Cationic ruthenium hydrido-carbonyls derived from metallocene-based pincers: Unusual rearrangements and H2 evolution with formation of cationic ruthenium metallocenylidenes

Article abstract of DOI:10.1021/om100667p

Cationic ruthenium hydrido-carbonyls {RuH(CO)[^tBuP,CH,P ^M]}BAr^F₄ (M = Fe, 3; M = Ru, 4) (Ar ^F= 3,5-(CF₃)₂C₆H₃) obtained in the reaction of H₂ with RuCl(CO)[^tBuP,C,P ^M] (M = Fe, 1; M = Ru, 2) in the presence of NaBAr^F ₄ add CO smoothly, giving the corresponding dicarbonyl complexes {RuH(CO)₂[^tBuP,CH,P^M]}BAr^F ₄ (M = Fe, 11; M = Ru, 12). According to X-ray analysis of 11 and 12, addition of extra CO to complexes 3 and 4 leads to strengthening of the C(1)-H(1)...Ru(1) agostic interaction. Simultaneously, CO addition to 3 and 4 triggers a sequence (up to three steps) of unprecedented intramolecular rearrangements including migration of H atoms. First metallocenylidene complexes {Ru(CO)₂[^tBuP,C,P^M]}BAr^F ₄ (M = Fe, 15; M = Ru, 16) were obtained in the course of these rearrangements accompanied by H₂ evolution.

Full text of DOI:10.1021/om100667p

4360 Organometallics 2010, 29, 4360–4368
DOI: 10.1021/om100667p
Cationic Ruthenium Hydrido-Carbonyls Derived from
Metallocene-Based Pincers: Unusual Rearrangements and H2 Evolution
with Formation of Cationic Ruthenium Metallocenylidenes
Avthandil A. Koridze,*,†,‡ Alexander V. Polezhaev,† Sergey V. Safronov,†
Alexey M. Sheloumov,† Fedor M. Dolgushin,† Mariam G. Ezernitskaya,†
Boris V. Lokshin,† Pavel V. Petrovskii,† and Alexander S. Peregudov†
†A.N. Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, 28 Vavilov Street,
119991 Moscow, Russian Federation, and ‡Institute of Organometallic Chemistry,
I. Javakhishvili Tbilisi State University, 3 Chavchavadze Avenue, 0128 Tbilisi, Georgia
Received July 8, 2010
Cationic ruthenium hydrido-carbonyls {RuH(CO)[tBuP,CH,PM]}BArF4 (M = Fe, 3; M = Ru, 4)
(ArF = 3,5-(CF3)2C6H3) obtained in the reaction of H2 with RuCl(CO)[tBuP,C,PM] (M = Fe, 1; M =
Ru, 2) in the presence of NaBArF4 add CO smoothly, giving the corresponding dicarbonyl complexes
{RuH(CO)2[tBuP,CH,PM]}BArF4 (M = Fe, 11; M = Ru, 12). According to X-ray analysis of 11 and
12, addition of extra CO to complexes 3 and 4 leads to strengthening of the C(1)-H(1) Ru(1)
3 3 3
agostic interaction. Simultaneously, CO addition to 3 and 4 triggers a sequence (up to three steps) of
unprecedented intramolecular rearrangements including migration of H atoms. First metallocenyl-
idene complexes {Ru(CO)2[tBuP,C,PM]}BArF4 (M = Fe, 15; M = Ru, 16) were obtained in the course
of these rearrangements accompanied by H2 evolution.
Introduction
intramolecular exchange of the hydrogen atoms between tran-
sition metal and carbon (nitrogen) atoms3 in connection with
processes such as hydrogenation, isomerization, dehydro-
genation, and other chemical transformations of organics.
Herein we describe an unusual set of intramolecular rear-
rangements that we observed in the course of the study of
reactions of H2 with cationic ruthenium carbonyl pincer com-
plexes based on ferrocene and ruthenocene. After addition of
H2 to these complexes, coordination of an extra CO ligand to
the ruthenium atom promotes a sequence (up to three steps) of
intramolecular rearrangements, leading to H2 evolution and
formation of the first (cationic) metallocenylidene complexes.
Formation and cleavage of dihydrogen on transition metal
complexes has attracted much attention in recent years
because of their application in hydrogenation and relevance
to hydrogen production.1 The dihydrogen activation may
occur via two different mechanisms, homolytic and hetero-
lytic, and the latter is of great importance in biological
systems1d and catalytic ionic hydrogenations.1e,f One of the
remarkable properties of transition metal hydride complexes
formed is the ability to undergo intramolecular rearrange-
ments; well-known examples of hydrogen migration are
fluxionality of hydride ligands in transition metal clusters.2
Even more interesting in view of chemical reactivity is
Results and Discussion
Formation of Complexes {RuH(CO)[{2,5-(tBu2PCH2)2C5H3}-
M(C5H5)]}BArF4 (M = Fe, 3; M = Ru, 4). Complexes 3and 4
were obtained quantitatively in clean reactions of the corres-
ponding chloro-carbonyl ruthenium complexes RuCl(CO)[{2,5-
(tBu2PCH2)2C5H2}M(C5H5)] (M = Fe, 1; M = Ru, 2) with H2
in the presence of NaBArF4 (ArF = 3,5-(CF3)2C6H3) (eq 1).
*Corresponding author. E-mail: koridze@ineos.ac.ru. Fax: (þ)7(499)-
135-50-85.
(1) (a) Kubas, G. J. Chem. Rev. 2007, 107, 4152. (b) Esswein, A. J.;
Nocera, D. G. Chem. Rev. 2007, 107, 4022. (c) Crabtree, R. H. The Organo-
metallic Chemistry of the Transition Metals, 4th ed.; Wiley-Interscience:
NJ, 2005. (d) Evans, D. J.; Pickett, C. J. Chem. Soc. Rev. 2003, 32, 268.
(e) Noyori, R.; Yamakawa, M.; Hashiguchi, S. J. Org. Chem. 2001, 66,
7931. (f) Rautenstrauch, V.; Hoang-Cong, X.; Churfand, K. A-R.; Morris,
R. H. Chem.;Eur. J. 2003, 9, 4954.
(2) (a) Knox, S. A. R.; Kaesz, H. D. J. Am. Chem. Soc. 1971, 93, 4594.
(b) Constable, E. C.; Johnson, B. F. G.; Lewis, J.; Pain, G. N.; Taylor, M. J.
J. Chem. Soc., Chem. Commun. 1982, 754. (c) Predieri, G.; Tiripicchio, A.;
Vignali, C.; Sappa, E. J. Organomet. Chem. 1988, 342, C33.
(3) (a) Calvert, R. B.; Shapley, J. R. J. Am. Chem. Soc. 1978, 100,
7726. (b) Koridze, A. A.; Astakhova, N. M.; Petrovskii, P. V.; Dolgushin,
F. M.; Yanovsky, A. I.; Struchkov, Yu. T. J. Organomet. Chem. 1994, 481,
€
247. (c) Jensen, C. M. Chem. Commun. 1999, 2443. (d) Kab, M.; Friedrich,
A.; Dress, M.; Schneider, S. Angew. Chem., Int. Ed. 2009, 48, 905.
(e) Stepovska, D.; Jiang, H.; Song, D. Chem. Commun. 2010, 556.
(f) Deeming, A. J.; Hasso, S. J. Organomet. Chem. 1976, 114, 313.
(g) Gunanathan, C.; Ben-David, Y.; Milstein, D. Science 2007, 317, 790.
(h) Iron, M. A.; Ben-Ari, E.; Cohen, R.; Milstein, D. Dalton Trans. 2009,
9433. (i) Bullock, R. M. Angew. Chem., Int. Ed. 2007, 46, 7360.
Complex 3 was characterized by a single-crystal X-ray
diffraction. The molecular structure of 3 is illustrated in
r
pubs.acs.org/Organometallics
Published on Web 09/09/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 19, 2010 4361
species related to the arenium ions, in contrast to catio-
nic species oxidized at the iron atom, for which the term
“ferricenium” seems to be more appropriate.) Noteworthy,
the hydrogen atom H(1) is almost equally remote from both
metal atoms: the H(1) Fe(1) and H(1) Ru(1) distances
3 3 3
3 3 3
˚
are 2.50 and 2.58 A, respectively. These distances are too
large to suppose marked (if any) interaction between the
hydrogen atom H(1) and any of the above metal atoms,
though it is reported that the distances for intramolecular
˚
M
HC bonds can vary in a wide range,4a from 1.84 A for a
3 3 3
manganese complex4b to 2.92 A for a palladium complex.
