Homogeneous hydrogenation of alkynes

Article abstract of DOI:10.1016/S0022-328X(98)00896-1

The title complexes, containing phosphido or phosphinidene bridges, catalyze the hydrogenation of alkynes and of 1,4-cyclohexadiene (1,4-CHD). The benzyne-substituted clusters Ru₃(CO)₇(PPh₂)₂(C₆H ₄) (1) and Ru₄(CO)₁₁(PPh)(C₆H₄) (2) show the highest hydrogenation activity yet observed for substituted metal carbonyl clusters towards alkynes; the activity is related to the nature of the alkyne substrate, C₂Et₂2Ph2Ph₂. The alkyne complexes Ru₃(CO)₇(PPh₂)₂(HC₂Ph) (3) and Ru₄(CO)₁₁(PPh)(C₂Ph₂) (4), structurally closely related to 1 and 2, have also been examined in comparable reactions; complex 3 shows very high activity, especially towards 1,4-CHD. Organometallic intermediates could not be isolated but direct and indirect evidence supporting a reaction pathway based on cluster catalysis was obtained; this will require the formation of an active site, dihydrogen activation and insertion of the substrate into M-H bonds. Possible alternative mechanisms are also discussed.

Full text of DOI:10.1016/S0022-328X(98)00896-1

Journal of Organometallic Chemistry 571 (1998) 251–260
Homogeneous hydrogenation of alkynes and of 1,4-cyclohexadiene in
the presence of the clusters Ru (CO) (v-PPh ) (C H ),
3
7
2 2
6
4
Ru (CO) (v -PPh)(C H ), Ru (CO) (v-PPh ) (HC Ph) and
4
11
4
6
4
3
7
2 2
2
Ru (CO) (v -PPh)(C Ph )
4
11
4
2
2
a,1
b
b
c
Mario Castiglioni , Stefano Deabate , Roberto Giordano , Philip J. King ,
c
b,
Selby A.R. Knox , Enrico Sappa *
a
Dipartimento di Chimica Generale ed Organica Applicata, Uni6ersit a´ di Torino, Corso Massimo d’Azeglio 48, I-10125 Torino, Italy
b
Dipartimento di Chimica Inorganica, Chimica Fisica e Chimica dei Materiali, Uni6ersit a´ di Torino, Via Pietro Giuria 7, I-10125 Torino, Italy
c
School of Chemistry, Uni6ersity of Bristol, Bristol BS8 1TS, UK
Received 8 June 1998; received in revised form 23 July 1998
Abstract
The title complexes, containing phosphido or phosphinidene bridges, catalyze the hydrogenation of alkynes and of 1,4-cyclo-
hexadiene (1,4-CHD). The benzyne-substituted clusters Ru (CO) (PPh ) (C H ) (1) and Ru (CO) (PPh)(C H ) (2) show the
3
7
2 2
6
4
4
11
6
4
highest hydrogenation activity yet observed for substituted metal carbonyl clusters towards alkynes; the activity is related to the
nature of the alkyne substrate, C Et BEtC PhBC Ph .
2
2
2
2
2
The alkyne complexes Ru (CO) (PPh ) (HC Ph) (3) and Ru (CO) (PPh)(C Ph ) (4), structurally closely related to 1 and 2,
3
7
2 2
2
4
11
2
2
have also been examined in comparable reactions; complex 3 shows very high activity, especially towards 1,4-CHD.
Organometallic intermediates could not be isolated but direct and indirect evidence supporting a reaction pathway based on
cluster catalysis was obtained; this will require the formation of an active site, dihydrogen activation and insertion of the substrate
into M–H bonds. Possible alternative mechanisms are also discussed. © 1998 Elsevier Science S.A. All rights reserved.
