Reproducible high-yield syntheses of [Ru3(CO)12], [H4Ru4(CO)12], and [Ru6C(CO)16]2- by a convenient two-step methodology involving controlled reduction in ethylene glycol of RuCl3·nH2O

Article abstract of DOI:10.1016/S0022-328X(02)02215-5

Ru3CO12 [H₄Ru₄(CO)₁₂], and [Ru₆C(CO)₁₆]^2- have been synthesized in reproducible high yields and under mild conditions (1 atm) by a two-step methodology involving (i) first carbonylation of RuCl₃·nH₂O dissolved in ethylene glycol to give a mixture of tri- and di-carbonyl ruthenium(II) species, probably of the kind [Ru(CO)₃Cl₂(ethylene glycol)] and [Ru(CO)₂Cl₂(ethylene glycol)_x ] (x = 1, 2), and (ii) addition of specific amounts of alkali carbonates and further reductive carbonylation to give the desired ruthenium carbonyl cluster. The selectivity of the second step is controlled by the: (i) nature and quantity of the alkali carbonate (Na₂CO₃ or K₂CO₃); (ii) gas-phase composition (CO or CO+H₂); (iii) temperature.

Full text of DOI:10.1016/S0022-328X(02)02215-5

Journal of Organometallic Chemistry 669 (2003) 44Á47
/
www.elsevier.com/locate/jorganchem
Reproducible high-yield syntheses of [Ru (CO) ], [H Ru (CO) ],
3
12
4
4
12
2
and [Ru C(CO) ] by a convenient two-step methodology
ꢀ
6
16
involving controlled reduction in ethylene glycol of RuCl ×
/
nH O
2
3
a
b
b
b,
Elena Lucenti , Elena Cariati , Claudia Dragonetti , Dominique Roberto *
a
Istituto di Scienze e Tecnologie Molecolari del CNR (ISTM), Via C. Golgi 19, 20133 Milan, Italy
b
Dipartimento di Chimica Inorganica, Metallorganica e Analitica dell’Universit a` di Milano and Istituto di Scienze e Tecnologie Molecolari del CNR
(
ISTM), via Venezian 21, 20133 Milan, Italy
Received 3 April 2002; received in revised form 10 December 2002; accepted 15 December 2002
Abstract
2
ꢀ
[
Ru
atm) by a two-step methodology involving (i) first carbonylation of RuCl
tri- and di-carbonyl ruthenium(II) species, probably of the kind [Ru(CO)
xꢁ1, 2), and (ii) addition of specific amounts of alkali carbonates and further reductive carbonylation to give the desired
ruthenium carbonyl cluster. The selectivity of the second step is controlled by the: (i) nature and quantity of the alkali carbonate
Na CO or K CO ); (ii) gas-phase composition (CO or COꢂ ); (iii) temperature.
2003 Elsevier Science B.V. All rights reserved.
3
(CO)12], [H
4
Ru
4
(CO)12], and [Ru
6
C(CO)16
]
have been synthesized in reproducible high yields and under mild conditions (1
nH O dissolved in ethylene glycol to give a mixture of
x
Cl (ethylene glycol)] and [Ru(CO) Cl (ethylene glycol) ]
3
×
/
2
3
2
2
2
(
/
(
2
3
2
3
/H
2
#
Keywords: Ruthenium carbonyl clusters synthesis; Ruthenium trichloride reductive carbonylation
1
. Introduction
solvent carrying non acidic OH groups that could mimic
the role and polarity of the OH groups of the silica
surface [5,6]. The selective and facile direct reductive
Although the reported syntheses in solution of high
nuclearity ruthenium and osmium carbonyl clusters
usually require [M (CO) ] (MꢁRu, Os) as starting
material, it appeared that various neutral or anionic
ruthenium [1] and osmium [2,3] carbonyl clusters can be
synthesized in high yields and rather easily by controlled
reduction at atmospheric pressure of silica-supported
carbonylation of MCl3×
cularly convenient on a synthetic point of view but quite
nH O gives
/
nH O (Mꢁ
/
Ru, Os) was parti-
2
/
3
12
unexpected because silica-supported MCl3×
/
2
a low selectivity and very low yields when reduced under
similar conditions in the presence of alkali carbonates
[1Á
/
3]. However, it appeared that the syntheses of
2
ꢀ
[
M(CO) Cl ] or silica-bound [M(CO) Cl (HOSiÅ
/
)]
generated in situ by controlled reductive carbonylation
nH O) [4] in the presence of
[Ru (CO) ], [H Ru (CO) ], and [Ru C(CO) ] by
nH O dissolved in ethylene
3
2 2
3
2
3
12
4
4
12
6
16
(
direct reduction of RuCl3×
/
2
of silica-supported MCl3×
/
glycol are not always reproducible, probably due to
various factors such as amount of water present in the
reaction mixture and nature of the commercial salt [7].