Spectroscopic data obtained for 3 in solution are in
accordance with its solid-state X-ray crystal structure. In
the 1H NMR spectrum (CD2Cl2), there are resonances from
the H(1) atom at δ 2.03 ppm (s) and the hydride atom at
δ -26.22 ppm (t, J(H-P)=15.9 Hz). Such a high-field hydride
resonance is typical of a hydride ligand trans to a vacant
coordination site. In the 13C{1H} NMR spectrum, a reso-
nance from the C(1) carbon is observed at δ 37.92 ppm (s).
Spectral characteristics for the diruthenium complex 4 are
similar to those for complex 3.
4c
˚
Figure 1. Molecular structure of 3 (ellipsoids at the 50% prob-
ability level; hydrogen atoms except H(1M) and H(1) and BArF
˚
anion are omitted for clarity). Selected bond lengths [A]
The DFT calculations for complexes 3 and 4 gave geo-
metrical parameters close to those obtained by the X-ray
diffraction method for complex 3 (see Table 1 in the SI), thus
confirming that the above peculiarities of the structure are
associated with the molecular properties rather than with
crystal-packing forces.
The IR spectrum of complex 3 in solution contains a
strong band at 1943 cm-1 belonging to ν(CO) and a very
weak band at 2020 cm-1 belonging to ν(Ru-H) (see Figure 1
in the SI). The intensity ratio is in accordance with the DFT
data, which show that the former band is 40-fold more
intense than the latter. It should be noted that the calcu-
lated ν(CO) frequency matches very well the experimen-
tal frequency, but the calculated ν(MH) is higher due to
anharmonicity.
Consideration of the structure of complex 3 does not
provide information on the mechanism of its formation; it
is evident, however, from the transoid arrangement of C(1)-H
and Ru(1)-H hydrogen atoms along the C(1)-Ru(1) bond
that after H2 addition some rearrangement(s) take place,
giving finally complex 3. At present, a mechanistic pathway
can be proposed involving H2 activation/hydrogenolysis of
the C(1)-Ru(1) bond followed by exo attack of the cyclo-
pentadienyl C(1)-H(1) carbon atom with the cationic
(chelated) ruthenium hydrido-carbonyl fragment.
As for formation of complexes 3 and 4, it is interesting
that they can also be generated by protonation of carbonyl
hydrides RuH(CO)[{2,5-(tBu2PCH2)2C5H2}M(C5H5)], 5
(M = Fe) and 6 (M = Ru), respectively. The latter can be
obtained from 3 and 4, respectively, by the abstraction of the
H(1) atom as a proton under the action of amines (NEt3 or
Py); however, several unidentified complexes are simulta-
neously formed in these reactions. Pure complex 5, contain-
ing the hydride ligand in the exo position, was isolated in the
reaction of complex 1 with LiAlH4 (eq 2). Its protonation
with CF3COOH gave the cationic complex 3* (eq 3). Chemi-
cal shifts of the C(1)-H and Ru(1)-H hydrogen atoms for
complex 3* are practically the same as those observed for
4
and angles [deg]: Ru(1)-P(1) 2.3847(6), Ru(1)-P(2) 2.3856(6),
Ru(1)-C(1) 2.326(2), Ru(1)-C(29) 1.809(3), Ru(1)-H(1M)
1.64, Fe(1)-C(1) 2.035(2), Fe(1)-C(2-10) 2.023(2)-2.065(3),
C(1)-C(2) 1.447(3), C(1)-C(5) 1.458(3), C(1)-H(1) 0.91,
C(2)-C(3) 1.424(4), C(3)-C(4) 1.431(4), C(4)-C(5) 1.417(3),
P(1)-Ru(1)-P(2) 166.92(2), C(29)-Ru(1)-C(1) 171.1(1), C(2)-
C(1)-C(5) 108.0(2), C(2)-C(1)-Ru(1) 97.4(1), C(5)-C(1)-
Ru(1) 96.5(2), Ru(1)-C(1)-H(1) 96.
Figure 1. As is seen, the structure is unusual for a ferrocene
derivative. The complex bears two substituents on the C(1)
atom of the cyclopentadienyl ring: the hydrogen H(1) and the
ruthenium Ru(1) atoms arranged in the endo and exo posi-
tions, respectively, relative to the metallocene central atom
Fe(1). The C(1)-H(1) bond forms an angle of 16.5° with the
ring plane, and the Ru(1)-C(1) vector forms an angle of
79.1° with the same plane. The second hydrogen atom from
the reacting H2 molecule occupies the apical position at the
pentacoordinated Ru(1) atom; the H(1M) and H(1) hydro-
gen atoms are in the transoid orientation along the Ru(1)-
C(1) bond. Thus, the Ru(1) atom has a distorted tetragonal
pyramid configuration; Ru(1), P(1), P(2), and C(29) atoms
˚
lie in the basal plane (maximum deviation of 0.014 A for
Ru(1)), whereas the C(1) atom deviates from this plane by
˚
0.40 A in the opposite direction from the H(1M) atom, and
the Ru(1)-H(1M) bond forms an angle of 85.2° with this
plane. The structure of complex 3 does not reveal any agostic
Ru(1) H-C interaction with the participation of C-H sp3
bonds of the tert-butyls, as well as close contacts with the
3 3 3
BArF - counterion.
In complex 3, the substituted carbocyclic ligand is almost
planar; the C(1) atom deviates from the plane of the other
4
˚
carbon atoms toward the Ru(1) atom only by 0.02 A. The
C(1)-C(2) and C(1)-C(5) distances (1.447(3) and 1.458(3)
A, respectively) are somewhat longer (by about 0.03 A)
˚
˚
than the C(2)-C(3) and C(4)-C(5) distances (1.424(3) and
˚
1.417(3) A, respectively), whereas the C(3)-C(4) bond has
˚
an intermediate length (1.431(4) A). The differences in C-C
(4) (a) Youngs, W. J.; Kinder, J. D.; Bradshaw, J. D.; Tessier, C. A.
Organometallics 1993, 12, 2406. (b) Brookhart, M.; Lamanna, W.; Humphrey,
M. B. J. Am. Chem. Soc. 1982, 104, 2117. (c) Dehand, J.; Fischer, J.; Pfeffer,
M.; Mitschler, A.; Zinsius, M. Inorg. Chem. 1976, 15, 2675.
bond lengths in the five-membered carbocycle can be con-
sidered as evidence for the contribution of the ferrocenium
ion structure. (Here the term “ferrocenium” means cationic

4362 Organometallics, Vol. 29, No. 19, 2010
Koridze et al.
methods and single-crystal X-ray diffraction (see Figures 3
and 4). As expected, the incoming CO occupies a vacant
coordination position at the chelated ruthenium atom in
complexes 3 and 4. On the whole, the structures of complexes
11 and 12 are similar to those of their precursors 3 and 4.
The most remarkable are the distances between the hydro-
gen atom H(1) and the chelated Ru(1) and metallocene
central atoms in complexes 11 and 12. In complex 11, they
˚
are 2.36 and 2.49 A, respectively; that is, on going from
monocarbonyl to dicarbonyl complex, the H(1) Fe(1) dis-
3 3 3
tance does not virtually change, whereas the H(1) Ru(1)
3 3 3
˚
distance is somewhat decreased (compare 2.58 and 2.36 A).
˚
A close value of the H(1) Ru(1) distance (2.39 A) is
3 3 3
observed for complex 12 with the ruthenocene backbone,
whereas the distance between the same hydrogen and the
˚
ruthenocene central atom, Ru(2), is 2.57 A. Although these
Figure 2. Benzene-based cationic pincer complexes.
changes in the distances are relatively small, the trends
in their changes can indicate the existence of a C(1)-
complex 3, indicating the absence of hydrogen bonding with
-
the anion CF3CO2
.
H(1) Ru(1) agostic interaction. This interaction is weak,
3 3 3
as follows from the large distances H(1) Ru(1).
