Synthesis, structures, and properties of the phosphonium-1-indenylide (PHIN) Ligands 1-C9H6PPh3, 1-C 9H6PMePh2, and 1-C9H 6PMe2Ph and of the corresponding ruthenium(II) Complexes [Ru(η5-C5H5)(η5-PHIN)] PF6

Article abstract of DOI:10.1021/om200545j

Syntheses of the phosphonium-1-indenylide (PHIN) ligands triphenylphosphonium-1-indenylide (1-C₉H₆PPh₃, I), methyldiphenylphosphonium-1-indenylide (1-C₉H ₆PMePh₂, II), and dimethylphenylphosphonium-1-indenylide (1-C₉H₆PMe₂Ph, III) are reported, as are syntheses of the corresponding planar chiral ruthenium(II) complexes [Ru(η⁵-C₅H₅)(η⁵-1-C ₉H₆PPh₃)]PF₆ (IV), [Ru(η⁵-C₅H₅)(η⁵-1-C ₉H₆PMePh₂)]PF₆ (V), and [Ru(η⁵-C₅H₅)(η⁵-1-C ₉H₆PMe₂Ph)]PF₆ (VI). The ruthenium complexes have been characterized by ¹H, ¹³C, and ³¹P NMR spectroscopy, by X-ray crystallography, and by extensive DFT calculations, which produce optimized geometries consistent with the crystallographic data. The PHIN-Ru bond strengths are calculated to be ～20 kcal/mol greater than the corresponding benzene-Ru bond strength of [Ru(η⁵-C₅H₅)(η⁶-C ₆H₆)]⁺ and are compatible with the observed configurational stability of the complexes. That interconversion of enantiomers via interfacial exchange of the η⁵-bound ligands does not occur is demonstrated by the observation of diastereotopic phenyl groups in the ¹H NMR spectrum of V and of diastereotopic methyl groups in the ¹H NMR spectrum of VI.

Full text of DOI:10.1021/om200545j

Article
pubs.acs.org/Organometallics
Synthesis, Structures, and Properties of the Phosphonium-
1-indenylide (PHIN) Ligands 1-C9H6PPh3, 1-C9H6PMePh2, and
1-C9H6PMe2Ph and of the Corresponding Ruthenium(II) Complexes
[Ru(η5-C5H5)(η5-PHIN)]PF6
Kevin G. Fowler, Shalyn L. Littlefield, and Michael C. Baird*
Department of Chemistry, Queen’s University, Kingston, Ontario K7L 3N6, Canada
Peter H. M. Budzelaar*
Department of Chemistry, University of Manitoba, Winnipeg, Manitoba R3T 2N2, Canada
S
* Supporting Information
ABSTRACT: Syntheses of the phosphonium-1-indenylide (PHIN) ligands triphenylphosphonium-1-
indenylide (1-C9H6PPh3, I), methyldiphenylphosphonium-1-indenylide (1-C9H6PMePh2, II), and
dimethylphenylphosphonium-1-indenylide (1-C9H6PMe2Ph, III) are reported, as are syntheses of the
corresponding planar chiral ruthenium(II) complexes [Ru(η 5-C5H5)(η5-1-C9H6PPh3)]PF6 (IV),
[Ru(η5-C5H5)(η5-1-C9H6PMePh2)]PF6 (V), and [Ru(η5-C5H5)(η5-1-C9H6PMe2Ph)]PF6 (VI). The
1
ruthenium complexes have been characterized by H, 13C, and 31P NMR spectroscopy, by X-ray
crystallography, and by extensive DFT calculations, which produce optimized geometries consistent
with the crystallographic data. The PHIN−Ru bond strengths are calculated to be ∼20 kcal/mol
greater than the corresponding benzene−Ru bond strength of [Ru(η 5-C5H5)(η6-C6H6)]+ and are
compatible with the observed configurational stability of the complexes. That interconversion of
enantiomers via interfacial exchange of the η5-bound ligands does not occur is demonstrated by the observation of diastereotopic
1
1
phenyl groups in the H NMR spectrum of V and of diastereotopic methyl groups in the H NMR spectrum of VI.
little systematic attention, in spite of its clear potential. We
therefore initiated an investigation of transition-metal coordi-
nation complexes of phosphonium cyclopentadienylides and
reported the synthesis and reactivity of the group 6 compounds
M(η5-C5H4PMePh2)(CO)3 (M = Cr, Mo, W), containing
C5H4PMePh2, methyldiphenylphosphonium cyclopentadieny-
lide.3 This ligand was chosen over the Ramirez ylide because
much less had been done with it and the methyl group
provided a useful NMR “handle”. We found, inter alia, that the
donor properties of C5H4PMePh2 fall between those of
benzene and the cyclopentadienyl (η5-C5H5) ligand,3 con-
sistent with the presence of a partial negative charge on the five-
membered ring and suggesting that there would be interesting
comparisons and contrasts to be made with the coordination
chemistry of similar arene and cyclopentadienyl complexes.
We also investigated the electronic structures of
C5H4PMePh2 and Cr(η5-C5H4PMePh2)(CO)3 using ab initio
methodologies, finding that the near-degenerate HOMO and
HOMO-1 orbitals of the free ylide exhibit symmetries very
similar to those of the corresponding, doubly degenerate
HOMO of the cyclopentadienyl anion (E1 symmetry). We also
found that a lower energy, almost fully symmetric orbital
corresponds to the fully symmetric bonding A1 MO of the
he compound triphenylphosphonium cyclopentadienylide,
C5H4PPh3 (Figure 1), was first reported in 1956 by
T
Figure 1. Resonance structures of C5H4PPh3.
Ramirez and Levy,1 who found inter alia that this ylide is
unusually inert. Unlike normal ylides, for instance, it does not
react with ketones. This unusual stability was attributed to the
electron delocalization implied by the zwitterionic resonance
structure b, consistent with the relatively high dipole moment
of 7.0 D.1c
Further evidence for the relevance of b was found
crystallographically in the P−C5H4 bond length and the C−C
distances in the five-membered ring1g and in the 13C NMR
spectrum, which exhibited an unusually high field chemical shift
for the ylide carbon and a P−C(ylide) coupling constant typical
of an aliphatic carbon−P bond.1h Phosphonium cyclopentadie-
nylides of this type are thus isoelectronic with neutral arene and
anionic cyclopentadienyl ligands and are expected to exhibit an
extensive coordination chemistry.2
That said, however, as we noted in 2007,3a the coordination
chemistry of this class of very interesting ligands has received
Received: June 25, 2011
Published: October 24, 2011
© 2011 American Chemical Society
6098
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107

Organometallics
Article
cyclopentadienyl anion. In none of these three orbitals did
there appear to be significant π involvement with an orbital on
the phosphorus atom, consistent with the low P−C bond order
implicit in zwitterionic structure b. The primary ylide−metal
interactions in Cr(η5-C5H4PMePh2)(CO)3 were found to
involve donation of the HOMO and HOMO-1 orbitals into
the dxz and dyz orbitals of the chromium, respectively, while the
calculated ylide−Cr(CO)3 bond dissociation energy was about
30% higher than the analogous ring−metal bond dissociation
energy calculated for (η6-C6H6)Cr(CO)3.