For this reason and because the syntheses starting from
[Ru(CO) Cl ] dissolved in ethylene glycol were well
2
specific amounts of alkali carbonates. These silica-
mediated syntheses were the springboard of new con-
venient syntheses of [M (CO) ], [H M (CO) ],
3
12
2ꢀ
4
4
12
ꢀ
[
H M (CO) ] , [Ru C(CO) ]
3
,
and [H Os -
4 10
4
12
6
16
3
2 2
2
ꢀ
( starting from [M(CO) Cl ] or MCl ×nH O
and working in ethylene glycol solution, a high boiling
CO) ]
3 /
reproducible [5], we investigated a modified two-step
methodology, inspired by the two-step route to convert
24
3
2 2
2
silica-supported RuCl3×
/
nH O into ruthenium carbonyl
2
clusters via [Ru(CO) Cl (HOSiÅ
/
)] [1], involving (i)
preparation of tri- and di-carbonyl ruthenium(II) species
by carbonylation of RuCl3×nH O at 110 8C in ethylene
3
2
*
E-mail address: dominique.roberto@unimi.it (D. Roberto).
Corresponding author. Fax: ꢂ
/
39-02-5031-4405.
/
2
0022-328X/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0022-328X(02)02215-5

E. Lucenti et al. / Journal of Organometallic Chemistry 669 (2003) 44Á
/
47
45
glycol, as previously reported [5] and (ii) addition of
specific amounts of alkali carbonates and further
reductive carbonylation to give selectively the desired
neutral or anionic cluster.
It is worth pointing out that the new one-pot two-step
synthesis in solution of [Ru (CO) ] and [H Ru (CO) ]
3
12
4
4
12
must be carried out under well defined conditions with
the stream of CO at the top of the condenser in order to
obtain reproducible high yields of the clusters. In fact, if
the reaction is carried out by bubbling CO in the flask, a
non reproducible quantity of HCl is lost during the first
step working at 110 8C; consequently, during the second
step, a considerable amount of anionic ruthenium
carbonyl complexes can be formed as by-products
2
. Results and discussion
When, in a flask equipped with a condenser, RuCl3×/
nH O (ca. 0.2 g, see Section 3) dissolved in ethylene
2
even using a molar ratio Na CO :Ruꢁ
/
3:2. These
2
3
glycol is heated at 110 8C for 2 h under 1 atm of CO a
mixture of tri- and di-carbonyl ruthenium(II) species
probably of the kind [Ru(CO) Cl (ethylene glycol)]
results confirm that, in the case of the [Ru (CO) ] and
12
3
[H Ru (CO) ] syntheses, the reduction must be carried
4 4 12
out in the presence of the stoichiometric amount of
3
2
ꢀ
1
(
nCO
ꢁ
/
2136 (m) and 2065 (s) cm ) and [Ru(CO)2-
Cl (ethylene glycol) ] (xꢁ1, 2; nCO 2067 (s) and 1999
s) cm ) is formed, as evidenced by infrared spectro-
scopy [5,8]. Addition at 25Á40 8C of Na CO (molar
ratio Na CO :Ruꢁ3:2) followed by further reaction
Na CO necessary to remove the chloro ligands from
3
2
/
ꢁ
/
2
x
the coordination sphere of ruthenium (molar ratio
Na:Clꢁ1:1), because, when working in solution, use
of even a small excess of Na CO leads to the formation
ꢀ
1
(
/
/
2
3
2
3
/
2
3
of anionic carbonyl clusters as by-products [5]. There-
fore, before the preparation of the clusters it is crucial to
control the exact Ru content of the salt which is very
hygroscopic. Besides, in our preparation of [Ru (CO) ],
