Synthesis, structures, and reactions of coordinatively unsaturated trinuclear ruthenium polyhydrido complexes, [{Ru(C(5)Me (5))}(3)(μ-H)(6)](Y) (Y = BF(4), CF (3)SO(3), 1/2(SO(4)), C(6)H (5)CO(2), CH(3)CO(2), B(C(6)H (5))(4), PF(6)) and [{Ru(C(5)M (5))}(3)(μ-H)(3)(μ(3)-H)(2)]

ꢀ 2005 The Chemical Society of Japan

Bull. Chem. Soc. Jpn., 78, 67–87 (2005)

67

BCSJ Award Article

Synthesis, Structures, and Reactions of Coordinatively

Unsaturated Trinuclear Ruthenium Polyhydrido Complexes,

[{Ru(C₅Me₅)}₃(ꢀ-H)₆](Y) (Y ¼ BF₄, CF₃SO₃, 1/2(SO₄), C₆H₅CO₂,

CH₃CO₂, B(C₆H₅)₄, PF₆) and [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂]

ꢀ

Hiroharu Suzuki, Takeaki Kakigano, Ken-ichi Tada, Minoru Igarashi,

Kouki Matsubara,¹Akiko Inagaki,²Masato Oshima,³and Toshiro Takao

Department of Applied Chemistry, Graduate School of Science and Engineering, Tokyo Institute of Technology

and CREST, Japan Science and Technology Corporation (JST), O-okayama, Meguro-ku, Tokyo 152-8552

¹Department of Chemistry, Fukuoka University, 8-19-1 Nanakuma, Jonan-ku, Fukuoka 814-0180

²Chemical Resources Laboratory, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8503

³Department of Applied Chemistry, Faculty of Engineering, Tokyo Institute of Polytechnics, 1583 Iiyama, Atsugi,

Kanagawa 243-0297

Received June 18, 2004; E-mail: hiroharu@n.cc.titech.ac.jp

Cationic triruthenium hexahydrido complexes, [{Ru(C₅Me₅)}₃(ꢀ-H)₆](Y) (Y ¼ BF₄, CF₃SO₃, 1/2(SO₄),

C₆H₅CO₂, CH₃CO₂), have been synthesized by the reaction of diruthenium tetrahydride [{Ru(C₅Me₅)}₂(ꢀ-H)₄]

with the corresponding acid, HBF₄, CF₃CO₂H, H₂SO₄, C₆H₅CO₂H, and CH₃CO₂H. The trimeric structure of these

complexes has been conﬁrmed by a crystallographic study of [{Ru(C₅Me₅)}₃(ꢀ-H)₆](PF₆). The treatment of

[{Ru(C₅Me₅)}₃(ꢀ-H)₆](Y) with CH₃ONa in methanol selectively aﬀorded a neutral trinuclear pentahydrido complex,

[{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂], which underwent an intermolecular H/D exchange reaction with benzene-d₆, tol-

uene-d₈, or o-xylene-d₁₀to give [{Ru(C₅Me₅)}₃(ꢀ-D)₃(ꢀ₃-D)₂] as the result of arene C–H bond activation via an

ꢁ²-arene intermediate complex. A cationic hydrido complex, [{Ru(C₅Me₅)}₃(ꢀ-H)₆](Y), having a carboxylate as a

counter anion was equilibrated with [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] in solution, and the equilibrium constant depended

on the counter anion. An X-ray diﬀraction study showed that [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] has a triangular reaction

ﬁeld surrounded by three C₅Me₅ligands. The reaction of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] with 1 equiv of O₂proceeded

with the retention of the Ru₃framework to yield an 80/20 mixture of a mono-ꢀ₃-oxo complex [{Ru(C₅Me₅)}₃(ꢀ-

H)₃(ꢀ₃-O)] and a di-ꢀ₃-oxo complex [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-O)₂]. A novel trinuclear ꢀ₃-iodo-tetra-ꢀ-hydrido

complex, [{Ru(C₅Me₅)}₃(ꢀ₃-H)(ꢀ-H)₃(ꢀ₃-I)] was formed upon the treatment of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] with

1 equiv of CH₃I in tetrahydrofuran. The treatment of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] with carbon monoxide generated a

paramagnetic trinuclear tetracarbonyl complex, [{Ru(C₅Me₅)}₃(ꢀ-CO)₃(ꢀ₃-CO)].

The reactivity of transition metal cluster complexes has

been of special interest in the area of recent organometallic

chemistry in anticipation of their remarkable reactivity due

to the synergy of adjacent metal centers. In a multimetallic re-

action site of the cluster complex, substrates are often eﬀec-

tively activated in a manner diﬀerent from that achieved by

a mononuclear complex. This is most probably due to a coop-

erative action of the metal centers. A large number of studies

on the reaction chemistry of clusters have thus far been report-

ed.¹However, most of the metal clusters used in these studies

have been carbonylmetal complexes, or clusters directly de-

rived from them. In contrast to metal carbonyl clusters, metal

polyhydrido clusters have thus far been relatively unexplored,

probably due to their instability, and the lack of a rational syn-

thetic method applicable to the many ligand/metal combina-

tions, although they are expected to be much more active than

the polycarbonyl clusters in the oxidative addition of the sub-

strates.

Among the intrinsic properties of metal clusters, their capa-

bility for multiple-coordination is probably the most essential

point concerning the multimetallic activation, namely, multi-

ple-coordination seems likely to be an origin of the ‘‘cluster ef-

fect’’ and should give rise to cooperation of the metal centers

in the activation of the substrate. Therefore, we adopted a met-

al polyhydrido cluster, especially a cluster having bridging hy-

drido ligands, as a precursor of the active species for multime-

Published on the web January 13, 2005; DOI 10.1246/bcsj.78.67

68

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

BCSJ AWARD ARTICLE

tallic activation. The liberation of bridging hydrido ligands

from the cluster generates, spontaneously, multiple vacant co-

ordination sites on the adjacent metal centers, all together.

Although there have, thus far, been several reported exam-

ples of multinuclear polyhydrido-bridged complexes,^2–16

their reaction chemistry is not as remarkable or unique, un-

fortunately.

As a part of our early eﬀorts directed toward preparing the

polynuclear polyhydrido complexes, we reported on the syn-

thesis and reactions of a dinuclear tetrahydrido-bridged com-

plex of ruthenium, [{Ru(C₅Me₅)}₂(ꢀ-H)₄] (1), and demon-

strated several typical examples of bimetallic activation by

using 1 as a precursor of the active species.^17–31

room temperature revealed a broad resonance signal around

ꢂ 5.9 (2H) of H₂SO₄, incorporated into the molecule together

with those of C₅Me₅(ꢂ 2.00, 45H) and hydrido (ꢂ ꢃ11:23, 6H)

ligands. A strong and broad absorption for the O–H stretching

vibration of H₂SO₄with a hydrogen-bonding interaction ap-

peared in the range of 3000–3500 cm^ꢃ1in the IR spectrum.

The amount of sulfuric acid incorporated in 2c changed over

a wide range along with a change in the amount of acid used

in the reaction. However, excess sulfuric acid incorporated in

2c could hardly be removed, despite evacuation or washing

with diethyl ether.

The reaction of 1 with an acid having a nucleophilic conju-

gate base caused a more complicated result. The reaction of 1

with carboxylic acid, such as benzoic acid or acetic acid, yield-

ed the corresponding cationic hexahydrido complex, 2d

(Y ¼ C₆H₅CO₂) or 2e (Y ¼ CH₃CO₂), respectively, together

with a dinuclear di-ꢀ-carboxylato complex, 3d or 3e (vide

infra).³²

Here, we describe in full detail the synthesis, characteriza-

tion, and some reactions of trinuclear ruthenium polyhydrido

complexes, [{Ru(C₅Me₅)}₃(ꢀ-H)₆](Y) (Y ¼ BF₄, CF₃SO₃,

1/2(SO₄), C₆H₅CO₂, CH₃CO₂, BPh₄, and PF₆) and [{Ru-

(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂].

Trinuclear hexahydrido complexes 2a–c are stable in solu-

tion, even in chloroform, at room temperature, and can be kept

in chloroform for several hours without decomposition, while

2d and 2e are labile in solution. Since single crystals suitable

for X-ray diﬀraction could not be obtained by the recrystalli-

zation of 2a–c from tetrahydrofuran, an exchange of the coun-

ter anion was attempted. The counter anion of the cationi_ꢃc

complex 2c, 1/2SO₄^ꢃ, was readily replaced by BPh₄^ꢃor PF₆

upon a treatment with NaBPh₄or NH₄PF₆. The addition of an

excess amount of NaBPh₄or NH₄PF₆to a suspension of 2c in

tetrahydrofuran gave the corresponding cationic complex,

[{Ru(C₅Me₅)}₃(ꢀ-H)₆](Y) (Y ¼ BPh₄(2f) or PF₆(2g)), in

good yield (Eq. 2).

Results and Discussion

Synthesis of Cationic Triruthenium Hexahydrido Com-

plexes, [{Ru(C₅Me₅)}₃(ꢀ-H)₆](Y) (Y ¼ BF₄, CF₃SO₃, 1/2-

(SO₄), C₆H₅CO₂, CH₃CO₂, BPh₄, and PF₆). The addition

of an equimolar amount of HBF₄OEt₂to a suspension of a

ꢁ

diruthenium tetrahydrido complex, [{Ru(C₅Me₅)}₂(ꢀ-H)₄]

(1), in tetrahydrofuran gave a cationic trinuclear hydrido com-

plex, [{Ru(C₅Me₅)}₃(ꢀ-H)₆](BF₄) (2a), in excellent yield

(Eq. 1).

Ru

H

HY

H

Ru

(Y)

Ru

H

ð1Þ

Ru

H

AY

H

1

H

(1/2)SO₄

(Y)

2a: Y = BF

4

Ru

ð2Þ

H

2b: Y = CF SO

3

2c: Y = (1/2)SO

4

AY: NaBPh₄, NH₄PF₆

2c

2f: Y = BPh

4

Although the protonation of 1 proceeded very slowly below

^ꢂC, a mixture of a red suspension of the starting hydrido

2g: Y = PF

6

0

complex 1 and the acid gradually turned brown-purple with

the evolution of dihydrogen upon warming to room tempera-

ture. A workup gave 2a as a purple microcrystalline solid in

Complexes 2f and 2g are stable both in the solid state and in

solution. Single crystals of 2f and 2g suitable for X-ray diﬀrac-

tion studies were obtained from tetrahydrofuran and 1,2-dime-

thoxyethane, respectively.

88% yield. When HBF₄OEt₂was added to 1 at room temper-

ꢁ

ature, the reaction completed instantly with vigorous evolution

of dihydrogen, resulting in a quantitative formation of 2a.

Analogous procedures using acids having a non-coordinat-

ing, less-nucleophilic conjugate base led to the corresponding

cationic trinuclear hexahydrido complexes 2. For example, the

reaction of 1 with excess CF₃SO₃H smoothly proceeded to

give the corresponding cationic complex, [{Ru(C₅Me₅)}₃(ꢀ-

H)₆](CF₃SO₃) (2b). In contrast to these reactions, the use of

sulfuric acid brought about a somewhat complicated result.

When compound 1 was treated with an equimolar amoumt

of H₂SO₄, a purple-brown solid of a cationic complex,

[{Ru(C₅Me₅)}₃(ꢀ-H)₆](1/2SO₄+ H₂SO₄) (2c), was formed,

in which excess H₂SO₄was incorporated into the product

through a hydrogen bond to the sulfate ion. The incorporation

The cationic hexahydrido complexes 2a–g were character-

ized based on the ¹H and ¹³C NMR and IR spectroscopies,

as well as the electric conductivity. While cationic hydrido

complexes, 2d and 2e, immediately decomposed in chloro-

form-d₃, they remained unchanged for at least several minutes

in dichloromethane. Their NMR spectra were, therefore, re-

corded in dichloromethane-d₂.

1

In the H NMR spectra of hexahydrido complexes 2a–g, a

signal of the hydrido ligands appeared at around ꢂ ꢃ11:2 as

1

a singlet peak. A variable-temperature H NMR measurement

of 2a in acetone-d₆showed that the signals of the C₅Me₅

groups and the hydrido ligands were observed as a singlet

peak,_ꢂrespectively, in the temperature range from 50 ^ꢂC to

ꢃ90 C. The half-height width of the signal of the hydrido li-

gands slightly broadened upon lowering the temperature from

2.2 Hz (at 50 ^ꢂC) to 4.0 Hz (ꢃ90 ^ꢂC). This result implied that

1

of H₂SO₄was conﬁrmed by means of H NMR and IR spec-

1

troscopy. The H NMR spectrum of 2c recorded in CDCl₃at

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

69

the cationic hexahydrido complex has a pseudo-threefold sym-

metry axis; also the six hydrido ligands were observed to be

indistinguishable. A structure having two sets of three identical

hydrides, namely, three bridging hydrides and three terminal

hydrides, cannot be ruled out altogether, if they rapidly ex-

change each coordination site within the NMR time-scale. A

structure having six bridging hydrides was, however, support-

ed by an ab initio calculation,³³and was conﬁrmed by an X-ray

diﬀraction study of 2g (vide infra).

Synthesis and Structure Determination of the Di-ꢀ-car-

boxylato-di-ꢀ-hydrido-bis[(pentamethyl-ꢁ⁵-cyclopentadi-

enyl)ruthenium], [{Ru(C₅Me₅)}₂(ꢀ-H)₂(ꢀ-OCOR)₂] (R ¼

CF₃, H, and Adamantyl). As mentioned above, an acid bear-

ing a bulky and less-nucleophilic conjugate base, such as

HBF₄and CH₃SO₃H, reacts with 1 to form the cationic trinu-

clear hexahydrido complex 2. The reaction of 1 with a series of

carboxylic acids, however, leads to the formation of neutral di-

nuclear di-ꢀ-carboxylato complexes 3, as the result of a nucle-

ophilic attack of the carboxylato anion at the cationic metal

centers of an intermediary dicationic diruthenium species,

[{Ru(C₅Me₅)}₂(ꢀ-H)₄]^2þ(vide infra). The treatment of 1 with

excess carboxylic acid, such as CF₃CO₂H, HCO₂H, or

AdCO₂H (Ad = adamantyl), in tetrahydrofuran exclusively

gave a neutral dinuclear di-ꢀ-carboxylato-di-ꢀ-hydrido com-

plex, [{Ru(C₅Me₅)}₂(ꢀ-H)₂(ꢀ-OCOR)₂] (R ¼ CF₃(3a), H

(3b), and Ad (3c)), in excellent yields along with the release

of H₂(Eq. 4).³²Benzoic acid or acetic acid, however, pro-

duced a moderate result between those brought about by min-

eral acids, such as HBF₄or H₂SO₄, and carboxylic acid, such

as triﬂuoroacetic acid, formic acid, and 1-adamantanecarbox-

ylic acid. The reaction of 1 with excess benzoic acid or acetic

acid gave a mixture of the corresponding di-ꢀ-carboxylato-di-

ꢀ-hydrido complex 3 (R ¼ C₆H₅(3d) and CH₃(3e)) and the

cationic hexahydrido complex 2 (Eq. 5). The product distribu-

tion between two hydrido complexes 3/2 was about 3/1 in

each case.

In order to clarify the bonding mode of the hydrido ligands

in 2a, ꢁ²-H₂or ꢁ¹-H, we measured the relaxation time (T₁) by

using standard JEOL programs. The T₁values for the reso-

nance signal at ꢂ ꢃ11:23, measured in CD₂Cl₂at 500 MHz,

ꢂ

were found to be 385 ms (at ꢃ50 C) and 387 ms (at ꢃ80

^ꢂC). These values are larger than that of the hydride signal

in the dihydrogen complexes, and lies within the range of those

for the classical hydrido complexes.³⁴Consequently, we con-

cluded that not only 2a, but also a series of the cationic triru-

thenium hexahydrido complexes 2, are the classical hydrido

complexes. This result is consistent with the relatively long

distance between the two hydrogen atoms bridging the same

Ru–Ru bond observed in the X-ray structure of 2g (vide infra).

The cationic complex 2b was alternatively prepared by the

treatment of 1 with an ester, CF₃SO₃CH₃, instead of the acid,

CF₃SO₃H. In this case, as anticipated, the formation of 2b was

accompanied by the evolution of methane.

Whereas the reaction conducted in tetrahydrofuran generat-

ed the cationic complex 2, the reaction in the presence of

CH₃CN or benzene yielded a cationic mononuclear coordina-

tively saturated complex, [Ru(C₅Me₅)(CH₃CN)₃](Y)³⁵or

[Ru(C₅Me₅)(C₆H₆)](Y),^17,23respectively.

An ethyltetramethylcyclopentadienyl analogue, [{Ru-

(C₅Me₄Et)}₃(ꢀ-H)₆]1/2(SO₄+ H₂SO₄) (2c⁰), was obtained

in excellent yield in the reaction of [{Ru(C₅Me₄Et)}₂-

(ꢀ-H)₄]³⁶with sulfuric acid,_ꢃand the sulfate ion of 2c⁰was

readily exchanged with BPh₄to form [{Ru(C₅Me₄Et)}₃(ꢀ-

H)₆](BPh₄) (2f⁰) by a treatment of 2c⁰with NaBPh₄in tetrahy-

drofuran.

R

O

H

O

H

O

H

RCO₂H

Ru

H

ð4Þ

1

3a: R = CF

3

3b: R = H

3c: R = Ad

R

Ru

H

O

H

O

H

O

RCO₂H

H

Ru

+

(Y)

Ru

H

Hydrido ligands that exhibit an acidic character in polar sol-

vents can readily undergo an intermolecular hydrogen ex-

change reaction with a protic solvent, such as alcohol. An ex-

change of the hydrido ligands in 2 with methanol was exam-

ined by means of an isotopic labeling experiment using a deu-

terated solvent. Monitoring the H/D exchange reaction

between 2c and methanol-d₄by ¹H NMR spectroscopy showed

that the reaction did not take place at room temperature, but

3d: R = C H

6

1

5

2d: Y = C H CO

2

3e: R = CH

3

6

5

ð5Þ

2e: Y = CH CO

3

2

The di-ꢀ-carboxylato-di-ꢀ-hydrido complexes 3 are stable

in the solid state and are very soluble in diethyl ether or tol-

uene, whereas the cationic complex 2 is sparingly soluble in

such less-polar solvents. Complexes 2 and 3 formed in the re-

action could, therefore, be readily separated from each other

by extraction with diethyl ether.