In order to extend the chemistry of this class of ligand,
we subsequently began an investigation of the analogous
phosphonium-1-indenylide (PHIN) ligands, beginning with
methyldiphenylphosphonium-1-indenylide, 1-C9H6PMePh2
(Figure 2).4
Figure 4. Ruthenium complexes studied.
RESULTS AND DISCUSSION
■
Ligand Syntheses and NMR Characterization. We
previously prepared II via a procedure which involved the
synthesis of P(1-indenyl)Ph2 via reaction of lithium indenylide
with chlorodiphenylphosphine, followed by methylation and
deprotonation steps as shown in Figure 5.
Figure 2. Resonance structures of 1-C9H6PMePh2.
The first such ligand, triphenylphosphonium-1-indenylide (1-
C9H6PPh3), was reported in the 1960s, although it was not
characterized by NMR spectroscopy or crystallography.5 Two
much better characterized analogues, 1-C9H6P(CH2Ph)Ph2 and
1-C9H6P(CH2C6F5)Ph2, were subsequently reported in 2004,6
but the coordination chemistry of these three aromatic ligands
has not been investigated and, indeed, to our knowledge no
other indenyl-derived phosphorus ylides have been reported.
Adding to our interest, ligands of the type 1-C9H6PR3 are
planar prochiral, with the result that their coordination
compounds exhibit planar chirality (Figure 3). Thus, as an
Figure 5. Previously utilized procedure for the synthesis of II.
This procedure is not of general utility, however, as few
dialkyl- or diarylmonochlorophosphines are readily available.
We have therefore developed the more general route shown in
Figure 6, which also shows the ring atom numbering scheme
Figure 6. General procedure for the synthesis of phosphonium-1-
indenylides (R = methyl, phenyl) and numbering scheme for the ring
sites.
Figure 3. Enantiomers of the chiral complex Cr(η5-1-C9H6PMePh2)-
used below in discussions of the NMR spectra and the crystal
structures of the PHIN ligands I−III and their coordination
complexes IV−VI.
(CO)3.
example, we reported the synthesis and characterization,
including crystallography, of the planar chiral chromium
compound Cr(η5-1-C9H6PMePh2)(CO)3.4 On the basis of
the IR spectrum in the carbonyl region, the indenylide ligand
was found to exhibit donor properties very similar to those of
its cyclopentadienylide analogue.4
We have now extended the list of well-characterized
phosphonium-1-indenylide (PHIN) ligands and report herein
the synthesis and full characterization of the aforementioned
triphenylphosphonium-1-indenylide (1-C9H6PPh3, I), a new
procedure for the synthesis of the known 1-C9H6PMePh2 (II),
and the synthesis and characterization of the new ligand
dimethylphenylphosphonium-1-indenylide (1-C9H6PMe2Ph,
III). Furthermore, with a view to extending and investigating
the coordination chemistry of PHIN ligands to complexes of
metals in oxidation states higher than 0, we also describe the
synthesis, structures, and mode of bonding of the ruthenium-
(II) complexes [Ru(η5-C5H5)(η5-1-C9H6PPh3)]PF6 (IV), [Ru-
(η 5-C5H5)(η 5-1-C9H6PMePh2)]PF6 (V), and [Ru(η 5-
C5H5)(η5-1-C9H6PMe2Ph)]PF6 (VI) (Figure 4).
Thus, 1-bromoindene, prepared via the bromination of 1-
trimethylsilylindene by dioxane dibromide,7 was reacted with
the tertiary phosphines PPh3, PMePh2, and PMe2Ph as shown
in Figure 6 to form the corresponding phosphonium salts as
mixtures of regioisomers A and B as indicated. The product
1
mixtures were characterized by H and 31P NMR spectroscopy,
and as an example, the 1H NMR spectrum of (1-C9H7PMe2Ph)
Br is shown in Figure 7.
As can be seen, the olefinic and aromatic resonances of the
pairs of regioisomers overlap considerably in the region δ 6.3−
8.3, and assignments of these resonances were not attempted.
In contrast, the methylene resonance of isomer A and the
methyl resonances of both A and B are readily identified.
Interestingly, since isomer B contains a chiral center at C(1),
the two methyl groups are diastereotopic and exhibit distinct
1H resonances as shown.
The 1H NMR spectra of (1-C9H7PPh3)Br and (1-
C9H7PMePh2)Br (Figures S1 and S2, respectively, in the
6099
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107

Organometallics
Article
(1)
The reactions were generally completed within a few minutes
on stirring equimolar amounts of [Ru(η 5-C5H5)(MeCN)3]PF6
and a PHIN ligand in THF at room temperature, and
crystallographic quality crystals of the complexes [Ru(η 5-
C5H5)(η 5-1-C9H6PPh3)]PF6 (IV), [Ru(η 5-C5H5)(η 5-1-
C9H6PMePh2)]PF6 (V), and [Ru(η 5-C5H5)(η 5-1-
C9H6PMe2Ph)]PF6 (VI) were obtained by slow evaporation
from saturated CH2Cl2 solutions (see below). Unfortunately,
bulk purification of the complexes was not readily accom-
plished, as attempted recrystallization from a variety of solvents
and solvent mixtures generally failed to produce pure materials.
Washing the products with benzene was found to remove some
nonpolar impurities but, judging from the elemental analyses
1
Figure 7. H NMR spectrum of (1-C9H7PMe2Ph)Br.
Supporting Information) were also complex in the olefinic and
aromatic regions, but the chemical shifts of the various proton
environments are clearly similar to the corresponding chemical
shifts of (1-C9H7PMe2Ph)Br. The 31P NMR spectra of each of
the three phosphonium salts exhibited two distinct 31P
resonances (see the Experimental Section), and thus the ratios
of regioisomers were readily obtained. The relative amounts
were found to vary seemingly randomly from batch to batch,
but we have not attempted to ascertain the reason(s).
The free PHIN ligands 1-C9H6PPh3 (I), 1-C9H6PMePh2
(II), and 1-C9H6PMe2Ph (III) were readily synthesized in good
yields from their respective phosphonium salts by deprotona-
tion using an excess of NaH in THF. Ylide II has been
previously characterized by NMR spectroscopy and X-ray
crystallography,4 and ylides I and III have now also been
characterized by crystallography (see below) and by NMR
spectroscopy utilizing COSY, NOESY, HSQC, and
1
and the H NMR spectra, attempts to remove of all traces of
impurities were generally unsuccessful.
Nonetheless, complexes IV−VI were all unambiguously
characterized as such by X-ray crystallography, by electrospray
mass spectrometry, and by 1H, 31P, and 13C NMR spectroscopy.
Crystal structures of the three compounds are discussed below,
while experimental and simulated ES mass spectra of VI are
shown in Figure 9 and those of IV and V in Figure S5 of the
Supporting Information. The isotopic patterns observed for all
three complex cations are in excellent agreement with the
calculated spectra.
Chemical shift and coupling constant data are given in Tables
1−3, comparisons with the corresponding data for the free
1
PHIN ligands being readily obvious. The H spectrum of VI is
shown in Figure 10 and those of IV and V are shown
respectively in Figures S6 and S7 of the Supporting
Information.