under CO (1 atm) at 80 8C for 7 h affords an orange
material which precipitates in the reaction flask whereas,
in parallel, some orange powder sublimes on the cold
walls of the condenser. Extraction of the sublimate and
reaction mixture with dichloromethane affords
3
12
it is worth pointing out that if the second step is carried
out in the presence of a molar ratio Na:Cl inferior to 1, a
mixture of [Ru (CO) ] and unreacted tri- and di-
[
Ru (CO) ] in reproducible excellent yields (91%). By
3 12
3
12
working on a high preparative scale (starting from ca. 2
g of RuCl3×nH O), the second reduction step is slightly
slower affording after 15 h [Ru (CO) ] in 61% yield;
carbonyl ruthenium(II) species is obtained. This result
is in contrast with a recently reported similar two-step
preparation of [Ru (CO) ] (86% yield) involving (i)
/
2
3
12
3
12
under these conditions the formation of small amounts
ꢀ
reductive carbonylation of RuCl nH O in 2-methox-
×
/
3 2
of anionic clusters such as [H Ru (CO) ]
ꢀ
and
HRu (CO) ] [1] cannot be avoided (see Section 3).
3
4
12
yethanol at 125 8C followed by addition of ca. one
equivalent of KOH per Ru and successive reductive
carbonylation at 85 8C [9], but it was then observed that
it is better to use a higher amount of KOH (ca. two
equivalents per Ru) in this latter synthesis [10].
[
When the second reduction step is carried out with the
6
18
same amount of Na CO at 95 8C for 7 h under a
2
3
mixture of COꢂ
/
H (molar ratioꢁ
/1:3; 1 atm) instead of
2
CO, some [H Ru (CO) ] precipitates in the reaction
flask whereas, in parallel, some [H Ru (CO) ] sublimes
4
4
12
In contrast to the syntheses of [Ru (CO) ] and
3
12
4
4
12
2
ꢀ
[
H Ru (CO) ] that of [Ru C(CO) ]
6
can be carried
4
4
12
16
on the cold walls of the condenser. Besides, the infrared
spectrum of the ethylene glycol solution shows the
out under CO bubbling since it occurs in the presence of
a large excess of alkali carbonate and, therefore, it is not
sensitive to the exact quantity of base. Thus, when CO
ꢀ
presence of [H Ru (CO) ] (n
CO
ꢁ2036 (m), 2016 (s),
/
3
4
12
ꢀ
1
998 (s) and 1977 (m, br) cm ). Acidification of the
1
ethylene glycol solution with H SO leads to the
(
1 atm) is bubbled through a solution of RuCl3×
dissolved in ethylene glycol in a three necked flask at
10 8C for 3Á4 h a mixture of tri- and di-carbonyl
/
nH O
2
2
4
precipitation of more [H Ru (CO) ]. Successive extrac-
4
4
12
1
/
tion of the sublimate and the reaction mixture with
dichloromethane at room temperature, gives [H Ru -
ruthenium(II) species, probably of the kind [Ru(CO)3-
4
4
Cl (ethylene glycol)] and [Ru(CO) Cl (ethylene glycol) ]
2
(
scale). Such yields of [Ru (CO) ] and [H Ru (CO) ]
CO) ] in 73% yield (working on a high preparative
2
2
x
12
(
xꢁ/1, 2), [5,8] is formed. Addition of K CO (molar
2 3
3
12
4
4
12
ratio K CO :Ruꢁ
bling at 160 8C for 8 h, affords a dark red solution which
infrared spectrum shows carbonyl stretching bands at
/
10:1) followed by CO (1 atm) bub-
are similar to those obtained in ethylene glycol starting