Keywords: Metal carbonyl clusters; Phosphido, phosphinidene bridges; Alkyne, benzyne ligands; Homogeneous catalysis;
Hydrogenation
1. Introduction
behave as intermediates in hydrogenation reactions
1,2,4]. It was also established that phosphido- and
[
Homo- and hetero-metallic ruthenium carbonyl clus-
phosphinidene-bridged ruthenium clusters display cata-
lytic activity and reaction intermediates were character-
ized [5].
ters catalyze the homogeneous hydrogenation of alky-
nes [1,2] and of linear and cyclic dienes [3]. We are
interested in the structure/reactivity relationships rele-
vant to the behaviour of these complexes. We found,
for example, that some clusters with alkynes coordi-
nated in parallel fashion with respect to a cluster edge
Here we report on the behaviour of the benzyne
clusters
Ru (CO) (v-PPh ) (C H )
(1)
and
3
7
2 2
6
4
Ru (CO) (v-PPh)(C H ) (2) [6] as homogeneous cata-
4
11
6
4
lysts for the hydrogenation of C Et , PhC Et, C Ph
2
2
2
2
2
and 1,4-CHD; these clusters are very good hydrogena-
tion catalysts, and the turnover numbers (TON) ob-
served are the highest reported for these types of
compound. The recovery of unreacted parent clusters,
*
Corresponding author. Fax: +39 11 6707855; e-mail:
sappa@ch.unito.it.
1
RC R%; R=R%=Ph, Et; R=Ph, R%=Et.
2
0
022-328X/98/$ - see front matter © 1998 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(98)00896-1

2
52
M. Castiglioni et al. / Journal of Organometallic Chemistry 571 (1998) 251–260
the observed ease of hydrogenation of the alkynes
C Et BPhC EtBC Ph ), together with the observed
yellow; the IR spectra of the clear solutions showed the
presence of about 80% of unreacted 1, small amounts
of the metallacyclic isomeric complexes Ru (CO) (v-
(
2
2
2
2
2
effect of dihydrogen pressure and catalyst concentra-
tion, agree with the concept of ‘cluster catalysis’.
We have also studied the behaviour of Ru (CO) (v-
3
5
PPh ) {C H (RC R%)} (5a,b) [10] and trace amounts of
2
2
6
4
2
3
7
1,3-CHD. Complex 1 was reacted with C Ph (sub-
2
2
PPh ) (HC Ph) (3) [7] and Ru (CO) (v -PPh)(C Ph )
strate/cluster ratio=1840.1) in a sealed 250 ml vial
2
2
2
4
11
4
2
2
(
4) [8], bearing an alkyne instead of a benzyne ligand.
under 1 atm H for 50 min (experiment A); a conver-
2
Cluster 3 is isostructural with complex 1, and the
structures of 4 and 2 are closely related. The structures
of complexes 1–4 are given in Fig. 1.
sion of 31.7 and a TON of 584.0 were observed. About
9
0% of the cluster could be recovered and, among the
organic products, about 1% of 1,3-CHD was observed.
The products from the reactions of 1,4-CHD could not
be identified. With benzene, a clear brown solution was
obtained containing unreacted 1 and an unidentified
product (IR bands at 2045 s, 2029 s, 1957 s, 1810 vs
2. Experimental
2.1. General experimental details and materials
−1
cm ).
Clusters 1, 2 [6], 3 [7] and 4 [8] were obtained by
2
.5. Reactions of complex 2 with alkynes
established procedures. The alkynes and 1,4-CHD were
commercial products and were used as received after
GLC purity tests. High-purity benzene was tested by
GLC and GC-MS. Gases (N , H ) were dehydrated and
No colour changes were observed and clear solutions
were always obtained. IR spectra showed the presence
of unreacted 2 and of very small amounts of the
complexes Ru (CO) (v -PPh)(C H )(RC R%) (6) [10].
In the reaction of 2 with C Ph (after 15 min) a whitish
precipitate was observed whose H-NMR and mass
2
2
solvents (n-octane, toluene, chloroform) were dried be-
fore use.
4
10
4
6
4
2
2
2
1
2
.2. Catalytic reactions of complexes 1–4 and
gas-chromatographic identification of the organic
products
spectra were consistent with a formula H C Ph (x=1,
x
2
3
2
; tentative attribution triphenyl-ethene or ethane) [11].
The reactions were performed, as previously de-
scribed [1], in 25 ml sealed vials containing 2.0 ml
n-octane solution of the catalyst and of the substrate
under 1 atm pressure of H , if not otherwise specified.
2
After warming at 120°C for the required reaction time,
the solutions were analyzed with a Carlo Erba 4200 gas
chromatograph operated under the following condi-
tions: n-octane/Porasil C (100/120 mesh) column, N
2
−
1
−1
flow 25 ml min , 60°C (6 min) then 3°C min
up to
1
35°C for C Et and 1,4-CHD; SE 30 5% on Chromo-
2
2
sorb WAW (60/80 mesh) column, N flow 46 ml
2
−
1
−1
min , 60°C (6 min) then 10°C min
up to 240°C for
EtC Ph and C Ph . Column dimensions 2 m×0.6 mm
2
2
2
(
i.d.).