3 3 3
In addition, we note that the C(1)-H(1) bond forms an
angle of 17.3° and 19.6° (for 11 and 12, respectively) with the
ring plane, and the Ru(1)-C(1) vector forms an angle of
64.1° (11) and 63.1° (12) with the same plane. In comparison
to complex 3, vector C(1)-H(1) does not change its direction
relative to the five-membered ring, whereas vector C(1)-
Ru(1) markedly deviates from perpendicular to this plane. It
can also indicate some agostic C(1)-H(1) Ru(1) interac-
3 3 3
tion. The DFT-calculated structures of complexes 11 and 12
are similar and close to those obtained by X-ray diffraction.
IR spectra of dicarbonyl complexes 11 and 12 look quite
exceptional. Usually the bands of M-H stretching vibration
are very weak; therefore, two intense bands of ν(CO) and one
weak band of ν(MH) are expected in the range of 2000 cm-1
.
However, absorptions at 2057, 2001, and 1958 cm-1 of approxi-
mately equal intensities (intensity ratio is 1:1.16:0.81) are
observed for complex 11 (Figure 1 in the SI); the bands
at 2001 and 1958 cm-1 mainly belong to symmetrical and
asymmetrical ν(CO) stretching, and the band at 2057 cm-1 is
mainly associated with the Ru-H mode. The unusually high
intensity of the ν(MH) band results from strong coupling of
C-O and Ru-H stretching vibrations due to the mutual
trans arrangement of the CO and H ligands, which leads to
the redistribution of band intensities. The DFT calculation
predicts for complex 11 three intense bands at 2020, 1975,
and 1924 cm-1 with the intensity ratio of 1:1:0.66.
Two extreme canonical forms, metallocenium ion (A) and
cationic agostic complex (B), can be considered for com-
plexes 11 and 12 (Scheme 2). Alternation of C-C distances in
the five-membered ring of complexes 11 and 12 (see Figures 3
and 4 for structural parameters), similar to those observed
for complexes 3 and 16 (see Scheme 1), implies that struc-
ture A has a substantial contribution. Nevertheless, although
It should be noted that several cationic organometallic
complexes with structures similar to those of 3 and 4 gene-
rated from benzene-based pincer complexes are described
in the literature5-8 (see Figure 2). While platinum 75 and ruthe-
nium 86 complexes are considered essentially as transition-
metal-stabilized arenium ions (mainly on the basis of X-ray
diffraction data), rhodium 9a,b7 and ruthenium 108 com-
plexes are thought to be “agostic”, containing the C(1)-
H(1) M bond with an insignificant contribution of the
3 3 3
arenium structure (for example, in complex 9a the C(1)-
˚
H
Rh distance is only 1.950 A).
3 3 3
Formation of Complexes {RuH(CO)2[{2,5-(tBu2PCH2)2-
C5H3}M(C5H5)]}BArF4 (M = Fe, 11; M = Ru, 12) and Their
Rearrangements. Ruthenium complexes 3 and 4 are coordi-
natively unsaturated 16-electron compounds and add readily
CO to form dicarbonyl derivatives 11 and 12, respectively
(Scheme 1). They were fully characterized by spectroscopic
˚
the C(1)H(1) Ru(1) distances (2.58, 2.36, and 2.39 A) and
3 3 3
H(1)-C(1)-Ru(1) angles (96°, 82°, and 83°) for complexes
3, 11, and 12, respectively, do not allow us to delineate
“agostic” or anagostic”10 interactions, they suggest that in
(5) Tarheijden, J.; van Koten, G.; Vinke, I. C.; Spek, A. L. J. Am.
Chem. Soc. 1985, 107, 2891.
these complexes some C(1)-H(1) Ru(1) interaction takes
3 3 3
(6) Zhang, J.; Barakat, K. A.; Cundari, T. R.; Gunnoe, T. B.; Boyle,
P. D.; Petersen, J. L.; Day, C. S. Inorg. Chem. 2005, 44, 8379.
(7) (a) Vigalok, A.; Uzan, O.; Shimon, L. J. W.; Ben-David, Y.;
Martin, J. M. L.; Milstein, D. J. Am. Chem. Soc. 1998, 120, 12539.
(b) Montag, M.; Schwartsburd, L.; Cohen, R.; Leitus, G.; Ben-David, Y.;
Martin, J. M. L.; Milstein, D. Angew. Chem., Int. Ed. 2007, 46, 1901.
(8) Dani, P.; Toorneman, M. A. M.; van Klink, G. P. M.; van Koten,
G. Organometallics 2000, 19, 5287.
place.
(9) Sheloumov, A. M.; Dolgushin, F. M.; Kondrashov, M. V.;
Petrovskii, P. V.; Barbakadze, Kh.A.; Lekashvili, O. I.; Koridze, A. A.
Russ. Chem. Bull. Int. Ed. 2007, 56, 1757.
(10) Brookhart, M.; Green, M. L. H.; Parkin, G. Proc. Natl. Acad.
Sci. U. S. A. 2007, 104, 6908.

Article
Organometallics, Vol. 29, No. 19, 2010 4363
Scheme 1. Reactions of 3 and 4 with CO Result in Hydrido Dicarbonyls 11 and 12a
a Further isomerization of the latter complexes into 13 and 14 and subsequent expulsion of H2 afford metallocenylidene complexes 15 and 16 and
agostic complex 17.9
In CD2Cl2 solution, complex 11 slowly rearranges into
isomer 13, where the hydrogen atoms C(1)-H and Ru(1)-H
are cis arranged along the C(1)-Ru(1) line (Scheme 1). The
conclusion on the structure of 13 is drawn from spectro-
scopic data. Thus, in the 1H NMR spectrum of complex 13,
the hydride signal at δ -5.08 (dt) undergoes splitting with
J(H-H) ≈ 1.0 and J(P-H) = 15.5 Hz, while in the case of
its precursor, 11, the signal at δ -6.08 (dt) is split with J(H-H) =
2.7 and J(P-H) =15.5 Hz. The lower value of J(H-H) for 13
indicates mutual cisoid arrangement of the hydrogen atoms
C(1)-H and Ru(1)-H in this complex. (In contrast to 13, its
diruthenium analogue, 14, appears unstable under the same
conditions and was not fully characterized.)
Milstein et al. have reported11 the formation of benzene-
based ruthenium pincer complex {RuH(CO)2[2,6-(tBu2PCH2)2-
C6H4]}þ, with a structure similar to that of complex 13. This
complex, unstable at room temperature, was obtained by
reversible addition of H2 to cationic dicarbonyl complex
{Ru(CO)2[2,6-(tBu2PCH2)2C6H3]}þ. Although the complex
was not studied by X-ray diffraction, it is thought to contain
the agostic C(1)-H Ru bond. In contrast to {Ru(CO)2-
3 3 3 þ
[2,6-(tBu2PCH2)2C6H3]} , the metallocene-derived dicarbonyl
complex 17 as well as 15 and 16 (see Scheme 1) do not react with
H2 under the same conditions (1 atm H2, 25 °C).
After isomerization of complexes 11 and 12 into com-
plexes 13 and 14, respectively, the latter, while being kept in
CD2Cl2 solutions, release dihydrogen to form dicarbonyl
complexes 15 and 16, respectively (Scheme 1). (It should be
noted that reaction of 1 and 2 with CO in the presence
The rearrangement might include the rupture of the
˚
relatively weak C(1)-Ru(1) bond (2.315(2) and 2.326(3) A
for 11 and 12, respectively), accompanied by migration of the
Ru(1)-H hydrogen atom to the C(1) carbon and the C(1)-H
hydrogen atom to the Ru(1), and subsequent exo electro-
philic attack of ruthenium at the ring. The C(1)-Ru(1)-
H(1M) and H(1)-C(1)-Ru(1) angle values for 11 (87° and
82°, respectively) and 12 (78° and 83°, respectively) can be
considered as possible indications that such migration of H(1)
and H(1M) atoms includes the rearrangement of the transoid
complexes 11 and 12 into the corresponding cisoid isomers.
For monocarbonyl complex 3, the C(1)-Ru(1)-H(1M) and
H(1)-C(1)-Ru(1) angles are 99° and 96°, respectively.
of NaBArF afforded complexes 179 and 16, respectively.)
