (R =
Ph, tBu) from Fp+ and SiR2Me– were undertaken.
As described above, a number of [FpSiR3] complexes
have been reported in the literature. Due to the low elec-
tronegativity of silicon, these reactions generally proceed
with silyl chlorides. This, in essence, represents the umpo-
lung of the procedure used with phosphanyl borohydrides
2a and 2b. The synthesis of [FpSiPh2Me] (13) by this
method has been described.[57] When KFp (10) is treated
with ClSiPh2Me in thf, the desired complex is formed in
nearly quantitative yield (Scheme 1). Impurities can be re-
moved by filtration through silica gel using hexane/CH2Cl2
(2:1) as the eluent. Recrystallization from hexane at –25 °C
yields the complex as light-brown plates.
Considering the success achieved with the parent com-
pounds, it was logical to explore the behavior of the chal-
cogen derivatives KOPPh2BH3 (K5a), NaOSiPh2Me (Na7),
Scheme 2. Complexes of a phosphane sulfide and its isoelectronic
analogs, 15PF6–17.
and KSPPh2BH3 (K6a) with the Fp+ fragment (the silyl
thiolate NaSSiPh2Me is not readily accessible).
When FpI was treated with K5a in thf, a colorless solid
In the hope of isolating a crystalline product, the same
was slowly deposited over the course of several days. NMR reaction was carried out using the tert-butyl derivative K6b
analysis of the resulting mixture revealed conversion of FpI rather than the phenyl derivative K6a. In this case, filtration
to Fp2 and the appearance of unidentified components in of the crude product through silica gel and recrystallization
the 31P and 11B NMR spectra. Upon closer inspection of from benzene yielded orange X-ray quality needles of
the hydride region of the 1H NMR spectrum, a small [FpSPtBu2BH3] (16b).
amount of FpH[68] could be identified, which is known to
Since the isosteric silyl thiolate NaSSitBu2Me is not eas-
be an unstable intermediate that decomposes to Fp2 and ily accessible, we turned to the sterically more hindered
H2.[69,70] This finding suggests that rather than forming an NaSSitBu3 (Na8). This compound reacts readily with
Fe–O bond, the iron atom interacts with the borohydride [Fp(thf)]BF4 (14BF4) in CH2Cl2 at room temperature to
moiety of the ligand, ultimately resulting in hydride trans- give the desired complex [FpSSitBu3] (17; Scheme 2).
fer. Although the fate of the ligand after hydride transfer Recrystallization from a concentrated pentane solution
remains unclear, it seems likely that oligomeric or polymeric yielded X-ray quality crystals.
compounds of the type [OPPh2BH2]n are formed.
The complex of the analogous phosphane sulfide 4,
The observation of a hydride-transfer product with K5a [FpSPtBu3]BF4 (15BF4), has been described in the litera-
–
is particularly interesting in light of a comparison with the ture.[60] Using the PF6 salt 14PF6, we were able to isolate
siloxide Na7, since it is unlikely that the methyl moiety X-ray quality crystals of the compound by gas-phase dif-
would engage in the same kind of behavior. However, when fusion of pentane into a solution of [FpSPtBu3]PF6 (15PF6)
a thf solution of FpI is treated with Na7, Fp2 is the only in CHCl3 (Scheme 2).
Cp-containing product. In the 29Si NMR spectrum,
It was thus possible to prepare two series of three isoelec-
O(SiPh2Me)2 can be identified as the major silicon-contain- tronic compounds each in which all six compounds were
1
ing product (its H, 13C, and 29Si NMR shifts correspond characterized by X-ray crystal structure analysis and NMR
to an authentic sample prepared from ClSiPh2Me and and IR spectroscopy. Complexes 11I, 12a, and 13 are de-
NaOSiPh2Me).[71–73] A similar reaction to that with FpI rived from the phosphane PPh2Me, while complexes 15PF6,
was observed when the alternative starting material 16b, and 17 are related to the phosphane sulfide SPtBu3. A
[Fp(thf)]PF6 was treated with K5a and Na7. However, in comparison of these two sets of isoelectronic complexes
the latter case a significant amount of ferrocene was formed grants insight into the electronic differences between the li-
in addition to Fp2.
gands.
1992
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 1989–2003

Broadening the Scope of Ancillary Phosphane-Type Ligands
FULL PAPER
Structural and Spectroscopic Comparison of 11I, 12a, 12b,
and 13
hand, are slightly shorter in 12b (Table 1). Furthermore, the
B–P–Fe–COG torsion angles (COG = center of gravity of
the Cp ring) differ significantly: the torsion angle is 56.8°
for 12a and –175.9° for 12b. The only other structurally
characterized Fp complex with an anionic phosphorus/bo-
ron ligand was reported by Fu and co-workers and contains
coordinated diphenylphosphidoboratabenzene.[30] This
complex has an Fe–P bond of 2.276(2) Å, which is similar
to that in 12a [2.2705(6) Å]. The Fe–CO distances in Fu’s
compound [av. 1.742(10) Å] are slightly shorter than in 12a
[av. 1.772(2) Å] but similar to those in 12b [av. 1.756(3) Å].
The C–O distances [av. 1.169(9) Å] are longer than in both
of the phosphanyl borohydride complexes [12a: av.
1.141(2) Å; 12b: av. 1.142(3) Å].
Selected bond lengths and angles for 11I, 12a, 12b, and
13 are compiled in Table 1. Details of the X-ray crystal
structure analyses of these compounds are summarized in
Table 5.
Table 1. Selected bond lengths [Å] for compounds 11I–13; calcu-
lated values in italics.
Fe–P/Si
Fe–CO (av.)
C–O (av.)
11I
calc.
12a
calc.
12b
13
2.236(2)
2.338
2.2705(6)
2.367
2.3404(6)
2.3353(5)
2.389
1.779(9)
1.787
1.772(2)
1.772
1.756(3)
1.755(2)
1.757
1.146(10)
1.158
1.141(2)
1.163
1.142(3)
1.160(3)
1.168
calc.
Phosphane compound 11I crystallizes in the triclinic
¯
space group P1 with well-separated complex cations and
iodide anions (Figure 3). The Fe–P bond [2.236(2) Å] is in
the range of Fe–P distances found for other [FpPR3]+ com-
plex cations.[74]
Figure 4. Solid-state structure of 12a. Ellipsoids are drawn at the
50% probability level. H atoms have been omitted for clarity.
Figure 3. Solid-state structure of the complex cation of 11I. Ther-
mal ellipsoids are drawn at the 50% probability level. H atoms have
been omitted for clarity.
Both phosphanyl borohydride complexes 12a and 12b
crystallize in the orthorhombic space group P212121 (Fig-
ures 4 and 5, respectively). Their structural parameters are
similar, although 12b shows signs of the higher steric de-
mand of the tert-butyl substituents: the Fe–P bond
[2.3404(6) Å] is somewhat longer than in the diphenyl deriv-
Figure 5. Solid-state structure of 12b. Ellipsoids are drawn at the
ative 12a [2.2705(6) Å]. The Fe–CO bonds, on the other 50% probability level. H atoms have been omitted for clarity.
Eur. J. Inorg. Chem. 2007, 1989–2003
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1993

M. C. Holthausen, M. Wagner et al.
FULL PAPER
Complex 13 (orthorhombic, P212121; Figure 6) has the carbonyl complex leads to a shift of the carbonyl carbon
longest Fe–Si bond [2.3353(5) Å] of the structurally charac- atom resonance to lower field.[41,49,77] The carbonyl carbon
terized Fp-silyl complexes in the literature, with the excep- atom of silyl complex 13 resonates at δ = 215.8 ppm. The
tion of those complexes with donor silicon atoms bound to corresponding signals for 12a and 11I are found at δ =
silyl residues without electron-withdrawing substituents (cf. 213.5 and 210.4 ppm, respectively. The 13C NMR chemical
[FpSiMe2SiPh3]: Fe–Si = 2.346(1) Å;[75] [FpSiPh(SiMe3)2]: shifts of the cyclopentadienido ligands in metal carbonyl
Fe–Si = 2.366(1) and 2.355(1) Å in two crystallographically complexes also correlate to the electron density at the metal
independent molecules[76]). The Fe–Si bond in 13 cannot, center, although in this case higher electron density corre-
of course, be compared directly to the Fe–P bonds in com- sponds to a shift to higher field.[42,49,77] Phosphane complex
plexes 11I, 12a, and 12b. However, the difference between 11I contains the Cp ring with the most deshielded carbon
the Fe–P bond length in 12a and the Fe–Si bond length in nuclei (δ = 89.1 ppm). The cyclopentadienido ring of 12a
13 is roughly equivalent to the difference in the covalent resonates at δ = 86.8 ppm (12b: δ = 86.5 ppm), and the cor-
radii[71] of P and Si. The Fe–CO distances in 13 are compar- responding signal for silyl complex 13 can be seen at δ =
able to those found in other Fp-silyl complexes[74] but 84.4 ppm. In summary, the NMR spectroscopic data also
shorter than those in complexes 11I and 12a (Table 1). The support the conclusion that the following trend can be es-
C–O bonds found in silyl complex 13 [av. 1.160(3) Å] are tablished for the donor strengths of the isoelectronic, iso-
roughly 0.015 Å longer than in the complexes 11I and 12. steric ligands: 3a Ͼ 2a Ͼ 1.