2.3. Identification of organic and organometallic
products
The reaction solutions were examined by IR spec-
troscopy (Brucker Equinox 55, KBr cells) and subjected
to TLC (Kieselgel PF Merck; eluants: mixtures of light
petroleum and diethyl ether in variable v/v proportions
according to the nature of the products). The products
were analyzed also with a quadrupolar Finnigan Mat
TSQ-700 mass spectrometer [9].
2.4. Reactions of cluster 1 with alkynes
Fig. 1. Structures of complexes 1–4 as determined by X-ray diffrac-
Reaction mixtures changed from purple to orange-
tion (Refs. [6–8]).

M. Castiglioni et al. / Journal of Organometallic Chemistry 571 (1998) 251–260
253
Table 1
Hydrogenation of C Et , PhC Et, C Ph in the presence of clusters 1–4a
2
2
2
2
2
Cluster and experiment
Reaction time (min)
Conversion
TON
Selectivity to
C H R
cis-C H R
trans-C H R
2 2 2
2
4
2
2
2
2
Hex-3-yne(C Et )
2
2
1
2
3
4
[C]
[D]
[E]
[F]
30
0.2
2.1
4.9
0.3
0.3
0.6
0.2
0.5
0.2
0.2
0.3
0.5
1.6
15.4
35.3
1.4
1.5
2.4
4.1
9.4
4.3
1.4
2.1
3.5
17.4
2.8
2.6
–
–
–
–
21.8
–
–
–
–
82.6
90.7
90.7
45.4
50.0
75.0
83.3
74.5
88.0
100.0
100.0
100.0
4
6
5
0
6.5
6.7
15
54.6
50.0
25.0
16.7
3.6
12.0
–
3
4
0
5
15
3
4
0
5
15
3
4
0
5
–
–
–
1
-Phenyl-1-butyne (EtC Ph)
2
1
[G]
[H]
[I]
15
6.1
10.6
11.7
0.2
0.8
1.3
0.2
0.3
1.5
0.2
0.5
1.9
27.3
47.1
51.9
0.8
2.9
4.5
2.6
3.7
20.8
3.9
8.0
–
–
–
–
–
–
–
–
–
–
–
–
4.4
4.6
4.4
–
–
–
42.1
33.3
4.6
–
–
–
95.6
95.4
95.6
100.0
100.0
100.0
57.9
66.7
95.4
100.0
100.0
100.0
3
4
0
5
2
3
4
15
3
4
0
5
15
3
4
0
5
[J]
15
3
4
0
5
32.1
Diphenylacetylene (C Ph )
2
2
1
2
3
4
[K]
15
4.5
9.2
31.7
3.3
83.0
169.5
584.0
74.5
161.4
235.2
57.8
90.6
188.1
12.7
18.4
4.1
1.3
7.5
3.8
2.4
30.5
4.3
2.0
2.9
6.4
2.5
11.3
51.1
57.3
39.8
59.9
59.5
35.4
49.7
47.5
12.5
34.0
57.5
70.3
44.7
41.5
52.7
36.3
38.1
34.1
46.0
50.5
79.6
59.6
40.0
3
5
5
0
[L]
15
3
5
5
0
7.2
10.5
10.3
16.2
33.6
5.3
[M]
[N]
15
3
4
0
5
15
3
4
0
5
7.2
13.0
17.2
30.9
a
Always t=120°C, H =1 atm. Substrate/cluster (S/C): [C]=715.4; [D]=430.3; [E]=1708.7; [F]=755.4; [G]=444.8; [H]=344.7; [I]=1368.9;
2
[
J]=1696.7; [R]=1840.1; [L]=2244.5; [M]=559.2; [N]=238.2.