ꢂ

was completed upon heating at 80 C for 4 h. The half-height

width of the hydrido signal signiﬁcantly broadened with the

progress of the exchange reaction. This process is, of course,

reversible and the starting hexahydrido complex 2c is quanti-

tatively recovered when 2c-d₆is heated in methanol-h₄at 80

^ꢂC for 4 h (Eq. 3).

Complexes 3a–3e were characterized on the basis of ¹H, ¹³C,

and ¹⁹F NMR and infrared spectra. In the ¹H NMR spectra,

a singlet resonance signal of intensity 2H for the bridging

hydrides appeared in the range of ꢂ ꢃ4:50 to ꢃ4:10 ppm, which

is signiﬁcantly low ﬁeld compared to those for the hydrido

ligands in the dinuclear tetrahydride 1 (ꢂ ꢃ13:99 in C₆D₆)^17,23

or the cationic trinuclear complexes 2 (ꢂ ca. ꢃ11:2). The longi-

tudinal relaxation time (T₁) for the hydrido ligands in 3a was

Ru

CD₃OD

CH₃OH

H

D

H

D

(1/2)SO₄

Ru

H

D

ꢂ

determined to be 251 ms at ꢃ50 C by using the inversion–

recovery method. The observed T₁value is suﬃcient to charac-

terize the complex as a classical dihydrido complex.

ð3Þ

2c

2c-d

6

70

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

BCSJ AWARD ARTICLE

The two C₅Me₅groups were observed to be equivalent, as

them. The liberation of the dihydrogen was conﬁrmed when

the reaction of 1 with an equimolar amount of H₂SO₄in tetra-

demonstrated by the presence of a singlet peak of intensity

30H in the ꢂ 1.70 to 1.84 region. There is evidently a symme-

try element that relates the two ruthenium atoms; most likely,

this is a mirror plane that bisects the ruthenium–ruthenium

vector and passes through the two bridging hydrides and the

two acetoxy carbon atoms. The infrared spectra of complexes

3 revealed strong absorption bands characteristic of a bidentate

carboxylato ligand in the 1660 to 1560 cm^ꢃ1region.³⁸

The molecular structure of 3a was conﬁrmed based on an X-

ray diﬀraction study.³⁹The structure is fully consistent with

the spectral data mentioned above. The two ruthenium atoms

are quadruply bridged by the two carboxylato ligands and

the two hydrides. The interatomic distances of the two ruthe-

ꢂ

1

hydrofuran-d₈at 10 C was monitored by means of H NMR

spectroscopy. A resonance signal of the liberated dihydrogen

appeared at ꢂ 4.54 as a sharp singlet peak, although the integral

intensity of the signal was somewhat weak because of its poor

solubility in tetrahydrofuran-d₈. The chemical shift of ꢂ 4.54

agreed well with that of dihydrogen saturated in tetrahydrofu-

ran-d₈, and is also fully consistent with that reported for free H₂

liberated from W(CO)₃(PR₃)₂(ꢁ²-H₂) (ꢂ 4.72).⁴²In solution,

the dicationic species B must be in equilibrium with a highly

unsaturated mononuclear cationic species C (14e), which im-

mediately reacts with the unprotonated dinuclear tetrahydrido

complex 1 to form the monocationic trinuclear hexahydrido

complex 2. The reaction of 1 with C would be much faster

than that of 1 with the dinuclear species B, due to sterical rea-

sons. The formation of the trinuclear monocationic complex 2

is, as a result, predominant over that of a tetranuclear dication-

ic complex resulting from the reaction of 1 with B.⁴³In fact,

the tetranuclear dicationic complex was not detected at all un-

der these conditions. The generation of the intermediate mon-

onuclear cationic species C is supported by the fact that treat-

ment of 1 with CF₃SO₃H in benzene quantitatively aﬀords a

coordinatively saturated cationic mononuclear complex having

an ꢁ⁶-C₆H₆ligand, [Ru(C₅Me₅)(C₆H₆)](CF₃SO₃) (4),^17,23,35as

a result of the coordinative addition of benzene to C, followed

by elimination of the two hydrido ligands as a molecular hy-

drogen. The mononuclear cation C was also quenched by ace-

tonitrile molecules to lead to a known cationic complex,

[Ru(C₅Me₅)(CH₃CN)₃](BF₄) (5),³⁵when 1 was treated with

ꢀ

nium atoms is 2.846(2) A, comparable to that observed in

the analogous dinuclear ꢀ-carboxylato complexes, [{Ru-

ꢀ

(C₆H₆)}₂(ꢀ-H)(ꢀ-OH)(ꢀ-OCOCH₃)](ClO₄) (2.785 A) and

40

ꢀ

[{Ru(C₆H₆)}₂(ꢀ-H)₂{ꢀ-OCOPh(OH)}](ClO₄) (2.674 A).

The EAN rule applied to the di-ꢀ-carboxylato-di-ꢀ-hydrido-

diruthenium 3 requires a metal–metal ꢃ bond between the

two ruthenium atoms; this is clearly reﬂected in the above-

mentioned Ru–Ru bond length.

Formation Mechanism of the Cationic Hexahydridotri-

ruthenium 2 and the Di-ꢀ-carboxylato-di-ꢀ-hydridodiru-

thenium 3. The formation of the trinuclear cationic hexahy-

drido complex 2 would be reasonably elucidated by the reac-

tion sequence illustrated in Scheme 1.

At the initial stage of the reaction, two protons successively

attack the ruthenium centers of 1 to yield a dinuclear dicationic

hexahydrido species A (30e), which would undergo a reductive

elimination of dihydrogen to generate an intermediary highly

unsaturated dinuclear dicationic tetrahydride complex B

(28e). This protonation–H₂liberation step is rationalized by

a calculation performed by Morokuma and Koga.⁴¹Shortening

of the interatomic distance between the two adjacent hydrido

ligands in the hexahydrido species A in comparison to 1 is re-

sponsible for giving rise to a bonding interaction between

HBF₄(OEt₂) in diethyl ether in the presence of a large excess

amount of acetonitrile.

ꢁ

In this mechanism, the dinuclear dicationic species B would

be captured by the counter anion in the case that the counter

anion is less bulky and suﬃciently nucleophilic. In the reaction

of 1 with carboxylic acid, di-ꢀ-carboxylato complex 3 is pre-

dominantly formed due to a nucleophilic attack of the carbox-

H

2 HY

(Y)

2

RuH

Ru

6

Ru

1

A

1

Ru

O

-

O

H

RCO

2

H

(Y)

RuH

2

RuH

Ru

4

Ru

2

Ru

-

H

(Y = RCO

)

2

3

B

C

2

CH CN

3

C H

6 6

6

Ru(CH CN)

3

(Y)

Ru(η -C H ) (Y)

6 6

3

5

4

Scheme 1.

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

71

ylato anion at the metal centers, whereas the reaction of 1

with an acid with a noncoordinating, a less nucleophilic anion,

such as HBF₄, CF₃SO₃H, or H₂SO₄, exclusively yields 2.

These facts are also reasonably explicable according to this

mechanism.

Ru

CH₃ONa

H

(1/2)SO₄

ð7Þ

CH₃OH

Ru

H

Synthesis of a Neutral Pentahydridotriruthenium Clus-

6

2c

ter, [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂].

The stability of the

cationic hexahydrido complexes 2 depends on the counter

ꢃ

anions. Complexes 2d (Y ¼ C₆H₅CO₂

)

and 2e (Y ¼

The ethyltetramethylcyclopentadienyl analogue of 6, [{Ru-

(C₅Me₄Et)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6⁰), was derived from the corre-

sponding cationic hexahydrido complex, [{Ru(C₅Me₄Et)}₃(ꢀ-

H)₆](BPh₄) (2f⁰), in a similar manner.

Compound 6 is soluble in tetrahydrofuran, benzene, and tol-

uene, and is less soluble in diethyl ether, pentane, and hexane.

Although 6 is fairly stable in the solid state, even in air, it is

extremely sensitive towards oxygen in solution. When a solu-

tion of 6 was exposed to atmospheric oxygen at ambient tem-

perature, mono- and dioxygenated products, [{Ru(C₅Me₅)}₃-

(ꢀ-H)₃(ꢀ₃-O)] (9) and [{Ru(C₅Me₅)}₃(ꢀ-H)(ꢀ₃-O)₂] (10),

were consecutively formed within a few seconds (vide infra).

CH₃CO₂^ꢃ) are labile in both solid state and solution, while

2a (Y ¼ BF₄^ꢃ) and 2b (Y ¼ CF₃SO₃^ꢃ) are insensitive toward

air and moisture, and do not decompose even in solution at

ambient temperature. The decomposition process of 2e in solu-

tion was elucidated by means of ¹H NMR spectroscopy. When

2e was dissolved in benzene-d₆or tetrahydrofuran-d₈and

allowed to stand at ambient temperature, it decomposed within

several minutes to result in the quantitative formation of a

neutral trinuclear pentahydrido complex, [{Ru(C₅Me₅)}₃(ꢀ-

H)₃(ꢀ₃-H)₂] (6). Compound 6 was isolated as a brown-black

crystalline solid when the solvent was removed under reduced

pressure from the resultant solution. On the contrary, the

addition of excess acetic acid to a tetrahydrofuran solution

of isolated 6 resulted in a quantitative regeneration of 2e.

These results show that 2e and 6 are equilibrated in solution

(Eq. 6).

1

The H NMR spectra of 6, measured in C₆D₆at room tem-

perature, exhibits two sharp singlet peaks at ꢂ 2.04 and ꢃ7:22

for the three C₅Me₅groups and the ﬁve hydrido ligands, re-

spectively. The ¹H NMR spectrum is temperature-independent

ꢂ

in the temperature range from ꢃ120 C to room temperature.

A variable-temperature ¹H NMR analysis of 6 in tetrahydrofu-

ran-d₈/toluene-d₈(5/1) performed at 500 MHz proved that

resonances for ﬁve hydrido ligands were observed_ꢂto be equiv-

alent in the temperature range from 20 to ꢃ120 C. The line

shape and intensity of the signal assignable to the hydrido li-

gands remained unchanged over this temperature range, al-

though the signal shifted downﬁeld from ꢂ ꢃ7:55 at 20 ^ꢂC

Ru

- RCO₂H

+ RCO₂H

H

(Y)

H

Ru

H

ð6Þ

H

6

2d: Y = C H CO

2

6

5

ꢂ

2e: Y = CH CO

to ꢂ ꢃ6:37 at ꢃ120 C. This is probably due to a rapid ex-

3

2

change of the hydrides among the coordination sites, ꢀ₂-

and ꢀ₃-sites, on an NMR timescale. The species in which

the triply bridging hydride has a bonding interaction with the

doubly bridging hydride is one of the possible intermediates

or transition states of this dynamic process. Two hydrides

would change the coordination sites with each other via rota-

tion around the coordination bond between the resulting ꢁ²-

H₂ligand and the ruthenium center. Site exchange of the hy-

drido ligands via an intermediary ꢁ²-H₂complex is proposed

for the mononuclear polyhydrido complex.^44–50

The equilibrium process between 2d and 6 in solution was

conﬁrmed in a similar way. The equilibrium constant K, where

K ¼ ½2dꢄ=ð½6ꢄ½C₆H₅CO₂HꢄÞ, can be estimated directly based

1

on the H NMR integration of the two separate hydride reso-

nance signals of 2d and 6 observed at ꢂ ꢃ11:26 and ꢃ7:22, re-

spectively. The results show that the equilibrium constant (K)

changes from 0.76 [M^ꢃ1] at 293 K to 0.14 [M^ꢃ1] at 333 K. A

least-squares ﬁt to the equation RT ln K ¼ ꢃꢄH=T þ ꢄS

shows that ꢄH ¼ ꢃ7:5 ꢅ 0:3 kcal mol^ꢃ1and ꢄS ¼ ꢃ26:5 ꢅ

1:0 cal mol^ꢃ1K^ꢃ1for the protonation of 6 with benzoic acid

to form 2d.

If an appropriate basic reagent is added to the equilibrated

mixture of 2 and 6, the acid generated in the solution can be

trapped. As a result, the equilibrium shifts to the right hand,

and a neutral trinuclear pentahydrido complex, 6, must there-

fore be obtained in a reasonable yield.

Compound 2c seems to be the most versatile starting

material for the synthesis of 6, because 2c can be obtained

in excellent yield with simple procedures, and is remarkably

stable. Among the reactions of 2c with various basic reagents,

the best result was obtained in a reaction with CH₃ONa. The

treatment of 2c with excess CH₃ONa in methanol at room

temperature for 10 min, followed by a chromatographic work-

up, aﬀorded 6 in 96% yield as a black-brown microcrystalline

solid (Eq. 7).

In order to obtain information on the bonding mode of the

hydrido ligands in 6, the relaxation time (T₁) was measured

in tetrahydrofuran-d₈at 500 MHz. The T₁values of 4.33 s

ꢂ

(at ꢃ50 C) and 2.55 s (at ꢃ80 C) for the hydride signal of

6 are signiﬁcantly large compared with that of the hydride sig-

nal of 2a. Elongation of the relaxation time indicates that there

is no bonding interaction among the hydrido ligands in 6.

Complex 6 was, therefore, concluded to be a classical hydrido

complex, as well as 2. A lack of the bonding interaction among

the hydrido ligands in 6 is entirely consistent with the long

H–H separation determined by an X-ray diﬀraction study (vide

infra).

As anticipated from the equilibrium between 2e and 6

(Eq. 6), the treatment of 6 with an acid, such as HBF₄,

CF₃SO₃H, H₂SO₄, or HPF₆, generates the corresponding

cationic hexahydrido complex 2 in a quantitative yield (Eq. 8).

72

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

BCSJ AWARD ARTICLE

Fig. 2. Thermal ellipsoid representation of [{Ru(C₅Me₅)}₃-

(ꢀ-H)₃(ꢀ₃-H)₂] (6) showing 30% probability ellipsoids.

CP is centroid of the C₅Me₅ring.

Fig. 1. Thermal ellipsoid representation of [{Ru(C₅Me₅)-

(ꢀ-H)₂}₃](PF₆) (2g) showing 30% probability ellipsoids.

The anion, PF₆, and dioxane molecules are omitted for

clarity. CP is centroid of the C₅Me₅ring.

ꢀ

Table 2. Selected Bond Distances (A) and Angles (deg) in

[{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6)

ꢀ

Table 1. Selected Bond Distances (A) and Angles (deg) in

Bond distances

[{Ru(C₅Me₅)(ꢀ-H)₂}₃](PF₆) (2g)

Ru1–Ru2

Ru1–Ru3

Ru2–Ru3

Ru–ꢀ-H

2.7534(8)

2.7503(8)

2.7453(6)

1.66 (av)

Ru1–CP1

Ru2–CP2

Ru3–CP3

Ru–ꢀ₃-H

1.808

1.811

1.808

1.94 (av)

Bond distances

Ru1–Ru2

Ru1–Ru3

Ru2–Ru3

Ru–H

2.7033(8)

2.7078(10)

2.7079(11)

1.71 (av)

Ru1–CP1

Ru2–CP2

Ru3–CP3

1.836

1.851

1.836

Angles

Ru2–Ru3–Ru1

Ru1–Ru2–Ru3

Ru3–Ru1–Ru2

60.02(2)

59.844(18)

60.135(18)

Angles

Ru1–Ru2–Ru3

Ru3–Ru1–Ru2

60.06(3)

60.05(2)

Ru2–Ru3–Ru1

59.89(2)

distances and angles are listed in Table 1. Compound 2g con-

sists of a nearly equilateral triangular cluster of the three ruthe-

nium atoms with the six doubly bridging hydrido ligands. Each

Ru–Ru bond is bridged by the two hydrogen atoms; one of

them is up and the other is down with respect to the Ru₃plane.

The separation of the two hydrogen atoms bridging the same

Ru

H

HY

H

(Y)

H

Ru

H

ð8Þ

H

ꢀ

Ru–Ru bond, av. 1.74 A, is larger than the H–H distance re-

51

ꢀ

ported for the dihydrogen ligands (ca. 0.8–1.05 A) in the

2

6

mononuclear complexes; this clearly shows that 2g is a classi-

cal hydrido complex. The average Ru–Ru distance of

(Y = BF₄^-, CF₃SO₃^-, (1/2)SO₄^2-, PF₆

)

-

ꢀ

2.7063(10) A is somewhat shorter than that reported for the

X-ray Crystal Structures of [{Ru(C₅Me₅)}₃(ꢀ-H)₆](PF₆)

(2g) and [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6). Single crystals

of 2a and 2f with reasonable size were obtained from tetrahy-

drofuran at ꢃ20 ^ꢂC. However, we could not collect diﬀraction

data with suﬃcient quality because of a loss of the solvent

molecules included in the crystal during data collection. After

several attempts to crystallize 2a, 2f, and 2g from solvents

with higher boiling points, we successfully obtained single

crystals of 2g suitable for an X-ray structural characterization

from the mixed solvent of dioxane and 1,2-dimethoxyethane

(1/1) at 5 ^ꢂC. Data collection and reﬁnement proceeded

without problems; the reﬁnement converged with residuals

of R₁¼ 0:045 (wR₂¼ 0:118). In a unit cell, two cations [{Ru-

(C₅Me₅)}₃(ꢀ-H)₆]^þ, two counter anions, PF₆^ꢃ, and six diox-

ane molecules are included. Hydrogen atoms bonded to the

ruthenium atoms were located by a diﬀerence Fourier synthe-

sis, and were reﬁned isotropically. The structure of the cationic

part of 2g is shown in Fig. 1. The counter anions and dioxane

molecules are disregarded for clarity. Selected intramolecular

Ru–Ru single bond. The EAN rule applied to 2g requires a for-

mal Ru–Ru bond order of 5/3. Shortening of the Ru–Ru bond

is most likely due to the two three-center, two-electron Ru–H–

Ru bonds.

The structure of the neutral cluster 6 was determined by X-

ray crystallography using a dark-brown prismatic single crystal

that was obtained from a 4:1 mixed solvent of tetrahydrofuran

and pentane at ꢃ20 ^ꢂC. Diﬀraction data were collected at ꢃ55

^ꢂC. The positions of the hydrogen atoms directly bonded to the

ruthenium atoms were located by a sequential diﬀerence Four-

ier synthesis, but were not reﬁned. Figure 2 displays the molec-

ular structure of 6, and Table 2 lists the selected bond distances

and angles. The structure of 6 is approximately in the C_3vsym-

metry, except for the array of C₅Me₅rings. Three doubly

bridging hydrogen atoms and two triply bridging ones bond

three ruthenium atoms, which occupy the vertices of a nearly

equilateral triangle. The average interatomic distance between

ꢀ

the ꢀ₃-H and ꢀ-H atoms in 6 is 2.07 A, which is larger than

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

73

ꢀ

the value (av. 1.74 A) for the two hydrogen atoms bridging the

same Ru–Ru bond in 2g. This fact shows that there is no bond-

ing interaction among the hydrido ligands, and that 6 is a typi-

cal classical hydrido complex. The average Ru–Ru distance of

Ru

H

Ru

rt

140 °C

H

7

H

6

H

6

ꢀ

2.7497(7) A is signiﬁcantly longer than that observed in the

cationic hexahydrido complex 2g, although both 6 and 2g

are 44-electron complexes. This is probably due to the diﬀer-

ence in the numbers of the bridging hydrides, two ꢀ₃-hydro-

gens and three ꢀ-hydrogens in 6 vs six ꢀ-hydrogens in 2g.