In the 31P NMR spectrum, the resonance of the P atom
in 12b is found at δ = 81.4 ppm, a shift to considerably
lower field compared to 11I (δ = 48.6 ppm) and 12a (δ =
34.3 ppm). This is attributable to the difference in steric de-
mand between the two complexes. This effect has already
been observed for potassium phosphanyl borohydrides K2a
and K2b[78] as well as for the parent phosphanes PPh3 and
This indicates that the π-backbonding from the iron center
to the CO ligands is most pronounced in the silyl complex
and suggests that 3a is the strongest donor of the isoelec-
tronic series 1, 2a, and 3a.
1
PtBu3.[79] The JP,C coupling constants can be correlated to
the degree of s character in the P–C bonding orbital.[79]
1
For the two isosteric compounds 11I and 12a, JP,C (C-i) is
significantly larger for the phosphane complex (11I:
52.9 Hz; 12a: 27.8 Hz), thereby indicating a larger degree of
s character in the P–C-i bonding orbital of 11I. A similar
relationship between the 1JP,C values is observed for the free
ligands and can easily be explained by Bent’s rule.[58,80]
Comparison of the carbonyl IR stretching frequencies is,
of course, also an important measure of donor strength in
carbonyl complexes (Table 2). In accordance with the in-
terpretation of the 13C NMR spectra, complex 11I is found
to have the highest CO stretching frequencies (2055 and
2011 cm–1 in MeCN solution). These values compare favor-
ably with those found in the literature for similar com-
plexes.[81–83] Introduction of the BH3 moiety and conse-
quently of a negative charge in 12a leads to a marked de-
crease in the frequencies of these vibrations to 2030 and
Figure 6. Solid-state structure of 13. Ellipsoids are drawn at the
50% probability level. H atoms have been omitted for clarity.
This interpretation of relative donor strengths is sup-
ported by the 13C NMR chemical shifts of the carbonyl
carbon atoms. A summary of spectroscopic data can be
found in Table 2. It has been shown in the literature that an 1982 cm–1 (MeCN solution). The relatively low-energy
increase in the donor capacity of an ancillary ligand in a bands found for complex 13 (1994 and 1938 cm–1; MeCN
Table 2. Selected spectroscopic parameters for compounds 11I–13; calculated values in italics.
ν(CO)[a]
δ(13C) (CO)[b]
(2JP,C [Hz])
δ(13C) (Cp)[b]
δ(13C) (C-i)[b]
(1JP,C [Hz])
δ(31P/29Si)[b]
˜
[cm–1]
11I
calc.
12a
calc.
12b
13
2055, 2011 (2033)
2029[c]
210.4 (24.2)
205.7
213.5 (18.0)
208.6
215.5 (17.0)
215.8
89.1
89.1
86.8
88.6
86.5
84.4
85.6
134.1 (52.9)
128.9
140.2 (27.8)
139.2
48.6
–
34.3
–
81.4
35.1
–
2030, 1982 (2006)
2002[c]
2029, 1982
–
1994, 1938 (1966)
144.6
142.9
calc.
1977[c]
208.5
[a] In MeCN; (ν + ν )/2 given in parentheses. [b] 11I in CD CN and 12–13 in C D . [c] (ν + ν )/2 scaled by 0.9884 to correct for
˜
˜
˜
˜
as
s
3
6
6
as
s
systematic errors.
1994
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 1989–2003

Broadening the Scope of Ancillary Phosphane-Type Ligands
Table 3. Selected bond lengths [Å] and angles [°] for compounds 15PF6, 16b, and 17; calculated values in italics.
FULL PAPER
Fe–S
S–P/Si
Fe–CO (av.)
C–O (av.)
OC–Fe–CO
Fe–S–P/Si
15PF6
calc.
16b
calc.
17
calc.
2.3469(4)
2.434
2.3125(8)
2.399
2.345(2)
2.387
2.0473(5)
2.072
2.087(1)
2.104
2.147(2)
2.194
1.799(2)
1.791
1.784(3)
1.784
1.776(5)
1.774
1.139(3)
1.158
1.144(4)
1.159
1.151(6)
1.162
93.35(9)
95.3
98.0(2)
97.0
93.1(2)
94.6
127.88(2)
129.4
115.86(4)
118.6
129.01(7)
127.9
solution) are consistent with the NMR spectroscopic data, 2.3125(8) Å. This complex also displays an OC–Fe–CO an-
which indicate that the silyl ligand is the strongest donor of gle of 98.0(2)°, nearly 5° larger than those found for the
the three.
two related complexes with three tBu residues.
When moving from 1 to 2a to 3a within the Fp+ system,
it is also interesting to consider the effect of exchanging the
–
Cp ligand for Cp* (Cp* = C5Me5 ) while leaving the other
ligands unchanged. While the Cp* silyl complex has not
been described in the literature, both the phosphane and
phosphanyl borohydride adducts have been reported by
Malisch and co-workers.[29] In CH2Cl2 solution, the CO
stretching frequencies are observed at 2029 and 1987 cm–1 for
the phosphane complex [Cp*Fe(CO)2(PPh2Me)]I, thereby
demonstrating that the effect of exchanging Cp for Cp* is
comparable to exchanging the phosphane ligand for the cor-
responding phosphanyl borohydride. Bands attributable to
the CO ligands in [Cp*Fe(CO)2(PPh2BH3)], on the other
hand, are found at 2007 and 1959 cm–1 (CH2Cl2 solution).
These frequencies lie above those seen for complex 13, thus
indicating that moving from the phosphanyl borohydride to
the silyl ligand has a more dramatic effect on the electronic
structure of the compound than replacing Cp with Cp*.
Taken together, these results suggest that, when viewed elec-
tronically, the difference between the phosphane and phos-
phanyl borohydride ligands is somewhat smaller than be-
tween the phosphanyl borohydride and silyl ligands.
Figure 7. Solid-state structure of the complex cation of 15PF6. El-
lipsoids are drawn at the 50% probability level. H atoms have been
omitted for clarity.
Structural and Spectroscopic Comparison of 15PF6, 16b,
and 17
Selected bond lengths and angles for 15PF6, 16b, and 17
are compiled in Table 3. Details of the X-ray crystal struc-
ture analyses of these compounds are summarized in
Table 6.
The three iron–sulfur complexes [FpSPtBu3]PF6 (15PF6;
orthorhombic, P212121, Figure 7), [FpSPtBu2BH3] (16b; or-
thorhombic, Pbca, Figure 8) and [FpSSitBu3] (17; triclinic,
¯
P1, Figure 9) all adopt the expected structure with mono-
dentate, end-on coordination of the sulfide ligands. Sepa-
rated ion pairs are found in 15PF6. The effect of the in-
creased steric bulk of 15PF6 and 17 compared to 16b is
most noticeable in the Fe–S–P/Si bond angle, which is ne-
arly 130° in the tri-tert-butyl derivatives 15PF6 [127.88(2)°]
and 17 [129.01(7)°], compared to 115.86(4)° in 16b. The Fe–
S bond is also affected by the steric strain: 15PF6 and 17
possess the longest reported Fe–S bonds of all Fp com-
plexes with terminal sulfur ligands [2.3469(4) and
Figure 8. Solid-state structure of 16b. Ellipsoids are drawn at the
50% probability level. H atoms have been omitted for clarity.
The iron–carbonyl bond lengths and the C–O distances
2.345(2) Å, respectively].[74] The less sterically hindered in the carbonyl ligands, however, seem to follow trends that
complex 16b, on the other hand, has an Fe–S bond of are independent of steric hindrance. Of complexes 15PF6,
Eur. J. Inorg. Chem. 2007, 1989–2003
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1995

M. C. Holthausen, M. Wagner et al.
FULL PAPER
[FpSSiiPr3]: δ = 214.7;[62] [FpSPPh3]PF6: δ = 211.3 ppm[61]).
In contrast to complexes 11I–13, the differences in the
chemical shifts of the carbon atoms in the Cp rings of com-
plexes 15PF6–17 are too small to discern any meaningful
trends.
When considering the IR data (Table 4), silyl thiolate
complex 17 is found to have CO vibrations at 2030 and
1982 cm–1, compared to 2040 and 1994 cm–1 for 16b and
2054 and 2010 cm–1 for 15BF4. The steric influence on the
IR stretching frequencies appears to be minimal, as
[FpSPPh2BH3] (16a) displays bands at 2042 and 1998 cm–1,
which vary only slightly from those of the corresponding
tert-butyl species 16b. These results correspond well with
published data.[60,62] It can thus be concluded that the CO
stretching frequencies in the IR spectra clearly indicate that
the silyl thiolate is the strongest donor of the three, while
the phosphane sulfide is the weakest.
In summary, X-ray, IR, and NMR spectroscopic data
indicate that differences in the electron densities of the non-
chalcogen complexes 11I–13 on the one hand and the thio
complexes 15PF6–17 on the other follow the same qualita-
tive trend. However, the differences found for 15PF6–17 are
less than half as large as those for 11I–13.
Figure 9. Solid-state structure of 17. Ellipsoids are drawn at the
50% probability level. H atoms have been omitted for clarity.