4
Apparently, the mutual cisoid orientation of two hydrogen
atoms in complexes 13 and 14 favors H2 elimination. In
addition, attractive interaction of acidic C(1)-H and hydridic
Ru(1)-H hydrogens can also lower the barrier for elimination.
(11) van der Boom, M. E.; Iron, M. A.; Atasoylu, O.; Shimon,
L. J. W.; Rozenberg, H.; Ben-David, Y.; Konstantinovski, L.; Martin,
J. M. L.; Milstein, D. Inorg. Chim. Acta 2004, 357, 1854.

4364 Organometallics, Vol. 29, No. 19, 2010
Koridze et al.
Figure 3. Molecular structure of 11 (ellipsoids at 30% prob-
ability level; hydrogen atoms except H(1M) and H(1) and BArF
˚
anion are omitted for clarity). Selected bond lengths [A] and
Figure 4. Molecular structure of 12 (ellipsoids at 30% prob-
ability level; hydrogen atoms except H(1M) and H(1) and BArF
˚
anion are omitted for clarity). Selected bond lengths [A] and
4
4
angles [deg]: Ru(1)-P(1) 2.4152(7), Ru(1)-P(2) 2.449(8), Ru(1)-
C(1) 2.315(2), Ru(1)-C(29) 1.847(3), Ru(1)-C(30) 1.970(3),
Ru(1)-H(1M) 1.68, Fe(1)-C(1) 2.050(2), Fe(1)-C(2-10) 2.023-
(3)-2.074(3), C(1)-C(2) 1.454(4), C(1)-C(5) 1.453(3), C(1)-
H(1) 0.90, C(2)-C(3) 1.423(4), C(3)-C(4) 1.425(4), C(4)-C(5)
1.423(3),P(1)-Ru(1)-P(2) 156.2(2), C(29)-Ru(1)-C(1) 172.8(1),
C(30)-Ru(1)-H(1M) 175, C(2)-C(1)-C(5) 107.7(2), C(2)-
C(1)-Ru(1) 104.8(2), C(5)-C(1)-Ru(1) 106.7(2), Ru(1)-C(1)-
H(1) 82.
angles [deg]: Ru(1)-P(1) 2.419(1), Ru(1)-P(2) 2.414(1), Ru(1)-
C(1) 2.326(3), Ru(1)-C(29) 1.844(4), Ru(1)-C(30) 1.972(4),
Ru(1)-H(1M) 1.70, Ru(2)-C(1) 2.181(3), Ru(2)-C(2-10)
2.153(3)-2.209(3), C(1)-C(2) 1.444(5), C(1)-C(5) 1.453(5),
C(1)-H(1) 0.93, C(2)-C(3) 1.419(5), C(3)-C(4) 1.429(5), C(4)-
C(5) 1.420(5), P(1)-Ru(1)-P(2) 155.95(3), C(29)-Ru(1)-C(1)
170.6(2), C(30)-Ru(1)-H(1M) 176, C(2)-C(1)-C(5) 108.1(3),
C(2)-C(1)-Ru(1) 105.0(2), C(5)-C(1)-Ru(1) 106.7(2), Ru-
(1)-C(1)-H(1) 83.
The structure of complex 16 was established by X-ray diffrac-
tion study (Figure 5).
Scheme 2. Description of Complexes 11 and 12 as Two Extreme
Forms, Ferrocenium A and Agostic B
It is apparent at first glance that the structure of complex
16 resembles those of R-metallocenyl carbocations.12 It is
known that when the latter are formed from metallocene
derivatives (upon reactions of R-metallocenylethylenes or
R-metallocenylcarbinols with acids), the substituted cyclo-
pentadienyl ligand transforms into pentafulvene; the latter
is bound to the metal atom via the six carbon atoms (the
strength of the M-CR bond depends on the electronic and
steric properties of the substituents at the CR atom; the same
factors determine the bend of the exocyclic bond C(1)-CR
toward M in the given metallocenyl carbocation). Similar to
R-metallocenyl carbocations, in 16 the exocyclic C(1)-Ru(1)
bond is bent toward the metallocene central atom Ru(2) by
29.5°. For comparison, the bending angle of the exocyclic
bond out of the C5 ring in the R-ruthenocenyl carbocation12c
[RcCH2][CF3SO3] (Rc = ruthenocenyl) is 42.6°. Similar
strong bending of the exocyclic bond was found for tran-
sition metal fulvene compexes.13 In complex 16, five-
membered carbocycles are nonparallel: the angle between
the rings is 19.2°. This strong deviation from parallel, in
comparison with R-metallocenyl carbocations12 (for com-
plex [RcCH2][CF3SO3] the ring tilt is only 7.1°), is apparently
associated with steric contacts between the pseudoequatorial
tert-butyl groups and the nonmetalated cyclopentadienyl
ring: the effect of these contacts on the ring tilt in metallo-
cene-derived pincer complexes of palladium was unambigu-
ously demonstrated earlier.14 Thus, a comparative X-ray
diffraction study of complexes PdCl[{2,5-(tBu2PCH2)2-
C5H2}Ru(Cp0)] (Cp0 = C5H5, 18; Cp0 = C5Me5, 19) showed
that deviations of the cyclopentadienyl rings in complexes 18
and 19 from parallel are 2.6° and 10.1°, respectively.14b
We suggest that unless there is steric contact between the
pseudoequatorial tert-butyl groups and the nonmetalated
cyclopentadienyl ring, stronger bending of the exocyclic bond
and a shorter Ru(1)-Ru(2) distance should be expected. The
consequence of the above-mentioned steric contacts in 16 is
the hindrance to the approach of the Ru(1) atom to Ru(2);
this may be the reason for the long Ru(1)-Ru(2) distance
(12) (a) Watts, W. E. In Comprehensive Organometallic Chemistry;
Wilkinson, G.; Stone, F. G. A.; Abel, E. W., Eds.; Pergamon: London, 1988;
Vol. 8. (b) Koridze, A. A. Russ. Chem. Rev. 1986, 55, 113. (c) Barlow, S.;
Cowley, A.; Green, J. C.; Brunker, T. J.; Hascall, T. Organometallics 2001,
20, 5351.
(13) (a) Klahn, A. H.; Moore, M. H.; Perutz, R. N. J. Chem. Soc.,
Chem. Commun. 1992, 1699. (b) Lubke, B.; Edelmann, F.; Behrens, H.
Chem. Ber. 1983, 116, 11.
(14) (a) Koridze, A. A.; Kuklin, S. A.; Sheloumov, A. M.; Dolgushin,
F. M.; Lagunova, V.Yu.; Petukhova, I. I.; Ezernitskaya, M. G.; Peregudov,
A. S.; Petrovskii, P. V.; Vorontsov, E. V.; Baya, M.; Poli, R. Organome-
tallics 2004, 23, 4585. (b) Kuklin, S. A.; Dolgushin, F. M.; Petrovskii, P. V.;
Koridze, A. A. Russ. Chem. Bull., Int. Ed. 2006, 55, 1950.

Article
Organometallics, Vol. 29, No. 19, 2010 4365
Scheme 3. Description of Complexes 15 and 16 as Two
Canonical Forms, C and D
As in the case of R-metallocenyl carbocations, two extreme
canonical forms can be considered for bonding of substituted
ligands in cationic complexes 15 and 16 (Scheme 3). The
structure C is similar to that of cationic η5-cyclopentadienyl-
η6-fulvene-metal(II) complexes, whereas structure D resem-
bles bonding of a fulvene ligand in early transition metal
complexes,16 where the η6-ligand has little fulvene character
and the metalated η5,η1-bonding description is more prefer-
able; in this case the formal oxidation state of the metallo-
cene central atom should be M(IV).