The ruthenium centers in 2g are, therefore, bound to each other

more tightly than those in 6. The average distance between Ru

d₆

Ru

D

Ru

d₆

ꢀ

and the centroid of the C₅Me₅ring is 1.809 A, and slightly

D

ꢀ

shorter than that observed in 2g (1.841 A). The diﬀerence in

7-d₉

the electron density at the metal centers between 6 and 2g is

most probably responsible for the diﬀerence in the Ru–cent-

roid distance. Lowering the electron density at the metal center

in cationic 2g reduces back-donation from the ruthenium to the

C₅Me₅group and, as a result, the Ru–centroid distance in 2g is

lengthened compared to that in the neutral complex 6.

Scheme 2.

served. Within the period of monitoring the reaction, the for-

mation of any product other than 6-d_n(n ¼ 1{5) was not ob-

served. Based on these observations, we can propose a couple

of plausible reaction paths involving an intermediary ꢁ²:ꢁ²:ꢁ²-

C₆D₆or ꢁ²-C₆D₆complex. Previously, we reported that the

pentahydride 6 reacted with 1,3- or 1,4-cyclohexadiene to form

a trinuclear ꢀ₃-ꢁ²:ꢁ²:ꢁ²-C₆H₆complex 7 via an allylic C–H

bond activation, followed by ꢅ-H elimination (Scheme 2).⁵²

The face-capping benzene complex 7 was alternative_ꢂly ob-

tained by heating the pentahydride 6 in benzene at 140 C for

17 h. The benzene ligand in 7 is tightly bound to the metal cen-

ters, and added benzene-d₆could not replace the coordinated

benzene-h₆to generate 7-d₆when isolated 7-h₆was heated

at 140 ^ꢂC in C₆D₆for 68 h. More noteworthy is that the hydro-

gen atoms in neither the C₆H₆ligand nor the hydrido ligands in

7 were exchanged with deuterium in C₆D₆under these condi-

tions. As mentioned above, the H/D exchange between 6 and

C₆D₆was realized under much milder conditions than those

which brought about the incorporation of a benzene molecule

in the Ru₃core of 6. These results altogether rule out the pos-

sibility that the H/D exchange between the hydrides in 6 and

C₆D₆proceeds via an intermediate ꢀ₃-ꢁ²:ꢁ²:ꢁ²-C₆D₆com-

plex 7. Another plausible path for the H/D exchange reaction

is shown in Scheme 3, which involves an ꢁ²-C₆D₆intermedi-

ate.^53–57

H/D Exchange Reaction of 6 with Benzene-d₆, Toluene-

d₈, o-Xylene-d₁₀, and D₂. The hydrido ligands of 6 undergo

an intermolecular exchange reaction with hydrogen atoms at-

tached to aromatic rings. A solution of pentahydride 6 in C₆D₆

ꢂ

(0.014 M) was heated in a sealed NMR sample tube at 80 C,

and the H/D exchange reaction was monitored by means of

¹H NMR spectroscopy. Upon heating for 5 h, ﬁve resonance

signals assignable to hydrides for the isotopomers 6-d₀, 6-d₁,

6-d₂, 6-d₃, and 6-d₄appeared at ꢂ ꢃ7:26, ꢃ7:85, ꢃ8:47,

ꢃ9:09, and ꢃ9:72, respectively. After an additional heating

for 17 h, 61% of the hydrido ligands were converted into the

deuteride, and the molar ratio of the isotopomers 6-d₀/6-d₁/

6-d₂/6-d₃/6-d₄/6-d₅became 2/9/23/31/21/14 (Fig. 3-B).

After 40 h, the ratio reached 0/3/12/28/32/25 at 73% conver-

sion (Fig. 3-C).

As the reaction proceeded, a progressive increase in the in-

tegral intensity of the ¹H signals in the aromatic region was ob-

The coordination of a C₆D₆molecule to the unsaturated

(44e) pentahydrido complex 6 yielded an intermediary ꢀ-ꢁ²-

C₆D₆complex I-1 (46e) and a subsequent oxidative addition

of the coordinated benzene-d₆to the adjacent ruthenium center

form a phenyl-d₅derivative I-2. Reductive coupling between

the phenyl-d₅group and the hydrido ligand, followed by the

liberation of C₆D₅H, would generate 6-d₁by way of a ꢀ-ꢁ²-

C₆D₅H complex I-3. Upon repetition of such an equilibrating

process, the hydrido ligands in 6 are exchanged for deuterium

to yield 6-d₆. As far as mononuclear late transition metals,

such as rhodium, iridium, and rhenium complexes, are con-

cerned, the intermediacy of the ꢁ²-arene complex in arene

C–H activation has been well established.³⁴If the H/D ex-

change reaction between 6 and aromatic compounds proceeds

via an ꢁ²-arene intermediate, toluene is expected to be an ap-

propriate probe compound. The H/D exchange at the para-

and meta-positions would precede that at the ortho-position,

because the steric repulsion between the C₅Me₅groups sur-

rounding the reaction site and the methyl group attached to

Fig. 3. ¹H NMR spectra of the hydride regions in the H/D

exchange reaction of 6 with C₆D₆at 80 C.

ꢂ

74

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

BCSJ AWARD ARTICLE

oxidative

addition

C₆D₆

d₆

d₅

reductive

elimination

- C₆D₆

H₅

H₅D

oxidative

addition

reductive

elimination

C₆D₅H

- C₆D₅H

d₅h₁

D₅

H₄D

= (C₅Me₅)Ru

Scheme 3.

ꢂ

Fig. 4. ¹H NMR spectra of the aromatic and the hydride regions in the H/D exchange reaction of 6 with toluene-d₈at 80 C.

1

the aromatic ring largely aﬀected the regiochemistry of the ꢁ²-

coordination of the toluene molecule. Thus, we examined the

H/D exchange reaction between 6 and toluene-d₈(Fig. 4). Af-

6.97 was monitored by H NMR spectroscopy (Fig. 5).

These two signals are attributable to the aromatic protons

contained in the deuterated xylene, and had equal intensity

in the initial stage. These two peaks at ꢂ 6.93 (2H) and 6.97

(2H) were unambiguously assigned to ring protons at C-4

and C-5, and those at C-3 and C-6, respectively, based on

the INADEQUATE-2D ¹³C and ¹H–¹³C HSC spectra. With

the elapse of time, the formation of the isotopomers 6-d_n

and an increase in the intensity of the two signals at ꢂ 6.93

and 6.97 were observed as a result of the H/D exchange reac-

tion. The methyl protons of o-xylene did not participate in the

H/D exchange reaction, just as the methyl protons of toluene

did not. The intensity of the signal at ꢂ 6.97 increased more

rapidly than that of the signal at ꢂ 6.93. The ratio of the ex-

change rate, k_C-4=k_C-3, derived from the increments of the sig-

nal intensity at ꢂ 6.97 and 6.93 was about 4.6/1. Based on

these observations, we propose that the H/D exchange reaction

between 6 and the arenes proceeds by way of an intermediacy

of the ꢁ²-arene cluster complex.

ꢂ

ter being heated at 80 C in toluene-d₈for 44.5 h, 59% of the

hydrides in 6 were exchanged with deuterium atoms, and

the isotopomer ratio 6-d₀/6-d₁/6-d₂/6-d₃/6-d₄/6-d₅reached

2/10/25/31/21/11.

Interestingly, only deuterium atoms attached to the aromatic

ring carbons of toluene participated in the exchange reaction,

and deuterium atoms at the benzylic carbon were not ex-

changed at all within the experimental errors. The bond-disso-

ciation energy of a benzylic C–H bond (ca. 88 kcal/mol) is

much smaller than that of an aromatic C–H bond (ca. 110

kcal/mol);³⁵the situation is, of course, almost the same for

the corresponding C–D bonds. This result, therefore, strongly

suggests that bond cleavage does not proceed via a homolytic

process, but via an intermediate arene complex, and that the

coordination of toluene via an aromatic C=C bond likely takes

place prior to C–D bond cleavage. In the H/D exchange reac-

tion in toluene-d₈, the exchange rate of the deuterium was on

the order of ortho- (ꢂ 7.10) < meta- (ꢂ 6.98) < para-position

(ꢂ 7.02) and the ratio k_o=k_m=k_pwas about 1/2/4. Such regio-

selectivity was also observed in the H/D exchange reaction be-

tween 6 and o-xylene-d₁₀. A solution of 6 in o-xylene-d₁₀was

The triruthenium pentahydride 6 also underwent H/D ex-

change with D₂, as did the cationic triruthenium hexahydride

2 and the dinuclear ruthenium tetrahydrido complex 1. Hy-

drides coordinated in 6 were completely exchanged for deuter-

ides to generate the corresponding pentadeuterido complex 6-

d₅when a stirred solution of 6 in tetrahydrofuran was exposed

ꢂ

heated at 80 C, and the strength of the signals at ꢂ 6.93 and

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

75

Fig. 5. ¹H NMR spectra of the aromatic and the hydride regions in the H/D exchange reaction of 6 with o-xylene-d₁₀.

to D₂at room temperature for 3 h (Eq. 9).

One is an associative path by way of a 46e species (B), and

the other is a dissociative one that involves a highly unsaturat-

ed 42e species (D). Regarding the mechanism of the site-ex-

change process of the hydrido ligands of 6, we assumed an in-

termediate species in which the triply bridging hydride has a

bonding interaction with the doubly bridging hydride. Such a

bonding interaction between the ꢀ₃-hydride and the ꢀ-hydride

forms an unsaturated species A. The coordination of a sub-

strate to the newly generated vacant coordination site leads

to a 46e species B, which undergoes the liberation of hydrogen

and subsequent bond activation of the substrate to yield C. On

the other hand, the liberation of hydrogen from A prior to co-

ordination of the substrate generates a highly unsaturated 42e

species D, which activates the substrate by way of a 44e spe-

cies E. The 42e species D is probably much more unstable than

the 46e species B because of its highly unsaturated character.

We, therefore, presumed that the activation of the substrate on

the reaction site of 6 was achieved by way of the associative

path from a stability standpoint.

Although we have, thus far, investigated the reaction chem-

istry of the trinuclear ruthenium pentahydrido complex 6

through the reaction with hydrocarbons, such as oleﬁns and di-

enes,^23,58we have not obtained conclusive evidence concern-

ing the associative pathway of the reaction. However, we for-

tunately obtained explicit evidence of an associative pathway

in the reaction of 6 with phosphines and phosphites.

The reaction of 6 with trimethylphosphine proceeded in tet-

rahydrofuran at room temperature to result in the formation of

a purple solid, which was unambiguously identiﬁed as a 1:1-

adduct [{Ru(C₅Me₅)}₃(ꢀ-H)₅(PMe₃)] (8a) based on NMR

spectroscopy as well as the analytical data (Eq. 10).

Ru

H

D

D₂

H

D

H

ð9Þ

Ru

THF

H

D

6

6-d₅

The coordination of C₆D₆or D₂to the vacant site, generated

as the result of a bonding interaction between the two hydrido

ligands, would lead to the oxidative addition of an aromatic C–

D or a D–D bond, respectively. While the D₂molecule is small

enough to enter freely into the reaction ﬁeld of 6, the molecu-

lar size of C₆D₆is much larger than that of D₂. We, therefore,

propose that the rate-determining step of the intermolecular H/

D exchange is most probably the step of capture of D₂or C₆D₆

in the triangular space surrounded by the three C₅Me₅groups.

Trirutheniumpentahydrido Complex 6 as a Precursor of

an Active Species for the ‘‘Trimetallic Activation’’. Reac-

tion of 6 with Small Molecules: A newly synthesized triru-

theniumpentahydrido complex 6 is an appropriate candidate

for the precursor of an active species for trimetallic activation.

Here, we use the term ‘‘trimetallic activation’’ to refer to one

of the activation modes of the substrate in which the synergy

of the three metal centers is achieved in the activation step.

It is one of the key features of 6 that no carbonyl ligand is in-

cluded in 6, irrespective of the bridging or terminal positions.

As mentioned previously, the electron density at the metal cen-

ters of 6 is, therefore, likely to be higher than that of the poly-

carbonyl metal clusters or polycarbonylhydrido metal clusters

due to coordination of the electron-releasing C₅Me₅groups.

Thus, complex 6 is expected to be much more active than

the metal clusters having carbonyl ligands with respect to both

coordinative and oxidative addition.

Ru

H

PR₃

Ru

H

PR₃

THF

H

Upon liberation of the hydrides, or making a bonding inter-

action between the two hydrido ligands to form an ꢁ²-H₂inter-

mediate, vacant coordination sites would be generated on the

Ru₃plane. Two possible modes of the interaction between

the Ru₃cluster and the substrate are illustrated in Scheme 4.

Ru

H

ð10Þ

H

6

8a (R = Me)

8b (R = OPh)

76

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

BCSJ AWARD ARTICLE

S

H H

S

- H₂

H

46e

44e

B

C

S

Associative

H H

H

44e

Dissociative

A

6

- H₂

= (η⁵-C₅Me₅)Ru

S

H

42e

44e

D

E

Scheme 4.

1

The H NMR spectrum measured at room temperature re-

vealed four resonance signals at ꢂ ꢃ11:82 (d, J ¼ 10:9 Hz),

1.06 (d, J ¼ 8:3 Hz), 1.87 (d, J ¼ 1:8 Hz), and 2.02 (s) ppm

in an intensity ratio of 5/9/15/30. These signals were assigned

to those of hydrides, coordinated trimethylphosphine, the

C₅Me₅group on the ruthenium atom bound to the phosphine

ligand, and two C₅Me₅groups on the rest of the ruthenium

centers, respectively. The ﬁve hydrides were observed to be

equivalent as a doublet peak with an intensity of 5H at room

temperature as the result of a rapid dynamic process, namely,

exchange among the ꢀ-coordination sites. The signal of the

hydrides broadened with a decrease of the temperature, and

split into three broad signals with an intensity ratio of 1/3/1

ꢂ

around ꢂ ꢃ3, ꢃ11, and ꢃ23:5, respectively, at ꢃ50 C. The

signal observed around ꢂ ꢃ11 with the intensity of 3H further

split into two signals as the temperature dropped. At ꢃ108 ^ꢂC,

the signals of the hydrides appeared at ꢂ ꢃ2:96 (s, 1H),

ꢃ10:90 (d, J_PH¼ 32:5 Hz, 2H), ꢃ11:06 (s, 1H), and ꢃ23:50

(s, 1H). The doublet signal at ꢂ ꢃ10:90 is assigned to the hy-

drides bound to the ruthenium atom, which has a direct bond-

ing interaction with the phosphine ligand. The coupling con-

stant of 32.5 Hz is close to the ²J_PHvalue observed in the mon-

onuclear hydridophosphineruthenium complex, Ru(C₅Me₅)-

(PMe₃)₂H (²J_PH¼ 38 Hz).⁵⁹The rest of the hydrido ligands

probably bridge the two ruthenium centers that have no bond-

ing interaction with the phosphine ligand. These results clearly

show that the reaction of 6 with trimethylphosphine proceeded

via an associative pathway to generate the 1:1-adduct 8a.

Complex 6 also reacted with triphenyl phosphite in a similar

manner to result in the formation of 8b. The ¹H NMR spectrum

of 8b showed a marked resemblance to that of 8a. The signals

of the C₅Me₅groups and the hydrido ligands appeared at ꢂ

2.06, 1.83 (d, J_PH¼ 2:3 Hz), and ꢃ11:13 (d, J_PH¼ 10:3 Hz),

Fig. 6. Thermal ellipsoid representation of [{Ru(C₅Me₅)}₃-

(ꢀ-H)₅{P(OPh)₃}] (8b) showing 30% probability ellip-

soids. CP is centroid of the C₅Me₅ring.

respectively, with an intensity ratio of 30/15/5. The resonance

signals for the coordinated triphenyl phosphite were observed

at ꢂ 6.84 (t, 3H), 7.08 (t, 6H), and 7.24 (d, 6H) in the ¹H NMR

spectrum and ꢂ 134.3 in the ³¹P NMR spectrum. These results

strongly indicate a structural analogy of 8b with 8a.

The structure of 8b was unequivocally established by an

X-ray diﬀraction study (Fig. 6). Some of the relevant bond

lengths and angles are listed in Table 3.

Figure 6 clearly demonstrates a structure that consists of a

triangular Ru₃core bearing ﬁve bridging hydrogen atoms

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

77

ꢀ

Table 3. Selected Bond Distances (A) and Angles (deg) in

[{Ru(C₅Me₅)}₃(ꢀ-H)₅{P(OPh)₃}] (8b)

in an autoclave. These results showed that 6 was stable enough

to undergo neither fragmentation of the Ru₃core nor the liber-

ation of hydrido ligands upon heating in solution.

Bond distances

We then examined the reaction of 6 with reactive small

molecules, such as dioxygen, CH₃I, and carbon monoxide.