16b, and 17, for example, [FpSPtBu3]PF6 (15PF6) displays
the longest Fe–CO and the shortest C–O distances [an
average of 1.799(2) and 1.139(3) Å, respectively]. The
equally sterically hindered silyl thiolate complex 17, on the
other hand, has the shortest Fe–CO and the longest C–O
bond [an average of 1.776(5) and 1.151(6) Å, respectively],
although the individual C–O distances for this complex are
Quantum Chemical Calculations
The experimental studies described thus far were aug-
quite different [1.139(6) and 1.163(6) Å]. The phosphanyl mented by quantum chemical calculations in order to learn
borohydride sulfide complex 16b displays intermediate more about qualitative differences in the charge density dis-
average bond lengths of 1.784(3) (Fe–CO) and 1.144(4) Å tribution between formally isoelectronic ligands as well as
(C–O). These results indicate that the donor-strength trends between the [CpFe(CO)2]+ complexes of these ligands and
observed for the non-chalcogen complexes 11I–13 also hold also about the hydride-donor capacities of phosphanyl
true for the thio derivatives. The silyl thiolate 8 appears to borohydrides compared to their chalcogen derivatives. The
be the strongest donor, causing increased Fe–CO back- latter computations were inspired by the experimental ob-
bonding, which leads to a shortening of the Fe–CO bonds servation that the parent phosphanyl borohydride
–
and lengthening of the C–O bonds, followed by 6b and fi- PtBu2BH3 (2b) and the corresponding thio derivative 6b
nally phosphane sulfide 4.
react with the CpFe(CO)2 fragment with formation of a co-
This general trend is also supported by the chemical ordinative Fe–P and Fe–S bond, respectively. In contrast,
–
shifts of the CO carbon atoms in the 13C NMR spectra the oxo derivative OPPh2BH3 (5a) does not form stable
(Table 4). As explained above, a more deshielded CO car- complexes. Instead, we identified the hydride species FpH
bon atom correlates to a stronger donor ligand being at- in the reaction mixture, which is most likely generated by
tached to the same metal center.[41,42,49,77] The silyl thiolate hydride transfer from the BH3 group.
complex 17 has the furthest downfield CO signal (δ =
For the calculation of complex structures and vibrational
215.1 ppm), while the carbonyl ligands of 16b and 15PF6 frequencies, we employed the BP86 functional in combina-
resonate at δ = 213.1 and 211.5 ppm, respectively (cf. tion with the SVP valence double-zeta basis of Ahlrichs and
Table 4. Selected spectroscopic parameters for compounds 15PF6–17; calculated values in italics.[a]
ν(CO) [cm–1]
δ(13C) (CO)
(3JP,C [Hz])
δ(13C)
(Cp)
δ(13C) (CMe3)
(1JP,C [Hz])
δ(31P/29Si)
˜
15PF6
calc.
16a
2054, 2010[60] (2032)
2028[b]
211.5(2.3)
211.1
212.1(4.4)
213.1(3.1)
212.6
215.1
213.0
87.4
89.5
86.3
86.5
87.5
86.9
88.4
44.1(24.4)
93.8
–
–
–
2042, 1998
48.3
86.8
–
16b
2040, 1994 (2017)
36.8(20.7)
calc.
17
calc.
2022[b]
–
25.3
–
2030, 1982 (2006)
2004[b]
29.2[c]
–
[a] IR spectra measured in CH Cl ; NMR spectra measured in CD Cl unless otherwise noted; (ν + ν )/2 given in parentheses. [b] (ν
˜
˜
˜
2
2
2
2
as
s
as
+ ν )/2 scaled by 0.9884 to correct for systematic errors. [c] In C D .
˜
s
6
6
1996
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 1989–2003

Broadening the Scope of Ancillary Phosphane-Type Ligands
FULL PAPER
co-workers.[84] The good agreement between computed and
experimental values for key geometric data (cf. Tables 1 and
3), as well as for characteristic spectroscopic parameters (cf.
Tables 2 and 4) validates this level of density functional
theory for an adequate description of electronic structures
of the class of compounds studied here.[85,86] For the com-
putation of thermodynamic data for the hypothetical hy-
dride transfer reaction sketched in Scheme 3, we employed
the more reliable B3LYP functional in combination with
the TZVP valence triple-zeta basis set of Ahlrichs et al.[87]
All computations were performed with the Gaussian03 pro-
gram.[88] Electrostatic potential surfaces were obtained
from analogous computations employing the SPARTAN
program.[89]
Figure 10. Electrostatic potential surfaces of the anionic ligands
–
–
–
PPh2BH3 (2a), SiPh2Me– (3a), OPtBu2BH3 (5b), SPtBu2BH3
–
–
(6b), OSitBu3 , and SSitBu3 (8); color code: red = negative poten-
tial; blue = positive potential.
Scheme 3. Isodesmic reactions to assess the relative hydride donor
strengths of PtBu2BH3 (2b), OPtBu2BH3 (5b), and SPtBu2BH3
(6b).
–
–
–
To test the relevance of a possible neighboring group ef-
fect, we calculated the molecular structures of the mono-
–
meric products of hydride abstraction from PtBu2BH3
–
–
An inspection of the electrostatic potential surfaces of (2b), OPtBu2BH3 (5b), and SPtBu2BH3 (6b) (Scheme 3).
PPh2BH3– (2a) and SiPh2Me– (3a; Figure 10) clearly reveals After geometry optimization, the P–B length in PtBu2BH2
an accumulation of negative charge concentrated on the (2b-H; 1.818 Å) indicates very little double-bond character
BH3 fragment for the phosphanyl borohydride and on the between the two atoms. In the case of the oxo derivative
silicon atom for the silyl ligand. In contrast, all of the chal- OPtBu2BH2, an interesting rearrangement is observed com-
–
–
cogen derivatives OPtBu2BH3 (5b), SPtBu2BH3 (6b), putationally which leads to a migration of the BH2 frag-
–
–
OSitBu3 , and SSitBu3 (8) bear the bulk of their negative ment from phosphorus to oxygen via the anticipated three-
charge on the chalcogen atom (Figure 10). membered P–O–B ring (Scheme 3). In the resulting isomer
Given this qualitative background, it is at first glance tBu2P–O–BH2 (5b-H), pronounced O–B π-bonding aids in
surprising to see that 2a and 2b react with FpI to give com- the stabilization of the BH2 fragment (O–B = 1.352 Å). To
plexes 12a and 12b, respectively, whereas treatment of FpI assess the relative stabilities of PtBu2BH2 and tBu2P–O–
with 5a leads to hydride transfer and formation of FpH BH2, we computed the free reaction energy for the iso-
–
–
and ultimately of Fp2. The reactivities of 2a/2b and 5a with desmic reaction PtBu2BH2 + OPtBu2BH3 Ǟ PtBu2BH3
respect to FpI are obviously not governed simply by electro- + tBu2P–O–BH2 and obtained a value of –11.4 kcalmol–1
static attraction between the positively charged iron center (Scheme 3). A neighboring group effect similar to the one
and the most negatively charged part of the ligand mole- suggested for hydride-transfer reactions of 5b may also be
cule; other factors appear to play an important role. We operative in the thio derivative 6b. This in turn raises the
suspected that in the case of ligand 5a, a neighboring group question as to why this ligand reacts with [Fp(thf)]+ to form
effect might be responsible for the facile hydride abstrac- [FpSPtBu2BH3] (16b) rather than [FpH]. To address this
tion, because the resulting borane OPPh2BH2 (5a-H) can point, we also calculated the molecular structure of
stabilize itself by intramolecular O–B adduct formation SPtBu2BH2 and found two minima, one containing a three-
leading to a three-membered P–O–B ring. In the case of membered P–S–B ring and the other being the open-chain
PR2BH2, the electron deficiency of the boron atom can only isomer (Scheme 3). Both isomers are similar in energy, with
be alleviated by P–B pπ–pπ bonding, which is not very ef- cyc-6b-H being less stable than the open chain species 6b-
ficient.[90]
H by merely 2.6 kcalmol–1. We interpret this result as fol-
Eur. J. Inorg. Chem. 2007, 1989–2003
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1997

M. C. Holthausen, M. Wagner et al.
FULL PAPER
lows: due to poor orbital overlap, S–B π-bonding is less the degree of charge transfer is different for the different
favorable than O–B π-bonding, and consequently S–B σ- ligand classes, with phosphanes being the weakest and silyl
donation becomes competitive, even though the resulting ligands being the strongest donors.
cyclic structure suffers from severe Baeyer strain. As before,
we computed the free reaction energy for the isodesmic re-
–
–
Conclusions
action PtBu2BH2 + SPtBu2BH3 Ǟ PtBu2BH3 + tBu2P–
S–BH2 (Scheme 3). Contrary to our results for the O-con-
taining analogs, we found this process to be endergonic by
+6.4 kcalmol–1 relative to the more stable open-chain iso-
mer. Hence, we conclude that hydride abstraction from
A series of isoelectronic and isosteric iron complexes
[CpFe(CO)2(EPh2XH3)]n+ (E = Si, X = C; E = P, X = C,
B; n = 0, 1) 11I, 12a, and 13 has been synthesized and
1
characterized by X-ray crystallography and H and hetero-
–
OPtBu2BH3 (5b) is significantly more facile (by
nuclear NMR and IR spectroscopy. A comparison of the
structural and spectroscopic properties of compounds 11I–
13 indicates that the silyl ligand 3a displays the highest pro-
pensity towards electron donation, followed by the anionic
phosphanyl borohydride ligand 2a. The phosphane ligand
1 is the weakest donor.