Another remarkable feature of the structure of complex 16
is that alternation of the carbon-carbon distances in the
substituted ligand is very similar to that observed for the
R-ruthenocenyl carbocation [RcCH2][CF3SO3].12c Thus, for
16 (crystallographic isomer A) the C-C distances are C(1)-
C(2) 1.454(6), C(1)-C(5) 1.451(5), C(2)-C(3) 1.419(6),
Figure 5. Molecular structure of one of the two crystallogra-
phically independent molecules of 16 (ellipsoids at 30% prob-
ability level; hydrogen atoms and BArF anion are omitted for
4
˚
clarity). Selected bond lengths [A] and angles [deg]: Ru(1)-Ru(2)
3.1092(5), Ru(1)-P(1) 2.434(1), Ru(1)-P(2) 2.442(1), Ru(1)-
C(1) 2.009(4), Ru(1)-C(29) 1.862(4), Ru(1)-C(30) 1.893(5),
Ru(2)-C(1) 2.194(4), Ru(2)-C(2-10) 2.161(4)-2.237(5), C(1)-
C(2) 1.454(6), C(1)-C(5) 1.451(5), C(2)-C(3) 1.419(6), C(3)-
C(4) 1.444(6), C(4)-C(5) 1.422(6), P(1)-Ru(1)-P(2) 152.34(4),
C(29)-Ru(1)-Ru(2) 174.0(1), C(30)-Ru(1)-C(1) 143.4(2),
C(29)-Ru(1)-C(1) 129.4(2), C(29)-Ru(1)-C(30) 87.2(2),
C(30)-Ru(1)-Ru(2) 98.8(2), C(1)-Ru(1)-Ru(2) 44.6(1).
˚
C(4)-C(5) 1.422(6), and C(3)-C(4) 1.444(6) A, whereas
for [RcCH2][CF3SO3], the corresponding distances are
˚
1.458(6), 1.459(6), 1,413(7), 1.412(6), and 1.429 A, respec-
tively. In the case of the ruthenocenyl carbocation, the
alternation of the C-C bond lengths was considered as
consistent with the description of the C5H4CH2 ligand as a
coordinated fulvene similar to some substituted (η6-fulvene)-
Cr(CO)6 derivatives13c. On the other hand, for structure D,
we may expect equalization of bond lengths in the η-cyclo-
pentadienyl ring and lengthening of the exocyclic bond. It
˚
(3.1092(5) A) in comparison with Ru-Ru bond distances,
which generally range from 2.65 to 2.90 A. However there are
˚
precedents for similar long Ru-Ru distances in ruthenium
complexes.15 The Ru-Ru distance of 3.141(2) A is longer
˚
˚
should be noted that the C(1)-Ru(1) distance (2.009(4) A) in
than the normal Ru-Ru bond length in diruthenium carbo-
nyl complexes,15a but it is still within the range of bonding
distances.15b In any case, strong bending of the exocyclic
C(1)-Ru(1) bond toward the Ru(2) atom might be explained
as a result of bonding interaction of the Ru(2) atom with the
exocyclic substituent.
complex 16 is shorter than the ordinary C(1)-Ru(1) bond in
ruthenium pincer complexes with the metallocene backbone;
for example, in isomeric cationic complex 179 (see Scheme 1)
˚
it is equal to 2.047(3) A.
Altogether, the comparison of the X-ray diffraction data
for 16 with the structural data for the known R-metallocenyl
carbocation and fulvene complexes gives reason to consider
complex 16 as essentially a cationic ruthenium ruthenocenyl-
idene. The formation of pincer complexes with the carbene
backbone was reported in the literature;17 however, to the
best of our knowledge, metallocenylidene complexes have
not been obtained.
Complex 15 is less stable than 16; in solution it rearranges
into the known agostic complex 179 (Scheme 1). Apparently,
in 15 the metallocene central atom is bound to the chelated
ruthenium atom weaker than in 16, which explains the higher
inclination of complex 15 to this rearrangement as compared
It does not seem possible to describe unambiguously the
Ru(1) polyhedron in complex 16. If this atom is assumed
as pentacoordinative (neglecting the Ru(1)-Ru(2) inter-
action), its coordination sphere can be considered as strongly
distorted trigonal bipyramidal with the phosphorus atoms in
the apical positions (P(1)-Ru(1)-P(2) angle is 152.34(4)°;
C(1), C(29), and C(30) atoms lie in the equatorial plane;
˚
Ru(1) lies in this plane with an accuracy of 0.004 A).
However the C(30)-Ru(1)-C(1) (143.4(2)°), C(29)-Ru(1)-
C(1) (129.4(2)°), and C(29)-Ru(1)-C(30) (87.2(2)°) angles
differ very strongly from the ideal value of 120° for a trigonal
bipyramid. Taking into account the Ru(1)-Ru(2) inter-
action, the Ru(1) atom has a coordinative number equal
to 6; however, the polyhedron is strongly distorted from the
octahedral geometry (see angle values in the caption to
Figure 5).
(16) (a) Schock, L. E.; Brock, C. P.; Marks, T. Organometallics 1987,
6, 232. (b) Bulls, A. R.; Schaefer, W. P.; Serfas, M.; Bercaw, J. E. Organo-
metallics 1987, 6, 1219.
(17) See, for example: (a) Crocker, C.; Empsall, N. D.; Errington,
R. J.; Hyde, E. M.; McDonald, W. S.; Markham, R.; Norton, M. C.;
Shaw, B. L.; Weeks, B. J. Chem. Soc., Dalton Trans. 1982, 1217.
(b) Kuznetsov, V. F.; Abdur-Rashid, K.; Lough, A. J.; Gusev, D. G. J. Am.
Chem. Soc. 2006, 128, 14388. (c) Weng, W.; Chen, C.-H.; Foxman, B. M.;
(15) (a) Matsubara, K.; Ryu, K.; Maki, T.; Iura, T.; Nagashima, H.
Organometallics 2002, 21, 3023–3032. (b) Hogarth, G.; Phillips, J. A.; Van
Gastel, F.; Taylor, N. J.; Marder, T. B.; Carty, A. J. J. Chem. Soc., Chem.
Commun. 1988, 1570. (c) Johnson, B. F. G.; Lewis, J.; Mace, J. M.; Raithby,
P. R.; Vargas, M. D. J. Organomet. Chem. 1987, 321, 409.
ꢀ
Ozerov, O. L. Organometallics 2007, 26, 3315. (d) Oulie, P.; Nebra, N.;
Saffon, N.; Maron, L.; Martin-Vaca, B.; Bourissou, D. J. Am. Chem. Soc.
2009, 131, 3493.

4366 Organometallics, Vol. 29, No. 19, 2010
Koridze et al.
to that of its diruthenium analogue 16. Earlier complex
85% solution of phosphoric acid in D2O. Complete signal
assignments were made using COSY, NOESY, HMQC, and
HMBC spectra (see Supporting Information). IR spectra were
recorded on a Magna 750 IR (Nicolet) FTIR spectrometer with
a resolution of 2 cm-1. Calculations were performed using the
PRIRODA program20 and PBE functional21 with 3z (includes
functions for elements up to Xe) and sbk (includes functions
for elements heavier than Xe up to radon) basis sets. Results
obtained with these basis sets are practically the same. No
scaling for calculated IR frequencies was applied. Elemental
analyses were performed at the A.N. Nesmeyanov Institute of
Organoelement Compounds of RAS. Satisfactory elemental
analysis for complexes 4 and 5 could not be obtained due to
their instability. The products, however, appear to be analyti-
cally pure, as indicated by NMR. The synthesis and spectral and
structural data for complexes 1 and 17 were reported earlier.9
Their diruthenium counterpart, 2, was synthesized by a similar
procedure. NaBArF4 was purchased from Aldrich.
[Fc*CHRc]BF4 (Fc* is ferrocenyl-57Fe) was studied by 13
C
and 57Fe NMR, and it was found that the iron atom cannot
compete with the ruthenium atom for binding with the
“carbocationic center” CR; as a result, the positive charge
is delocalized on the Rc fragment.18
DFT calculations for complexes 15 and 16 are also in
agreement with the X-ray data (see Table 2 in the SI). The
Ru(1) atom deviates from the ring plane toward the Fe (or
Ru) atoms of the metallocene backbone. The calculation
predicts two ν(CO) bands in the IR spectra of approximately
equal intensities, which is in accord with the experimental
values. The DFT calculations revealed that complex 15,
where the Ru(1) atom deviates from the ring plane toward
the Fe(1) atom by 28.9°, is 3.9 kcal/mol less favorable than
agostic complex 17, where the Ru(1) atom deviates to the
opposite side by 9.3°, which is in accordance with X-ray data
(10.5°) . This difference is small, but it can be the driving force
of the rearrangement of 15 into 17.