The reaction of 6 with 1.1 equiv of O₂in a mixed solvent of

tetrahydrofuran and methanol (23/1) smoothly proceeded un-

Ru1–Ru2

Ru1–Ru3

Ru2–Ru3

Ru3–P1

Ru1–H1

Ru2–H1

Ru3–H1

Ru1–H2

Ru2–H2

Ru2–H4

Ru3–H4

2.6504(7)

3.0262(6)

2.9969(6)

2.1714(14)

1.80(5)

1.84(5)

2.21(5)

1.64(6)

1.74(6)

Ru1–CP1

Ru2–CP2

Ru3–CP3

P1–O1

P1–O2

P1–O3

1.822

1.809

1.893

1.636(4)

1.630(5)

1.570(5)

ꢂ

der a temperature range from ꢃ78 to ꢃ60 C to form an 88/

12 mixture of a trinuclear mono-ꢀ₃-oxo ruthenium complex,

[{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-O)] (9), and a di-ꢀ₃-oxo complex,

[{Ru(C₅Me₅)}₃(ꢀ-H)(ꢀ₃-_ꢂO)₂] (10), while the reaction at

higher temperature (ꢃ30 C) formed a 1/1 mixture of these

complexes. The addition of methanol to the reaction system

was essential for preparing ꢀ₃-oxo complexes 9 and 10 in high

yields. Otherwise, the reaction became non-selective and was

accompanied by the formation of an unidentiﬁed insoluble

black solid. As mentioned above, no reaction other than an ex-

change between the hydrido ligands of 6 and the protonic hy-

drogen of methanol took place when 6 was treated with meth-

anol. This evidently shows that the ꢀ₃-oxo ligands in 9 and 10

do not stem from methanol, but from dioxygen. Although we

failed to synthesize the mono-ꢀ₃-oxo complex 9 selectively

in the reaction of 6 with dioxygen, we successfully separated

analytically pure 9 as dark-green crystals in 39% yield from

88/12 mixtures of 9 and 10 by crystallization from a cooled

(ꢃ30 ^ꢂC) tetrahydrofuran solution. The mono-ꢀ₃-oxo com-

plex 9, as well as 6, is very sensitive to dioxygen. Monitoring

the reaction of the isolated mono-ꢀ₃-oxo complex 9 with an

equimolar amount of dioxygen by ¹H NMR spectroscopy con-

ﬁrmed the immediate and quantitative formation of the di-ꢀ₃-

oxo complex 10. Compound 10 was, of course, the exclusive

product in the reaction of 6 with a large excess of dioxygen.

The results of the reaction of 6 with dioxygen are summarized

by the following equation (Eq. 11).

Ru1–H3

Ru2–H3

Ru1–H5

Ru3–H5

1.80(6)

1.58(6)

1.89(6)

1.59(6)

1.76(6)

1.64(6)

Angles

Ru1–Ru2–Ru3

Ru3–Ru1–Ru2

P1–Ru3–Ru2

64.46(2)

63.33(2)

92.78(4)

Ru2–Ru3–Ru1

P1–Ru3–Ru1

52.21(1)

100.01(4)

and a triphenyl phosphite molecule at one of the three rutheni-

um atoms. The structure of 8b is approximately in the C_ssym-

metry with a mirror plane passing through Ru3 and the mid-

point of Ru1 and Ru2. Hydrogen atom H4 bridges Ru2 and

Ru3 and H5 bridges Ru1 and Ru3, respectively. Two hydrogen

atoms, H2 and H3, doubly bind two ruthenium atoms, Ru1 and

Ru2. Hydrogen atom H1 is considered to have bonding inter-

action between the three ruthenium atoms, although one of the

ꢀ

three Ru–H distances, Ru3–H1 (2.21(6) A), is somewhat long-

ꢀ

er than the others (av. 1.82 A). The locations of the metal-

bound hydrogen atoms are fully consistent with the H NMR

data of the structurally analogous phosphine complex 8a re-

corded at ꢃ108 C. A Ru3–P1 distance of 2.1714(14) A is

in the range of those for the terminally coordinated phosphine

and phosphite complexes.

1

ꢂ

ꢀ

Ru

H

Thus, we proved that the reaction of 6 with tertiary phos-

phine or phosphite proceeded by way of an associative path-

way. The associative pathway is also supported by a theoreti-

cal study. Recently, Morokuma proposed an associative path-

way in the reaction of 6 with cyclopentadiene⁶⁰based on an ab

initio MO calculation.⁶¹

H

O₂

O

H

O

H

+

Ru

THF

H

6

9

10

ð11Þ

An ethyltetramethylcyclopentadienyl analogue of the di-

ꢀ₃-oxo complex, [{Ru(C₅Me₄Et)}₃(ꢀ-H)(ꢀ₃-O)₂] (10⁰), was

similarly obtained in quantitative yield upon the treatment of

6⁰with 2 equiv of dioxygen in tetrahydrofuran. In the reaction

of 6 with molecular oxygen, a single oxygen atom stemmed

from dioxygen was incorporated into the complex as a ꢀ₃-

oxo ligand, and the second oxygen atom was reduced by the

hydrido ligands to form water. The formation of water in the

reaction could not be conﬁrmed when the reaction of 6 with

2 moles of dioxygen in tetrahydrofuran-d₈was monitored by

means of ¹H NMR spectroscopy. However, the IR spectra

and the analytical data of the freshly obtained di-ꢀ₃-oxo com-

plexes 10 and 10⁰showed the inclusion of a water molecule in

them as the solvent of crystallization. As mentioned below, a

single-crystal X-ray diﬀraction study of 10⁰clearly showed

the presence of a water molecule held by hydrogen bonds.

The ¹H NMR spectrum of 9 revealed two singlet signals at-

tributable to the C₅Me₅ligand and the hydrides at ꢂ 1.80 and

In the case of the Ru₃framework of 6 being stable enough

to be retained during the reaction, the multiple vacant sites

generated as a result of H₂liberation would cooperate in the

activation of a substrate.

We conﬁrmed the thermal stability of 6, as described below.

The Ru₃framework never broke up into a mononuclear or di-

nuclear species at any signiﬁcant rate in reﬂuxing tetrahydro-

furan, even in the absence of substrates, such as alkenes and

alkynes. Furthermore, the reaction of 6 with linear alkanes,

ꢂ

such as hexane and octane, proceeded at 170 C with a reten-

tion of the Ru₃framework to yield a trinuclear closo-ruthena-

cyclopentadiene.⁶²

The reactivity of 6 toward alkane strongly depended on the

bulkiness of the alkane, and bulky 1,3,5-trimethylcyclohexane

put up strong resistance to the C–H bond activation. As a re-

sult, compound 6 was quantitatively recovered even when it

ꢂ

was heated at 180 C for 450 h in 1,3,5-trimethylcyclohexane

78

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

BCSJ AWARD ARTICLE

ꢃ11:51, respectively, in an intensity ratio of 45H/3H. Thus,

the three C₅Me₅groups in 9 were observed to be equivalent

in both the ¹H and ¹³C NMR spectra, whereas there were

two inequivalent resonances for the C₅Me₅ligands in 10,

which appeared at ꢂ 1.78 (15H) and 1.76 (30H). This implies

that compound 9 has a symmetry element that relates the three

ruthenium atoms; most likely, this is a three-fold rotation axis

that passes through the triply-bridging oxygen atom, and is

perpendicular to the plane of the three ruthenium atoms. In

contrast, a symmetry element included in the di-ꢀ₃-oxo com-

plex 10 is a mirror plane that bisects one of the Ru–Ru bonds

and passes through the two bridging oxygen atoms. The mo-

lecular structures of 9, 10, and 10⁰were conﬁrmed by means

of single-crystal X-ray diﬀraction studies.

ꢂ

Crystals of 9 were obtained from cooled (ꢃ30 C) tetrahy-

Fig. 7. Thermal ellipsoid representation of [{Ru(C₅Me₅)}₃-

(ꢀ-H)₃(ꢀ₃-O)] (9) showing 30% probability ellipsoids.

CP is centroid of the C₅Me₅ring.

drofuran. However, we encountered a bothersome problem;

namely, compounds 9 and 6 crystallized as mixed crystals

when ꢀ₃-oxo complex 9 contained a small amount of unreact-

ed 6 as an impurity. Because single crystals of 9 and 6 both

ꢁ

conform to the triclinic space group P1, and their cell param-

eters are very similar to each other, they often crystallizes as

mixed crystals. The purity of the crystals used for the X-ray

structure determination should, therefore, be checked by

1

means of H NMR spectroscopy prior to a diﬀraction study.

After repeated eﬀorts, we were fortunately able to obtain sin-

gle crystals of 9. Data collection and reﬁnement proceeded

without any problem, and the reﬁnement converged with a re-

sidual of R₁¼ 0:0315 (wR₂¼ 0:0702).

Reasonably-sized single crystals of di-ꢀ₃-_ꢂoxo complex 10

were obtained from tetrahydrofuran at ꢃ30 C. In this case,

the data collection, itself, proceeded without any problem,

but reﬁnement was unsuccessful because of the presence of

disordered water molecules that were bound to ꢀ₃-oxo ligands

with hydrogen bonding. Several attempts to reﬁne the structure

were in vain, but single crystals of 10 free of water molecules

were fortunately recovered from the reaction mixture of 10

with ethylene in tetrahydrofuran. The crystals were suitable

for X-ray structural determination, and data collection and re-

ﬁnement went well. In all cases, hydrogen atoms bonded to the

ruthenium atoms were not located in the diﬀerential Fourier

maps.

Fig. 8. Thermal ellipsoid representation of [{Ru(C₅Me₅)}₃-

(ꢀ-H)(ꢀ₃-O)₂] (10) showing 30% probability ellipsoids.

CP is centroid of the C₅Me₅ring.

The structures of 9, 10, and 10⁰are displayed in Figs. 7–9.

Tables 4–6 list some of the relevant bond distances and angles.

The triply bridging oxygen atom in the mono-ꢀ₃-oxo com-

plex 9 is disordered between the two positions, O1 and O2

(55.0/45.0). In the case of the di-ꢀ₃-oxo complex 10⁰, water

molecules are included in the unit cell, and the oxygen atom

is disordered between O3 and O4. The occupancies of O3

and O4 are both 0.5, and they are bound to O1 and O2, respec-

tively, by hydrogen bonding. The O1–O3 and O2–O4 dis-

ꢀ

tances of 2.822(6) and 2.738(6) A, respectively, are compara-

ble to the reported value for the hydrogen bonded O–H O

system.⁶³

ꢁꢁꢁ

The average Ru–Ru distances in the di-ꢀ₃-oxo complexes

0

ꢀ

10 (2.7135(10) A) and 10 (2.7164(4) A) are slightly longer

Fig. 9. Thermal ellipsoid representation of [{Ru-

(C₅Me₄Et)}₃(ꢀ-H)(ꢀ₃-O)₂] (10⁰) showing 30% probabil-

ity ellipsoids. Oxygen atom of hydrogen-bonded water is

disordered between two positions, O3 and O4 (50/50).

CP is centroid of the C₅Me₅ring.

than that observed in the mono-ꢀ₃-oxo complex

(2.7028(8) A). This can possibly be explained according to

the EAN rule applied to 10 (48e), 10⁰(48e), and 9 (46e).

9

ꢀ

The average Ru–O distances of 1.958(6) and 1.953(7) A for

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

79

aration method of the ꢀ₃-oxo complexes 9, 10, and 10⁰is es-

sentially similar to that for the polycarbonyl clusters, but is the

ﬁrst example of the oxygenation method applied to polyhydri-

dometal clusters.

The reaction of 6 with iodomethane in tetrahydrofuran

smoothly proceeded at room temperature to result in the gen-

eration of a novel trinuclear tetrahydrido complex, [{Ru-

(C₅Me₅)}₃(ꢀ₃-H)(ꢀ-H)₃(ꢀ₃-I)] (11), in which an iodido

ligand triply bridged ruthenium atoms (Eq. 12).

ꢀ

Table 4. Selected Bond Distances (A) and Angles (deg) in

[{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-O)] (9)

Bond distances

Ru1–Ru2

Ru1–Ru3

Ru2–Ru3

Ru1–O1

Ru2–O1

Ru3–O1

2.6989(9)

2.7005(6)

2.7089(8)

1.963(6)

1.961(6)

1.964(6)

Ru1–CP1

Ru2–CP2

Ru3–CP3

Ru1–O1A

Ru2–O1A

Ru3–O1A

1.827

1.826

1.833

1.949(7)

1.945(8)

1.950(7)

Angles

Ru

Ru1–Ru2–Ru3

Ru2–Ru3–Ru1

Ru3–Ru1–Ru2

Ru1–O1A–Ru2

Ru2–O1A–Ru3

Ru3–O1A–Ru1

59.92(2)

59.86(2)

60.23(2)

87.8(3)

88.1(3)

86.9(2)

Ru1–O1–Ru2

Ru2–O1–Ru3

Ru3–O1–Ru1

86.9(2)

87.3(2)

86.9(2)

H

MeI

I

H

ð12Þ

Ru

THF

H

6

11

Compound 11 was unambiguously identiﬁed on the basis of

¹H and ¹³C NMR spectral data as well as elemental analysis.

Resonance signals for the C₅Me₅groups and the hydrido li-

gands appeared as singlet peaks at ꢂ 2.12 (45H) and ꢃ2:25

(4H), respectively, in the ¹H NMR spectrum recorded at 25

^ꢂC. Resonance for the three C₅Me₅groups was observed to

be equivalent as a_ꢂsharp singlet peak in the temperature range

from 25 to ꢃ110 C, while the singlet signal for the hydride

ligands was signiﬁcantly broadened at low temperature. This

result implies that 11 has a threefold rotation axis in the mole-

cule. The signal for the three C₅Me₅groups are, therefore, ob-

served to be equivalent irrespective of the ﬂuxional behavior

of the hydrides. Temperature-dependent resonance signals

for the four hydrido ligands are most likely to be due to a rapid

exchange of the hydrides among the coordination sites, that is,

among the ꢀ₃-site and the three ꢀ₂-sites. The rapid site-ex-

change of the hydrido ligands was conﬁrmed by ¹H NMR

spectroscopy. With a decrease of temperature, the line width

at the half-height of the signal for the hydrido ligands signiﬁ-

cantly broadened. The line-widths were 9.0, 36.1, 180.1,

and 802.8 Hz at 27, ꢃ20, ꢃ55, and ꢃ80 ^ꢂC, respectively.

The signal for hydrides was ﬂattened near ꢃ110 ^ꢂC. We could

not, however, observe decoalescence of the hydrido resonance

into two signals with an integral intensity ratio of 3/1.

When the reaction of 6 with 0.9 equiv of iodomethane in tet-

rahydrofuran-d₈was monitored by means of ¹H NMR spec-

troscopy, no reaction intermediates were detected, and a sharp

singlet signal of the formed methane was observed at ꢂ 0.19

together with the signals of 11. The formation of methane

seems most likely to be due to the oxidative addition of iodo-

methane to one of the ruthenium centers and subsequent reduc-

tive coupling between the methyl ligand and the hydride. The

reductive elimination of methane from the triruthenium cluster

must be an energetically favored process compared to that of

dihydrogen or hydrogen iodide, as established in the chemistry

of the mononuclear complex.

ꢀ

Table 5. Selected Bond Distances (A) and Angles (deg) in

[{Ru(C₅Me₅)}₃(ꢀ-H)(ꢀ₃-O)₂] (10)

Bond distances

Ru1–Ru1⁰

Ru1–O1

O1–O2

2.7135(10)

2.013(6)

2.499

Ru1–CP1

Ru1–O2

1.828

1.995(5)

Angles

Ru1⁰–Ru1–Ru1⁰⁰

O1–Ru1–O2

60.0

77.2(3)

Ru1–O1–Ru1⁰

Ru1–O2–Ru1⁰⁰

84.8(3)

85.7(3)

ꢀ

Table 6. Selected Bond Distances (A) and Angles (deg) in

[{Ru(C₅Me₄Et)}₃(ꢀ-H)(ꢀ₃-O)₂] (10⁰)

Bond distances

Ru1–Ru2

Ru1–Ru3

Ru2–Ru3

Ru1–O1

Ru2–O1

Ru3–O1

O1–O2

2.7211(4)

2.7169(4)

2.7103(4)

2.0203(17)

2.0158(17)

2.0210(17)

2.543

Ru1–CP1

Ru2–CP2

Ru3–CP3

Ru1–O2

Ru2–O2

Ru3–O2

O1–O3

1.813

1.812

1.811

2.0172(16)

2.0156(16)

2.0239(16)

2.84

O2–O4

2.75

Angles

Ru1–Ru2–Ru3

Ru2–Ru3–Ru1

Ru3–Ru1–Ru2

Ru1–O2–Ru2

Ru2–O2–Ru3

Ru3–O2–Ru1

60.029(11)

60.183(11)

59.788(11)

84.87(6)

84.28(6)

84.49(6)

Ru1–O1–Ru2

Ru2–O1–Ru3

Ru3–O1–Ru1

84.78(6)

84.35(6)

84.49(6)

Ru–O1 and Ru–O1A in 9, respectively, are signiﬁcantly short-

er than those observed in the di-ꢀ₃-oxo complexes 10

In order to determine the structure of 11, an X-ray crystallo-

graphic study was carried out on a black-brown, needle-like

crystal obtained by cooling a tetrahydrofuran–pentane solution

of 11 to ꢃ20 ^ꢂC. The structure of 11 is shown as ORTEP

diagrams in Fig. 10. Selected bond distances and bond angles

are given in Table 7.

0

ꢀ

(2.004(6) A) and 10 (2.019(2) A). These values, however, still

lie within the range of those for the metal–oxygen bond in ꢀ₃-

oxo and di-ꢀ₃-oxo complexes of transition metals.^64–69

There are several precedents of trinuclear mono- and di-ꢀ₃-

oxo complexes of late transition metals derived from a poly-

carbonyl cluster by a reaction with dioxygen.^66,69–71The prep-

Notable features of the structure include a pseudo tetrahe-

80

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

BCSJ AWARD ARTICLE

as elemental analysis. The IR spectrum of 12 reveals strong ab-

sorption attributable to the stretching vibrations of the bridging

carbonyl ligands at 1782, 1735, and 1615 cm^ꢃ1. The absorp-

tion at 1615 cm^ꢃ1is due to the triply bridging carbonyl group.

This value is comparable to the reported values of 1689, 1627,

and 1626 cm^ꢃ1for [Ru₃(ꢀ₃-O)(ꢀ₃-CO)(CO)₅(ꢀ-dppm)₂],⁶⁶

[Ru₃(ꢀ₃-O)(ꢀ₃-CO)(CO)₃(ꢀ-dppm)₃],⁶⁹and [Ru₃(ꢀ₃-O)-

(ꢀ₃-CO)(CO)₃(ꢀ-dppm)]Cl,⁶⁹respectively. In the ﬁeld-de-

sorption mass spectrum of 12, the intensities of the obtained

isotopic peaks for C₃₄H₄₅O₄Ru₃were in agreement with the

calculated values within the experimental error. The proposed

molecular structure was conﬁrmed by means of X-ray crystal-

lography. The shapes of the black-green single crystals of 12,

obtained from tetrahydrofuran at ꢃ34 ^ꢂC, were thin plates, and

were not suitable for a diﬀraction study. Although the full-

matrix least-squares reﬁnement did not suﬃciently converge

(R₁¼ 0:148, wR₂¼ 0:386), a preliminary result clearly eluci-

dated the atom connectivity of the structure.