from [Ru(CO) Cl ] [5] or on the silica surface starting
2
3
3
2 2
from RuCl3×
quires a much longer reaction time (ca. 5 days) [1]. In
fact, both the reduction of RuCl3×nH O to Ru(II)
/
nH O via [Ru(CO) Cl (HOSiÅ
/)] that re-
2
3
2
n
ꢁ2034 (vw), 1977 (vs), 1953 (vw, sh), 1917 (w)
/
CO
ꢀ
1
2ꢀ
/
cm , typical for [Ru
of this anion under N
CH Cl affords [NBu
6
C(CO)16
]
. Repeated extraction
with a solution of [NBu ]I in
[Ru C(CO)16] in 80Á84% yields
2
carbonyl species and the subsequent formation of
carbonyl clusters occur much more rapidly in ethylene
glycol than on the silica surface due to a higher mobility
of reagents and intermediate species in solution than on
a solid surface [1,5].
2
2
4
]
/
2
2
4
6
(working on a high preparative scale), similar to the
yields obtained starting from [Ru(CO) Cl dissolved in
ethylene glycol [5] or on the silica surface starting from
]
2 2
3

46
E. Lucenti et al. / Journal of Organometallic Chemistry 669 (2003) 44Á47
/
RuCl3×
/
nH O via [Ru(CO) Cl (HOSiÅ
/
)] that requires a
respectively. During the second step some brown
powder (still unidentified complex, characterized by
2
3
2
much longer reaction time (ca. 5 days) [1].
ꢀ
1
In conclusion, by a spring of inspiration from silica-
mediated syntheses, we found a new convenient two-step
methodology to prepare [Ru (CO) ], [H Ru (CO) ],
nCO at 1998 (vs) and 1975 (sh) cm
in Nujol, which
is insoluble in organic solvents, donor solvents and even
aqueous 37% HCl) precipitated along with orange
[Ru (CO) ]. The work-up was carried out as follows.
3
12
4
4
12
2
ꢀ
and [Ru C(CO) ]
rapidly, in high yields and under
mild conditions by controlled reduction of a simple
starting material such as RuCl3×nH O dissolved in
6
16
3
12
Dichloromethane (ca. 150 ml) was added under stirring
to the reaction mixture, left to stand and separated in
order to extract [Ru (CO) ]. This treatment was
/
2
ethylene glycol. These new syntheses in solution are
reproducible and much more convenient than those
previously reported either in solution [5] or on the silica
surface as reaction medium [1].
3
12
repeated several times until no more [Ru (CO) ] was
3
12
present in the CH Cl phase. The CH Cl phase was
2
2
2
2
then filtered in order to eliminate the brown powder and
then dried to give [Ru (CO) ] which was washed with
3
12
water (3ꢃ
/
10 ml) and then pentane (3ꢃ10 ml), in order
/
3
. Experimental
to eliminate traces of ethylene glycol, affording spectro-
scopically pure [Ru (CO) ] (1.145 g; 1.790 mmol; 61%
3
12
3
.1. General comments
yield). Extraction of the brown ethylene glycol solution
with a solution of [NBu ]I in CH Cl afforded a mixture
4
2
2
ꢀ
RuCl3×
/
nH O (40.82Á
/
42.68 wt.% Ru) was purchased
of anionic carbonyl clusters such as [H Ru (CO) ]
3 4 12
2
ꢀ
1
from Engelhard. The exact content of water was
determined before each reaction by thermogravimetric
analysis. All the reactions were monitored by infrared
spectroscopy in the carbonyl region. Reaction products
and [HRu (CO) ] as identified by infrared and H-
6 18
NMR spectroscopies [1].