Complex 2 was reacted with C Ph in a sealed 250 ml
2.7. Reactions of complex 4 with alkynes
2
2
vial as previously described for 1 (experiment B); about
0% of the cluster was recovered, but no 1,3-CHD
9
Solutions turned from clear orange to pink or red
suspensions containing mostly decomposition. TLC
analysis of the solutions from the reactions with C Ph
showed the presence of some unreacted 4 and of
alkynic oligomers (mass spectrometry). With 1,4-CHD
no change in colour was observed and only unreacted 4
was recovered.
could be detected. With 1,4-CHD a clear orange solu-
tion was obtained, shown by IR to contain 2 and
additional bands at 1926 m–w, 1893 m and 1872 m
2
2
−
1
cm . With benzene IR bands at 1957 vs and 1811 vs
−
1
cm
were observed.
2.6. Reactions of complex 3 with alkynes
2.8. Hydrogenation reactions with clusters 5 and 6
Solutions turned from deep red to pale yellow. No
organometallic products could be detected.
The diphenylacetylene clusters 5 and 6 were prepared

2
54
M. Castiglioni et al. / Journal of Organometallic Chemistry 571 (1998) 251–260
Table 2
Effect of CO, of dihydrogen pressure and of substrate/cluster ratios on the hydrogenation of diphenylacetylene
a
Cluster and experiment
S/C
Conversion
TON
Selectivity to
C Ph H
cis-SB
trans-SB
2
2
4
1
2
1
3
1
1
2
2
[O]
[P]
[Q]
[R]
[S]
197.2b
179.7b
340.9e,f
344.8e,f
177.0c
10.1
20.0
9.1
5.1
–
–
–
37.8
40.4
61.2
62.2
59.6
38.8
4
2
.6
.6
7.1
12.7
6.1
3.3
–
–
–
27.4
23.1
70.8
72.6
76.8
29.2
3
1
.4
.8
7.3
24.8
17.6
29.7
52.8
33.3
36.0
29.8
31.6
20.7
17.3
35.1
43.3
5
8
.2
.7
9.0
30.9
18.1
20.5
27.9
24.8
51.1
17.4
18.7
25.2
54.7
56.5
23.7
5
5
.2
.9
5.6
3.5
1.9
10.0
9.4
10.0
22.2
17.9
26.6
25.2
34.1
23.4
52.6
48.0
50.0
2
5
65.9
29.2
[T]
[U]
[V]
132.6d
9.7
7.8
17.9
12.8
5.1
5.9
41.0
38.4
79.0
32.9
24.7
21.0
26.1
36.9
–
6
3
6.2
3.1
169.2c
6.2
9.2
14.7
10.6
23.3
73.7
47.6
35.1
30.2
30.4
34.9
34.6
21.9
30.0
35.1
2
5
54.3
01.0
254.3d
25.6
0.4
3.7
5.9
18.7
9.5
7.4
13.2
15.0
28.6
27.0
16.4
28.1
32.4
68.5
43.3
40.6
1
7
a
Always t=120°C and 30 min reaction time.
b
c
d
e
f
Decreasing H2 pressure: 1.0, 0.50, 0.25 atm.
Decreasing amount of cluster, keeping constant alkyne concentration, H =1 atm.
Decreasing amount of alkyne, keeping constant cluster concentration, H =1 atm.
2
2
Using 1 atm of a H /CO mixture (0.5/0.5 atm).
2
Increasing reaction times (15, 30, 45 min).
following established methods [10] and were reacted
with C Ph under H as reported above for clusters
for 2 than for 1. Cluster 3 is less active than 1 and 2 but
the trend observed for the three alkynes is the same.
Finally, cluster 4 shows a modest activity (TON=3.5,
4.5 and 30.9, respectively). The lower activity of 4 is
predictable when considering its synthetic route and its
role as by-product in the alkyne hydrogenation reac-
tions with Ru (CO) (v-PPh) [8]. For all the clusters the
2
2
2
1
–4. After the reactions only unreacted clusters 5 or 6
could be recovered.
3. Results and discussion
4
13
selectivity towards trans-derivatives is higher (and in-
3
.1. Hydrogenation of C Ph , PhC Et and C Et
2
2
2
2
2
creases with time) for C Et and PhC Et; in contrast,
2
2
2
C Ph gives higher selectivity for the cis-isomer, which
again increases with time.
When the dihydrogen pressure is decreased, the hy-
drogenation activity of clusters 1 and 2 drops sharply
2
2
The results of these experiments are given in Table 1.
Table 2 shows the effects of CO, of dihydrogen pressure
and of the substrate/cluster ratio on the hydrogenation
of C Ph .