Fig. 10. Thermal ellipsoid representation of [{Ru-

(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)(ꢀ₃-I)] (11) showing 30% proba-

bility ellipsoids. CP is centroid of the C₅Me₅ring.

Thus, the three ruthenium atoms in 6 appeared to be tightly

bound to each other so as not to split up into three mononu-

clear species in the reaction with reactive small molecules,

such as O₂, CH₃I, and CO.

ꢀ

Table 7. Selected Bond Distances (A) and Angles (deg) in

[{Ru(C₅Me₅)}₃(ꢀ₃-H)(ꢀ-H)₃(ꢀ₃-I)] (11)

Bond distances

In conclusion, we have synthesized a cationic triruthenium

hexahydrido complex, [{Ru(C₅Me₅)}₃(ꢀ-H)₆](Y) (2: Y ¼

BF₄, CF₃SO₃, 1/2(SO₄), C₆H₅CO₂, CH₃CO₂), by the treat-

ment of diruthenium tetrahydride [{Ru(C₅Me₅)}₂(ꢀ-H)₄] (1)

with the corresponding acid. The protonation of 1 and a subse-

quent liberation of dihydrogen from the resulting dicationic

hexahydrido species led to the intermediary dicationic tetrahy-

drido complex, [{Ru(C₅Me₅)}₂(ꢀ-H)₄]^2þ(Y)₂(B), which is in

equilibrium with a mononuclear monocationic dihydrido com-

plex, [Ru(C₅Me₅)(ꢀ-H)₂]^þ(Y) (C), in diethyl ether, and react-

ed instantly with 1 to yield the trinuclear complex 2. This

mechanism involving the highly unsaturated cationic species

C was proved by the reaction of 1 with acid HY in the pres-

ence of benzene or acetonitrile, which yielded the correspond-

ing mononuclear cationic complex, [Ru(C₅Me₅)(C₆H₆)](Y) or

[Ru(C₅Me₅)(CH₃CN)₃](Y), respectively.

The treatment of the trinuclear cationic complex 2 with a

base, such as CH₃ONa, generated the triruthenium pentahydri-

do complex 6, which has two ꢀ₃-H and three ꢀ-H ligands.

The hydrides in 6 rapidly exchange the coordination sites with

each other in solution, even at low temperature, probably by

way of an intermediary ꢁ²-H₂species generated as the result

of a bonding interaction between the two hydrido ligands.

The tendency to generate the intermediary ꢁ²-H₂species is

probably responsible for the high reactivity of 6.

Pentahydrido complex 6 readily undergoes an intermolecu-

lar H/D exchange reaction with D₂or C₆D₆. As the result of a

detailed analysis using ¹H NMR spectroscopy, it was proposed

that the H/D exchange reaction with deuterated arenes, such as

C₆D₆, C₆D₅CD₃, and C₆D₄(CD₃)₂, is likely to proceed via an

ꢁ²-arene intermediate.

Pentahydrido complex 6 is a coordinatively unsaturated 44e

complex and, which is a suitable and useful precursor of the

active species for trimetallic activation. To elucidate the initial

step of the reaction of 6, the reaction with trimethylphosphine

was examined. The exclusive formation of [{Ru(C₅Me₅)}₃(ꢀ-

H)₅(PMe₃)] (8a) in a reaction performed at room temperature

Ru1–Ru2

Ru1–Ru3

Ru2–Ru3

Ru1–I1

Ru2–I1

Ru3–I1

2.8665(12)

2.8642(12)

2.8651(13)

2.8500(13)

2.8461(13)

2.8486(14)

Ru1–CP1

Ru2–CP2

Ru3–CP3

Ru–ꢀ-H

Ru–ꢀ₃-H

1.803

1.802

1.789

1.70 (av)

2.08 (av)

Angles

Ru1–Ru2–Ru3

Ru2–Ru3–Ru1

Ru3–Ru1–Ru2

59.96(3)

60.04(2)

59.99(3)

Ru1–I1–Ru2

Ru2–I1–Ru3

Ru3–I1–Ru1

60.43(3)

60.41(3)

60.35(3)

dral Ru₃I core with average Ru–Ru and Ru–I distances of

ꢀ

2.8653(12) and 2.8482(13) A, respectively. Complex 11 is a

48e cluster, as are complexes 10 and 10⁰, and the EAN rule ap-

plied to these complexes requires a metal–metal single bond

between the ruthenium atoms. Thus, the interatomic distances

between the ruthenium atoms in 11 correspond to a Ru–Ru sin-

gle bond. The signiﬁcantly short Ru–Ru bond lengths in 10

and 10⁰, compared to those in 11, are probably due to the pres-

ence of two triply bridging 4e ligands, ꢀ₃-oxo groups, which

tightly bind the ruthenium atoms.

Pentahydrido complex 6 smoothly reacted with atmospheric

pressure of carbon monoxide to generate a paramagnetic tetra-

carbonyl complex, [{Ru(C₅Me₅)}₃(ꢀ-CO)₃(ꢀ₃-CO)] (12),

with a retention of the trinuclear metallic framework (Eq. 13).

Ru

OC

Ru

CO

Ru

H

CO

H

CO

ð13Þ

Ru

THF

C

O

H

6

12

Although we could not obtain a well-resolved ¹H NMR

spectrum of 12, because of its paramagnetic character, it was

unambiguously assigned on the basis of IR, FD-MS, as well

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

81

X-ray Data Collection and Reduction. X-ray-quality crystals

of 2g, 6, 9, and 10 were obtained as described in the text, and that

of 10⁰was obtained as described below. They were mounted on

glass ﬁbers. X-ray diﬀraction measurements were performed on

a Rigaku AFC-5R four-circle diﬀractometer with graphite-mono-

showed that the reaction proceeded via an associative path.

The above-mentioned H/D exchange reaction between C₆D₆

and the hydrides in 6 is also associative, and presumably be-

gins with a coordinative addition of C₆D₆to one of the ruthe-

nium centers in an ꢁ²-fashion.

Thus, the generation of a vacant site at the ruthenium center

and an intramolecular site-exchange of the hydride are closely

correlated by the formation of the intermediate ꢁ²-H₂species.

The activation parameters for the site-exchange of the hydrido

ligands is, therefore, one of the measures of the reactivity in

the intermolecular process of the polyhydrido cluster. Actual-

ly, we have recently proposed a tentative correlation between

the activation parameters for the site-exchange of the hydrido

ligands and that for the H/D exchange between the hydrides

and D₂in the heterobimetallic complex, [Re(C₅Me₅)(H)₂(ꢀ-

H)₃Ru(C₅Me₅)], and [Re{CH₃C(CH₂PPh₂)₃}(H)(ꢀ-H)₃Ru-

(C₅Me₅)].⁷²

The thermal stability of the Ru₃framework of 6 was also

demonstrated by heating 6 in 1,3,5-trimethylcyclohexane at

180 ^ꢂC. Complex 6 underwent neither fragmentation of the

Ru₃core nor liberation of the hydrido ligand. The trinuclear

framework of 6 is suﬃciently stable to react with reactive

small molecules, such as dioxygen, iodomethane, and carbon

monoxide with retention of the framework of the cluster.

Thus, complex 6 has electron-donating C₅Me₅groups, and

the multiple hydrido ligands suitable for generating the vacant

coordination site and its Ru₃framework is extremely stable in

solution, even at high temperature. Complex 6 is, therefore, ca-

pable of activating a substrate in a concerted manner, such that

the three metal centers in 6 are allotted respective roles as the

coordination site and the activation site in the substrate-activa-

tion step.

ꢀ

chromated Mo Kꢆ (ꢇ ¼ 0:71069 A) radiation. Intensity data were

collected using a !=2ꢈ scan technique; 3 standard reﬂections were

recorded every 150 reﬂections. The data were processed using the

TEXSAN crystal solution Package operating on an IRIS Indigo

computer.⁷³Neutral atom scattering factors were obtained from

the standard sources.⁷⁴In the reduction of the data, Lorentz/polar-

ization corrections and empirical absorption corrections based on

azimuthal scans were applied to the data for each structure.

Structure Solution and Reﬁnement for 2g, 6, 9, 10, and 10⁰.

The positions of the Ru atoms were determined using direct meth-

ods employing the PATTY direct-method routines.⁷⁵In each case,

the non-hydrogen atoms were located from successive diﬀerence

Fourier map calculations using the DIRDIF-92 programs,⁷⁵and

were reﬁned anisotropically by using full-matrix least-squares

techniques on F². In the case of 2g, 6, and 10⁰, the positions of

hydrogen atoms bonded to the Ru atoms were located by sequen-

tial diﬀerence Fourier synthesis, and were reﬁned isotropically.

The crystal data and the results of the analyses are listed in

Table 8.

Reaction of 1 with Tetraﬂuoroboric Acid. To a 50-mL

Schlenk tube charged with [{Ru(C₅Me₅)}₂(ꢀ-H)₄] (196 mg,

0.412 mmol) and tetrahydrofuran (10 mL), HBF₄Et₂O (85 wt

ꢁ

% in Et₂O) (84 mL, 0.49 mmol) was added at ꢃ78 ^ꢂC. The mixture

was gradually warmed to room temperature and stirred for 1 h.

The red suspension turned brown-purple with evolution of dihy-

drogen as the mixture was warmed. After removal of the solvent

under reduced pressure, the brown-purple residual solid was puri-

ﬁed by column chromatography on alumina with tetrahydrofuran

to give [{Ru(C₅Me₅)}₃(ꢀ-H)₆](BF₄) (2a) (207 mg, 94%) as a

brown-purple crystalline solid. Although this material appeared

Experimental

1

to be pure by H NMR, several attempts to obtain satisfactory el-

1

General. All of the manipulations were carried out under an

argon atmosphere using standard Schlenk techniques. Toluene,

toluene-d₈, benzene, benzene-d₆, tetrahydrofuran, tetrahydrofu-

ran-d₈, 1,2-dimethoxyethane, diethyl ether, and pentane were

dried over Na/benzophenone. Chloroform-d and dichlorometh-

emental analyses were unsuccessful. 2a: H NMR (270 MHz, 23

^ꢂC, CDCl₃) ꢂ ꢃ11:23 (s, 6H, Ru–H–Ru), 1.99 (s, 45H, Cp–Me).

ꢂ

¹³C{¹H} NMR (67.5 MHz, 23 C, CDCl₃) ꢂ 12.1 (Cp–Me), 97.7

(Cp-ring). T₁(500 MHz, CD₂Cl₂): 849 ms (at 20 ^ꢂC), 513 ms

(at ꢃ20 ^ꢂC), 274 ms (at ꢃ50 ^ꢂC), and 387 ms (at ꢃ80 ^ꢂC). IR

(KBr, cm^ꢃ1): 2983, 2901, 1489, 1459, 1424, 1378, 1091, 1046,

1018, 630.

ꢀ

ane-d₂were dried over calcium hydride and stored over 4 A mo-

lecular sieves. [{Ru(C₅Me₅)}₂(ꢀ-H)₄] was prepared according to

previously published procedures.²³All other chemicals were of re-

agent grade, and were puriﬁed according to standard procedures,

as necessary. The alumina (Merck Art. 1097) used for column

chromatography was deoxygenated under a vacuum prior to use.

Proton and ¹³C NMR spectra were recorded on JEOL GX-500,

JEOL EX-270, Varian Gemini 300, and Varian Unity Inova 400

Fourier-transform spectrometers with tetramethylsilane as an in-

ternal standard. Variable-temperature ¹H NMR spectra were re-

corded on JEOL GX-500 and Varian Unity Inova 400 spectrome-

ters. The chemical shifts are reported in parts per million (ꢂ), the

coupling constants are reported in hertz (Hz) and the integrations

are reported in according to the number of protons. The IR spectra

were recorded on a JASCO FT/IR-5000 spectrophotometer, and

the frequencies were reported in cm^ꢃ1. Field-desorption mass

spectra were recorded on a Hitachi GC-MS M80 high-resolution

mass spectrometer. Mass-spectral results were reported in m=z us-

ing the most intense peak in each envelope. Elemental analyses

were performed by the Analytical Facility at the Research Labo-

ratory of Resources Utilization, Tokyo Institute of Technology.

Reaction of 1 with Triﬂuoromethanesulfonic Acid. To a

suspension of [{Ru(C₅Me₅)₂(ꢀ-H)₄}] (339 mg, 0.712 mmol) in

tethahydrofuran (10 mL) was added CF₃SO₃H (1.0 M in Et₂O)

ꢂ

(1.03 mL, 1.03 mmol) at 25 C with vigorous stirring. The color

of the solution changed from red to deep purple along with the

evolution of hydrogen. After being stirred for 2 h at room temper-

ature, the solvent was removed under reduced pressure from the

reaction mixture. Washing the brown-purple residual solid with

diethyl ether (20 mL ꢆ 5) yielded [{Ru(C₅Me₅)(ꢀ-H)₂}₃]-

(CF₃SO₃) (2b) (375 mg, 92%) as a brown-purple crystalline solid.

1

ꢂ

2b: H NMR (270 MHz, 23 C, CDCl₃) ꢂ ꢃ11:24 (s, 6H, Ru–H–

Ru), 1.99 (s, 45H, Cp–Me). ¹³C{¹H} NMR (67.5 MHz, 23 ^ꢂC,

CDCl₃) ꢂ 12.1 (Cp–Me), 97.7 (Cp-ring), 120.5 (q, J_CF¼ 320:9

Hz, CF₃SO₃). IR (KBr, cm^ꢃ1): 2951, 2894, 1457, 1424, 1373,

1254, 1217, 1137, 1085, 1021, 795, 748, 632, 570, 518. Anal.

Calcd for C₃₁H₅₁O₃F₃SRu₃: C, 43.07; H, 5.95%. Found: C,

42.86; H, 6.14%.

Reaction of 1 with Sulfuric Acid.

To a suspension of

[{Ru(C₅Me₅)}₂(ꢀ-H)₄] (223 mg, 0.467 mmol) in tetrahydrofuran

Table 8. Crystallographic Data for 2g, 6, 8b, 9, 10, 10⁰, and 11

2g

6

8b

9

10

10⁰

11

Formula

C

₄₂H₇₅F₆O₆PRu₃

1124.20

C₃₀H₅₀Ru₃

713.91

C₅₅H₇₃O₃PRu₃

1116.35

0:30 ꢆ 0:20 ꢆ 0:20

monoclinic

P2₁=c

C₃₀H₄₈ORu₃

727.89

C₃₀H₄₆O₂Ru₃

740.87

C₃₃H₅₄O₃Ru₃

802.00

C₃₀H₄₉IRu₃

839.80

Formula weight

Cryst size/mm

Crystal system

Space group

0:30 ꢆ 0:25 ꢆ 0:20

0:40 ꢆ 0:40 ꢆ 0:20

0:40 ꢆ 0:20 ꢆ 0:20

0:30 ꢆ 0:20 ꢆ 0:20

0:30 ꢆ 0:30 ꢆ 0:20

0:20 ꢆ 0:15 ꢆ 0:10

triclinic

trigonal

triclinic

ꢁ

P1

ꢁ

P1

ꢁ

P1

ꢁ

R3

ꢁ

P1

ꢁ

P1

ꢀ

a/A

15.622(3)

16.274(4)

10.997(3)

109.492(18)

109.872(18)

91.162(19)

2450.1(10)

2

1.524

297

10.09

Rigaku AFC-5R

11.076(3)

15.342(3)

11.070(3)

108.551(19)

119.506(15)

73.258(18)

1532.3(6)

2

1.547

293

14.74

Rigaku AFC-5R

17.973(3)

16.0387(15)

18.761(4)

90

119.580(13)

90

5162.0(15)

4

1.436

293

9.38

11.055(3)

15.463(4)

11.042(2)

108.756(18)

119.580(13)

72.92(2)

1533.4(6)

2

1.576

292

14.77

Rigaku AFC-5R

18.555(3)

15.756(7)

90

120

4698(2)

6

1.571

297

1.451

Rigaku AFC-5R

12.6920(17)

13.5363(17)

11.3209(17)

108.927(11)

104.644(11)

98.585(11)

1722.3(4)

2

1.547

293

13.27

Rigaku AFC-5R

11.0807(17)

18.150(4)

8.4955(19)

95.324(18)

108.763(14)

89.969(16)

1610.0(5)

2

1.732

297

23.63

Rigaku AFC-5R

ꢀ

b/A

ꢀ

c/A

ꢆ/deg

ꢅ/deg

ꢇ/deg

ꢀ ³

V/A

Z

D_calcd/g cm^ꢃ1

Temp/K

ꢀ(Mo Kꢆ)/cm^ꢃ1

Diﬀractometer

Radiation

Monochromator

2ꢈ_max/deg

No. of rﬂns collected

Rigaku AFC-5R

Mo Kꢆ (ꢉ ¼ 0:71069) Mo Kꢆ (ꢉ ¼ 0:71069) Mo Kꢆ (ꢉ ¼ 0:71069) Mo Kꢆ (ꢉ ¼ 0:71069) Mo Kꢆ (ꢉ ¼ 0:71069) Mo Kꢆ (ꢉ ¼ 0:71069) Mo Kꢆ (ꢉ ¼ 0:71069)

graphite

50

8923

8582

graphite

55

9143

8717

6558

graphite

50

9365

9049

6911

graphite

55

5675

5371

4386

graphite

55

2913

1818

1282

graphite

55

6348

6054

5448

graphite

55

7564

7333

3130

No. of indep rﬂns

No. of rﬂns obsd (>2ꢃ) 6635

R1 (I > 2ꢃðIÞ)

wR2 (I > 2ꢃðIÞ)

R1 (all data)

wR2 (all data)

No. of params

GOF

0.0453

0.0356

0.0707

0.0635

0.0773

331

1.024

0.799; ꢃ0:689

0.0370

0.0867

0.0637

0.1025

595

1.051

1.124; ꢃ1:002

0.0315

0.0702

0.0476

0.0753

318

1.035

0.405; ꢃ0:466

0.0502

0.1190

0.0874

0.1332

107

0.979

1.382; ꢃ0:537

0.0206

0.0508

0.0259

0.0530

387

1.038

0.504; ꢃ0:437

0.0509

0.1151

0.2002

0.1541

276

0.933

1.076; ꢃ0:777

0.1176

0.0700

0.1296

498

1.021

Max and min. peaks in 1.150; ꢃ0:519

ﬁnal diﬀ map/e^ꢃ

A )

ꢀ ^ꢃ3

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

83

1

ꢂ

(10 mL) was added H₂SO₄(0.20 M in Et₂O) (2.5 mL, 0.50 mmol)

at 25 ^ꢂC with vigorous stirring. After being stirred for 3 h at room

temperature, the solvent was removed under reduced pressure

from the reaction mixture. Washing the brown-purple residual sol-

id with diethyl ether (20 mL ꢆ 5) aﬀorded [{Ru(C₅Me₅)}₃(ꢀ-

H)₆](1/2SO₄+ H₂SO₄) (2c) (261 mg, 97%) as a brown-purple

crystalline solid. The incorporation of sulfuric acid in 2c was con-

ﬁrmed by ¹H NMR. 2c: ¹H NMR (270 MHz, 23 ^ꢂC, CDCl₃) ꢂ

ꢃ11:23 (s, 6H, Ru–H–Ru), 2.00 (s, 45H, Cp–Me), 5.9 (br, 2H,

solid state. 3e: H NMR (270 MHz, 23 C, THF-d₈) ꢂ ꢃ4:41 (s,

2H, Ru–H–Ru), 1.69 (s, 30H, Cp–Me), 1.81 (s, 6H, CH₃CO₂).