3.3. Synthesis of [H Ru (CO) ]
4
4
12
1
were identified by infrared, H-NMR (when appropri-
ate), and mass spectroscopies, their spectra being
compared with those of pure samples. Their purity
was controlled by thin-layer chromatography (when
possible) and by elemental analysis.
RuCl3×
/
nH O (40.88 wt.% Ru; 0.998 g, 4.038 mmol)
2
was dissolved in ethylene glycol (250 ml) in a three
necked flask equipped with a condenser and thermo-
meter. CO (1 atm) was bubbled through the resulting
dark brown solution for 5 min and than was allowed to
flow only on the top of the condenser. The reaction
mixture was heated under CO at 110 8C for 5 h. After
cooling to 25 8C, Na CO (0.639 g, 6.031 mmol; molar
3
.2. Synthesis of [Ru (CO) ]
3 12
RuCl3×/nH O (40.82 wt.% Ru; 0.231 g, 0.933 mmol)
2
2
3
was dissolved in ethylene glycol (60 ml) in a three necked
flask equipped with a condenser and thermometer. CO
ratio Na CO :Ruꢁ
/
3:2) was added and a mixture of
H (molar ratioꢁ1:3; 1 atm) was allowed to flow
on the top of the condenser. The solution was heated at
95 8C for 6Á7 h. During the reaction a yellow precipitate
of [H Ru (CO) ] was formed whereas, in parallel, some
2
3
COꢂ
/
/
2
(
1 atm) was bubbled through the resulting dark brown
solution for 5 min and than was allowed to flow only on
the top of the condenser. The reaction mixture was
heated under CO at 110 8C for 2 h. After cooling to
/
4
4
12
[H Ru (CO) ] sublimed on the cold walls of the
12
4
4
2
5 8C, Na CO (0.147 g, 1.389 mmol; molar ratio
condenser. Besides, in the ethylene glycol solution
ꢀ
2
3
Na CO :Ruꢁ
/3:2) was added and CO (1 atm) was
some [H Ru (CO) ] was present as confirmed by
12
2
3
3
4
allowed to flow on the top of the condenser. Thermal
treatment at 80 8C for 7 h afforded an orange material
infrared spectroscopy. After 7 h, acidification of the
ethylene glycol solution with aqueous (aq.) H SO (up
2
4
(
whereas, in parallel, some [Ru (CO) ] sublimed on the
[Ru (CO) ]) which precipitated in the reaction flask
to pHꢁ
/
1; ca. 5 ml) led to the precipitation of more
[H Ru (CO) ]. Successive extraction of the sublimate
3
12
3
12
4
4
12
cold walls of the condenser. Extraction of the sublimate
and reaction mixture with dichloromethane (4ꢃ50 ml)
at room temperature (r.t.), followed by evaporation to
and the reaction mixture with dichloromethane (4ꢃ80
ml) at r.t., followed by evaporation to dryness of the
dichloromethane phase, gave a yellow solid residue
/
/
dryness of the dichloromethane phase, gave an orange
5
which was washed twice with water (2ꢃ
then pentane (4ꢃ5 ml) affording spectroscopically pure
[H Ru (CO) ] (0.550 g; 0.738 mmol; 73% yield).
/
10 ml) and
solid residue which was washed twice with water (2ꢃ
/
/
ml) and then pentane (2ꢃ5 ml), in order to eliminate
/
4
4
12
traces of ethylene glycol, affording spectroscopically
pure [Ru (CO) ] (0.181 mg; 0.283 mmol; 91% yield).