2
2
(
Table 2). This indicates that the reactions are depen-
Clusters 1 and 2 show modest hydrogenation activity
dent on dihydrogen. One should bear in mind that,
when Ru (CO) (PPh ) was used as a hydrogenation
towards C Et (TON 35.3 and 2.4, respectively),
2
2
3
9
3 3
medium activity towards PhC Et (TON 51.9 and 4.5)
2
catalyst for cyclic olefins [12], the hydrido-carbonyls
H Ru (CO) (PPh ) were formed in the reaction mix-
and high activity towards C Ph (TON 584.0 and 235.2)
2
2
(
Table 1). The selectivity towards monoenes is higher
4
4
12–x
3 x

M. Castiglioni et al. / Journal of Organometallic Chemistry 571 (1998) 251–260
255
Fig. 2. Structures of complexes 5 (isomers), 6, 7 and 8 as determined by X-ray diffraction (Refs. [10,14,19]).
tures; increasing the dihydrogen pressure had no effect
on the yields of hydrogenated products, but led to the
preferential formation of H Ru (CO) (PPh ) . By rising
[14]. The formation of these kind of complexes could
hence modify the hydrogenation activity, as observed
for 1. The formation of these complexes under CO, and
of complexes 5 and 6 with alkynes, shows once more
the role of phosphido and phosphinidene bridges in
keeping together metal fragments.
4
4
8
3 4
the temperature complex 1 was obtained [13]; it was
considered a catalytic hydrogenation intermediate. Fi-
nally, pyrolysis of Ru (CO) (PPh ) yielded the binu-
3
9
3 3
clear Ru (CO) (PPh )(v-PPh )(v-OꢀCPh); this complex
In order to evaluate the effect of the cluster and
substrate concentration the following procedure was
adopted: for each cluster two experiments were per-
formed, one in which the cluster concentration was held
constant and that of the substrate C Ph was varied
(experiments T, V), and one in which the cluster con-
centration was varied and that of the substrate held
constant (experiments S, U). With cluster 1, decreasing
the cluster concentration, and therefore increasing the
substrate/cluster ratio from 177 to 529, results in nearly
constant TON values and formation of a considerable
amount of fully hydrogenated products. However,
when the cluster concentration was held constant (with
the consequent decrease of the substrate/cluster ratio
from 133 to 33), the TON decreased and the amount of
2
5
3
2
acted as a catalyst precursor.
The hydrogenation of C Ph with clusters 1 and 3
2
2
using a mixture of H /CO (0.5/0.5 atm) was also at-
2
tempted (experiments Q, R; Table 2); the presence of
CO lowers the TON. Worthy of note is the high
amount of fully hydrogenated products obtained in
these reactions; it is also interesting to observe that the
selectivities to fully and partially hydrogenated prod-
ucts shown by 1 and 3 follow different trends. This
behaviour can be explained by considering that cluster
2
2
1
reacts with CO under mild conditions forming com-
plex 7 with a CO inserted into a Ru-benzyne bond and
that cluster 2 also reacts with CO, albeit under more
drastic conditions, to give complex 8 (Fig. 2, below)

2
56
M. Castiglioni et al. / Journal of Organometallic Chemistry 571 (1998) 251–260
Table 3
Hydrogenation of diphenylacetylene on clusters 5 and 6
a
Cluster and experiment
Reaction time (min)
Conversion
TON
Selectivity to
C H R
cis-SB
trans-SB
2
4
2
5
6
[X]
[Y]
15
13.8
25.9
47.4
1.1
2.1
3.8
36.7
11.0
6.8
28.3
52.5
51.2
35.0
36.4
42.0
3
4
0
5
15
6.6
7.9
4.5
1.8
2.2
1.3
23.7
36.9
33.3
38.3
32.5
16.3
38.0
30.6
50.4
3
4
0
5
a
Always t=120°C, H =1 atm. [X] S/C=8.1; [Y] S/C=28.1.
2
fully hydrogenated product increased with time. With
cluster 2 decrease of cluster concentration (substrate/
cluster ratio increasing from 169 to 501) results in an
increase of the TON and a decrease of fully hydro-
genated products during time; a decrease of the sub-
strate concentration (substrate/cluster from 254 to 70)
results in nearly constant TON values. These results
indicate that the reaction rate is dependent on the
substrate concentration rather than on that of the
cluster.