¹³C{¹H} NMR (67.5 MHz, 23 ^ꢂC, THF-d₈) ꢂ 10.5 (Cp–Me),

22.0 (CH₃CO₂), 87.9 (Cp-ring), 178.5 (CH₃–CO–). IR (KBr,

cm^ꢃ1): 2970, 2893, 1571, 1412, 1372, 1340, 1257, 1154, 1072,

1043, 990, 843, 710, 673, 618, 583, 431, 392. Anal. Calcd for

C₂₄H₃₈O₄Ru₂: C, 48.64; H, 6.46%. Found: C, 48.70; H, 6.58%.

Reaction of 1 with Triﬂuoroacetic Acid. To a suspension of

[{Ru(C₅Me₅)}₂(ꢀ-H)₄] (344 mg, 0.72 mmol) in tetrahydrofuran

(10 mL) was added CF₃CO₂H (2.16 M in THF) (0.67 mL, 1.45

mmol) at room temperature with vigorous stirring. After being

stirred for 4 h at room temperature, the solvent was removed under

reduced pressure from the reaction mixture. Extraction from the

residual brown solid with diethyl ether (20 mL ꢆ 5) aﬀorded

[{Ru(C₅Me₅)}₂(ꢀ-H)₂(ꢀ-OCOCF₃)₂] (3a) (495 mg, 98%) as a

red-orange crystalline solid. 3a: ¹H NMR (270 MHz, 23 ^ꢂC,

CDCl₃) ꢂ ꢃ4:50 (s, 2H, Ru–H–Ru), 1.72 (s, 30H, Cp–Me).

ꢂ

H₂SO₄). ¹³C{¹H} NMR (67.5 MHz, 23 C, CDCl₃) ꢂ 12.1 (Cp–

Me), 97.6 (Cp-ring). IR (KBr, cm^ꢃ1): 3500–3000 (ꢊ_{OH O}),

ꢁꢁꢁ

2974, 2955, 2898, 1463, 1373, 1357, 1253 (ꢊ_SO), 1228 (ꢊ_SO),

1102, 1062, 1013, 850, 759, 718, 636, 607, 576, 425. The amount

of sulfuric acid incorporated in 2c was changed according to the

amount used in the reaction. Therefore, several attempts to obtain

satisfactory elemental analyses were unsuccessful.

Reaction of 1 with Benzoic Acid. To a 50-mL Schlenk tube

charged with [{Ru(C₅Me₅)}₂(ꢀ-H)₄] (159 mg, 0.333 mmol) and

C₆H₅COOH (82 mg, 0.67 mmol) was added tethahydrofuran (10

mL). After stirring for 3 days at room temperature, the mixture be-

came red-orange and homogeneous. The solvent was removed in

vacuo. The resulting brown-red residue was extracted with diethyl

ether (10 mL ꢆ 5) to leave [{Ru(C₅Me₅)}₃(ꢀ-H)₆](C₆H₅CO₂)

(2d) (48 mg, 26%) as a brown-purple solid. The resulting solution

was ﬁltered through a Celite pad, and dried in vacuo to aﬀord

[{Ru(C₅Me₅)}₂(ꢀ-H)₂(ꢀ-OCOC₆H₅)₂] (3d) (170 mg, 71%) as a

red-orange solid. 2d: ¹H NMR (270 MHz, 23 ^ꢂC, CD₂Cl₂) ꢂ

ꢃ11:26 (s, 6H, Ru–H–Ru), 1.98 (s, 45H, Cp–Me), 7.30–7.54

(m, 3H, Ph), 8.11 (d, J ¼ 7:1 Hz, 2H, Ph). ¹³C{¹H} NMR (67.5

MHz, 23 ^ꢂC, CD₂Cl₂) ꢂ 12.2 (Cp–Me), 97.7 (Cp-ring), 120.3

(Ph), 128.3 (Ph), 130.1 (Ph), 133.0 (Ph), 169.9 (Ph–CO–). IR

(KBr, cm^ꢃ1): 2996, 2948, 2891, 2856, 1699 (ꢊ_CO), 1592, 1444,

1371, 1359, 1307, 1256, 1167, 1091, 1063, 1015, 854, 805, 781,

712, 649, 641. Although this material appeared to be pure by

¹H NMR, several attempts to obtain satisfactory elemental analy-

ses were unsuccessful because of its instability, even in the solid

state. 3d: ¹H NMR (270 MHz, 23 ^ꢂC, THF-d₈) ꢂ ꢃ4:10 (s, 2H,

Ru–H–Ru), 1.84 (s, 30H, Cp–Me), 7.1–8.0 (m, 10H, Ph).

¹³C{¹H} NMR (67.5 MHz, 23 ^ꢂC, THF-d₈) ꢂ 10.7 (Cp–Me),

88.5 (Cp-ring), 128.0 (Ph), 130.2 (Ph), 130.6 (Ph), 133.6 (Ph),

174.5 (Ph–CO–). IR (KBr, cm^ꢃ1): 2950, 2899, 1596, 1447,

1402, 1390, 1171, 1067, 1025, 995, 866, 852, 714, 706, 686,

476. Anal. Calcd for C₃₄H₄₂O₄Ru₂: C, 56.96; H, 5.91%. Found:

C, 56.82; H, 5.90%.

¹³C{¹H} NMR (67.5 MHz, 23 C, CDCl₃) ꢂ 10.4 (Cp–Me), 89.2

ꢂ

2

(Cp-ring), 113.7 (q, J_CF¼ 286:8 Hz, CF₃CO₂), 165.8 (q, J_CF

¼

37:5 Hz, CF₃CO₂). T₁(270 MHz, ꢃ50 ^ꢂC, CD₂Cl₂): 251 ms.

IR (KBr, cm^ꢃ1): 2910, 1656 (ꢊ_CO), 1446, 1382, 1199, 1146,

1079, 1050, 993, 868, 858, 784, 732. Anal. Calcd for

C₂₄H₃₂F₆O₄Ru₂: C, 41.14; H, 4.60%. Found: C, 41.34; H, 4.42%.

Reaction of 1 with Formic Acid.

To a suspension of

[{Ru(C₅Me₅)}₂(ꢀ-H)₄] (180 mg, 0.38 mmol) in tetrahydrofuran

(10 mL) was added formic acid (0.15 mL, 4.0 mmol) at room tem-

perature with vigorous stirring. After being stirred for 1 h at room

temperature, the solvent was removed under reduced pressure

from the reaction mixture. Extraction from the residual brown sol-

id with diethyl ether (10 mL ꢆ 3) aﬀorded [{Ru(C₅Me₅)}₂(ꢀ-

H)₂(ꢀ-OCOH)₂] (3b) (173 mg, 81%) as a red-orange crystalline

solid. 3b: H NMR (270 MHz, 23 C, THF-d₈) ꢂ ꢃ4:40 (s, 2H,

Ru–H–Ru), 1.72 (s, 30H, Cp–Me), 6.75 (s, 1H, –OCOH).

¹³C{¹H} NMR (67.5 MHz, 23 ^ꢂC, THF-d₈) ꢂ 10.6 (Cp–Me),

88.5 (Cp-ring), 170.4 (–OCOH). IR (KBr, cm^ꢃ1): 2952, 2905,

2835, 1593 (ꢊ_CO), 1463, 1374, 1360, 1344, 1264, 1078, 1026,

987, 846, 790, 587. Anal. Calcd for C₂₂H₃₄O₄Ru₂: C, 46.80; H,

6.07%. Found: C, 46.67; H, 6.13%.

Reaction of 1 with 1-Adamantanecarboxylic Acid. To a 50-

mL Schlenk tube charged with [{Ru(C₅Me₅)}₂(ꢀ-H)₄] (71 mg,

0.149 mmol) and 1-adamantanecarboxylic acid (54 mg, 0.300

mmol) was added tetrahydrofuran (10 mL). After stirring for 40

h at room temperature, the mixture became red-orange and homo-

geneous. The solvent and excess acid were removed in vacuo. Ex-

traction from the resulting brown-red residue with diethyl ether

(10 mL ꢆ 5) followed by removal of solvent under reduced pres-

sure gave [{Ru(C₅Me₅)}₂(ꢀ-H)₂(ꢀ-OCOC₁₀H₁₅)₂] (3c) (110 mg,

89%) as a red-orange crystalline solid. 3c: ¹H NMR (270 MHz, 23

^ꢂC, THF-d₈) ꢂ ꢃ4:49 (s, 2H, Ru–H–Ru), 1.72 (s, 30H, Cp–Me),

1.6–1.7 (m, adamantyl), 1.8–1.9 (m, adamantyl). ¹³C{¹H} NMR

(67.5 MHz, 23 ^ꢂC, THF-d₈) ꢂ 10.6 (Cp–Me), 29.8 (adamantyl),

37.9 (adamantyl), 41.7 (adamantyl), 87.9 (Cp-ring), 184.9

(C₁₀H₁₅–COO). The resonance for the quarternary carbon of ada-

mantyl group could not be observed. IR (KBr, cm^ꢃ1): 2893, 2843,

1562 (ꢊ_CO), 1451, 1398, 1371, 1344, 1311, 1089, 993, 840, 764,

674. Anal. Calcd for C₄₂H₆₂O₄Ru₂: C, 60.55; H, 7.50%. Found:

C, 60.26; H, 7.43%.

1

ꢂ

Reaction of 1 with Acetic Acid. To a 50-mL Schlenk tube

charged with [{Ru(C₅Me₅)}₂(ꢀ-H)₄] (251 mg, 0.527 mmol) and

tethahydrofuran (10 mL) was added CH₃CO₂H (5.0 mL, 87

mmol). After stirring for 36 h at room temperature, the mixture

became red-orange and homogeneous. The solvent and excess

acid were removed in vacuo. The resulting brown-red residue

was extracted with diethyl ether (10 mL ꢆ 5) to leave

[{Ru(C₅Me₅)}₃(ꢀ-H)₆](CH₃CO₂) (2e) (71 mg, 26%) as

a

brown-purple solid. The resulting solution was ﬁltered through a

Celite pad, and dried in vacuo to aﬀord [{Ru(C₅Me₅)}₂(ꢀ-

H)₂(ꢀ-OCOCH₃)₂] (3e) (227 mg, 73%) as a red-orange solid.

2e: ¹H NMR (270 MHz, 23 ^ꢂC, CD₂Cl₂) ꢂ ꢃ11:21 (s, 6H, Ru–

H–Ru), 1.73 (s, 3H, CH₃CO₂), 1.98 (s, 45H, Cp–Me).

¹³C{¹H} NMR (67.5 MHz, 23 ^ꢂC, CD₂Cl₂) ꢂ 12.3 (Cp–Me),

22.3 (CH₃CO₂), 98.1 (Cp-ring), 175.3 (CH₃CO₂). IR (KBr,

cm^ꢃ1): 1563 (ꢊ_CO). Although this material appeared to be pure

by ¹H NMR, several attempts to obtain satisfactory elemental

analyses were unsuccessful because of its instability, even in the

Reaction of 2c with NaBPh₄. To a stirred suspension of

[{Ru(C₅Me₅)}₃(ꢀ-H)₆] (1/2SO₄+ H₂SO₄) (100 mg, 0.116

mmol) in tetrahydrofuran (10 mL) was added NaBPh₄(198 mg,

0.580 mmol) at 25 ^ꢂC. After being stirred for 6 h at room temper-

ature, the mixture was passed through a column packed with neu-

84

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

BCSJ AWARD ARTICLE

¹³C NMR (75 MHz, 23 ^ꢂC, CDCl₃) ꢂ 11.9 (q, J_CH¼ 128:4 Hz,

tral alumina. Removal of the solvent from the eluent under re-

duced pressure aﬀorded [{Ru(C₅Me₅)}₃(ꢀ-H)₆](BPh₄) (2f) (118

mg, 99%) as a red-purple crystalli_ꢂne solid. Slow cooling of tetra-

hydrofuran solution of 2f to ꢃ20 C gave single crystals that in-

cluded tetrahydrofuran as a solvent of crystallization. 2f: ¹H NMR

Cp–Me), 12.1 (q, J_CH¼ 128:5 Hz, Cp–Me), 14.7 (qt, J_CH

¼

127:8, 4.3 Hz, Cp–CH₂–Me), 20.2 (tq, J_CH¼ 129:1, 4.1 Hz,

Cp–CH₂–Me), 97.1 (s, Cp-ring), 98.4 (s, Cp-ring), 102.6 (s, Cp-

ring), 121.7 (dt, J_CH¼ 156:7, 7.8 Hz, Ph-para), 125.6 (d, J_CH

¼

ꢂ

(500 MHz, 24 C, THF-d₈) ꢂ ꢃ11:28 (s, 6H, Ru–H–Ru), 1.93 (s,

153:0 Hz, Ph-meta), 136.6 (dt, J_CH¼ 154:1, 7.3 Hz, Ph-

ortho), 163.6 (s, Ph-ipso), 164.3 (s, Ph-ipso), 165.0 (s, Ph-ipso),

165.6 (s, Ph-ipso). IR (KBr, cm^ꢃ1): 3056, 3036, 2927, 2912,

2874, 1580, 1456, 1428, 1378, 1085, 1052, 1028, 844, 734, 702,

634, 613. Anal. Calcd for C₅₇H₇₇BRu₃: C, 63.61; H, 7.21%.

Found: C, 63.87; H, 7.51%.

45H, Cp–Me), 6.87 (t, J ¼ 7:5 Hz, 4H, Ph), 7.03 (t, J ¼ 7:5 Hz,

8H, Ph), 7.42 (m, 8H, Ph). ¹³C NMR (125 MHz, 24 ^ꢂC, THF-

d₈) ꢂ 12.1 (Cp–Me), 97.6 (Cp-ring), 121.5 (Ph), 125.4 (Ph),

136.4 (Ph), 164.4 (q, J_BC¼ 49 Hz, Ph). IR (KBr, cm^ꢃ1): 2988,

2910, 2866, 1580, 1460, 1428, 1378, 1068, 1024, 845, 734, 704,

640, 612, 534, 516, 482, 470, 429, 410. Anal. Calcd for

C₅₄H₇₁BRu₃+ C₄H₈O: C, 62.97; H, 7.20%. Found: C, 62.99;

H, 7.19%.

Reaction of 1 with Triﬂuoromethanesulfonic Acid in the

Presence of Benzene. To a suspension of [{Ru(C₅Me₅)}₂(ꢀ-

H)₄] (70 mg, 0.15 mmol) in a mixed solvent of tetrahydrofuran

(5 mL) and benzene (2 mL) was added CF₃SO₃H (1.0 M in Et₂O)

(0.5 mL, 0.5 mmol) at 25 ^ꢂC with vigorous stirring. Immediately,

the color of the solution changed from red to green-brown with the

evolution of hydrogen. After being stirred for 1 h, the solution was

allowed to stand. The colorless solid of [Ru(C₅Me₅)(C₆H₆)]-

(CF₃SO₃) (4) (133 mg, 97%) precipitated from the solution was

collected on a glass frit. 4: ¹H NMR (270 MHz, 23 ^ꢂC,

CD₃COCD₃) ꢂ 2.04 (s, 15H, Cp–Me), 5.97 (s, 6H, C₆H₆).

¹³C{¹H} NMR (67.5 MHz, 23 ^ꢂC, CD₃COCD₃) ꢂ 9.74 (Cp–

Me), 87.0 (C₆H₆), 96.5 (Cp-ring). IR (KBr, cm^ꢃ1): 3072, 2981,

2909, 1258, 1170, 1041, 645, 581, 528. FD-MS: m=z 316. Anal.

Calcd for C₁₇H₂₁F₃O₃SRu: C, 44.07; H, 4.57%. Found: C,

43.96; H, 4.65%.

Reaction of 2c with NH₄PF₆. To a stirred suspension of

[{Ru(C₅Me₅)}₃(ꢀ-H)₆](1/2SO₄+ H₂SO₄) (175 mg, 0.203

mmol) in tetrahydrofuran (15 mL) was added NH₄PF₆(342 mg,

2.10 mmol) at 25 ^ꢂC. After being stirred for 40 min at room tem-

perature, the mixture was passed through a column packed with

alumina. Removal of the solvent from the eluent under reduced

pressure aﬀorded [{Ru(C₅Me₅)}₃(ꢀ-H)₆](PF₆) (2g) (123 mg,

70%) as an analytically pure red-purple crystalline solid. Slow

cooling of a saturated dioxane/dimethoxyethane (1/1) solution

of 2g to 5 ^ꢂC gave single crystals that included three dioxane

molecules in a unit cell as a solvent of crystallization. 2g:

¹H NMR (500 MHz, 24 ^ꢂC, CDCl₃) ꢂ ꢃ11:24 (s, 6H, Ru–H–

Ru), 1.99 (s, 45H, Cp–Me). ¹³C NMR (125 MHz, 24 ^ꢂC, THF-

d₈) ꢂ 12.1 (Cp–Me), 97.8 (Cp-ring). IR (KBr, cm^ꢃ1): 2990,

2916, 1466, 1381, 1079, 1024, 875, 843, 798, 642, 613, 559,

536, 513, 467, 435, 406. Anal. Calcd for C₃₀H₅₁F₆PRu₃: C,

41.90; H, 5.98%. Found: C, 41.70; H, 6.49%.

Reaction of [{Ru(C₅Me₄Et)}₂(ꢀ-H)₄] with Sulfuric Acid.

To a suspension of [{Ru(C₅Me₄Et)}₂(ꢀ-H)₄] (421 mg, 0.833

mmol) in tetrahydrofuran (20 mL) was added H₂SO₄(0.20 M in

Et₂O) (4.2 mL, 0.84 mmol) with vigorous stirring at room temper-

ature. After being stirred for 12 h at room temperature, the solvent

was removed under reduced pressure from the reaction mixture.