Synthesis of RuCl(CO)[tBuP,C,PRu] (2). RuCl2(DMSO)4
(560 mg, 1.156 mmol) and NEt3 (230 mg, 2.277 mmol) were
added to a solution of {1,3-(CH2PtBu2)2C5H3}Ru(C5H5) (610
mg, 1.113 mmol) in 50 mL of 2-methoxyethanol. The reaction
mixture was stirred at 115 °C for 4 h. After cooling, the sol-
vent was removed under vacuum. The residue was purified by
column chromatography on neutral Al2O3. The crimson-red
fraction was eluated with a hexane/ethyl acetate mixture (7:1).
The solvent was removed under vacuum, and the residue was
crystallized from a CH2Cl2/hexane mixture. Crimson-red crys-
talline powder, yield: 570 mg (72%). 1H NMR (400 MHz,
CDCl3): δ 4.73 (s, 2H, C5H2), 4.44 (s, 5H, C5H5), 2.98 (dt, 2H,
Concluding Remarks
Novel cationic ruthenium complexes were obtained by the
reaction of ruthenium chloro-carbonyl metallocene-based
pincer complexes with H2 in the presence of NaBArF4. Addi-
tionally, it was demonstrated that the same complexes can be
obtained by the protonation of the appropriate ruthenium
hydrido-carbonyls. The structures of cationic ruthenium
hydrido-carbonyls and the ways of their formation are inter-
esting in light of the mechanism of electrophilic substitution/
protonation in iron group metallocenes.19 After CO addi-
tion, these complexes undergo a sequence (up to three steps)
of unprecedented intramolecular rearrangements. First
metallocenylidene complexes were obtained in the course
of these rearrangements accompanied by dihydrogen evolu-
tion. The mechanism of formation of complexes 3 and 4
and reactivity of cationic ruthenium hydrido-carbonyls and
metallocenylidene complexes are currently under study.
J
H-H = 16.7 Hz, JH-P = 3.3 Hz, CH2), 2.86 (dt, 2H, JH-H =
16.7 Hz, JH-P = 4.6 Hz, CH2), 1.44 (vt, JH-P = 7.0 Hz, 18H,
C(CH3)3), 1.19 (vt, JH-P = 6.4 Hz, 18H, C(CH3)3). 31P{1H}
NMR (162 MHz, CDCl3): δ 83.14. IR (CDCl3): 1923 cm-1
(νCO). Anal. Calcd for C29H47ClOP2Ru2: C, 48.97; H, 6.67.
Found: C, 48.82; H, 6.39.
Synthesis of {RuH(CO)[tBuP,CH,PFe]}þ{BArF }- (3). NaBArF
4
4
(30.0 mg, 0.034 mmol) was added to a solution of 1 (20 mg,
0.03 mmol) in 5 mL of CH2Cl2 in a stream of dry hydrogen at
room temperature. After 3 min the color turned from dark green
to orange, and both NaCl and excess NaBArF4 were filtered off.
The solvent was removed by hydrogen stream, and a yellow-
orange precipitate of 3 was formed. Yellow-orange crystalline
1
Experimental Section
powder, quantitative yield. H NMR (600 MHz, CD2Cl2): δ
7.75 (s, 8H, ArF), 7.59 (s, 4H, ArF), 4.45 (s, 5H, C5H5), 4.29 (s,
General Considerations. All reactions were carried out under a
purified argon atmosphere. All solvents were refluxed and
distilled over appropriate reagents under an argon atmosphere.
Deuterated solvents were degassed with argon. The NMR spec-
tra were recorded on Bruker Avance 300, 400, and 600 MHz
2H, C5H2), 3.22 (dt, 2H, J
= 16.4 Hz, JH-P = 3.6 Hz,
H-H
CH2), 3.09 (dt, 2H, JH-H = 16.4 Hz, JH-P = 3.6 Hz, CH2), 1.97
(s, 1H, C(1)-H), 1.32 (vt, JH-P = 5.5 Hz, 18H, C(CH3)3), 1.24
(vt, JH-P = 7.3 Hz, 18H, C(CH3)3), -26.22 (t, 1H JH-P = 15.9
Hz, RuH). 31P{1H} NMR (243 MHz, CD2Cl2): δ 67.9. 13C{1H}
1
spectrometers. H and 13C{1H} NMR chemical shifts are re-
NMR (151 MHz, CD2Cl2): δ 200.69 (s, CO), 161.65 (m, JC-B
=
=
49.8 Hz, Cipso-ArF), 134.70 (s, Cortho-ArF), 128.72 (q, JC-F
ported in parts per million downfield from tetramethylsilane; for
1H NMR spectra residual signals of deuterated solvents were
used as references (7.26 ppm for CDCl3, 7.16 ppm for C6D6, 5.32
ppm for CD2Cl2) (s, singlet; br s, broad singlet; d, doublet; t,
triplet; vt, virtual triplet). In 13C{1H} NMR measurements the
signals of CD2Cl2 (53.7 ppm) or CDCl3 (77.0 ppm) were used as
references. 31P{1H} NMR chemical shifts are reported in parts
per million downfield from H3PO4 and referenced to an external
32.0 Hz, Cmeta-ArF), 124.50 (q, JC-F = 273.0 Hz, CF3-ArF),
117.38 (s, Cpara-ArF), 98.51 (t, JC-P = 7.7 Hz, 2,5-C5H2), 71.30
(s, C5H5), 37.82 (s, C(1)H), 36.64 (t, C(CH3)3, JC-P = 6.4 Hz),
35.37 (t, C(CH3)3, JC-P = 6.4 Hz), 28.90 (s, C(CH3)3,), 27.94 (s,
C(CH3)3), 21.16 (t, CH2, JC-P = 6.6 Hz). IR (CH2Cl2): 1943
(νCO), 2020 cm-1 (νRuH). Anal. Calcd for C61H61BF24FeOP2Ru:
C, 48.98; H, 4.11. Found: C, 49.35; H, 4.17. A single crystal
suitable for X-ray analysis was obtained from CH2Cl2/hexane
at 0-4 °C.
(18) Koridze, A. A.; Astakhova, N. M.; Petrovskii, P. V. J. Organo-
met. Chem. 1983, 254, 345.
Synthesis of RuH(CO)[tBuP,CH,PRu]þ{BArF - (4). NaBArF
}
4 4
(19) (a) Cunningham, A. F., Jr. Organometallics 1997, 16, 1114.
(b) Cunningham, A. F., Jr. Organometallics 1998, 17, 4983. (c) Curphey,
T. J.; Santer, J. O.; Rosenblum, M.; Richards, J. H. J. Am. Chem. Soc. 1960,
82, 5249. (d) Bitterwolf, T. E.; Ling, A. C. J. Organomet. Chem. 1972, 40,
197. (e) Mueller-Westerhoff, U. T.; Haas, T. J.; Swiegers, G. F.; Leipert, T. K.
J. Organomet. Chem. 1994, 472, 229. (f) Karlsson, A.; Broo, A.; Ahlberg, P.
Can. J. Chem. 1999, 77, 628. (g) Borisov, Yu. A.; Ustynyuk, N. A. Russ.
Chem. Bull., Int. Ed. 2002, 51, 1900. (h) Wrackmayer, B.; Tok, O. L.;
Koridze, A. A. Magn. Reson. Chem. 2004, 42, 750. (i) B€uhl, M.; Grigoleit,
S. Organometallics 2005, 24, 1516.
(20.4 mg, 0.023 mmol) was added to a solution of 2 (15 mg, 0.021
mmol) in 5 mL of CH2Cl2 in a stream of dry hydrogen at room
temperature. After 3 min the color turned from dark red to
orange, and both NaCl and excess NaBArF were filtered off.
4
(20) Laikov, D. N. Chem. Phys. Lett. 1997, 281, 151.
(21) (a) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996,
77, 3865. (b) Ernzerhof, M.; Scuseria., G. E. J. Chem. Phys. 1999, 110, 5029.

Article
Organometallics, Vol. 29, No. 19, 2010 4367
The solvent was removed by hydrogen stream, and a yellow-
orange precipitate of 4 was formed. Yellow-orange crystalline
crystalline powder. 1H NMR (400 MHz, CD2Cl2): δ 7.74 (s, 8H,
ArF), 7.58 (s, 4H, ArF), 4.88 (s, 2H, C5H2), 4.25 (s, 5H, C5H5),
3.53 (dt, 2H, JH-H = 17.0 Hz, JH-P = 3.0 Hz, CH2), 3.18 (dt,
2H, JH-H = 17.0 Hz, JH-P = 4.3 Hz, CH2), 1.53 (vt, JH-P = 7.5
Hz, 18H, C(CH3)3), 0.94 (vt, JH-P = 6.5 Hz, 18H, C(CH3)3).