3.3. Comments on the catalytic acti6ity of the
benzyne complexes 1 and 2
A comparison of the catalytic activities of complexes
1–4 with those of other ruthenium clusters (electroni-
cally unsaturated, or bearing parallel alkynes) is given
in Table 5. The activity of complexes 1 and 2 is
remarkable; to our knowledge it is the highest reported
for ruthenium clusters.
While studying the homogeneous hydrogenation of
Finally, the data collected in Table 3 indicate that
clusters 5 and 6 (which were found in small amounts in
the reaction mixtures of 1 and 2) show only very
modest hydrogenation activity and therefore should be
considered inactive by-products and not intermediates
in these reactions.
1
,4- and 1,3-CHD using Ru (CO) , H Ru (CO) ,
3 12 4 4 12
H Ru (CO) and related compounds [2] we isolated
2
4
13
from the reaction mixtures complexes 9–12, substituted
with facial benzene or ‘parallel’ cyclohexadiene ligands,
shown in Fig. 3.
Complexes 9–12 were inactive in the hydrogenation
of 1,4-CHD. This was explained by considering that
these derivatives were presumably formed, in an early
reaction step, by oxidative addition of 1,3- or 1,4-CHD
to the catalysts and that in the presence of external H2
they could further react to release the parent substrate.
Clusters 1 and 2, formed upon oxidative addition and
P–C bond cleavage of PPh on Ru (CO) [16], have
been considered as models for the dissociative
chemisorption of benzene on metal surfaces [6]. They
contain ‘parallel’ benzyne ligands, which can be readily
compared with ‘parallel’ alkyne ligands [1,4,17]. It was
previously observed that clusters with parallel alkynes
can act as active intermediates in the homogeneous
hydrogenation of alkynes [1,4].
Complexes 5 and 6 are obtained upon reaction of 1
or 2, respectively, with alkynes under stoichiometric
conditions [10]. The isomeric complexes 5 contain a
metallacyclic ring formed by the linking of C Ph with
2
2
the coordinated benzyne on 1, while complexes 6 con-
tain a C R ligand coordinated in independent fashion.
2
2
The structures of 5 and 6, and those of 7 and 8 are
given in Fig. 2.
3
3
12
3.2. Hydrogenation-isomerization of 1,4-CHD and
attempted hydrogenation of benzene
The results of these experiments are compiled in
Table 4 [15].
Clusters 1 and 4 are very active in the hydrogenation
of 1,4-CHD; however, the selectivity towards cyclohex-
ene is only 50–60%. Cluster 3 shows very high activity
but, once again, the selectivity towards cyclohexene is
relatively low. Cluster 2 shows low activity but better
selectivity for cyclohexene. The considerable amounts
of 1,3-CHD formed indicate that isomerization occurs
in competition with (or prior to) hydrogenation. It is
noteworthy that the alkyne clusters 3 and 4 are more
active than the benzyne clusters 1 and 2 in these
reactions, opposite to the behaviour observed for the
hydrogenation of alkynes.
3.4. Possible reaction pathways
The results obtained indicate that clusters 1 and 2 are
active catalysts. From the reactions of 1 and 2 with
alkynes under dihydrogen large amounts of unreacted
parent clusters were recovered and very small yields of
complexes 5 or 6 isolated, respectively. However, no
organometallic intermediates or by-products could be
isolated, which were helpful for establishing unequivo-
cally a reaction pathway.

M. Castiglioni et al. / Journal of Organometallic Chemistry 571 (1998) 251–260
257
Table 4
Hydrogenation of 1,4-CHD on clusters 1–4
a
Cluster and experiment
Reaction time (min)
Conversion
TON
Selectivity to
cis-hexene
1,3-CHD
1
2
3
4
[W]
[Z]
15
0.4
1.7
7.2
3.3
15.1
62.1
60.5
61.5
50.2
39.5
38.5
49.8
3
4
0
5
15
0.6
0.7
0.4
8.0
9.2
4.6
34.9
34.7
27.8
65.1
65.3
72.2
3
4
0
5
[ZA]
[ZB]
15
2.3
4.7
23.9
71.1
146.2
742.0
41.3
28.1
36.6
53.5b
71.9
63.4
3
4
0
5
15
0.3
0.6
3.7
7.7
15.1
95.9
50.0
35.6
39.0
50.0
64.4
61.0
3
4
0
5
a
Always t=120°C, H =1 atm. [W] S/C=865.8; [Z] S/C=1275.4; [ZA] S/C=3091.4; [ZB] S/C=2563.2.