Washing the residual solid with diethyl ether (20 mL ꢆ 5) aﬀord-

ed [{Ru(C₅Me₄Et)}₃(ꢀ-H)₆](1/2SO₄+ H₂SO₄) (2c⁰) (467 mg,

93%) as a red-purple crystalline solid. Several attempts to obtain

satisfactory elemental analyses were unsuccessful. 2c⁰: ¹H NMR

Reaction of 1 with Tetraﬂuoroboric Acid in the Presence of

Acetonitrile. To a 50-mL Schlenk tube charged with acetonitrile

(1 mL), diethyl ether (5 mL), methanol (5 mL), and HBF₄(85 wt

% in Et₂O) (75 mL), a solution of [{Ru(C₅Me₅)}₂(ꢀ-H)₄] (126

mg, 0.27 mmol) in diethyl ether (10 mL) was slowly added

ꢂ

through a dropping funnel at 25 C with vigorous stirring. After

being stirred for 1 h at ambient temperature, the solvent was re-

moved under reduced pressure from the reaction mixture; washing

the residual yellow crystalline solid with diethyl ether (5 mL ꢆ 5)

yielded [Ru(C₅Me₅)(CH₃CN)₃](BF₄) (5) (410 mg, 98%). 5:

¹H NMR (300 MHz, 23 ^ꢂC, CD₃COCD₃) ꢂ 1.61 (s, 15H, Cp–

Me), 2.37 (s, 9H, CH₃CN). Anal. Calcd for C₁₆H₂₄N₃BF₄Ru: C,

43.06; H, 5.42; N, 9.42%. Found: C, 43.23; H, 5.33; N, 9.56%.

Determination of Equilibrium Constant among 2d, 6, and

ꢂ

(300 MHz, 23 C, CDCl₃) ꢂ ꢃ11:25 (s, 6H, Ru–H–Ru), 1.12 (t,

Benzoic Acid.

Benzoic acid (112 mg, 0.92 mmol) and

J ¼ 7:6 Hz, 9H, Cp–CH₂–Me), 1.99 (s, 36H, Cp–Me), 2.32 (q,

J ¼ 7:6 Hz, 6H, Cp–CH₂–Me). ¹³C NMR (75 MHz, 23 ^ꢂC,

[{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6) (30 mg, 0.042 mmol) were

dissolved in tetrahydrofuran-d₈(0.60 mL). After the solution

was charged in an NMR tube, the tube was sealed. The equilibri-

um constant K, where K ¼ ½2dꢄ=ð½6ꢄ½PhCO₂Hꢄ, was determined

based on ¹H NMR integration of the hydride signals for 2d and

6. The constant K was 0.76, 0.47, 0.36, 0.31, 0.27, 0.21, 0.19,

and 0.14 [M^ꢃ1] at 293, 303, 308, 313, 318, 323, 328, and 333

K, respectively.

CDCl₃) ꢂ 11.9 (q, J_CH¼ 128:5 Hz, Cp–Me), 12.1 (q, J_CH

¼

128:5 Hz, Cp–Me), 14.8 (qt, J_CH¼ 127:7, 4.7 Hz, Cp–CH₂–

Me), 20.2 (tq, J_CH¼ 129:1, 3.9 Hz, Cp–CH₂–Me), 97.1 (s, Cp-

ring), 98.5 (s, Cp-ring), 102.6 (s, Cp-ring).

Reaction of 2c⁰with NaBPh₄.

To a stirred solution of

[{Ru(C₅Me₄Et)}₃(ꢀ-H)₆](1/2SO₄+ H₂SO₄) (320 mg, 0.355

mmol) in tetrahydrofuran (20 mL) was added NaBPh₄(610 mg,

1.78 mmol) at room temperature. After stirring for 30 min at room

temperature, the mixture was passed through a column packed

with alumina. Removal of solvent from the eluent under reduced

pressure gave [{Ru(C₅Me₄Et)}₃(ꢀ-H)₆](BPh₄) (2f⁰) (345 mg,

90%) as a red-purple crystalline solid. Slow cooling of a saturated

Preparation of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6). To a stir-

red solution of [{Ru(C₅Me₅)}₃(ꢀ-H)₆] 1/2(SO₄+ H₂SO₄)

ꢁ

(1.211 g, 1.405 mmol) in methanol (30 mL) was added CH₃ONa

(302 mg, 5.60 mmol) at 25 ^ꢂC. After being stirred for 10 min, the

solvent was removed under reduced pressure. A black-brown res-

idue was extracted with tetrahydrofuran (30 mL ꢆ 3), ﬁltered

through a column packed with alumina and the volatile material

was removed to give 6 (964 mg, 96%) as a black-brown solid.

6: ¹H NMR (500 MHz, 24 ^ꢂC, C₆D₆) ꢂ ꢃ7:22 (s, 5H, Ru–H–

Ru), 2.04 (s, 45H, Cp–Me). ¹³C NMR (125 MHz, 24 ^ꢂC, C₆D₆)

ꢂ 13.1 (q, J_CH¼ 126:2 Hz, Cp–Me), 85.7 (s, Cp-ring). T₁(500

tetrahydrofuran/diethyl ether solution of 2f⁰to ꢃ30 C aﬀorded

ꢂ

red-purple single crystals. 2f⁰: ¹H NMR (300 MHz, 23 ^ꢂC, CDCl₃)

ꢂ ꢃ11:30 (s, 6H, Ru–H–Ru), 1.09 (t, J ¼ 7:6 Hz, 9H, Cp–CH₂–

Me), 1.90 (s, 18H, Cp–Me), 1.92 (s, 18H, Cp–Me), 2.27 (q, J ¼

7:6 Hz, 6H, Cp–CH₂–Me), 6.86 (t, J ¼ 7:1 Hz, 4H, Ph-para),

7.03 (t, J ¼ 7:3 Hz, 8H, Ph-meta), 7.42 (br, 8H, Ph-ortho).

ꢂ

MHz, THF-d₈): 8.63 s (at 20 ^ꢂC), 4.33 s (at ꢃ50 C), and 2.55

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

85

ꢂ

s (at ꢃ80 C). Anal. Calcd for C₃₀H₅₀Ru₃: C, 50.47; H, 7.06%.

126 Hz, P–Me), 82.9 (s, Cp-ring), 90.0 (d, J_PC¼ 2:6 Hz, Cp-ring).

³¹P{¹H} NMR (109.4 MHz, 24 ^ꢂC, C₆D₆–H₃PO₄) ꢂ 12.0. Anal.

Calcd for C₃₃H₅₉PRu₃: C, 50.17; H, 7.53%. Found: C, 49.79; H,

7.52%.

Found: C, 50.27; H, 7.02%.

Preparation of [{Ru(C₅Me₄Et)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6⁰). To a

stirred solution of [{Ru(C₅Me₄Et)}₃(ꢀ-H)₆](BPh₄) (248 mg,

0.231 mmol) in methanol (10 mL) was added CH₃ONa (46 mg,

0.852 mmol) dissolved in 5 mL of methanol at room temperature.

After stirring for 15 min at room temperature, the solvent was re-

moved in vacuo. A black-brown residue was extracted with tetra-

hydrofuran (10 mL ꢆ 3) and ﬁltered through a column packed

with alumina. Removal of the solvent from the eluent to dryness

gave 6⁰(162 mg, 93%) as a red-brown crystalline solid. 6⁰:

¹H NMR (500 MHz, 24 ^ꢂC, C₆D₆) ꢂ ꢃ7:13 (s, 5H, Ru–H–Ru),

1.07 (t, J ¼ 7:5 Hz, 9H, Cp–CH₂–Me), 2.03 (s, 18H, Cp–Me),

2.10 (s, 18H, Cp–Me), 2.53 (q, J ¼ 7:5 Hz, 6H, Cp–CH₂–Me).

¹³C NMR (125 MHz, 24 ^ꢂC, C₆D₆) ꢂ 13.1 (q, J_CH¼ 126:4 Hz,

Reaction of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6) with Tri-

phenyl Phosphite.

Preparation of [{Ru(C₅Me₅)}₃(ꢀ-H)₅-

{P(OC₆H₅)₃}] (8b): To a 50-mL Schlenk tube charged with

133 mg (0.186 mmol) of 6 and tetrahydrofuran (2.8 mL) was add-

ed triphenyl phosphite (98.0 mL, 0.370 mmol) at room tempera-

ture. The color of the reaction mixture immediately turned from

dark-red to red-purple. Removal of the solvent under reduced

pressure, followed by crystallization from tetrahydrofuran at

ꢃ30 ^ꢂC, aﬀorded 8b (171 mg, 82%). The slow cooling of saturat-

ꢂ

ed tetrahydrofuran solution of 8b to ꢃ30 C aﬀorded red-purple

1

prisms suitable for an X-ray diﬀraction study. 8b: H NMR (300

Cp–Me), 13.2 (q, J_CH¼ 126:4 Hz, Cp–Me), 15.9 (qt, J_CH

¼

MHz, 24 ^ꢂC, C₆D₆) ꢂ ꢃ11:13 (d, J_PH¼ 10:3 Hz, 5H, Ru–H–

Ru), 1.83 (d, J_PH¼ 1:8 Hz, 15H, Cp–Me), 2.06 (s, 30H, Cp–

Me), 6.84 (t, J ¼ 7 Hz, 3H, para), 7.08 (t, J ¼ 7 Hz, 6H, meta),

7.24 (t, J ¼ 7 Hz, 6H, ortho). ¹³C NMR (75.5 MHz, 24 ^ꢂC, C₆D₆)

ꢂ 12.4 (q, J_CH¼ 127 Hz, Cp–Me), 13.0 (q, J_CH¼ 127 Hz, Cp–

Me), 85.7 (s, Cp-ring), 92.6 (d, J_PC¼ 4:1 Hz, Cp-ring), 121.6

(dddd, J ¼ 162:8, 7.3, 4.0, 4.0 Hz, ortho), 122.6 (dt, J ¼ 162:2,

7.1 Hz, para), 128.8 (dd, J ¼ 161:6, 8.3 Hz, meta), xxx.x (dt,

J ¼ 13:0, 9 Hz, ipso). ³¹P{¹H} NMR (109.4 MHz, 24 ^ꢂC,

C₆D₆–H₃PO₄) ꢂ 134. Anal. Calcd for C₅₅H₇₃O₃PRu₃: C, 58.80;

H, 6.70%. Found: C, 59.17; H, 6.59%.

126:2, 4.3 Hz, Cp–CH₂–Me), 21.5 (tq, J_CH¼ 127:5, 4.3 Hz,

Cp–CH₂–Me), 85.5 (s, Cp-ring), 86.4 (s, Cp-ring), 91.9 (s, Cp-

ring). Anal. Calcd for C₃₃H₅₆Ru₃: C, 52.43; H, 7.47%. Found:

C, 52.07; H, 7.15%.

H/D Exchange between [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6)

and Benzene-d₆, Toluene-d₈, or o-Xylene-d₁₀. An NMR sample

tube (5 ꢋ), in which the atmosphere was replaced with argon, was

charged with 6.0 mg (0.0084 mmol) of 6 and 0.62 mL of benzene-

d₆. The tube was sealed after the solution was frozen by using a

dry ice–methanol bath. The solution was then heated at 80 ^ꢂC,

1

and the H NMR spectra were recorded at appropriate time inter-

vals. In a similar manner, NMR samples were prepared using tol-

Reaction of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6) with Di-

oxygen.

Preparation of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-O)] (9):

uene-d₈and o-xylene-d₁₀as solvents; the spectra were recorded at

ꢂ

A 50-mL Schlenk tube was charged with 93 mg (0.13 mmol) of

6, 7 mL of tetrahydrofuran, and 0.3 mL of methanol. The Schlenk

tube was evacuated with cooling at ꢃ78 ^ꢂC. Then, 1.1 equiv (0.14

mmol) of dioxygen dried over P₂O₅was admitted to the reactor by

use of a vacuum line. The mixture was stirred for 1.5 h under the

80 C.

Reaction of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6) with Ben-

zene. Preparation of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-C₆H₆)] (7):

A 50-mL glass autoclave, in which the atmosphere was replaced

with argon, was charged with 36.8 mg (0.052 mmol) of 6 and 5

mL of benzene. The mixture was heated at 140 ^ꢂC for 17 h.

The color of the solution changed from brown to red-purple. After

removal of the solvent under reduced pressure, the residual dark-

purple microcrystalline solid was washed three times with metha-

nol. Drying in vacuo gave 7 as a dark-purple crystalline solid

(39.6 mg, 97%). The product was identiﬁed by a comparison of

ꢂ

temperature range from ꢃ78 to ꢃ60 C. The color changed from

dark-brown to black-green. Immediately after removal of the cool-

ing bath, the solvent was removed in vacuo to aﬀord an 88/12

mixture of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-O)] (9) and [{Ru-

(C₅Me₅)}₃(ꢀ-H)(ꢀ₃-O)₂] (10) as a black-green solid (94 mg) in

a quantitative yield. Slow cooling of a tetrahydrofuran solution

of the resulting mixture to ꢃ30 ^ꢂC gave analytically pure 9 as

black-green crystals. 9: ¹H NMR (300 MHz, 23 ^ꢂC, C₆D₆) ꢂ

ꢃ11:51 (s, 3H, Ru–H–Ru), 1.80 (s, 45H, Cp–Me). ¹³C NMR (75

MHz, 23 ^ꢂC, C₆D₆) ꢂ 12.5 (q, J_CH¼ 126:6 Hz, Cp–Me), 92.0

(s, Cp-ring). Anal. Calcd for C₃₀H₄₈ORu₃: C, 49.50; H, 6.65%.

Found: C, 49.12; H, 6.72%.

1

its H NMR spectrum with that of authentic sample.

Reaction of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6) with Tri-

methylphosphine.

Preparation of [{Ru(C₅Me₅)}₃(ꢀ-H)₅-

{P(CH₃)₃}] (8a): To a 50-mL Schlenk tube charged with 285

mg (0.400 mmol) of 6 and tetrahydrofuran (10 mL) was added

0.51 mL (0.800 mmol) of a solution of trimethylphosphine in tol-

uene (1.57 M) at room temperature. The color of the reaction mix-

ture immediately turned from dark-red to purple. After being stir-

red for 10 min at ambient temperature, the solvent was removed

under reduced pressure from the reaction mixture to yield 8a

(300 mg, 95%). The slow cooling of a saturated tetrahydrofuran

Reaction of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6) with Di-

oxygen. Preparation of [{Ru(C₅Me₅)}₃(ꢀ-H)(ꢀ₃-O)₂] (10):

A 50-mL Schlenk tube was charged with 100 mg (0.140 mmol)

of 6, 10 mL of tetrahydrofuran, and 2 mL of methanol. The solu-

ꢂ

tion was cooled to ꢃ78 C and the Schlenk tube was evacuated.

Then, 2 equiv (0.29 mmol) of dioxygen dried over P₂O₅was ad-

mitted to the reactor by using a vacuum line. The_ꢂmixture was stir-

ꢂ

solution of 8a to ꢃ30 C aﬀorded purple prisms suitable for an

1

ꢂ

X-ray diﬀraction study. 8a: H NMR (300 MHz, 24 C, C₆D₆) ꢂ

ꢃ11:82 (d, J_PH¼ 10:9 Hz, 5H, Ru–H–Ru), 1.06 (d, J_PH¼ 8:3

Hz, 9H, P–Me), 1.87 (d, J_PH¼ 1:8 Hz, 15H, Cp–Me), 2.02 (s,

30H, Cp–Me). ¹H NMR (500 MHz, ꢃ108 ^ꢂC, tetrahydrofuran-

d₈/toluene-d₈) ꢂ ꢃ23:50 (s, 1H, Ru–H–Ru), ꢃ11:06 (s, 1H,

Ru–H–Ru), ꢃ10:90 (d, J_PH¼ 32:5 Hz, 2H, Ru–H–Ru), ꢃ2:96

(s, 1H, Ru–H–Ru), 1.06 (d, J_PH¼ 8:3 Hz, 9H, P–Me), 1.87 (d,

J_PH¼ 1:8 Hz, 15H, Cp–Me), 2.02 (s, 30H, Cp–Me). ¹³C NMR

(75.5 MHz, 24 ^ꢂC, C₆D₆) ꢂ 12.3 (q, J_CH¼ 126 Hz, Cp–Me),

red for 50 min at ꢃ50 C and for 10 min at 0 C. The resulting

mixture was passed through a column packed with alumina, and

the solvent was removed in vacuo to give [{Ru(C₅Me₅)}₃(ꢀ-

H)(ꢀ₃-O)₂] (10) as a brown solid. The slow cooling of a tetrahy-

ꢂ

drofuran-diethyl ether solution of 10 to ꢃ30 C gave analytically

pure 10+H₂O (84.8 mg, 80%) as black-brown needles. 10:

ꢂ

¹H NMR (300 MHz, 23 C, C₆D₆) ꢂ ꢃ15:84 (s, 1H, Ru–H–Ru),

1.76 (s, 30H, Cp–Me), 1.78 (s, 15H, Cp–Me). ¹³C{¹H} NMR

(75 MHz, 23 ^ꢂC, C₆D₆) ꢂ 11.8 (Cp–Me), 12.2 (Cp–Me), 90.9

(Cp-ring), 91.3 (Cp-ring). IR (KBr, cm^ꢃ1): 2984, 2960, 2906,

13.4 (q, J_CH¼ 126 Hz, Cp–Me), 24.5 (dq, J_PC¼ 27:6 Hz, J_CH

¼

86

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

BCSJ AWARD ARTICLE

2858, 1431, 1404, 1377, 1079, 1067, 1027. Anal. Calcd for

C₃₀H₄₆O₂Ru₃+ H₂O: C, 47.41; H, 6.36%. Found: C, 47.44; H,

6.42%.

room temperature, and was then passed through a column packed

with Celite. Removal of the solvent from the eluent under reduced

pressure gave 12 (209 mg, 85%) as a green-black solid. Slow cool-

ing of a tetrahydrofuran solution of 12 to ꢃ34 ^ꢂC aﬀorded green-

black single crystals. 12: IR (KBr, cm^ꢃ1): 2944, 2888, 2821, 1782,

1735, 1615, 1473, 1447, 1422, 1370, 1068, 1022, 515. Anal. Calcd

for C₃₄H₄₅O₄Ru₃: C, 49.74; H, 5.53%. Found: C, 49.67; H,

5.43%. FD-MS: 813 (8), 814 (16), 815 (20), 816 (33), 817 (45),

818 (59), 819 (78), 820 (88), 821 (95), 822 (199), 823 (83), 824

(70), 825 (55), 826 (28), 827 (22), 828 (10).