In addition, the reactivity of phosphane chalcogenides,
phosphanyl borohydride chalcogenides, and silyl chalcogen-
olates with the [CpFe(CO)2]+ fragment has been investi-
17.8 kcalmol–1) than from SPtBu2BH3– (6b). These theoret-
ical findings agree very well with our experimental observa-
tions.
In all three compounds [11]+, 12a, and 13, the CpFe-
(CO)2 fragment bears a positive charge q (obtained from
NPA analyses) which is largest in the cationic phosphane
complex [11]+ (q = + 0.486), intermediate in the phosphanyl
borohydride complex 12a [q = +0.302; ∆q([11]+ – 12a) =
0.184], and smallest in the silyl complex 13 [q = +0.072;
∆q(12a – 13) = 0.230]. This finding supports the conclusion
drawn from inspection of the IR and NMR spectroscopic
data that silyl ligand 3a is a significantly stronger electron
donor than phosphanyl borohydride ligand 2a. A similar
trend is also visible for the series of complexes [15]+ (q =
+0.576), 16b [q = +0.435; ∆q([15]+ – 16b) = 0.141], and 17
[q = +0.353; ∆q(16b – 17) = 0.082], even though the differ-
ences in the computed NPA charges on the three CpFe-
(CO)2 fragments are less pronounced with a sulfur atom
between the iron center and the P/Si atom. Again, these
results are nicely reflected by the changes of the CO stretch-
ing frequencies: The highest wavenumbers are found for
[15]+, while a shift to lower wavenumbers is observed for 17
(Table 4), although this difference is much smaller than in
the case of [11]+ and 13 (Table 2). An inspection of the
HOMOs of the three isoelectronic ligands 1, 2a, and 3a
reveals very similar topologies, all of which correspond to
the lone pair being located on the donor atoms P or Si
(Figure 11). This finding provides convincing evidence that
the interaction of these ligands with transition metal ions
will be of essentially the same nature and implies that these
ligands are truly isolobal species. Quantitatively, however,
gated.
A
second series of isoelectronic complexes
[CpFe(CO)2(SEtBu2XH3)]n+ (E = Si, X = C; E = P, X = C,
B; n = 0, 1) 15PF6–17 has been isolated with tert-butyl
rather than phenyl residues and fully characterized. The
conclusion that the silyl species is the strongest donor of
the sulfur-containing ligands can be drawn, but this effect,
now mediated by the sulfur atom, is not as pronounced as
for the non-chalcogen species 11I–13, and the differences in
the CO stretching frequencies in the IR spectra and the 13
C
NMR resonances of the CO carbon atoms are about half
as large in 15PF6–17 as in 11I–13. The molecular structures
of all crystallographically characterized compounds as well
as key spectroscopic parameters have been calculated by
DFT. Overall, an excellent agreement was found between
experimentally and theoretically derived data. In the reac-
–
tion with [CpFe(CO)2]+, the oxo derivative OPPh2BH3
transfers a hydride ion rather than forming a stable
[CpFe(CO)2(OPPh2BH3)] complex. The tendency to un-
dergo hydride transfer reactions has been studied by DFT
–
–
calculations for the series PtBu2BH3 , OPtBu2BH3 , and
–
–
SPtBu2BH3 . The results reveal that OPtBu2BH3 is the
–
strongest and SPtBu2BH3 the weakest hydride donor, in
agreement with experimental observations. Further theoret-
ical analysis indicates that the three derivatives PPh2Me,
–
PPh2BH3 , and SiPh2Me– are truly isolobal species, despite
variations in their charge distributions.
Experimental Section
General Considerations: All reactions and manipulations of air-sen-
sitive compounds were carried out under an atmosphere of dry
nitrogen using standard Schlenk techniques. Solvents were freshly
distilled under argon from sodium/benzophenone (thf, diethyl
ether, toluene, C6D6) or sodium-lead alloy (pentane, hexane) prior
to use. MeCN, CD3CN, CH2Cl2, CD2Cl2, and CDCl3 were dried
with molecular sieves (4 Å) and degassed. NMR spectra were re-
corded with Bruker AMX 250, DPX 250, Avance 300, or Avance
400 spectrometers. 1H and 13C chemical shifts are reported relative
to tetramethylsilane and were referenced against residual solvent
Figure 11. HOMOs of the series of isoelectronic ligands PPh2Me
(1), PPh2BH3 (2a), and SiPh2Me– (3a).
–
1998
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 1989–2003

Broadening the Scope of Ancillary Phosphane-Type Ligands
FULL PAPER
peaks (C6D5H: δ = 7.16; C6D6: δ = 128.06; CHCl3: δ = 7.26;
CDCl3: δ = 77.16 ppm).[91] 11B NMR spectra were referenced
against external BF3·OEt2. 31P NMR spectra are reported relative
to external H3PO4 (85%). 29Si NMR spectra were referenced
against external TMS and recorded using the INEPT pulse se-
quence with empirically optimized parameters for polarization
transfer from the methyl and phenyl substituents (13) or the tBu
substituents (Na3b and 17). Abbreviations: Fp = CpFe(CO)2, n.r.
= multiplet expected but not resolved. IR spectra were recorded
with a Perkin–Elmer 1650 FTIR spectrophotometer or a Jasco FT/
IR-420 spectrophotometer. Elemental analyses were performed by
Synthesis of [FpPPh2BH3] (12a): A solution of KPPh2BH3 (K2a;
0.239 g, 1.00 mmol) in thf (10 mL) was added with stirring to a
solution of FpI (9, 0.304 g, 1.00 mmol) in thf (10 mL) at –78 °C.
Slowly warming to ambient temperature overnight resulted in a
color change from black to light brown and formation of a color-
less precipitate (KI). The mixture was filtered and the solvent re-
moved from the filtrate in vacuo. The resulting brown solid residue
was washed with pentane (5 mL) and extracted with toluene
(10 mL). The toluene was removed from the filtrate in vacuo to
give the product as a yellow crystalline material. Yield: 0.274 g
(73%). X-ray quality crystals were obtained by gas-phase diffusion
the microanalytical laboratory of the J. W. Goethe University, of pentane into a solution of 12a in toluene. 1H NMR (C6D6,
3
Frankfurt(Main).
250.13 MHz): δ = 2.3 (m, 3 H, BH3), 4.16 (d, JP,H = 1.6 Hz, 5 H,
Cp), 6.97–7.16 (m, 6 H, H-m,p), 7.93–8.02 ppm (m, 4 H, H-o).
Materials: Li, Na, KH, nBuLi solution (ca. 1.6  in hexane),
BH3·thf solution (1  in thf), PPh3, ClSiPh2Me, biphenyl, FpI, and
MeI were obtained from commercial sources (Aldrich, Fluka,
Acros or Chemetall) and used as received. HPPh2 was obtained by
reductive cleavage of PPh3 with Li powder in thf and subsequent
hydrolysis and distillation, as described by Bianco and Doronzo.[92]
Deprotonation of HPPh2 with nBuLi yields LiPPh2;[93] employing
hexane leads to precipitation of solvent-free LiPPh2. Bromination
of HSitBu2Me[94] according to the literature method[95] yields
BrSitBu2Me. KFp (10),[96] KPPh2BH3 (K2a),[12,58] and KPtBu2BH3
(K2b)[12,58] were prepared by literature methods. [Fp(thf)]PF6
(14PF6) was obtained by treating [Cp2Fe]PF6 with Fp2 as reported
by Catheline and Astruc.[97] [Fp(thf)]BF4 (14BF4) was prepared
from FpI and AgBF4.[98]
1
11B{1H} NMR (C6D6, 128.37 MHz): δ = –28.0 ppm (d, JP,B
=
44 Hz). 13C{1H} NMR (C6D6, 62.90 MHz): δ = 86.8 (n.r., Cp),
128.3 (d, 3JP,C = 8.7 Hz, C-m), 129.3 (d, 4JP,C = 2.3 Hz, C-p), 133.3
2
1
(d, JP,C = 8.0 Hz, C-o), 140.2 (d, JP,C = 27.8 Hz, C-i), 213.5 ppm
(d, JP,C = 18.0 Hz, CO). 31P{1H} NMR (C6D6, 161.98 MHz): δ =
2
34.3 ppm (m). IR (MeCN):
ν =
˜
2030, 1982 cm–1 (CO).
C19H18BFeO2P (375.96): calcd. C 60.74, H 4.83; found C 60.58, H
4.73.
Synthesis of [FpPtBu2BH3] (12b): A solution of KPtBu2BH3 (K2b,
0.291 g, 1.47 mmol) in thf (10 mL) was added with stirring to a
solution of FpI (9, 0.446 g, 1.47 mmol) in thf (10 mL) at –78 °C.
Slowly warming to ambient temperature overnight resulted in a
color change from black to yellow and formation of a colorless
precipitate (KI). The mixture was filtered and the solvent removed
from the filtrate in vacuo. The resulting brown solid residue was
washed with pentane (5 mL) and extracted with toluene (10 mL).