Selectivity to cyclohexane 5.2%.
2
b
Clusters 5 and 6 show very modest hydrogenation
This could explain why 1 and 3 (three M–C bonds) are
usually more active than 2 and 4 (four bonds). Finally,
the observed (preferential) formation of cis-stilbene
when C Ph is hydrogenated on clusters 1–4 is also
activity (Table 3). This indicates that a mechanism
involving addition or substitution of the alkynes in
complexes 1 and 2 as the first reaction step, followed by
uptake of hydrogen, should be disregarded. However,
we hypothesize that cluster catalysis occurs on the basis
of the following direct or indirect pieces of evidence. It
is important to note that phosphido or phosphinidene
bridges such as those found in clusters 1–4 are thought
to prevent the fragmentation of clusters under CO or
2
2
consistent with cluster catalysis [2].
Another interesting aspect of the above reactions is
the behaviour of the different alkyne substrates. It can
be seen that, for all of the clusters 1–4, the ease of
hydrogenation of the alkynes is:
H , thus favouring cluster catalysis. These bridges
2
C Et BPhC EtBC Ph
2
2
2
2
2
would also allow reversible M–M bond cleavage [18]
and hence formation of active vacant sites. Alterna-
tively these could be generated by release of a coordi-
nated carbonyl. However, the reaction of CO with
cluster 1, discussed above, leads to complex 7 which, in
the presence of dihydrogen, forms the known
H Ru (CO) (PPh ) and releases benzaldehyde [19].
This could indicate that a further step of the reaction is
the coordination of the alkyne and its insertion into an
M–H bond; hydrogenation would therefore occur
faster if the entering alkyne coordinated easily onto the
cluster. In previous work we have observed that C Ph
2
3
8
2 2
2
2
This could explain the lower TON observed in the
presence of CO.
The occurrence of cluster catalysis is also supported
by the recovery of very large amounts of clusters 1 and
is hydrogenated faster than C Et ([4]b, [22]) and this
2 2
was explained by the better coordinating properties of
the former. On the other hand, experiments A and B
indicate that there is no alkyne–benzyne exchange on 1
(or 2) during hydrogenation.
2
after the hydrogenation experiments. The observed
dependence of the catalytic activity on the cluster and
substrate concentration also suggests that cluster catal-
ysis occurs [20].
It is also worth noting that the ease of hydrogena-
tion of 1,4-CHD on clusters 1 and 2 is: C Et B
2
2
PhC EtB1,4-CHDBC Ph , whereas on clusters 3 and
2
2
2
The dependence of the catalytic activity on dihydro-
gen pressure indicates that uptake and activation of
dihydrogen is one of the rate-determining steps, per-
haps the first reaction step. Addition of dihydrogen
would also result in the transformation of the coordi-
nated ligand (benzyne, alkyne) into a vinyl [2,21]; the
ease of this process would depend on the number of
M–C bonds existing between the ligand and the cluster.
4 the order becomes C Et BPhC EtBC Ph B1,4-
2
2
2
2
2
CHD (Table 4). Thus, the alkyne clusters are more
active in these reactions; the reverse behaviour is ob-
served in the hydrogenation of C Ph , where 1 and 2
2
2
are the most active. Once again this can be explained by
considering the relative ease of coordination of 1,4-
CHD, compared with C Ph on the benzyne or alkyne
2
2
clusters.

2
58
M. Castiglioni et al. / Journal of Organometallic Chemistry 571 (1998) 251–260
Benzene is not hydrogenated under the conditions
employed, presumably because of its lower coordinat-
ing ability.
Finally a non-negligible aspect of these reactions is
the release of small amounts of 1,3-CHD when cluster
1
is reacted with alkynes and the formation of
triphenyl-ethene (or ethane) from cluster 2 when re-
acted with C Ph . This is consistent with the proposed
2
2
mechanism; the formation and release of 1,3-CHD
could occur at the expense of the coordinated benzyne
after uptake of dihydrogen and in competition with the
hydrogenation of the alkynes. Triphenyl-ethene (or
ethane) could be formed by coupling of the coordinated
Table 5
Comparison of the activity of clusters 1–4 with that of selected
ruthenium-containing clusters
Cluster
S/C ratioa TOFb Selectivity to Ref.