1
HMBC experiments. The H NMR spectrum of III is shown
in Figure 8, NMR spectroscopic data (1H, 13C) for all three
Of specific interest, coordination of the three planar-prochiral
PHIN ligands resulted in the formation of enantiomeric pairs,
depending on which face of the five-membered rings
coordinates (as in Figure 3). The result is that the methyl
groups of VI and the phenyl groups of V are diastereotopic, as
with regioisomer a of the phosphonium salts (1-C9H7PMe2Ph)
Br and (1-C9H7PMePh2)Br, respectively (see above). There-
fore, in contrast to the 1H NMR spectrum of free 1-
1
C9H7PMe2Ph (III; Figure 8), the H NMR spectrum of VI
exhibited two methyl resonances (Figure 10) while that of IV
exhibited two sets of phenyl resonances (although these were
difficult to distinguish). A corollary of these findings is, of
course, that inter- and intramolecular interfacial exchange of the
coordinated PHIN ligands (enantiomeric exchange) must be
slow on at least the NMR time scale.
1
Figure 8. H NMR spectrum of 1-C9H6PMe2Ph (III).
ylides are given in Tables 1−3, and 1H NMR spectra of I and II
are shown in Figures S3 and S4, respectively, of the Supporting
Information.
As can be seen from Figure 8, the methyl resonance of III is
observed as the expected doublet, with JP−H = 13.2 Hz, and the
resonances of the 1-indenyl group, H(2)−H(8), are upfield of
and well separated from those of the phenyl group. Similar
upfield shifts are observed for I and II (Tables 1 and 2, Figures
S3 and S4), and the JP−H(Me) values of II and III are very similar
to that of C5H4PMePh2.3a
As is also clear from Tables 1−3 and Figures 8 and 10 and
Figures S3, S4, S6, and S7, formation of complexes IV−VI
results in significant upfield coordination shifts of the
resonances of H(2) and H(3), about 2 and 1 ppm, respectively.
Very similar coordination shifts were observed for the
4
compound Cr(η5-1-C9H6PMePh2)(CO)3 and also for H(2,5)
and H(3,4) of the five-membered rings in the group 6
methyldiphenylphosphonium cyclopentadienylide complexes
M(η5-C5H4PMePh2)(CO)3 (M = Cr, Mo, W).3a In contrast,
the resonances of H(5)−H(8) and of the phenyl groups in the
NMR spectra of compounds IV−VI change relatively little,
consistent with metal−ylide bonding involving only the five-
membered rings.
Ruthenium Complex Syntheses and NMR Character-
ization. Ruthenium coordination complexes of the three
phosphonium-1-indenylides were obtained via substitution
8
reactions of the labile complex [Ru(η5-C5H5)(MeCN)3]PF6
(eq 1).
6100
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107

Organometallics
Article
1
Table 1. H and 13C NMR Data for I and IV
I
IV
H, C position
δ(1H)
δ(13C)
δ(1H)
δ (13C)
1
1
1
66.52 (d, JP−C = 123)
57.7 (d, JC−P = 104.8 Hz)
3
2
2
6.61 (t, JH−H = JP−H = 4.53)
6.56 (t, JH−H = JP−H = 4.16)
128.2 (d, JP−C = 16.0)
4.66 (s)
5.86 (s)
79.2 (d, JC−P = 14.5 Hz)
2
3
3
106.4 (d, JP−C = 16.0)
71.6 (d, JC−P = 11.2 Hz)
2
3
4
135.7 (d, JP−C = 13.5)
96.2 (d, JC−P = 12.3 Hz)
3
5
7.66 (?, obscured)
120.4 (s)
117.7 (s)
118.0 (s)
117.2 (s)
6.84 (d, JH−H = 8.9 Hz)
123.6 (s)
127.1 (s)
126.0 (s)
127.7 (s)
3
3
6
6.95 (t, JH−H = 7.18)
7.03 (td, JH−H = 7.3 Hz)
3
3
7
6.76 (t, JH−H = 7.18)
7.14 (td, JH−H = 7.7 Hz)
3
3
8
6.89 (d, JH−H = 7.93)
7.67 (d, JH−H = 8.5 Hz)
3
2
9
137.8 (d, JP−C = 14.8)
97.7 (d, JC−P = 9.7 Hz)
1
1
ipso-C
o-C
m-C
p-C
C5H5
125.8 (d, JP−C = 89.8)
120.3 (d, JC−P = 92.2 Hz)
3
2
7.68−7.65 (m)
133.8 (d, JP−C = 9.84)
7.82 (m)
7.76 (m)
8.00 (m)
4.35 (s)
134.4 (d, JC−P = 10.9 Hz)
2
3
7.51 (td, JH−H = 3.02, 7.93)
129.1 (d, JP−C = 12.3)
130.6 (d, JC−P = 12.4 Hz)
3
4
4
7.62 (t, JH−H = 7.55)
133.6 (d, JP−C = 3.7)
135.8 (d, JC−P = 3.8 Hz)
73.8 (s)
1
Table 2. H and 13C NMR Data for II and V
II
V
H, C position
δ(1H)
δ(13C)
δ(1H)
δ(13C)
1
1
1
66.14 (d, JP−C = 120.8)
58.2 (d, JC−P = 103.4 Hz)
2
2
2
6.74 (t, JH−H = JP−H = 4.5)
6.46 (t, JH−H = JP−H = 4.2)
126.30 (d, JP−C = 17.6)
4.64 (s)
5.79 (s)
77.8 (d, JC−P = 14.4 Hz)
3
1
3
105.00 (d, JP−C = 15.4)
70.9 (d, JC−P = 9.6 Hz)
2
3
4
135.42 (d, JP−C = 14.3)
94.3 (d, JC−P = 13.2 Hz)
3
5
7.68 (m)
120.82 (s)
117.28 (s)
117.91 (s)
117.36 (s)
7.59 (t, JH−H = 8.9 Hz)
126.7 (s)
123.1 (s)
125.4 (s)
126.2 (s)
3
3
6
6.97 (t, JH−H = 7.2)
6.93 (d, JH−H = 8.0 Hz)
3
3
7
6.84 (t, JH−H = 6.8)
7.08 (d, JH−H = 7.9 Hz)
3
3
8
7.04 (d, JH−H = 7.9)
6.98 (d, JH−H = 6.0 Hz)
3
2
9
137.79 (d, JP−C = 15.4)
97.0 (d, JC−P = 8.2 Hz)
2
1
2
1
P-Me
ipso-C
o-C
m-C
p-C
C5H5
2.50 (d, JP−H = 12.6)
12.97 (d, JP−C = 62.6)
2.88 (d, JP−H = 13.3 Hz)
13.3 (d, JC−P = 61.9 Hz)
1
1
127.13 (d, JP−C = 87.8)
119.8 (d, JC−P = 40.8 Hz)
2
2
7.55−7.52
7.67−7.63
7.67−7.53
129.45 (d, JP−C = 12.1)
7.71 (m)
132.6 (d, JC−P = 10.9 Hz)
3
3
132.68 (d, JP−C = 11.0)
7.64 (m)
130.1, 130.3 (d, JC−P = 12.8 Hz)
4
4
132.93 (d, JP−C = 3.3)
7.83 (dd, JH−H = 8.9 Hz)
4.45 (s)
135.1, 135.2 (d, JC−P = 3.2 Hz)
73.2
1
Table 3. H and 13C NMR Data for III and VI
III
VI
H, C
position
δ(1H)
δ(13C)
δ(1H)
δ(13C)
1
1
1
2
3
4
5
6
7
8
9
66.69 (d, JP−C = 121)
58.4 (d, JP−C = 102)
2
2
6.98 (t, JH−H = JP−H = 4.53) 124.2 (d, JP−C = 16.5) 5.00 (s)
77.0 (d, JP−C = 14.5)
3
3
6.64 (t, JH−H = JP−H = 4.53) 105.8 (d, JP−C = 14.8) 5.83 (s)