The toluene was removed from the filtrate in vacuo to give the
product as a yellow crystalline material. Yield: 0.355 g (72%). X-
ray quality crystals were obtained by gas-phase diffusion of pen-
Synthesis of (thf)xNaSitBu2Me (Na3b): Na (2.0 g, 87.0 mmol) and
biphenyl (1.29 g, 8.4 mmol) were stirred overnight in thf (20 mL)
to give a dark green solution surrounding blocks of unreacted so-
dium. A solution of BrSitBu2Me (3.38 g, 14.2 mmol) in thf (15 mL)
was then added dropwise to this mixture over 2.5 h. After stirring
for 16 h, all volatiles were removed in vacuo, and the dark brown
residue was extracted with pentane (25 mL). After filtration, the
filter cake was washed with pentane (10 mL). Volatiles were re-
moved in vacuo overnight from the dark brown filtrate, and the
resulting brown residue was dissolved in thf (30 mL) to give a
0.305  solution of the product in thf (9.15 mmol, 64%). 1H NMR
(C6D6, 250.13 MHz): δ = 0.33 (s, 3 H, SiCH3), 1.35 ppm (s, 18 H,
CCH3). 13C{1H} NMR (C6D6, 62.90 MHz): δ = 0.3 (SiCH3), 21.3
(CCH3), 32.2 ppm (s, CCH3). 29Si{1H} NMR (C6D6, 49.7 MHz): δ
= 14.4 ppm.
tane into
a
solution of 12b in toluene. 1H NMR (C6D6,
3
250.13 MHz): δ = 1.35 (d, JP,H = 11.9 Hz, 18 H, CH3), 1.60 (q,
1JB,H = 93.6 Hz, 3 H, BH3), 4.33 ppm (s, 5 H, Cp). 11B{1H} NMR
(C6D6, 128.37 MHz): δ = –30.0 ppm (d, JP,B = 49 Hz). 13C{1H}
1
2
NMR (C6D6, 62.90 MHz): δ = 31.2 (d, JP,C = 2.0 Hz, CH3), 37.4
1
2
(d, JP,C = 6.6 Hz, CCH3), 86.5 (n.r., Cp), 215.5 ppm (d, JP,C
=
17.0 Hz, CO). 31P{1H} NMR (C6D6, 161.98 MHz): δ = 81.4 ppm
(m). IR (MeCN): ν = 2029, 1982 cm–1 (CO). C15H26BFeO2P
˜
(335.99): calcd. C 53.62, H 7.80; found C 53.48, H 7.69.
Synthesis of [FpSiPh2Me] (13): A solution of ClSiPh2Me (0.33 mL,
0.367 g, 1.57 mmol) in thf (8 mL) was added dropwise to a suspen-
sion of KFp (0.307 g, 1.42 mmol) in thf (10 mL). After stirring
overnight, volatiles were removed in vacuo, then the brown residue
was extracted with pentane (4ϫ10 mL) and filtered. Removal of
the solvent yielded the crude product as a brown oil (0.392 g, 74%).
The crude product was purified by filtration through silica gel un-
der an inert atmosphere using hexane/CH2Cl2 (2:1) as the eluent,
which yielded the product as a yellow solid after removal of the
volatiles in vacuo. X-ray quality crystals were obtained from hex-
Synthesis of [FpPPh2Me]I (11I): A solution of LiPPh2 (0.147 g,
0.77 mmol) in thf (6 mL) was added with stirring to a solution of
FpI (9, 0.233 g, 0.77 mmol) in thf (8 mL) at –78 °C. Slowly warm-
ing to ambient temperature overnight resulted in a color change
from black to dark red and the formation of a colorless precipitate
(LiI). The mixture was filtered and the filtrate treated with a solu-
tion of MeI (0.112 g, 0.79 mmol) in thf (3 mL). The resulting dark
yellow precipitate was washed with thf (2ϫ5 mL) and dried in
vacuo. Yield: 0.243 g (63%). X-ray quality crystals were obtained
by gas-phase diffusion of diethyl ether into a solution of 11I in
1
ane solution at –25 °C. H NMR (300.0 MHz, C6D6): δ = 0.95 (s,
2
MeCN. 1H NMR (CD3CN, 400.13 MHz): δ = 2.38 (d, JP,H
=
3 H, CH3), 3.97 (s, 5 H, Cp), 7.21, 7.69 ppm (2ϫm, 6 H, 4 H, H-
o,m,p). 13C{1H} NMR (75.4 MHz, C6D6): δ = 5.5 (CH3), 84.4 (s,
Cp), 128.0 (C-m or C-o), 128.4 (s, C-p), 134.6 (C-m or C-o), 144.6
(C-i), 215.8 ppm (CO). 29Si{1H} NMR (59.6 MHz, C6D6): δ =
3
10.5 Hz, 3 H, CH3), 5.38 (d, JP,H = 1.6 Hz, 5 H, Cp), 7.50–
7.64 ppm (m, 10 H, H-o,m,p). 13C{1H} NMR (CD3CN,
1
2
100.63 MHz): δ = 19.5 (d, JP,C = 35.4 Hz, CH3), 89.1 (d, JP,C
=
=
=
3
2
0.5 Hz, Cp), 130.4 (d, JP,C = 11.0 Hz, C-m), 132.3 (d, JP,C
35.1 ppm. IR (MeCN): ν = 1994, 1938 cm–1 (CO). C20H18FeO2Si
˜
4
1
10.2 Hz, C-o), 132.9 (d, JP,C = 2.8 Hz, C-p), 134.1 (d, JP,C
(374.28): calcd. C 64.18, H 4.85; found C 64.14, H 4.88.
52.9 Hz, C-i), 210.4 ppm (d, JP,C = 24.2 Hz, CO). 31P{1H} NMR
2
(CD CN, 161.98 MHz): δ = 48.6 ppm (s). IR (MeCN): ν = 2055,
Reaction of KOPPh2BH3 (K5a) with FpI (9): An NMR tube was
˜
3
2011 cm–1 (CO). C20H18FeIO2P (504.06): calcd. C 47.65, H 3.60; charged with K5a (0.05 g, 0.2 mmol), FpI (9; 0.06 g, 0.2 mmol),
found C 47.46, H 3.65.
and [D8]thf (0.6 mL), and the tube was flame-sealed and shaken
Eur. J. Inorg. Chem. 2007, 1989–2003
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1999

M. C. Holthausen, M. Wagner et al.
FULL PAPER
until all the solids had dissolved. Slow deposition of a colorless
precipitate (KI) took place over several days. Analysis by NMR
spectroscopy showed slow formation of Fp2 and several unidenti-
fied phosphorus- and boron-containing products. The integrals of
the 1H NMR Cp resonances of Fp2 and FpI account for over 95%
of the total Cp region integral. Main phosphorus-containing prod-
Synthesis of [FpSPPh2BH3] (16a): A solution of KSPPh2BH3 (K6a;
0.098 g, 0.36 mmol) in thf (3 mL) was added to a solution of
[Fp(thf)]PF6 (14PF6; 0.143 g, 0.36 mmol) in CH2Cl2 (30 mL) with
stirring at –78 °C. As the mixture warmed to ambient temperature
overnight, the color lightened from dark red to orange. The solvent
was removed in vacuo and the resulting red-brown solid residue
uct, approx. 60% yield by 31P NMR: 31P{1H} NMR ([D8]thf, washed with pentane (2ϫ5 mL) and extracted with toluene
161.98 MHz): δ = 80.8 ppm (m); 11B NMR ([D8]thf, 128.37 MHz):
(2ϫ10 mL). The extracts were filtered through silica gel, and the
toluene was removed from the filtrate in vacuo to give the product
as an orange oil. Yield: 0.106 g (72%). 1H NMR (CD2Cl2,
300.03 MHz): δ = 1.25 (br. m, 3 H, BH3), 5.08 (s, 5 H, Cp), 7.1–
7.9 ppm (m, 10 H, H-o,m,p). 11B{1H} NMR (CD2Cl2, 96.26 MHz):
1
δ = –37.5 ppm (m). Analysis of the hydride region of the H NMR
spectrum showed small amounts of FpH at δ = –11.7 ppm.[68]
Reaction of NaOSiPh2Me (Na7) with FpI (9): A Schlenk flask was
charged with FpI (9; 0.220 g, 0.72 mmol) in thf (5 mL), cooled to
–78 °C, and a solution of Na7 (0.170 g, 0.72 mmol) in thf (5 mL)
was added dropwise to this solution. The mixture was then warmed
to room temperature overnight. All volatiles were removed in
vacuo, and the brown residue was extracted with toluene and fil-
tered. The only major Cp-containing product in the filtrate was
Fp2 and the only major Si-containing product O(SiPh2Me)2, both
of which were identified by comparison with authentic samples.
Red crystals of [(NaOSiPh2Me)4(Fp2)2] were isolated by removing
toluene from the filtrate, washing the residue with benzene, and
recrystallizing from toluene. It should be noted that the solubility
of the adduct is significantly higher in toluene than in benzene. IR
1
δ = –33.8 ppm (d, JP,B = 55 Hz). 13C{1H} NMR (CD2Cl2,
3
75.45 MHz): δ = 86.3 (s, Cp), 128.5 (d, JP,C = 9.9 Hz, C-m), 130.6
4
2
(d, JP,C = 2.5 Hz, C-p), 132.0 (d, JP,C = 9.9 Hz, C-o), 136.2 (d,
1JP,C = 46.4 Hz, C-i), 212.1 ppm (d, JP,C = 4.4 Hz, CO). 31P{1H}
3
NMR (CD2Cl2, 121.46 MHz): δ = 48.3 ppm (br. m). IR (CH2Cl2):
ν = 2042, 1998 cm–1 (CO). A correct elemental analysis could not
˜
be obtained due to residual solvent and an unidentified impurity
that could not be removed.