31P{1H} NMR (162 MHz, CD2Cl2): δ 98.54. IR (CH2Cl2): 2003,
1949 cm-1 (νCO).
1
powder, quantitative yield. H NMR (400 MHz, CD2Cl2): δ
7.73 (s, 8H, ArF), 7.57 (s, 4H, ArF), 4.78 (s, 5H, C5H5), 4.61 (s,
2H, C5H2), 3.13 (dt, 2H, J JH-H = 15.5 Hz, JH-P = 4.0 Hz,
CH2), 2.98 (dt, 2H, JH-H = 15.5 Hz, JH-P = 3.5 Hz, CH2), 2.95
(s, 1H, C(1)-H), 1.40-1.21 (m, 36H, C(CH3)3), -25.88 (t, 1H
H-P = 15.9 Hz, RuH). 31P{1H} NMR(162 MHz, CD2Cl2): δ
Synthesis of RuH(CO)2[tBuP,CH,PRu]þ{BArF }- (12). NaBArF
J
4
4
66.9. 13C{1H} NMR (151 MHz, CD2Cl2): δ 201.71 (t, JC-P
10.1 Hz, CO), 161.58 (m, JC-B = 49.9 Hz, Cipso-ArF), 134.62 (s,
ottho-ArF), 128.68 (q, JC-F = 31.4 Hz, Cmeta-ArF), 124.42 (q,
JC-F = 272.4 Hz, CF3-ArF), 117.31 (s, Cpara-ArF), 97.58 (t,
C-P = 9.7 Hz, 2,5-C5H2), 73.68 (s, C5H5), 73.11 (s, 3,4-C5H2),
44.17 (s, C(1)H), 36.69 (t, JC-P =7.2 Hz, C(CH3)3), 35.46 (t,
C-P=7.2 Hz, C(CH3)3), 28.83 (s, C(CH3)3,), 27.91 (s, C(CH3)3),
=
(35.0 mg, 0.038 mmol) was added to a solution of 2 (25 mg, 0.035
mmol) in 5 mL of CH2Cl2 in a stream of dry hydrogen at room
temperature. After 3 min the color turned from dark red to
orange. Then CO was bubbled through the solution until the
C
color turned to light yellow (4-5 min). Excess NaBArF and
J
4
NaCl were filtered off, and the solvent was removed under
reduced pressure. The spectra were recorded immediately after
the reaction because of fast elimination of dihydrogen with
formation of 16. Yellow crystalline powder; yield 95%. 1H
NMR (600 MHz, CDCl3): δ 7.76 (s, 8H, ArF), 7.58 (s, 4H, ArF),
4.83 (d, JH-P = 2.0 Hz, 2H, C5H2), 4.66 (s, 5H, C5H5), 2.98 (dt,
2H, JH-H = 15.3 Hz, JH-P = 4.0 Hz, CH2), 2.75 (dt, 2H, JH-H
= 15.3 Hz, JH-P = 3.6 Hz, CH2), 2.39 (s, 1H, C(1)H), 1.37 (vt,
J
21.14 (t, JC-P = 6.4 Hz, CH2). IR (CH2Cl2): 1945 (νCO), 2021
cm-1 (νRuH).
-
Synthesis of “trans”-RuH(CO)2[tBuP,CH,PFe]þ{BArF
}
4
(11). NaBArF4 (30.0 mg, 0.034 mmol) was added to a solution
of 1 (20 mg, 0.03 mmol) in 5 mL of CH2Cl2 in a stream of dry
hydrogen at room temperature. After 3 min the color turned
from dark green to orange; then CO was bubbled through the
solution until the color turned to dark violet (4-5 min). Excess
J
H-P = 4.0 Hz, 18H, C(CH3)3), 1.33 (vt, JH-P = 4.0 Hz, 18H,
C(CH3)3), -5.44 (dt, JH-P = 15.5 Hz, JH-H = 2.4 Hz, 1H,
NaBArF and NaCl were filtered off, and the solution was
RuH). 31P{1H} NMR (243 MHz, CDCl3): δ 75.06. 13C{1H}
4
concentrated to 1 mL. Hexane was accurately added, and the
solution was kept in a freezer at -18 °C. A strawberry-red,
crystalline powder was formed and filtered after 12 h in 82%
yield. 1H NMR (400 MHz, CD2Cl2): δ 7.74 (s, 8H, ArF), 7.58 (s,
4H, ArF), 4.63 (d, JH-P = 2.0 Hz, 2H, C5H2), 4.37 (s, 5H, C5H5),
3.15 (dt, 2H, JH-H = 15.1 Hz, JH-P = 4.0 Hz, CH2), 2.91 (dt,
2H, JH-H = 15.1 Hz, JH-P = 3.7 Hz, CH2), 1.43 (vt, JH-P = 7.0
Hz, 18H, C(CH3)3), 1.32 (vt, JH-P = 7.6 Hz, 18Hþ1H, C(1)-H,
C(CH3)3), -6.02 (dt, JH-P = 15.4 Hz, JH-H = 3.1 Hz, 1H,
RuH). 31P{1H} NMR (162 MHz, CD2Cl2): δ 75.7. 13C{1H}
NMR (100 MHz, CD2Cl2): δ 201.15 (s, CO), 198.51 (s, CO),
163.16 (m, JC-B = 50.3 Hz, Cipso-ArF), 136.44 (s, Cortho-ArF),
130.55 (q, JC-F = 32.7 Hz, Cmeta-ArF), 126.23 (q, JC-F = 273.7
Hz, CF3-ArF), 119.10 (s, Cpara-ArF), 107.43 (t, JC-P = 6.9 Hz,
NMR (150 MHz, CDCl3): δ 198.96 (t, CO), 197.27 (t, JC-P
11.4 Hz, CO), 161.70 (m, JC-B = 48.9 Hz, Cipso-ArF), 134.80 (s,
ortho-ArF), 128.9 (q, JC-F = 33.4 Hz, Cmeta-ArF), 124.66 (q,
C-F = 269.0 Hz, ArF), 117.50 (s, Cpara-ArF), 105.06 (t, JC-P
7.6 Hz, 2,5-C5H2), 75.52 (s, 3,4-C5H2), 73.52 (s, C5H5), 37.54 (t,
C-P = 8.2 Hz, C(CH3)3), 36.14 (t, JC-P = 6.0 Hz, C(CH3)3),
29.90 (s, C(CH3)3,), 29.39 (s, C(CH3)3), 22.28 (t, JC-P = 8.3 Hz,
CH2), 17.08 (s, C(1)H). IR (CH2Cl2): 1960, 2003 (νCO), 2057
cm-1(νRu-H). Anal. Calcd for C62H61BF24O2P2Ru2: C, 47.46;
H, 3.92. Found: C, 47.72; H, 3.90. A single crystal suitable for
=
C
J
=
J
X-ray analysis was obtained from CH2Cl2/hexane at 0-4 °C.
-
Synthesis of Ru(CO)2[tBuP,C,PRu]þ{BArF
}
(16). NaBArF
4
4
(35.0 mg, 0.038 mmol) was added to a solution of 2 (25 mg, 0.035
mmol) in 5 mL of CH2Cl2 in a stream of CO at room tempera-
ture, and the color of the reaction mixture turned from dark red
2,5-C5H2), 75.04 (s, C5H5), 72.34 (s, 3,4-C5H2), 39.06 (t, JC-P
=
to yellow. After 3 min the excess NaBArF and NaCl were
7.2 Hz, C(CH3)3), 37.88 (t, JC-P = 6.1 Hz, C(CH3)3), 31.41 (s,
C(CH3)3), 31.11 (s, C(CH3)3), 24.27 (t, JC-P = 7.0 Hz, CH2),
8.82 (s, C(1)H). IR (CH2Cl2): 1958, 2001 (νCO), 2057 cm-1
(νRu-H). Anal. Calcd for C62H61BF24FeO2P2Ru: C, 48.87; H,
4.03. Found: C, 48.38; H, 3.93. A single crystal suitable for
X-ray analysis was obtained from CH2Cl2/hexane at -18 °C.