71.5 (d, JP−C = 9.6)
2
3
134.4 (d, JP−C = 14.8)
94.3 (d, JP−C = 14.5)
3
4
3
7.68 (d, JH−H = 8.69)
120.6 (d, JP−C = 1.85) 7.61 (d, JH−H = 8.38)
126.8 (s)
125.9 (s)
125.7 (s)
123.1 (s)
3
3
6.95 (t, JH−H = 7.55)
116.9 (s)
117.6 (s)
7.13 (d, JH−H = 7.50)
3
3
6.87 (t, JH−H = 7.55)
7.10 (d, JH−H = 7.50)
3
3
3
7.18 (d, JH−H = 7.93)
116.8 (d, JP−C = 1.85) 7.21 (d, JH−H = 8.52)
2
2
137.1 (d, JP−C = 13.0)
97.3 (d, JP−C = 8.0)
2
2
P-Me
ipso-C
o-C
2.21 (d, JP−H = 13.2)
12.92 (d, JP−C = 61.0) 2.70 (d, 2JP−H = 13.2), 2.44 (d, 2JP−H = 13.2) 11.8 (d, 2JP−C = 62.1), 13.7 (d, 2JP−C = 62.1)
128.3 (1JP−C = 84.2)
121.4 (1JP−C = 86.5)
2
3
7.64 (dd, JH−H = 7.30, 13.0) 131.0 (d, JP−C = 11.1) 7.77 (dd, JH−H = 5.58, 12.71)
131.5 (d, JP−C = 10.6)
3
2
m-C
7.49 (td, JH−H = 2.27, 7.56) 129.3 (d, JP−C = 12.0) 7.67 (td, JH−H = 3.25, 7.59)
130.6 (d, JP−C = 12.4)
4
4
p-C
7.58 (t, JH−H = 7.93)
132.6 (d, JP−C = 2.77) 7.37 (m)
135.1 (d, JP−C = 3.1)
C5H5
4.51 (s)
73.2 (s)
We note that the η5-C5H5 resonances of complexes IV (δ
4.35), V (δ 4.45), and VI (δ 4.55) correlate with the electron-
donating abilities of the PHIN ligands expected on the basis of
the substituents on phosphorus and are all somewhat
deshielded relative to the corresponding resonance of their
precursor, [Ru(η5-C5H5)(MeCN)3]PF6 (δ 4.29). The 31P
6101
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107

Organometallics
Article
Figure 9. Experimental (a) and simulated (b) mass spectra of VI.
Figure 12. Molecular structure of 1-C9H6PPh3 (III). Displacement
ellipsoids for non-H atoms are shown at the 50% probability level, and
H atoms are represented by circles of arbitrary size.
were obtained by slow evaporation of saturated CH2Cl2
solutions and the structures are shown in Figures 13−15,
Figure 10. 1H NMR spectrum of [Ru(η 5-C5H5)(η 5-1-
C9H6PMe2Ph)]PF6 (VI).
chemical shifts of I−VI are δ 1.60, 4.83, 10.4, 22.4, 23.3, and
24.3, respectively, and thus coordination of the ylides to Ru(II)
results in downfield coordination shifts of about 14−21 ppm.
These changes compare with downfield shifts of ∼14 ppm for
4
Cr(η 5-1-C9H6PMePh2)(CO)3 and 10−12 ppm for the
compounds M(η5-C5H4PMePh2)(CO)3 (M = Cr, Mo, W).3a
Crystallographic Characterization of Ligands and
Ruthenium Complexes. X-ray-quality crystals of I and III
were obtained by recrystallization from CH2Cl2 solution
layered with hexanes, and the structures are shown in Figures
11 and 12, respectively. Similarly, crystals of compounds IV−VI
Figure 13. Molecular structure of [Ru(η 5-C5H5)(η5-1-C9H6PPh3)]PF6
(IV). Displacement ellipsoids for non-H atoms are shown at the 50%
probability level, and H atoms are represented by circles of arbitrary size.
respectively. Important bond lengths and angles for compounds
I−VI are shown in Table 4, and full crystallographic
information is available in the Supporting Information.
With reference to Figure 2, an important structural parameter
for assessing the relative contributions of the resonance
structures a and b to the overall electronic structure involves
the bond lengths and angles of the 1-indenyl-P moiety. The P−
C9H6 bond lengths of I−III are respectively 1.7284(19),
1.711(2), and 1.727(3) Å, all significantly shorter than the
corresponding P−Ph bonds. The latter average 1.810, 1.788,
and 1.817 Å, respectively, all falling within the typical range for
C−Ph single bonds in phenyl phosphonium ylides.9 On the
other hand, the P−C9H6 bond lengths of I−III are significantly
longer than the PCH2 bond in Ph3PCH2 (1.66 Å);9 the
latter is a typical example of a non-resonance-stabilized ylide,
and its PCH2 bond is believed to contain considerable
double-bond character.9a Thus, the P−C9H6 bond lengths of I−
III are consistent with significant contributions from both the
Figure 11. Molecular structure of 1-C9H6PPh3 (I). Displacement
ellipsoids for non-H atoms are shown at the 50% probability level, and
H atoms are represented by circles of arbitrary size.
6102
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107

Organometallics
Article
somewhat on coordination, and thus the zwitterionic structure
b apparently becomes more significant in the complexes than
is the case for the free ligands. Similar behavior was observed
with group 6 metal compounds of C5H4PMePh2 and 1-
3,4
C9H6PMePh2 and presumably reflects enhanced aromatiza-
tion of the cyclopentadienyl-like ligands on coordination to the
low-spin d6 metal systems.
There are, however, marked differences in the lengths of the
bonds between the ruthenium atoms and the ylidic carbon
atoms of complexes IV−VI. While the Ru−C(1), Ru−C(2),
and Ru−C(3) bond lengths are similar in each case and only
slightly longer (averages 2.17−2.19 Å) than the averages of the
Ru−C5H5 bond lengths, the Ru−C(4) and Ru−C(9) bonds are
significantly longer (averages 2.23−2.25 Å). Similar behavior
was observed with the aforementioned C5H4PMePh2 and 1-
C9H6PMePh2 group 6 metal compounds.3,4 As a result of these
differences, the five-membered ylidic rings of IV−VI are tilted
at angles of 5.63, 6.90, and 4.94°, respectively, relative to the
corresponding C5H5 rings.