Synthesis of [FpSPtBu2BH3] (16b): A solution of KSPtBu2BH3
(K6b; 0.137 g, 0.60 mmol) in thf (3 mL) was added to a solution of
[Fp(thf)]PF6 (14PF6; 0.185 g, 0.47 mmol) in CH2Cl2 (40 mL) with
stirring at –78 °C. As the mixture warmed to ambient temperature
overnight, the color lightened from dark red to orange, and a color-
less precipitate (KPF6) formed. The solvent was removed in vacuo
and the resulting red-brown semi-solid residue was extracted with
pentane (2ϫ10 mL). The extracts were filtered through silica gel
(approx. 1 cm) on a Schlenk frit. The product adsorbed completely
on the silica gel and was subsequently washed with 30 mL of pen-
tane. It was eluted with benzene, and all volatiles were removed in
[(NaOSiPh Me) (Fp ) ] (KBr): ν = 2005 (m), 1957 (s), 1932 (m),
˜
2
4
2 2
1756 (m), 1735 (s) cm–1 (CO).
Synthesis of [FpSPtBu3]PF6 (15PF6): The synthesis and NMR and
IR data of the complex cation as its BF4 salt have been reported
by Kuhn and Schumann (data given in Table 4 are for the PF6
salt).[60] We modified their procedure by using [Fp(thf)]PF6 instead
of the BF4– salt, as described in the literature.[97] X-ray quality crys-
tals were obtained by gas-phase diffusion of pentane into a solution
of 15PF6 in CHCl3. C19H32F6FeO2P2S (556.30): calcd. C 41.02, H vacuo to give a dark red oil. Upon standing overnight, fine orange
5.80; found C 40.97, H 5.81.
needles of 16b formed in X-ray quality. Yield: 0.100 g (58%). 1H
Table 5. Crystallographic data for compounds 11I–13.
11I
12a
12b
13
Empirical formula
Formula mass
Color, shape
Temperature [K]
Crystal system
Space group
a [Å]
b [Å]
c [Å]
α [°]
β [°]
C20H18FeIO2P
504.06
orange, block
173(2)
C19H18BFeO2P
375.96
yellow, block
203(2)
orthorhombic
P212121
8.488(1)
13.727(2)
15.195(2)
90
C15H26BFeO2P
335.99
yellow, block
293(2)
orthorhombic
P212121
7.7785(5)
14.802(1)
14.991(1)
90
C20H18FeO2Si
374.28
light-brown, plate
173(2)
orthorhombic
P212121
7.2717(6)
15.532(1)
16.055(1)
90
triclinic
¯
P1
9.768(1)
12.026(2)
18.416(3)
108.78(1)
98.99(1)
90
90
90
γ [°]
91.41(1)
2016.6(5)
4
1.660
992
90
1770.4(4)
4
1.411
776
90
1725.9(2)
4
1.293
712
90
1813.3(3)
4
1.371
776
V [Å3]
Z
Dcalcd. [gcm–3]
F(000)
µ [mm–1]
2.368
0.949
0.964
0.906
Crystal size [mm]
Reflections collected
Independent reflns (Rint
0.19ϫ0.15ϫ0.14
15825
7125 (0.0792)
7125/0/451
1.108
0.32ϫ0.22ϫ0.19
29550
3679 (0.0583)
3679/0/217
0.857
0.48ϫ0.45ϫ0.34
17282
3246 (0.0590)
3246/0/182
0.973
0.32ϫ0.26ϫ0.11
9046
3321 (0.0292)
3321/0/218
1.005
)
Data/restraints/parameters
GOOF on F2
R1, wR2 [I Ͼ 2σ(I)]
R1 = 0.0666,
wR2 = 0.0631
R1 = 0.1274,
wR2 = 0.0684
1.613/–2.791
R1 = 0.0215,
wR2 = 0.0440
R1 = 0.0271,
wR2 = 0.0448
0.290/–0.150
R1 = 0.0270,
wR2 = 0.0650
R1 = 0.0307,
wR2 = 0.0662
0.188/–0.162
R1 = 0.0223
wR2 = 0.0512
R1 = 0.0259
wR2 = 0.0522
0.187/–0.195
R1, wR2 (all data)
Largest difference peak/hole [eÅ–3]
2000
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 1989–2003

Broadening the Scope of Ancillary Phosphane-Type Ligands
FULL PAPER
Table 6. Crystallographic data for compounds 15PF6, 16b, and 17.
15PF6
16b
17
Empirical formula
Formula mass
Color, shape
Temperature [K]
Crystal system
Space group
a [Å]
b [Å]
c [Å]
α [°]
C19H32F6FeO2P2S
556.30
red, plate
173(2)
orthorhombic
P212121
8.3294(3)
14.2375(5)
20.7808(9)
90
C15H26BFeO2PS
368.05
orange, needle
173(2)
orthorhombic
Pbca
12.3927(8)
15.1586(8)
19.310(1)
90
C19H32FeO2SSi
408.45
orange, plate
173(2)
triclinic
¯
P1
8.671(1)
8.707(1)
16.215(2)
86.25(1)
β [°]
90
90
2464.4(2)
4
1.499
1152
0.885
90
90
3627.5(4)
8
1.348
1552
1.035
88.78(1)
62.46(1)
1083.0(3)
2
1.252
436
0.856
γ [°]
V [Å3]
Z
Dcalcd. [gcm–3]
F(000)
µ [mm–1]
Crystal size [mm]
Reflections collected
Independent reflns (Rint
0.34ϫ0.23ϫ0.12
45095
6885 (0.0563)
6885/0/281
1.053
0.17ϫ0.09ϫ0.08
31420
3330 (0.0622)
3330/0/191
0.934
0.18ϫ0.15ϫ0.08
10454
3997 (0.0996)
3997/60/213
0.952
)
Data/restraints/parameters
GOOF on F2
R1, wR2 [I Ͼ 2σ(I)]
R1 = 0.0295,
wR2 = 0.0789
R1 = 0.0303,
wR2 = 0.0794
0.447/–0.323
R1 = 0.0397,
wR2 = 0.0742
R1 = 0.0671,
wR2 = 0.0805
0.325/–0.521
R1 = 0.0699,
wR2 = 0.1704
R1 = 0.0986,
wR2 = 0.1888
0.530/–0.744
R1, wR2 (all data)
Largest difference peak/hole [eÅ–3]
NMR (CD2Cl2, 300.03 MHz): δ = 0.68 (m, 3 H, BH3), 1.27 (d, positions and refined with fixed isotropic displacement parameters
3JP,H = 13.2 Hz, 18 H, CH3), 5.12 ppm (s, 5 H, Cp). 11B{1H} NMR using a riding model. Complex 17: Two of the three tBu substitu-
1
(CD2Cl2, 96.26 MHz): δ = –37.1 ppm (d, JP,B = 55 Hz). 13C{1H} ents are disordered over two sites and were refined isotropically.
2
NMR (CD2Cl2, 75.45 MHz): δ = 28.4 (d, JP,C = 2.3 Hz, CH3), In order to keep the geometric parameters in reasonable ranges,
1
3
36.8 (d, JP,C = 20.7 Hz, CCH3), 86.5 (s, Cp), 213.1 ppm (d, JP,C
similarity restraints were applied for bond lengths and bond angles.
= 3.1 Hz, CO). 31P{1H} NMR (CD2Cl2, 121.46 MHz): δ = Details of the X-ray crystal structure analyses of compounds 11I,
86.8 ppm (br. m). IR (CH Cl ): ν = 2040, 1994 cm–1 (CO). 12a, 12b, and 13 are summarized in Table 5 and compounds 15PF6,
˜
2
2
C15H26BFeO2PS (368.05): calcd. C 48.95, H 7.12; found C 48.06,
H 7.12.
16b, and 17 in Table 6.
CCDC-628090 (11I), -628087 (12a), -628088 (12b), -628091 (13),
-628094 (15PF6), -628093 (16b), -628092 (17), and -628089 ([(18-
crown-6)(thf)2Na][4,4Ј-(SitBu2Me)2(C6H4)2]) contain the supple-
mentary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic
Data Center via www.ccdc.cam.ac.uk/data_request/cif.
Synthesis of [FpSSitBu3] (17): Solid (thf)2NaSSitBu3 (Na8, 0.155 g,
0.39 mmol) was added in one portion to a solution of [Fp(thf)]BF4
(14BF4; 0.131 g, 0.39 mmol) in CH2Cl2 (12 mL). After 4 h stirring,
the volatiles were removed from the reaction mixture and the brown
residue was extracted with pentane (5 mL) and filtered through Ce-
lite® on a frit. The filter cake was then washed with pentane
(2ϫ5 mL). Slow concentration of the filtrate yielded a dark brown,
microcrystalline solid. Yield: 0.156 g (98%). Recrystallization from
pentane yielded X-ray quality crystals. 1H NMR (C6D6,
400.13 MHz): δ = 1.39 (s, 27 H, CH3), 4.11 ppm (s, 5 H, Cp).
13C{1H} NMR (C6D6, 100.63 MHz): δ = 25.5 (CCH3), 31.6 (CH3),
86.2 (Cp), 215.1 ppm (CO). 29Si{1H} NMR (C6D6, 59.6 MHz): δ
= 29.2 ppm. 1H NMR (CD2Cl2, 300.0 MHz): δ = 1.13 (s, 27 H,
CH3), 5.02 ppm (s, 5 H, Cp). 13C{1H} NMR (CD2Cl2, 75.4 MHz):
δ = 25.3 (CCH3), 31.3 (CH3), 86.9 (Cp), 215.1 ppm (CO). IR (hex-
Supporting Information (see footnote on the first page of this arti-
cle): Calculated molecular structures of compounds 2b, 2b-H, 5b,
5b-H, 6b, and 6b-H (Figure S1).