2
ꢀ
3.4. Synthesis of [Ru C(CO) ]
16
3
12
6
Starting from a higher amount of RuCl3×
/
nH O (40.88
2
wt.% Ru; 2.173 g, 8.790 mmol) dissolved in ethylene
glycol (600 ml) the first (without base) and second (with
base) reductive steps were carried out for 5 and 15 h,
RuCl3×
/
nH O (41.2 wt.% Ru; 2.01 g, 8.17 mmol) was
2
dissolved in ethylene glycol (550 ml) in a three necked
flask equipped with a condenser and thermometer. CO

E. Lucenti et al. / Journal of Organometallic Chemistry 669 (2003) 44Á
/
47
47
(1 atm) was bubbled through the resulting dark brown
solution and the reaction mixture was heated under CO
Ministero dell’Istruzione, dell’Universit a` e della Ricerca
and by the Consiglio Nazionale delle Ricerche.
at 110 8C for 4 h. After cooling to r.t., K CO (11.29 g,
2
3
8
1.73 mmol; molar ratio K CO :Ruꢁ10:1) was added
/
2
3
and CO (1 atm) was bubbled through the resulting
mixture. The solution was heated under CO at 160 8C
for 8 h, then cooled to r.t. Repeated extraction under N2
References
[
[
1] D. Roberto, E. Cariati, E. Lucenti, M. Respini, R. Ugo,
Organometallics 16 (1997) 4531.
with a solution (3ꢃ
/
200 ml) of [NBu ]I (2.22 g, 6.03
4
mmol) in CH Cl (600 ml) gave a dark red solution. A
further extraction was carried out, till the CH Cl phase
2
2
2] D. Roberto, E. Cariati, R. Psaro, R. Ugo, Organometallics 13
(1994) 734.
2
2
was no longer colored, with 2ꢃ
/
50 ml of [NBu ]I (0.952
4
[3] D. Roberto, E. Cariati, R. Ugo, R. Psaro, Inorg. Chem. 35 (1996)
311.
2
g, 2.58 mmol) in CH Cl (100 ml). Evaporation of the
combined CH Cl solutions afforded an oily slurry that
2
2
[
[
4] D. Roberto, R. Psaro, R. Ugo, Organometallics 12 (1993) 2292.
5] C. Roveda, E. Cariati, E. Lucenti, D. Roberto, J. Organomet.
Chem. 580 (1999) 117 (and references therein).
2
2
was washed with H O (3ꢃ
/
50 ml) and pentane (2ꢃ
/
20
ml) in order to obtain spectroscopically pure [NBu ] [R-
2
4
2
[6] E. Lucenti, D. Roberto, C. Roveda, R. Ugo, E. Cariati, J. Cluster
Sci. 12 (2001) 113.
u C(CO) ] as a red powder (1.73 g, 1.11 mmol, 82%
6
16
[
7] Professor Peter Maitlis and Dr. Aurora Castro found that our
previously reported synthesis in ethylene glycol of
yield). Yields vary slightly from preparation to prepara-
tion and are typically in the range 80Á84%. Lower yields
were obtained by working with higher concentrations.
/
2
ꢀ
[
Ru
and we confirmed the non-reproducibility of the synthesis not
only of this anionic cluster but also that of [Ru (CO)12] and
Ru (CO)12].
8] D. Roberto, R. Psaro, R. Ugo, J. Organomet. Chem. 451 (1993)
23 (and references therein).
6
C(CO)16
]
3 2
starting from RuCl ×/nH O was not reproducible
3
[
H
4
4
[
[
Acknowledgements
1
9] (a) M. Faure, L. Maurette, B. Donnadieu, G. Lavigne, Angew.
Chem. Int. Ed. 38 (1999) 518;
We wish to thank deeply Professor Peter Maitlis and
Dr. Aurora Castro for fruitful discussions which are at
the origin of this Note. This work was supported by the
(
b) A.F. Hill, Angew. Chem. Int. Ed. 39 (2000) 130.
[10] G. Lavigne, personal communication.