Somewhat similar behavior is also observed in many
complexes containing coordinated indenyl ligands, where
indenyl “slippage” to what is essentially an η3 allyl-ene mode
of bonding is observed.10b In these cases, however, there is
considerable loss of planarity of the indenyl moiety such that
the angle between the C(1)−C(2)−C(3) plane and the C(1)−
C(3)−C(4)−C(9) plane (the hinge angle) is typically 20−30°
while the angle between the C(1)−C(2)−C(3) plane and the
C(4)−C(5)−C(6)−C(7)−C(8)−C(9) plane (the fold angle)
is typically >12°.10b In IV−VI, the hinge and fold angles are
2.91−3.72 and 4.32−5.89°, respectively, much more typical of
essentially planar η5-indenyl complexes.10b
Computational Studies. For purposes of comparison and
to gain insight into the nature of the metal−ligand bonding, we
have calculated structures for the free ligands C5H4PMe3,
C5H4PF3, 1-C9H6PMePh2 (II), 1-C9H6PMe2Ph (III), 1-
C9H6PF3, and C6H6, as well as for their corresponding
[Ru(η5-C5H5)]+ complexes. Geometries were optimized at
the b3-lyp/TZVP level, and improved single-point energies
were calculated at the b3-lyp/TZVPP level. Listings of the
Cartesian coordinates are available in the Supporting
Information (Table S1).
Figure 14. Molecular structure of [Ru(η5-C5H5)(η5-1-C5H4PMePh2)]PF6
(V). Displacement ellipsoids for non-H atoms are shown at the 50%
probability level, and H atoms are represented by circles of arbitrary size.
uncharged and the zwitterionic resonance structures a and b in
Figure 2.
Interestingly, although the C−C−C bond angles within the
five-membered rings of all three ylides are all very close to the
108° of a regular pentagon, moderately alternating patterns of
the C−C bond lengths seemingly provide evidence for a
degree of localization of π bond electron density. Thus, the
C(2)−C(3), C(5)−C(6), and C(7)−C(8) bonds of all three
ligands are “short” and the C(1)−C(2), C(3)−C(4), C(4)−
C(5), C(6)−C(7), and C(8)−C(9) bonds are “long”, as in a
and b of Figure 2. Thus, again there is evidence for a very
significant contribution from both the uncharged and the
zwitterionic resonance structures, although the C(4)−C(9)
bonds are “long”, as is the case with the analogous bond of
naphthalene.10a
The complexes IV−VI all assume sandwich structures, as
anticipated. The five-membered rings are almost exactly
eclipsed in all cases, and the phosphine moieties are oriented
such that a phenyl group is always trans to the η5-C5H5 ligand
while the other two groups (Me, Ph) straddle the η5-C5H5
ligand. The bond length alternation of the C(5)−C(6), C(6)−
C(7), and C(7)−C(8) bonds of the six-membered ring
becomes more pronounced on coordination, but the C(2)−
C(3) bonds lengthen such that the C−C bond lengths of the
five-membered rings are all rather similar, with the average
being ∼1.44 ± 0.02 Å. The P−C9H6 bond lengths also increase
The calculated geometries of both the 1-C9H6 ligands and
their complexes (Table 5) agree well with the observed X-ray
structures. In particular, the free ligands all show a “short”
Figure 15. Molecular structure of 1-C9H6PPh3 (VI). Displacement ellipsoids for non-H atoms are shown at the 50% probability level, and H atoms
are represented by circles of arbitrary size.
6103
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107

Organometallics
Article
a
Table 4. Selected Bond Lengths (Å) and Angles (deg) for Compounds I−VI
I
II
III
IV
V
VI
P(1)−C(1)
P−Me
P−Ph av
1.7284(19)
1.711(2)
1.787(2)
1.788
1.727(3)
1.785(4)
1.817(3)
1.424(4)
1.374(5)
1.417(5)
1.411(5)
1.368(5)
1.407(5)
1.372(5)
1.416(5)
1.437(4)
1.435(4)
1.768(5)
1.763(3)
1.787(3)
1.790
1.761(4)
1.786
1.8100
1.794
1.787(4)
1.446(5)
1.414(5)
1.435(6)
1.423(6)
1.365(6)
1.415(6)
1.367(6)
1.423(5)
1.439(6)
1.449(6)
2.151(4)
2.163(4)
2.187(4)
2.257(4)
2.246(4)
2.17
C(1)−C(2)
C(2)−C(3)
C(3)−C(4)
C(4)−C(5)
C(5)−C(6)
C(6)−C(7)
C(7)−C(8)
C(8)−C(9)
C(4)−C(9)
C(1)−C(9)
C(1)−Ru
1.427(3)
1.368(3)
1.429(3)
1.406(3)
1.374(3)
1.399(3)
1.372(3)
1.406(3)
1.438(3)
1.435(3)
1.420(3)
1.364(3)
1.421(3)
1.397(3)
1.362(3)
1.394(3)
1.369(3)
1.400(3)
1.423(3)
1.432(3)
1.444(7)
1.411(8)
1.418(7)
1.427(7)
1.350(8)
1.422(8)
1.355(7)
1.420(7)
1.445(7)
1.459(7)
2.163(5)
2.176(5)
2.176(5)
2.220(5)
2.240(5)
2.16
1.443(5)
1.412(5)
1.427(5)
1.431(5)
1.351(6)
1.427(5)
1.356(5)
1.423(5)
1.435(5)
1.457(5)
2.170(3)
2.172(4)
2.183(4)
2.230(4)
2.242(3)
2.17
C(2)−Ru
C(3)−Ru
C(4)−Ru
C(9)−Ru
C5H5−Ru av
C(1)−P−Me
C(1)−P−Ph av
Me−P−Ph av
111.9(10)
111.16
112.12
111.38(17)
109.2
110.8
110.80
112.31(15)
106.19
110.0
109.95(19)
108.0
107.0
109.5
Ph−P−Ph av
110.47
108.33
108.9
107.9
C(1)−C(2)−C(3)
C(2)−C(3)−C(4)
C(3)−C(4)−C(9)
C(4)−C(9)−C(1)
C(9)−C(1)−C(2)
110.12(17)
108.50(18)
107.47(17)
107.22(16)
106.67(16)
110.1(2)
108.2(2)
107.96(19)
107.09(18)
106.64(19)
110.1(3)
108.2(3)
108.1(3)
107.0(3)
106.6(3)
108.5
108.6(3)
108.5(3)
108.6(3)
107.0(3)
107.2(3)
108.0(3)
108.5(3)
108.6(4)
106.8(3)
108.0(3)
108.8(5)
108.8(5)
106.5(4)
107.3(5)
a
Data for II are from ref 4.