Acknowledgments
The authors are grateful to the Deutsche Forschungsgemeinschaft
(DFG) and the Fonds der Chemischen Industrie (FCI) for financial
support and to Chemetall for the donation of tBuLi. We gratefully
acknowledge computer time provided by the Frankfurt Center for
Scientific Computing.
ane): ν = 2035 (s), 1988 (s) cm–1 (CO). IR (CH Cl ): ν = 2030 (s),
˜
˜
2
2
1982 (s) cm–1 (CO). C19H32FeO2SSi (408.45): calcd. C 55.87, H
7.90; found C 54.59, H 7.85.
[1] C. A. Tolman, Chem. Rev. 1977, 77, 313–348.
X-ray Structural Characterization: Data were collected with a Stoe
IPDS-II two circle diffractometer with graphite-monochromated
Mo-Kα radiation. Empirical absorption corrections were performed
with the MULABS option[99] in the program PLATON.[100] Equiv-
alent reflections were averaged. The structures were solved by direct
methods[101] and refined with full-matrix least-squares on F2 using
the program SHELXL-97.[102] Hydrogen atoms were placed at ideal
[2] S. Trofimenko, Scorpionates: The Coordination Chemistry of
Polypyrazolylborate Ligands, Imperial College Press, 1999.
[3] B. Cornils, W. A. Herrmann, Applied Homogeneous Catalysis
with Organometallic Compounds, Wiley-VCH, Weinheim, 1996.
[4] A. Haghiri Ilkhechi, J. M. Mercero, I. Silanes, M. Bolte, M.
Scheibitz, H.-W. Lerner, J. M. Ugalde, M. Wagner, J. Am.
Chem. Soc. 2005, 127, 10656–10666.
Eur. J. Inorg. Chem. 2007, 1989–2003
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2001

M. C. Holthausen, M. Wagner et al.
FULL PAPER
[5] S. Bieller, M. Bolte, H.-W. Lerner, M. Wagner, Chem. Eur. J.
2006, 12, 4735–4742.
[6] S. Bieller, M. Bolte, H.-W. Lerner, M. Wagner, Inorg. Chem.
2005, 44, 9489–9496.
[7] S. Bieller, M. Bolte, H.-W. Lerner, M. Wagner, J. Organomet.
Chem. 2005, 690, 1935–1946.
[8] A. Haghiri Ilkhechi, M. Bolte, H.-W. Lerner, M. Wagner, J.
Organomet. Chem. 2005, 690, 1971–1977.
[9] S. Bieller, F. Zhang, M. Bolte, J. W. Bats, H.-W. Lerner, M.
Wagner, Organometallics 2004, 23, 2107–2113.
[10] A. Haghiri Ilkhechi, S. Guo, M. Bolte, M. Wagner, Dalton
Trans. 2003, 2303–2307.
[11] T. I. Kückmann, M. Hermsen, M. Bolte, M. Wagner, H.-W.
Lerner, Inorg. Chem. 2005, 44, 3449–3458.
[12] F. Dornhaus, M. Bolte, H.-W. Lerner, M. Wagner, Eur. J. Inorg.
Chem. 2006, 5138–5147.
[13] T. S. Lobana, in Coordination Chemistry of Phosphane Chalcog-
enides and Their Analytical and Catalytic Applications, Vol. 2
(Ed.: F. R. Hartley), John Wiley & Sons, Chichester, 1992, pp.
409–566.
[40]
[41]
[42]
[43]
[44]
[45]
[46]
F. A. Cotton, C. S. Kraihanzel, J. Am. Chem. Soc. 1962, 84,
4432–4438.
P. S. Braterman, D. W. Milne, E. W. Randall, E. Rosenberg, J.
Chem. Soc., Dalton Trans. 1973, 1027–1031.
B. E. Mann, B. F. Taylor, 13C NMR Data for Organometallic
Compounds, Academic Press, London, 1981, p. 326.
P. M. Treichel, D. A. Komar, J. Organomet. Chem. 1981, 206,
77–88.
H. Schumann, L. Eguren, J. Organomet. Chem. 1991, 403, 183–
193.
E. E. Isaacs, W. A. G. Graham, J. Organomet. Chem. 1976, 120,
407–421.
M. L. Brown, J. L. Cramer, J. A. Ferguson, T. J. Meyer, N. Win-
terton, J. Am. Chem. Soc. 1972, 94, 8707–8710.
A. S. Goldman, D. R. Tyler, Inorg. Chem. 1987, 26, 253–258.
E. Louattani, A. Lledós, J. Suades, Organometallics 1995, 14,
1053–1060.
H. Schumann, J. Organomet. Chem. 1987, 320, 145–162.
U. Schubert, G. Kraft, E. Walther, Z. Anorg. Allg. Chem. 1984,
519, 96–106.
W. Malisch, H. Jehle, S. Möller, G. Thum, J. Reising, A. Gbu-
reck, V. Nagel, C. Fickert, W. Kiefer, M. Nieger, Eur. J. Inorg.
Chem. 1999, 1597–1605.
W. Malisch, H. Jehle, D. Schumacher, M. Binnewies, N. Söger,
J. Organomet. Chem. 2003, 667, 35–41.
J. S. McIndoe, B. K. Nicholson, J. Organomet. Chem. 2002,
648, 237–245.
T. S. Piper, D. Lemal, G. Wilkinson, Naturwissenschaften 1956,
43, 129.
W. Malisch, M. Kuhn, Chem. Ber. 1974, 107, 979–995.
R. J. P. Corriu, W. E. Douglas, J. Organomet. Chem. 1973, 51,
C3–C4.
G. Cerveau, E. Colomer, R. Corriu, W. E. Douglas, J. Or-
ganomet. Chem. 1977, 135, 373–386.
F. Dornhaus, M. Bolte, H.-W. Lerner, M. Wagner, Eur. J. Inorg.
Chem. 2006, 1777–1785.
H. Nakazawa, Y. Ueda, K. Nakamura, K. Miyoshi, Organome-
tallics 1997, 16, 1562–1566.
N. Kuhn, H. Schumann, J. Organomet. Chem. 1986, 304, 181–
193.
M. Morán, I. Cuadrado, C. Pascual, J. R. Masaguer, J. Losada,
Inorg. Chim. Acta 1987, 132, 257–262.
I. Kovács, F. Bélanger-Gariépy, A. Shaver, Inorg. Chem. 2003,
42, 2988–2991.
N. J. Coville, E. A. Darling, A. W. Hearn, P. Johnston, J. Or-
ganomet. Chem. 1987, 328, 375–385.
F. W. G. Fearon, H. Gilman, J. Organomet. Chem. 1967, 9,
403–411.
[47]
[48]
[49]
[50]
[14] L. H. Gade, Koordinationschemie, Wiley-VCH, Weinheim,
1998.
[15] T. D. Tilley, in The Chemistry of Organic Silicon Compounds,
Vol. 1 (Eds.: S. Patai and Z. Rappoport), John Wiley & Sons,
New York, 1989, pp. 1415–1477.
[51]
[52]
[53]
[54]
[16] J. Y. Corey, J. Braddock-Wilking, Chem. Rev. 1999, 99, 175–
292.
[17] H. Nakazawa, K. Kamata, M. Itazaki, Chem. Commun. 2005,
4004–4006.
[18] M. Okazaki, S. Ohshitanai, M. Iwata, H. Tobita, H. Ogino,
Coord. Chem. Rev. 2002, 226, 167–178.
[19] R. D. Brost, G. C. Bruce, F. L. Joslin, S. R. Stobart, Organome-
tallics 1997, 16, 5669–5680.
[20] D. L. Lichtenberger, A. Rai-Chaudhuri, J. Am. Chem. Soc.
1991, 113, 2923–2930.
[55]
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
[66]
[21] J. Chatt, C. Eaborn, S. Ibekwe, Chem. Commun. (London)
1966, 700–701.
[22] H. Tobita, K. Hasegawa, J. J. G. Minglana, L.-S. Luh, M. Oka-
zaki, H. Ogino, Organometallics 1999, 18, 2058–2060.
[23] J.-M. Camus, J. Andrieu, P. Richard, R. Poli, C. Darcel, S.
Jugé, Tetrahedron: Asymmetry 2004, 15, 2061–2065.
[24] B. Wolfe, T. Livinghouse, J. Am. Chem. Soc. 1998, 120, 5116–
5117.
[25] T. Miura, H. Yamada, S.-I. Kikuchi, T. Imamoto, J. Org.
Chem. 2000, 65, 1877–1880.
[26] B. Wolfe, T. Livinghouse, J. Org. Chem. 2001, 66, 1514–1516.
[27] H. Dorn, C. A. Jaska, R. A. Singh, A. J. Lough, I. Manners,
Chem. Commun. 2000, 1041–1042.