Reaction of [{Ru(C₅Me₄Et)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6⁰) with

Dioxygen.

Preparation of [{Ru(C₅Me₄Et)}₃(ꢀ-H)(ꢀ₃-O)₂]

(10⁰): A 50-mL Schlenk tube was charged with 115 mg (0.152

mmol) of 6⁰, 10 mL of tetrahydr_ꢂofuran, and 2 mL of methanol.

The mixture was cooled to ꢃ78 C. After the Schlenk tube was

evacuated, 2 equiv of dioxygen dried over P₂O₅was admitted

to the tube by using a vacuum line. The mixture was stirred for

30 min at room temperature. The resulting mixture was passed

through a column packed with alumina. Removal of the solvent

under reduced pressure gave 10⁰+H₂O (112 mg, 92%) as a yel-

low-brown crystalline solid. Recrystallization from_ꢂthe mixed sol-

vent of tetrahydrofuran and diethyl ether at ꢃ20 C gave single

References

1

For example, ‘‘Catalysis by Di- and Polynuclear Metal

Cluster Complexes,’’ ed by R. Adams and F. A. Cotton, Wiley-

VCH, New York (1991), and references cited therein.

1

crystals of 10⁰as black-brown needles. 10⁰: H NMR (300 MHz,

ꢂ

23 C, C₆D₆) ꢂ ꢃ15:82 (s, 1H, Ru–H–Ru), 1.05 (t, J ¼ 7:6 Hz,

2

J. C. Huﬀman, M. A. Green, S. K. Kaiser, and K. G.

Caulton, J. Am. Chem. Soc., 107, 5111 (1985).

D. Alvarez, Jr., E. G. Lundquist, J. W. Ziller, W. J. Evans,

and K. G. Caulton, J. Am. Chem. Soc., 111, 8392 (1989).

J. W. Bruno, J. C. Huﬀman, M. A. Green, and K. G.

Caulton, J. Am. Chem. Soc., 106, 8310 (1984).

E. G. Lundquist, J. C. Huﬀman, and K. G. Caulton, J. Am.

Chem. Soc., 108, 8309 (1986).

J. T. Poulton, K. Folting, and K. G. Caulton, Organometal-

lics, 11, 1364 (1992).

6H, Cp–CH₂–Me), 1.16 (t, J ¼ 7:6 Hz, 3H, Cp–CH₂–Me), 1.75

(s, 12H, Cp–Me), 1.79 (s, 6H, Cp–Me), 1.808 (s, 12H, Cp–Me),

1.813 (s, 6H, Cp–Me), 2.18 (q, J ¼ 7:6 Hz, 2H, Cp–CH₂–Me),

2.22 (q, J ¼ 7:6 Hz, 4H, Cp–CH₂–Me). ¹³C NMR (75 MHz, 23

^ꢂC, C₆D₆) ꢂ 11.98 (q, J ¼ 126:7 Hz, Cp–Me), 12.04 (q, J ¼

126:7 Hz, Cp–Me), 12.4 (q, J ¼ 126:4 Hz, Cp–Me), 12.5 (q, J ¼

126:4 Hz, Cp–Me), 15.6 (tq, J ¼ 4:4, 126.4 Hz, Cp–CH₂–Me),

15.7 (tq, J ¼ 4:4, 126.4 Hz, Cp–CH₂–Me), 20.6 (qt, J ¼ 4:4,

127.4 Hz, Cp–CH₂–Me), 21.0 (qt, J ¼ 4:4, 127.2 Hz, Cp–CH₂–

Me), 90.4 (s, Cp-ring), 90.9 (s, Cp-ring), 91.3 (s, Cp-ring), 91.6

(s, Cp-ring), 97.0 (s, Cp-ring), 97.2 (s, Cp-ring). IR (KBr,

cm^ꢃ1): 3550, 3420, 2966, 2900, 2104, 1648, 1562, 1456, 1408,

1377, 1307, 1262, 1155, 1082, 1020, 963, 829, 670, 633, 526.

Anal. Calcd for C₃₃H₅₂O₂Ru₃+ H₂O: C, 49.42; H, 6.79%.

Found: C, 49.02; H, 6.98%.

3

4

5

6

7

D. J. Alvarez, K. G. Caulton, W. J. Evans, and J. W. Ziller,

J. Am. Chem. Soc., 112, 5674 (1990).

8

D. J. Alvarez, K. G. Caulton, W. J. Evans, and J. W. Ziller,

Inorg. Chem., 31, 5500 (1992).

Z. He, S. Nefedov, N. Lugan, D. Neibecker, and R.

9

Mathieu, Organometallics, 12, 3837 (1993).

10 I. Moldes, S. Nefedov, N. Lugan, and R. Mathieu, J. Orga-

nomet. Chem., 490, 11 (1995).

Reaction of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6) with Iodo-

methane.

Preparation of [{Ru(C₅Me₅)}₃(ꢀ₃-H)(ꢀ-H)₃(ꢀ₃-

I)] (11): To a stirred solution of 6 (253 mg, 0.355 mmol) in

20 mL of tetrahydrofuran was added iodomethane (26 mg,

0.382 mmol) at 25 ^ꢂC. After being stirred for 24 h at room temper-

ature, the resulting solution was ﬁltered through a Celite pad, and

dried in vacuo to aﬀord [{Ru(C₅Me₅)}₃(ꢀ₃-I)(ꢀ₃-H)(ꢀ-H)₃] (11)

as a black-brown solid (275 mg, 92%). The slow cooling of a sat-

urated tetrahydrofuran–pentane solution of 6 to ꢃ20 ^ꢂC gave sin-

gle crystals of 11 as black-brown needles. 11: ¹H NMR (270 MHz,

23 ^ꢂC, C₆D₆) ꢂ ꢃ2:25 (s, 4H, Ru–H), 2.12 (s, 45H, Cp–Me).

¹³C{¹H} NMR (67.5 MHz, 23 ^ꢂC, CDCl₃) ꢂ 13.2 (q, J ¼ 126:7

Hz, Cp–Me), 84.5 (s, Cp-ring). IR (KBr, cm^ꢃ1): 2962, 2888,

1448, 1369, 1146, 1067, 1020, 608, 580, 497, 432. Anal. Calcd

for C₃₀H₄₉IRu₃: C, 42.90; H, 5.88; I, 15.11%. Found: C, 42.71;

H, 5.74; I, 15.17%.

11 Z. He, D. Neibecker, N. Lugan, and R. Mathieu, Organo-

metallics, 11, 817 (1992).

12 M. Cazanoue, Z. He, D. Neibecker, and R. Mathieu,

J. Chem. Soc., Chem. Commun., 1991, 307.

13 Z. He, L. Plasseraud, I. Moldes, F. Dahan, D. Neibecker,

M. Etienne, and R. Mathieu, Angew. Chem., Int. Ed. Engl., 34,

916 (1995).

14 A. Albinati, A. Musco, R. Naegeli, and L. M. Venanzi,

Angew. Chem., Int. Ed. Engl., 20, 958 (1981).

15 J. L. Kersten, A. L. Reingold, K. H. Theopold, C. P. Casey,

R. A. Widenhoefer, and C. E. C. A. Hop, Angew. Chem., Int. Ed.

Engl., 31, 1341 (1992).

16 C. P. Casey, S. L. Hallenbeck, and R. A. Widenhoefer,

J. Am. Chem. Soc., 117, 4607 (1995).

Monitoring the Reaction of 6 with Iodomethane in Tetrahy-

drofuran-d₈. A ¹H NMR sample tube was charged with 6 (14.3

mg, 20 mmol) and tetrahydrofuran-d₈(0.5 mL). Iodomethane (1.1

17 H. Suzuki, H. Omori, D. H. Lee, Y. Yoshida, and Y.

Moro-oka, Organometallics, 7, 2243 (1988).

18 H. Suzuki, H. Omori, and Y. Moro-oka, Organometallics,

7, 2579 (1988).

19 H. Omori, H. Suzuki, and Y. Moro-oka, Organometallics,

8, 1576 (1989).

20 H. Omori, H. Suzuki, and Y. Moro-oka, Organometallics,

8, 2270 (1989).

21 H. Omori, H. Suzuki, T. Kakigano, and Y. Moro-oka,

Organometallics, 11, 989 (1992).

ꢂ

mL, 0.18 mmol) was then added to the cooled (ꢃ30 C) solution,

and the tube was sealed. Spectra were recorded at appropriate time

intervals. ¹H NMR (300 MHz, 23 ^ꢂC, THF-d₈) ꢂ ꢃ2:38 (s, 4H,

Ru–H), 2.06 (s, 45H, Cp–Me), 0.19 (s, CH₄).

Reaction of [{Ru(C₅Me₅)}₃(ꢀ-H)₃(ꢀ₃-H)₂] (6) with Carbon

Monoxide. Preparation of [{Ru(C₅Me₅)}₃(ꢀ-CO)₃(ꢀ₃-CO)]

(12): A 50-mL Schlenk tube, equipped with a rubber balloon ﬁll-

ed with CO, was charged with 214 mg (0.30 mmol) of 6 and_ꢂ10

mL of tetrahydrofuran. The mixture was then cooled to ꢃ78 C.

After the Schlenk tube was evacuated, carbon monoxide was ad-

mitted to the tube. The resulting mixture was stirred for 12 h at

22 H. Suzuki, T. Takao, M. Tanaka, and Y. Moro-oka,

J. Chem. Soc., Chem. Commun., 1992, 476.

23 H. Suzuki, H. Omori, D. H. Lee, Y. Yoshida, M.

Fukushima, M. Tanaka, and Y. Moro-oka, Organometallics, 13,

H. Suzuki et al.

Bull. Chem. Soc. Jpn., 78, No. 1 (2005)

87

1129 (1994).

24 T. Takao, H. Suzuki, and M. Tanaka, Organometallics, 13,

2554 (1994).

25 T. Takao, S. Yoshida, H. Suzuki, and M. Tanaka, Organo-

metallics, 14, 3855 (1995).

26 K. Tada, M. Oishi, H. Suzuki, and M. Tanaka, Organo-

metallics, 15, 2422 (1996).

27 Y. Ohki and H. Suzuki, Angew. Chem., Int. Ed., 39, 3463

(2000).

28 T. Takao, M. Amako, and H. Suzuki, Organometallics, 20,

3406 (2001).

29 T. Takao, S. Yoshida, and H. Suzuki, Chem. Lett., 2001,

1100.

30 Y. Ohki and H. Suzuki, Angew. Chem., Int. Ed., 41, 2994

(2002).

31 T. Takao, M. Amako, and H. Suzuki, Organometallics, 22,

3855 (2003).

and A. Sironi, J. Am. Chem. Soc., 124, 5117 (2002).

51 ‘‘Metal Dihydrogen and s-Bond Complexes. Structure,

Theory, and Reactivity,’’ ed by G. J. Kubas, Kluwer Academic/

Plenum Publishers, New York (2001), Chapter 4.

52 A. Inagaki, Y. Takaya, T. Takamura, M. Tanaka, M. Haga,

and H. Suzuki, J. Am. Chem. Soc., 119, 625 (1997).

53 W. D. Jones and F. J. Feher, J. Am. Chem. Soc., 104, 4240

(1982).

54 W. D. Jones and F. J. Feher, Organometallics, 2, 562

(1983).

55 W. D. Jones and F. J. Feher, Inorg. Chem., 23, 2376

(1984).

56 W. D. Jones and F. J. Feher, J. Am. Chem. Soc., 106, 1650

(1984).

57 W. D. Jones and F. J. Feher, J. Am. Chem. Soc., 108, 4814

(1986).

58 T. Takemori, H. Suzuki, and M. Tanaka, Organometallics,

21, 4346 (1996).

59 T. D. Tilley, R. H. Grubbs, and J. E. Bercaw, Organo-

metallics, 3, 274 (1984).

32 H. Suzuki, T. Kakigano, M. Igarashi, and Y. Moro-oka,

J. Chem. Soc., Chem. Commun., 1991, 283.

33 J.-F. Riehl, N. Koga, and K. Morokuma, Organometallics,

13, 4765 (1994).

34 D. G. Hamilton and R. H. Crabtree, J. Am. Chem. Soc.,

110, 4126 (1988).

60 H. Suzuki, Y. Takaya, T. Takemori, and M. Tanaka, J. Am.

Chem. Soc., 116, 10779 (1994).

61 D. V. Khoroshun, A. Inagaki, H. Suzuki, S. F.

Vyboishchikov, D. G. Musaev, and K. Morokuma, J. Am. Chem.

Soc., 125, 9910 (2003).

35 P. J. Fagan, M. D. Ward, and J. C. Calabrese, J. Am. Chem.

Soc., 111, 1698 (1989).

36 (C₅Me₄Et)Ru(ꢀ-H)₄Ru(C₅Me₄Et) was prepared according

to previously published procedures by using {(C₅Me₄Et)RuCl₂}₂

as a starting material.³⁷Dichloride complex {(C₅Me₄Et)RuCl₂}₂

was obtained in high yield from the reaction of ruthenium trichlo-

ride trihydrate with ethyltetramethylcyclopentadiene in ethanol.²³

37 N. Oshima, H. Suzuki, and Y. Moro-oka, Chem. Lett.,

1984, 1161.

62 A. Inagaki, T. Takemori, M. Tanaka, and H. Suzuki,

Angew. Chem., Int. Ed., 39, 404 (2000).

63 F. A. Cotton, G. Wilkinson, C. A. Murillo, and M.

Bochmann, ‘‘Advanced Inorganic Chemistry,’’ 6th ed, John Wiley

& Sons, Inc., New York (1999), pp. 54–56.

64 A. Uchtman and L. F. Dahl, J. Am. Chem. Soc., 91, 3763

(1969).

38 K. Nakamoto, ‘‘Infrared and Raman Spectra of Inorganic

and Coordination Compounds,’’ 3rd ed, John Wiley & Sons,

Inc., New York (1978).

65 A. Nutton, P. M. Bailey, N. C. Braund, J. Goodfellow, R.

S. Thompson, and P. M. Maitlis, J. Chem. Soc., Chem. Commun.,

1980, 631.

´

39 The molecular structure of 3a was determined by using a

single crystal obtained from the cooled (ꢃ40 ^ꢂC) Et₂O/pentane

solution. The structure and some relevant bond lengths and angles

are given in the previously reported communication (Ref. 18).

40 G. Meister, G. Rheinwald, H. Stoeckli-Evans, and Suss-

Fink, J. Organomet. Chem., 496, 197 (1995).

41 N. Koga and K. Morokuma, J. Mol. Struct., 300, 181

(1993).

42 G. J. Kubas, C. J. Burns, G. R. K. Khalsa, L. S. V. D.

Sluys, G. Kiss, and C. D. Hoﬀ, Organometallics, 11, 3390 (1992).

43 H. Suzuki, Y. Ito, J. Fujimoto, and N. Uehara, to be pub-

lished.

66 A. Colombie, J.-J. Bonnet, P. Fompeyrine, G. Lavigne, and

S. Sunshine, Organometallics, 5, 1154 (1986).

67 C. P. Gibson, J.-S. Huang, and L. F. Dahl, Organometal-

lics, 5, 1676 (1986).

68 S. E. Nefedov, A. A. Pasynskii, I. L. Eremenko, B.

Orazsakhatov, O. G. Ellert, V. M. Novotortsev, Y. T. Struchkov,

and A. I. Yanovsky, J. Organomet. Chem., 385, 277 (1990).

69 H. A. Mirza, J. J. Vittal, and R. J. Puddephatt, Inorg.

Chem., 34, 4239 (1995).

70 R. J. Goudsmit, B. F. G. Johnson, J. Lewis, P. R. Raithby,

and K. H. Whitmire, J. Chem. Soc., Chem. Commun., 1983,

246.

44 R. H. Crabtree, M. Lavin, and L. Bonneviot, J. Am. Chem.

Soc., 108, 4032 (1986).

45 V. Bakhmutov, T. Burgi, P. Bruger, U. Ruppli, and H.

¨

Berke, Organometallics, 13, 4203 (1994).

71 A. Ceriotti, L. Resconi, F. Demartin, G. Longoni, M.

Manassero, and M. Sansoni, J. Organomet. Chem., 249, C35

(1983).

72 J. Ito, T. Shima, and H. Suzuki, Organometallics, 23, 2447

(2004).

73 ‘‘TEXSAN: Crystal Structure Analysis Package,’’ Molecu-

lar Structure Corp., The Woodlands, TX, 1985 and 1992.

74 ‘‘Internatonal Tables for X-ray Crystallography,’’ Kynoch

Press, Birmingham, UK (1975), Vol. 4.

´

46 S. Camanyes, F. Maseras, M. Moreno, A. Lledos, J. M.

´

Lluch, and J. Bertran, J. Am. Chem. Soc., 118, 4617 (1996).

47 A. C. Cooper, O. Eisenstein, and K. G. Caulton, New J.

Chem., 22, 307 (1998).

48 A. Castillo, G. Barea, M. A. Esteruelas, F. J. Lahoz,

´

A. Lledos, F. Maseras, J. Modrego, E. Onate, L. A. Oro, N. Ruiz,

and E. Sola, Inorg. Chem., 38, 1814 (1999).

49 F. L. Taw, H. Mellows, P. S. White, F. J. Hollander, R. G.

Bergman, M. Brookhart, and D. M. Heinekey, J. Am. Chem. Soc.,

124, 5100 (2002).

75 P. T. Beurskens, G. Admiraal, G. Beurskens, W. P.

Bosman, S. Garcia-Granda, R. O. Gould, J. M. M. Smits, C.

Smykalla, ‘‘The DIRDIF Program System, Technical Report

of the Crystallography Laboratory,’’ University of Nijmegen,

Nijmegen, The Netherland (1992).

50 M. Bergamo, T. Beringhelli, G. D’Alfonso, P. Mercandelli,

Article Doi

Synthesis, structures, and reactions of coordinatively unsaturated trinuclear ruthenium polyhydrido complexes, [{Ru(C(5)Me (5))}(3)(μ-H)(6)](Y) (Y = BF(4), CF (3)SO(3), 1/2(SO(4)), C(6)H (5)CO(2), CH(3)CO(2), B(C(6)H (5))(4), PF(6)) and [{Ru(C(5)M (5))}(3)(μ-H)(3)(μ(3)-H)(2)]

DOI: 10.1246/bcsj.78.67

Source and publish data:

Authors:

Article abstract of DOI:10.1246/bcsj.78.67

Full text of DOI:10.1246/bcsj.78.67

Products guided by the article

R&D Labs maybe for 134148-49-5

Relevant to this article

Hot Product