Table 5. Calculated C−C Bond Lengths (Å) for C5H4 and 1-Indenylide Ligands and Their Complexes
free ylide
[Ru(η5-ylide)(η5-C5H5)]+
a
b
a
b
ligand
C(2)−C(3)
av C−C
rms dev
C(2)−C(3)
av C−C
rms dev
C5H4PMe3
1.387
1.370
1.376
1.376
1.375
1.375
1.360
1.419
1.428
1.437
1.438
1.439
1.439
1.449
0.021
0.038
0.025
0.025
0.026
0.026
0.038
1.421
1.415
1.417
1.417
1.417
1.416
1.410
1.434
1.439
1.447
1.447
1.447
1.448
1.452
0.011
0.019
0.014
0.014
0.014
0.015
0.020
C5H4PF3
1-C9H6PMe3
1-C9H6PMe2Ph
1-C9H6PMePh2
1-C9H6PPh3
1-C9H6PF3
a
b
Excluding C(2)−C(3). All C−C.
C(2)−C(3) bond of 1.37 Å while the other bonds in the five-
membered ring are all rather similar (1.42−1.44 Å).
Coordination of the ligands to [Ru(η 5-C5H5)]+ results in
“equalization” of the ring bonds, reducing the rms deviation
from the average ring C−C bond length by a factor of about 2,
which can be taken as evidence for the increased contribution
of resonance structure b in the ruthenium complexes.
The optimized geometries also reflect the difference in Ru−
C bond lengths for the five-membered rings: Ru−C(4) and
Ru−C(9) are systematically about 0.1 Å longer than the other
three Ru−C bonds, in contrast to the rather symmetrical
bonding of C5H4PR3 to [Ru(η5-C5H5)]+. Comparisons of the
relevant ligand donor orbitals of C5H4PR3 and 1-indenylidePR3
(see Figure 16) help to explain these trends. Donation from the
pair of orbitals HOMO, HOMO-1 is likely to be equally
effective for corresponding C5H4PR3 and (1-indenylide)PR3
ligands and would not result in very different bonding to
individual carbons. For C5H4PR3, the lower lying HOMO-2 is
also more or less evenly spread out over the C5H4 ring and, as a
result, the ruthenium complex shows nearly symmetrical
bonding (rms deviation 0.017 Å). In contrast, the correspond-
ing (1-indenylide)PMe3 HOMO-3 orbital is primarily con-
centrated on C(1)−C(3), and hence [Ru(η 5-C5H5)]+ binds
more strongly to these three carbons than to the remaining two
(rms deviation 0.04−0.05 Å).
The X-ray structures of complexes IV−VI exhibit some
tilting of the two five-membered rings (η 5-C5H5 and η5-1-
C9H6) relative to each other. The calculated tilt angles between
the two least-squares ring planes, 4−7° for the C5H4PMe3 and
(1-indenylide)PMexPh3‑x complexes, are essentially identical
with those observed. The tilting probably occurs because
6104
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107

Organometallics
Article
Figure 16. Donor orbitals of the ligands C5H4PMe3 and 1-indenylidePMe3.
neither the C5H4PR3 nor the (1-indenylide)PR3 ligands bind in
electronically symmetrical fashions to the metal.
does benzene. Replacement of the methyl groups at
phosphorus by phenyl groups might also be expected to reduce
the metal−ligand bond strength, since phenyl is more electron
withdrawing than is methyl, but we see no evidence of this. In
fact, the metal−ligand bond strengths increase by 2 kcal/mol for
each of the first two Ph groups and then only decrease by
2 kcal/mol for the third because of steric factors. It should be
noted that these calculations were done for gas-phase ions, and
the possibility of more extensive delocalization of the positive
charge may have artificially stabilized the more phenylated
species. Perhaps the safest conclusion is that the replacement of
methyl by phenyl groups has at most a very modest effect on
the ligand binding strength.
Table 6. Complex Stabilities (kcal/mol) Relative to C6H6
a
ligand
ΔG273
C5H4PMe3
−32.8
−6.1
C5H4PF3
1-C9H6PMe3
1-C9H6PMe2Ph
1-C9H6PMePh2
1-C9H6PPh3
1-C9H6PF3
−19.0
−20.6
−22.6
−20.4
4.7
a
ΔG for reaction 2, calculated for gas phase, 1 bar, 273 K, but entropy
scaled by 0.67 to correct for reduced freedom in solution.
CONCLUSIONS
■
A series of aromatic phosphonium-1-indenylide (PHIN)
ligands and the corresponding ruthenium complexes [Ru(η 5-
C5 H5 )(η 5-1-C9H6 PPh3 )]PF6, [Ru(η 5-C5 H5 )(η 5-1-
C9H6PMePh2)]PF6, and [Ru(η5-C5H5)(η5-1-C9H6PMe2Ph)]-
Given in Table 6 are calculated complex stabilities (ΔG273
)
relative to benzene, i.e. the free-energy change of the reaction
1
PF6 have been synthesized and characterized by H, 13C, and
(2)
31P NMR spectroscopy, by X-ray crystallography, and by
extensive DFT calculations. The calculations produce opti-
mized geometries consistent with the crystallographic data and
suggest that PHIN ligands bind considerably more strongly to
ruthenium(II) than do analogous arene ligands. Consistent
with this conclusion, NMR data show that interconversion of
enantiomers via interfacial exchange of the η5-bound ligands
does not occur: i.e., the complexes exhibit configurational
stability. Applications exploiting the planar chirality of the
complexes can be expected.
In agreement with a trend noted in earlier work on
isoelectronic Cr(CO)3 complexes,3,4 C5H4PMe3 was found to
be a much stronger donor than benzene (by about 33 kcal/
mol). As expected, introduction of electronegative fluorine
substituents on phosphorus strongly reduces the coordinating
power of the ligand (by 25 kcal/mol), presumably because of
inductive effects. Turning now to the 1-indenylide ligands, the
data in Table 6 demonstrate that these bind more weakly to
[Ru(η5-C5H5)]+ than their cyclopentadienyl analogues by
about 14 kcal/mol. A reasonable explanation is that one π
orbital is less available for donation to ruthenium because of
competing mixing with the benzene ring π orbitals (HOMO-x;
see Figure 16). Again, introduction of fluorines on phosphorus
reduces the ligand binding energy by about 25 kcal/mol, leading
now to a ligand that binds more weakly to [Ru(η 5-C5H5)]+ than
EXPERIMENTAL SECTION
■
All syntheses were carried out under dry, deoxygenated argon using
standard Schlenk line techniques. Argon was deoxygenated by passage
through a heated column of BASF copper catalyst and then dried by
passing through a column of activated 4A molecular sieves. NMR
spectra were recorded using a Bruker AV600 spectrometer, 1H and 13
C
6105
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107

Organometallics
Article
NMR data being referenced to TMS via the residual protons signals of
the deuterated solvent. [Ru(η 5-C5H5)(MeCN)3]PF6, was purchased
from Strem Chemicals, while 1-bromoindene was prepared via
cleavage of the carbon−silicon bond of 1-trimethylsilylindene by
dioxane dibromide.7
Syntheses of the Phosphonium Salts (1-C9H7PPh3)Br, (1-
C9H7PMePh2)Br and (1-C9H7PMe2Ph)Br. In a typical reaction, a
solution of 3.5 g of 1-bromoindene and 5.14 g of PPh3 (1:1 molar
ratio) in 50 mL of toluene was heated at 50 °C for 72 h to give a white
precipitate of the phosphonium salt [1-C9H7PPh3]Br. The product
was collected by filtration and dried under reduced pressure (4.0 g,
49% yield). Anal. Calcd for C27H22PBr: C, 70.95; H, 4.85. Found: C,
70.43; H, 4.93. The 1H NMR spectrum is shown in Figure S1
(Supporting Information), while a 31P NMR spectrum (CD2Cl2)
exhibited resonances at δ 12.5 (isomer A) and 26.6 (isomer B) in a 3:1
ratio.
crystallization from a concentrated CH2Cl2 solution layered with
hexanes.