[28] A.-C. Gaumont, M. B. Hursthouse, S. J. Coles, J. M. Brown,
Chem. Commun. 1999, 63–64.
A. Sekiguchi, T. Fukawa, M. Nakamoto, V. Y. Lee, M. Ichi-
nohe, J. Am. Chem. Soc. 2002, 124, 9865–9869.
N. Wiberg, K. Amelunxen, H.-W. Lerner, H. Schuster, H.
Nöth, I. Krossing, M. Schmidt-Amelunxen, T. Seifert, J. Or-
ganomet. Chem. 1997, 542, 1–18.
[29] W. Angerer, W. S. Sheldrick, W. Malisch, Chem. Ber. 1985, 118,
1261–1266.
[30] D. A. Hoic, W. M. Davis, G. C. Fu, J. Am. Chem. Soc. 1996,
118, 8176–8177.
[31] V. V. Grushin, Chem. Rev. 2004, 104, 1629–1662.
[32] R. Weber, U. Englert, B. Ganter, W. Keim, M. Möthrath,
Chem. Commun. 2000, 1419–1420.
[67]
Crystallographic
data
for
[(18-crown-6)(thf)2Na][4,4Ј-
(SitBu2Me)2(C6H4)2]: empirical formula C58H106NaO10Si2; for-
mula mass 1042.60; green block; T = 173(2) K, crystal system
monoclinic; space group P21/c; unit cell dimensions a =
13.1292(8), b = 18.3379(8), c = 14.4381(8) Å, β = 113.233(4)°;
[33] C. Abu-Gnim, I. Amer, J. Organomet. Chem. 1996, 516, 235–
243.
V
= Z = 2; Dcalcd. = µ =
3194.3(3) Å3; 1.084 gcm–3;
[34]
B. Marciniec, H. Maciejewski, Coord. Chem. Rev. 2001, 223,
301–335.
S. E. Denmark, J. D. Baird, Chem. Eur. J. 2006, 12, 4954–4963.
P. T. Wolczanski, Polyhedron 1995, 14, 3335–3362.
J. L. Bennett, P. T. Wolczanski, J. Am. Chem. Soc. 1994, 116,
2179–2180.
0.112 mm–1; F(000) = 1146; crystal size 0.40ϫ0.23ϫ0.22 mm;
reflections collected = 40036; independent reflections = 5846
[R(int)= 0.0665]; data/restraints/parameters = 5846/12/318;
GOOF on F2 = 1.086; final R [I Ͼ 2σ(I)]: R1 = 0.0999, wR2 =
0.2692; R (all data): R1 = 0.1169, wR2 = 0.2834; largest diff.
peak and hole 0.716 and –0.528 eÅ–3. The solvent thf molecule
is disordered over two positions and was refined isotropically.
In order to keep the geometric parameters in reasonable ranges,
bond length restraints of C–O = 1.40(1) and C–C = 1.50(1) Å
as well as restraints of 2.5(1) Å for 1,3-distances were applied.
[35]
[36]
[37]
[38]
[39]
K. F. Hirsekorn, A. S. Veige, M. P. Marshak, Y. Koldobskaya,
P. T. Wolczanski, T. R. Cundari, E. B. Lobkovsky, J. Am.
Chem. Soc. 2005, 127, 4809–4830.
B. Marciniec, P. Krzyzanowski, E. Walczuk-Gusciora, W.
Duczmal, J. Mol. Catal. A 1999, 144, 263–271.
2002
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 1989–2003

Broadening the Scope of Ancillary Phosphane-Type Ligands
FULL PAPER
[68] T. H. Whitesides, J. Shelly, J. Organomet. Chem. 1975, 92, 215–
226.
[69] S. B. Fergusson, L. J. Sanderson, T. A. Shackleton, M. C.
Baird, Inorg. Chim. Acta 1984, 83, L45–L47.
[70] M. L. H. Green, C. N. Street, G. Wilkinson, Z. Naturforsch.,
Teil B 1959, 14, 738.
[71] A. F. Holleman, N. Wiberg, Lehrbuch der Anorganischen
Chemie, Walter de Gruyter, Berlin, 1995, p. 2033.
perimentally determined frequencies revealed a systematic de-
viation. A scaling factor of 0.9884 was therefore employed to
improve the overall agreement.
[86] W. Koch, M. C. Holthausen, A Chemist’s Guide to Density
Functional Theory, Wiley-VCH, Weinheim, 2001.
[87] A. Schäfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100,
5829–5835.
[88] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. J. A. Montgomery, T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tom-
asi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratch-
ian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Sal-
vador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Ba-
boul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Lia-
shenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T.
Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M.
Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, C. Gonzalez, J. A. Pople, Gaussian 03, Revision B.3,
Gaussian, Inc., Pittsburgh, PA, 2003.
[72] It appears that rather than attacking the iron atom, the nucleo-
philic siloxide reacts with the iodide ligand to form an [Fp]–
fragment and intermediate IOSiPh2Me. The complex iron
anion subsequently reacts with one equivalent of FpI to give
Fp2. We speculate that the intermediate silicon-containing spe-
cies reacts with one equivalent of siloxide ligand to give disilox-
ane O(SiPh2Me)2 and NaIO. The disiloxane is also the major
product when Na7 is treated with I2. This can be thought of
as analogous to the reaction of I2 with OH–, which results in
IO3 , I–, and H2O.[71] When the same reaction is carried out
–
using FpBr rather than FpI, an identical outcome is observed.
However, the formation of Fp2 and O(SiPh2Me)2 from Na7
and FpI is slow, with about 50% of the starting material left
after stirring for two weeks at room temperature. This slow
reaction rate allows for the isolation of an adduct of Na7 and
Fp2 in which the siloxide sodium atoms are supported by do-
nor–acceptor interactions with the oxygen atoms of the bridg-
ing carbonyl ligands in Fp2. The details of this adduct forma-
tion have been discussed recently in another context.[73]
[73] T. I. Kückmann, M. Bolte, H.-W. Lerner, M. Wagner, Z. Anorg.
Allg. Chem. 2007, 633, 290–297.
[74] F. H. Allen, Acta Crystallogr., Sect. B 2002, 58, 380–388)
(Cambridge Crystallographic Database, CSD Version 5.27,
November 2005; update August 2006).
[75] L. Párkányi, K. H. Pannell, C. Hernandez, J. Organomet.
Chem. 1983, 252, 127–132.
[89] Spartan 06.1, Wavefunction, Inc., Irvine, CA, 2006.
[90] It is reasonable to assume that the final products of hydride
abstraction for both the oxide 5b and the phosphanyl borohyd-
ride 2b will be cyclic dimers or polymers of the corresponding
boranes. Nonetheless, the first intermediates following hydride
abstractions are key species along the reaction pathways, and
these are likely to be as described above.
[91] H. E. Gottlieb, V. Kotlyar, A. Nudelman, J. Org. Chem. 1997,
[76] R. West, E. K. Pham, J. Organomet. Chem. 1991, 402, 215–
220.
62, 7512–7515.
[92] V. D. Bianco, S. Doronzo, Inorg. Synth. 1976, 16, 161–163.
[93] G. Stieglitz, B. Neumüller, K. Dehnicke, Z. Naturforsch. Teil B
1993, 48, 156–160.
[77] N. Kuhn, H. Schumann, J. Organomet. Chem. 1984, 276, 55–
66.
[78] L. D. Quin, A Guide to Organophosphorus Chemistry, John
Wiley and Sons, Inc., New York, 2000, p. 394.
[94] T. J. Barton, C. R. Tully, J. Org. Chem. 1978, 43, 3649–3653.
[95] N. Wiberg, C.-K. Kim, Chem. Ber. 1986, 119, 2980–2994.
[96] T. Ohishi, Y. Shiotani, M. Yamashita, J. Org. Chem. 1994, 59,
250.
[97] D. Catheline, D. Astruc, Organometallics 1984, 3, 1094–1100.
[98] D. L. Reger, C. Coleman, J. Organomet. Chem. 1977, 131, 153–
162.
[79] S. Berger, S. Braun, H.-O. Kalinowski, 31P-NMR-Spektrosko-
pie, Thieme Verlag, Stuttgart, 1993.
[80] H. A. Bent, Chem. Rev. 1961, 61, 275–311.
[81] P. M. Treichel, W. K. Dean, W. M. Douglas, J. Organomet.
Chem. 1972, 42, 145–158.
[82] M. Green, D. J. Westlake, J. Chem. Soc. A. 1971, 367–371.
[83] P. M. Treichel, R. L. Shubkin, K. W. Barnett, D. Reichard, In-
org. Chem. 1966, 5, 1177–1181.
[84] A. Schäfer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97,
2571–2577.
[99] R. H. Blessing, Acta Crystallogr., Sect. A 1995, 51, 33–38.
[100] A. L. Spek, Acta Crystallogr., Sect. A 1990, 46, C34.
[101] G. M. Sheldrick, Acta Crystallogr., Sect. A 1990, 46, 467–473.
[102] G. M. Sheldrick, SHELXL-97, A Program for the Refinement
of Crystal Structures, Universität Göttingen, Germany, 1997.
Received: November 21, 2006
[85] A comparison of the averaged computed harmonic CO stretch-
ing frequencies [obtained as (ν + ν )/2] with the averaged ex-
Published Online: April 4, 2007
˜
˜
as
s
Eur. J. Inorg. Chem. 2007, 1989–2003
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
2003