Rearrangement of 11 into 13, 15, and 17. The solution of 11 in
CD2Cl2 obtained as described above was kept in an NMR tube
at room temperature for several days and monitored by NMR
spectra. After 24 h, the signals for 11:13:15:17 in a ratio of
4
filtered off through a short alumina column and the solvent was
removed under reduced pressure. Yellow crystalline powder,
yield: 91%. 1H NMR (400 MHz, CD2Cl2): δ 7.72 (s, 8H, ArF),
7.56 (s, 4H, ArF), 5.32 (d, 2H, C5H2), 4.90 (s, 5H, C5H5), 3.30
(dt, 2H, JH-H = 16.7 Hz, JH-P = 3.9 Hz, CH2), 2.34 (dt, 2H,
J
H-H = 16.7 Hz, JH-P = 2.7 Hz, CH2), 1.37 (vt, JH-P = 3.3 Hz,
18H, C(CH3)3), 1.35 (vt, JH-P = 3.9 Hz, 18H, C(CH3)3).
31P{1H} NMR (162 MHz, CD2Cl2): δ 110.87. IR (CH2Cl2):
2004, 1949 cm-1 (νCO). Anal. Calcd for C62H59BF24P2O2Ru2: C,
47.52; H, 3.80; P, 3.95. Found: C, 47.61; H, 3.87; P, 4.16. A single
crystal suitable for X-ray analysis was obtained from CH2Cl2/
hexane at room temperature.
1
10:10:1:6 were observed in the H and 31P{1H} NMR spectra.
After 72 h only 15 and 17 in a 2:1 ratio with 15% decomposition
products appeared in the 31P{1H} NMR spectrum. The solvent
was removed under reduced pressure, and the dark green resi-
due was separated by column chromatography on alumina
(CH2Cl2/hexane). Compound 179 was isolated as a single pro-
duct in 80% yield (from 1). Complex 15 was fully converted into
Synthesis of RuH(CO)[tBuP,C,PFe] (5). Complex 1 (100 mg,
0.15 mmol) in 3 mL of THF was added via canula to the
precooled, to -78 °C, suspension of 50 mg of LiAlH4 in 2 mL
of THF. The reaction mixture immediately changed color from
dark green to light yellow. After stirring for 3 min methanol
(0.5 mL) was added dropwise at -78 °C. The mixture was
filtered through Celite, and volatiles were removed under re-
duced pressure to yield 60 mg (63%) of an orange-red solid. 1H
NMR (600 MHz, CD2Cl2): δ 4.64 (d, 2H, C5H2), 3.94 (s, 5H,
C5H5), 2.97 (dt, 2H, JH-H = 17.0 Hz, JH-P = 3.2 Hz, CH2),
2.74 (dt, 2H, JH-H = 17.0 Hz, JH-P = 4.6 Hz, CH2), 1.44 (vt,
JH-P = 6.4 Hz, 18H, C(CH3)3), 1.19 (vt, JH-P = 6.3 Hz, 18H,
C(CH3)3), -26.65 (t, JH-P = 19.2 Hz, RuH). 31P{1H} NMR
(162 MHz, CD2Cl2): δ 117.88. IR (CH2Cl2): 1911 cm-1 (νCO).
Complex 5 was unstable both in solution and in solid state and
used immediately for protonation.
1
17 on Al2O3. Compound 13: H NMR (400 MHz, CD2Cl2): δ
7.74 (s, 8H, ArF), 7.58 (s, 4H, ArF), 4.63 (d, JH-P = 2.0 Hz,
2H, C5H2), 4.37 (s, 5H, C5H5), 3.15 (dt, 2H, JH-H = 15.1 Hz,
JH-P = 4.0 Hz, CH2), 2.91 (dt, 2H, JH-H = 15.1 Hz, JH-P = 3.7
Hz, CH2), 2.43 (d, JH-H = 1.3 Hz, 1H, C(1)-H), 1.43 (vt,
JH-P = 7.0 Hz, 18H, C(CH3)3), 1.34 (vt, JH-P = 7.6 Hz,
18H, C(CH3)3), -5.08 (dt, JH-P = 17.5 Hz, JH-H = 1.3 Hz,
1H, RuH). 31P{1H} NMR (162 MHz, CD2Cl2): δ 75.7. Com-
pound 15: 1H NMR (400 MHz, CD2Cl2): δ 7.74 (s, 8H, ArF), 7.58
(s, 4H, ArF), 4.49 (s, 2H, C5H2), 4.08 (s, 5H, C5H5), 3.02 (dt, 2H,
JH-H = 17.7 Hz, JH-P = 3.9 Hz, CH2), 2.71 (dt, 2H, JH-H
=
17.7 Hz, JH-P = 4.6 Hz, CH2), 1.70 (vt, JH-P = 7.4 Hz, 18H,
C(CH3)3), 1.36 (vt, JH-P = 7.6 Hz, 18H, C(CH3)3). 31P{1H}
NMR (162 MHz, CD2Cl2): δ 111.15. Compound 17:9 dark green,
Protonation of RuH(CO)[tBuP,C,PFe] (5) with CF3COOH.
One drop of TFA was added to a solution of 5 in C6D6 in an

4368 Organometallics, Vol. 29, No. 19, 2010
Koridze et al.
NMR tube; the spectra were immediately measured and showed
the formation of 3* in quantitative yield. 1H NMR (400 MHz,
C6D6): δ 4.45 (s, 5H, C5H5), 4.37 (s, 2H, C5H2), 3.05 (dt 2H,
the structures 3, 11, 12, and 16 were placed geometrically. All
hydrogen atoms were included in the structure factor calcula-
tions in the riding motion approximation. In the crystal struc-
tures of 11 and 12, disordered solvate molecules of hexane and
dichloromethane were located with 0.5/0.5 and 0.7/0.3 occu-
pancies, respectively. The tert-butyl groups at the P(2) atom in
11 are disordered over two positions with 0.67/0.33 occupancies.
In all structures, some of the CF3 groups in the anion are
disordered. In the crystal structure of 16, there are two crystal-
lographically independent cation-anion pairs with identical
geometrical parameters. The principal experimental and crystal-
lographic parameters are presented in Table 3 (see the SI).
JH-H = 15.8 Hz, CH2,), 2.76 (dt, 2H, JH-H = 15.8 Hz, CH2),
2.07 (s, 1H, C(1)-H), 1.45 (br s, 18H, C(CH3)3), 1.02 (br s, 18H,
C(CH3)3), -26.39 (t, 1H, JH-P = 16.6 Hz, RuH). 31P{1H}
NMR (162 MHz, C6D6): δ 67.3. Complex 3* was unstable in
solution, unlike 3, possibly because of intolerance of cationic
species to CF3COO- counterion.
X-ray Diffraction Study of 3, 11, 12, and 16. Single-crystal
X-ray diffraction experiments were carried out with a Bruker
SMART APEX II diffractometer (graphite-monochromated
ꢀ
˚
Mo KR radiation, λ = 0.71073 A, ω-scan technique, T =
100(2) K). APEX II software22 was used for collecting frames
of data, indexing reflections, determination of lattice constants,
integration of intensities of reflections, scaling, and absorption
correction, and SHELXTL23 for space group and structure
determination, refinements, graphics, and structure reporting.
The structures were solved by direct methods and refined by the
full-matrix least-squares technique against F2 with the aniso-
tropic thermal parameters for all non-hydrogen atoms. The
H(1M) and H(1) atoms in structures 3, 11, and 12 were located
in difference Fourier synthesis; the remaining hydrogen atoms in
Acknowledgment. This work was partly supported by
the Russian Academy of Sciences, OX-01 Program of
the Branch of Chemistry and Materials Sciences, the
Russian Foundation for Basic Research (Project Nos.
10-03-00505, 08-03-01020), and the International Science
and Technology Center (ISTC, Grant G 1361).
Supporting Information Available: Crystal structures for
complexes 3, 11, 12, and 16 (in CIF format), NMR and IR
spectra of complexes, and results of DFT calculations. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(22) APEX II software package; Bruker AXS Inc.: Madison, WI, 2005.
(23) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112.