Syntheses of [Ru(η5-C5H5)(η5-1-C9H6PPh3)]PF6 (IV), [Ru(η5-
C5H5)(η 5-1-C9H6PMePh2)]PF6 (V), and [Ru(η 5-C5H5)(η 5-1-
C9H6PMe2Ph)]PF6 (VI). In a typical reaction, a solution containing
0.13 g of [Ru(η5-C5H5)(MeCN)3]PF6 in 40 mL of CH2Cl2 and an
equimolar amount of I, II, or III was stirred for 15 min. X-ray-quality
crystals of IV−VI were obtained in all cases by the slow evaporation of
saturated CH2Cl2 solutions, but bulk purification of the complexes was
not readily accomplished. Compound IV was dissolved in a minimum
amount of THF, and the solution was layered with hexanes or ethyl
ether, but only impure product was obtained. Alternatively, IV was
dissolved in a minimum amount of hot acetonitrile and the solution
was cooled to −20 °C, but no precipitate formed. Similarly, compound
V was dissolved in a minimum amount of THF or acetone and the
solutions were layered with hexanes or toluene, respectively, but only
impure product was obtained. Attempts to purify VI involved similar
THF−hexanes, THF−ethyl ether, acetone−hexanes, and acetonitrile−
ethyl ether layering experiments, but in all cases only impure product
was obtained. Washing the impure products with benzene was found
to remove some impurities, but judging from the elemental analyses
In a similar manner, a white, air-sensitive precipitate of (1-
C9H7PMePh2)Br was prepared (8.90 g, 88% yield) by stirring a
solution of 4.84 mL of PMePh2 (25.7 mmol) and 5.03 g 1-
bromoindene (25.7 mmol) in 75 mL of toluene for 24 h at room
1
temperature. H NMR (Figure S2 (Supporting Information)) and 31P
NMR spectra (CD2Cl2) showed that the material was a mixture of
1
1
and the H NMR spectra, complete removal of all traces of solvent
regioisomers (Figure 7). H NMR (CD2Cl2, 500 MHz): δ 3.9 (br s,
2
molecules was difficult. Elemental analysis confirmed that pure VI was
obtained by washing with C6H6 followed by drying under vacuum for
several days (Anal. Found: C, 46.79; H, 3.86. Calcd: C, 46.90; H,
3.94). This approach did not give analytically pure material for
compounds IV (Anal. Found: C, 53.10; H, 4.04. Calcd: C, 55.90; H,
3.87) and V (Anal. Found: C, 49.48; H, 3.63. Calcd: C, 51.85; H,
3.81).
Computational Details. All geometry optimizations were carried
out with Turbomole11,12 using the TZVP basis13 (small-core
pseudopotential for Ru14) and the functional b3-lyp15−18 in combination
with an external optimizer (PQS OPTIMIZE).19,20 Vibrational analyses
were carried out for all stationary points to confirm their nature
(0 imaginary frequencies). Final energies were obtained using the
TZVPP basis,21 and these were combined with thermal corrections
(enthalpy and entropy, 273 K, 1 bar) from the TZVP vibrational
analyses to arrive at the final free energies. To account for the reduced
freedom of movement in solution, entropy contributions to the free
energies were scaled to two-thirds of their gas-phase values.22,23 Orbital
plots in Figure 16 were made using Molden.24 For a complete listing of
energies, geometrical details, and final coordinate files, see the
Supporting Information.
CH2CCHC of isomer A), 3.1 (d, PCH3 of isomer A, JP−H = 13.1
Hz), 2.8 (d, PCH3 of isomer B, 2JP−H = 13.1 Hz). 31P NMR (CD2Cl2):
1
δ 13.6 (isomer A), 25.9 (isomer B). The relative intensities of the H
doublets and the 31P singlets indicated for this sample a 42:58 ratio of
isomer A to isomer B.
In a similar fashion, a solution of 2.13 mL of PMe2Ph (15 mmol)
and 2.93 g of 1-bromoindene (15 mmol) in 50 mL of toluene was
stirred for 24 h at room temperature to yield white, air-sensitive (1-
C9H7PMe2Ph)Br (2.40 g, 48% yield). Anal. Founf for C17H18PBr: C,
1
58.97; H, 4.99. Calcd: C, 61.28; H, 5.44. H NMR (Figure 2) and 31
P
NMR spectra (CD2Cl2) showed that the material was a mixture of
regioisomers (Figure 7). 1H NMR (CD2Cl2): δ 3.9 (br s, CH2CCH
of isomer A), 2.9 (d, P(CH3)2Ph of isomer A, 2JP−H = 13.1 Hz), 2.5 (d,
PCH3 of isomer B, 2JP−H = 13.1 Hz), 2.4 (d, PCH3 of isomer B, 2JP−H
=
13.1 Hz). 31P NMR (CD2Cl2): δ 27.4 (isomer B), 13.4 (isomer A).
1
The relative intensities of the H NMR doublets and the 31P NMR
signals for this sample indicated a 32:68 ratio of isomer A to isomer B.
Syntheses of the Phosphonium-1-indenylide Ligands 1-
C9H6PPh3 (I), 1-C9H6PMePh2 (II), and 1-C9H6PMe2Ph (III). A
mixture of 4.0 g of [1-C9H7PPh3]Br and 0.63 g of NaH (3-fold molar
excess) in 60 mL of THF was stirred at room temperature for 48 h.
The deep green solution was then filtered through Celite, and the
solvent was removed under reduced pressure to give I as a dark green
solid (2.58 g, 65% yield) which could be stored in air without
decomposition. X-ray-quality crystals and analytically pure material
were obtained by crystallization from a concentrated CH2Cl2 solution
layered with hexanes. 1H and 13C NMR data are given in Table 1, and
the 1H NMR spectrum is given in Figure S3 (Supporting Information).
31P NMR (CD2Cl2): δ 10.39. Anal. Found for C27H21P: C, 84.74; H,
ASSOCIATED CONTENT
■
S
* Supporting Information
Text, figures, and tables giving the synthetic procedure for 1-
bromoindene, NMR spectroscopic data for compounds I and
III−VI and computational data and text, figures, tables, and
CIF files giving crystallographic details for 1-C9H6PPh3 (I), 1-
C9H6PMe2Ph (III), [Ru(η5-C5H5)(η5-1-C9H6PPh3)]PF6 (IV),
[Ru(η 5-C5H5)(η 5-1-C9H6PMePh2)]PF6 (V), and [Ru(η 5-
C5H5)(η 5-1-C9H6PMe2Ph)]PF6 (VI), including complete
numbering schemes, thermal ellipsoid figures, positional and
thermal parameters, bond lengths, and bond angles. This
material is available free of charge via the Internet at http://
pubs.acs.org. The crystallographic data for 1-C9H6PPh3
(CCDC 826560), C9H6PMe2Ph (CCDC 826559), [Ru(η 5-
C5H5)(η 5-1-C9H6PPh3)]PF6 (CCDC 826565), [Ru(η 5-
C5H5)(η5-1-C9H6PMePh2)]PF6 (CCDC 826564), and [Ru-
(η5-C5H5)(η5-1-C9H6PMe2Ph.)]PF6 (CCDC 826563) may
also be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif.
5.50. Calcd: C, 86.10; H, 5.62.
The syntheses of 1-C9H6PMePh2 (II) and 1-C9H6PMe2Ph (III)
were carried out similarly. Thus, a mixture of 7.85 g of [1-
C9H7PMePh2]Br (19.9 mmol) and 0.53 g of NaH (22.0 mmol) in
150 mL of THF was stirred for 24 h at room temperature to yield 5.32
g of the pale green, air-sensitive II (85% yield). Pure material (NMR)
could be obtained by crystallization from a concentrated CH2Cl2
solution layered with hexanes. A 1H NMR spectrum is shown in Figure
1
S4 (Supporting Information), while H and 13C NMR data are given
in Table 1; the data agree with literature values.4 31P NMR (CD2Cl2):
δ 5.69.
Similarly, a mixture of 2.39 g of [1-C9H7PMe2Ph]Br (7.2 mmol)
and 0.19 g of NaH (7.9 mmol) in 60 mL of THF was stirred for 24 h
at room temperature to yield, after solvent removal, 1.82 g of yellow-
1
green, air-sensitive III (94% yield). A H NMR spectrum is shown in
1
Figure 3, while H and 13C NMR data are given in Table 1. 31P NMR
(CD2Cl2): δ 1.78. Anal. Found for C17H17P: C, 79.38; H, 6.69. Calcd:
C, 80.93; H, 6.79. X-ray-quality crystals of I and III were obtained by
6106
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107

Organometallics
Article
ACKNOWLEDGMENTS
■
We thank the Natural Sciences and Engineering Research
Council (Discovery Grants to M.C.B., P.H.M.B.) and Johnson-
Matthey for a generous loan of RuCl3·3H2O.
REFERENCES
■
(1) (a) Ramirez, F.; Levy, S. J. Org. Chem. 1956, 21, 488.
(b) Ramirez, F.; Levy, S. J. Org. Chem. 1956, 21, 1333. (c) Ramirez,
F.; Levy, S. J. Am. Chem. Soc. 1957, 79, 67. (d) Ramirez, F.;
Dershowitz, S. J. Org. Chem. 1957, 22, 41. (e) Ramirez, F.; Levy, S.
J. Am. Chem. Soc. 1957, 79, 6167. (f) Ramirez, F.; Levy, S. J. Org. Chem.
1958, 23, 2035. (g) Ammon, H. L.; Wheeler, G. L.; Watts, P. H. Jr.
J. Am. Chem. Soc. 1973, 95, 6158. (h) Gray, G. A. J. Am. Chem. Soc.
1973, 95, 7736. (i) For a recent computational study, see: Laavanya,
P.; Krishnamoorthy, B. S.; Panchanatheswaran, K.; Manoharan, M.
J. Mol. Struct. (THEOCHEM) 2005, 716, 149.
(2) For a comprehensive review of phosphonium cyclopentadienides
and their coordination chemistry, see: Brownie, J. H.; Baird, M. C.
Coord. Chem. Rev. 2008, 252, 1734.
(3) (a) Brownie, J. H.; Baird, M. C.; Schmider, H. L. Organometallics
2007, 26, 1433. (b) Brownie, J. H.; Baird, M. C.; Laws, D. R.; Geiger,
W. E. Organometallics 2007, 26, 5890.
(4) Brownie, J. H.; Baird, M. C. J. Organomet. Chem. 2008, 693, 2812.
(5) (a) Crofts, P. C.; Williamson, M. P. J. Chem. Soc. C 1967, 1093.
(b) Ford, J. A. Tetrahedron Lett. 1968, 815.
(6) Rufanov, K. A.; Ziemer, B.; Hummert, M.; Schutte, S. Eur. J.
Inorg. Chem. 2004, 4759.
(7) Murphy, J. A.; Patterson, C. W. J. Chem Soc., Perkin Trans. 1
1993, 405.
(8) Trost, B. M.; Older, C. M. Organometallics 2002, 21, 2544.
́
(9) (a) Bart, J. C. J. J. Chem. Soc. B 1969, 350. (b) Pliva, J.; Johns, J.
W. C.; Goodman, L. J. Mol. Spectrosc. 1990, 140, 214.
(10) (a) Cruickshank, D. W. J.; Sparks, R. A. Proc. R. Soc. London, Ser.
́
A 1960, 258, 270. (b) Cadierno, V.; Diez, J.; Pilar Gamasa, M.;
Gimeno, J.; Lastra, E. Coord. Chem. Rev. 1999, 193−195, 147.
(11) Ahlrichs, R.; Bar, M.; Baron, H.-P.; Bauernschmitt, R.; Bocker,
̈
̈
S.; Ehrig, M.; Eichkorn, K.; Elliott, S.; Furche, F.; Haase, F.; Haser, M.;
̈
Hattig, C.; Horn, H.; Huber, C.; Huniar, U.; Kattannek, M.; Kohn, A.;
̈
̈
̈
Kolmel, C.; Kollwitz, M.; May, K.; Ochsenfeld, C.; Ohm, H.; Schafer,
̈
A.; Schneider, U.; Treutler, O.; Tsereteli, K.; Unterreiner, B.; Von
Arnim, M.; Weigend, F.; Weis, P.; Weiss, H. Turbomole Version 5;
Theoretical Chemistry Group, University of Karlsruhe, Karlsruhe,
Germany, 2002.
(12) Treutler, O.; Ahlrichs, R. J. Chem. Phys. 1995, 102, 346.
(13) Schafer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100,
̈
5829.
(14) Andrae, D.; Haussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
̈
Theor. Chim. Acta 1990, 77, 123.
(15) All Turbomole calculations were performed with the functional
“b3-lyp” of that package, which is similar (but not identical) to the
Gaussian “B3LYP” functional.
(16) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(17) Becke, A. D. J. Chem. Phys. 1993, 98, 1372.
(18) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(19) PQS version 2.4; Parallel Quantum Solutions, Fayetteville, AR,
2001 (the Baker optimizer is available separately from PQS upon
request).
(20) Baker, J. J. Comput. Chem. 1986, 7, 385.
(21) Weigend, F.; Furche, F.; Ahlrichs, R. J. Chem. Phys. 2003, 119,
12753.
(22) Tobisch, S.; Ziegler, T. J. Am. Chem. Soc. 2004, 126, 9059.
(23) Raucoules, R.; De Bruin, T.; Raybaud, P.; Adamo, C.
Organometallics 2009, 28, 5358.
(24) Schaftenaar, G.; Noordik, J. H. J. Comput.-Aided Mol. Design
2000, 14, 123.
6107
dx.doi.org/10.1021/om200545j|Organometallics 2011, 30, 6098−6107