Fig. 3. Structures of complexes 9–12 (see Ref. [3]).
monoenes
Hydrogenation of diphenylacetylene
Ru (CO)
49.4
60.3
94.3
108.0
100.0
24.8
194.6
274.5
559.2
238.2
97.4
117.3
121.0
90.1
100
100
[1]
[1]
3
12
H Ru (CO)
4
4
12
Cp Ru (CO) (C Ph )
98.5
100
99.3
100
98.7
97.6
81.1
97.5
[4]b
[5]b
[5]b
[5]c
t.w.
t.w.
t.w.
t.w.
benzyne with the incoming C Ph in the presence of
2 2
2
3
5
2
2
HRu (CO) (PPh )
dihydrogen; alternatively it could be formed by cou-
pling of a coordinated vinyl with C Ph .
3
7
2 3
HRu (CO) (PPh )
99.1
3
9
2
2
2
Ru (CO)13(PPh)
131.8
700.8
282.3
250.8
41.2
4
Although fragment catalysis cannot be completely
ruled out, the above evidence points to a ‘classic’
reaction pathway based on cluster catalysis requiring
formation of an active site, dihydrogen activation and
insertion of the substrate into a metal–hydrogen bond
[23]; this is tentatively shown in Scheme 1 for trinuclear
cluster percursors.
An alternative mechanism involving cluster catalysis
could also be envisaged; this would be represented by
the insertion of alkynes into M–P bonds ([4]b, [18]) as
the rate determining step, followed by uptake of dihy-
drogen and release of the hydrogenated products
through ‘vinylphosphinic’ intermediates [24]. However,
there is no evidence for this, more unlikely, mechanism.
The reactions reported in this paper represent a small
but interesting step in the development of structure/re-
activity relationships applied to homogeneous catalysis.
The high activity and the comparable behaviour ob-
served for the isostructural clusters 1 and 3 are indeed
a good example of a structure/reactivity relationship.
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Hydrogenation of 1-phenyl-1-butyne
Cluster 1
Cluster 2
Cluster 3
Cluster 4
444.8
344.7
1369.0
169.7
69.1
6.0
27.7
42.8
100
t.w.
t.w.
t.w.
t.w.
100
100
100
Hydrogenation of hex-3-yne
Ru (CO) 241.1
58.1
9.2
22.4
4.1
35.3
3.2
18.8
4.6
100
100
95.3
80.1
97.2
100
78.2
100
[21]
[4]b
[5]b
[5]b
t.w.
t.w.
t.w.
t.w.
3
12
Cp Ru (CO) (C Ph )
198.4
230.0
342.0
71S.4
430.3
559.2
755.4
2
3
5
2
2
HRu (CO) (PPh )
3
9
2
HRu (CO) (PPh )
3
7
2 3
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Hydrogenation of 1,4-CHD
Ru (CO) 429.4
123.5
48.9
229.5
82.8
6.1
989.3
128.8
32.6
41.3
83.7
50.2
27.8
36.6
[3]
[3]
3
12
H Ru (CO)
366.0
480.8
865.8
127.5
309.1
256.3
4
4
12
Cp Ru (CO) (C Ph )
[21]
t.w.
t.w.
t.w.
t.w.
2
3
5
2
2
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Acknowledgements
100
a
Financial support for this work has been obtained
from MURST (Rome). We thank Professor A.J. Carty
S/C=substrate/cluster.
Maximum TOF observed.
b
.

M. Castiglioni et al. / Journal of Organometallic Chemistry 571 (1998) 251–260
259
Scheme 1.
for helpful discussions. We are indebted to Drs F.
References
Van Gastel and J.F. Corrigan for a loan of com-
plexes 3 and 4. Fellowship (to P.J.K.) under the Eu-
ropean Scientific Exchange Programme (Royal Society
of Chemistry-Accademia dei Lincei) is also acknowl-
[
[
1] R. Giordano, E. Sappa, J. Organomet. Chem. 448 (1993) 157.
2] (a) R.D. Adams, T.S. Barnard, Z. Li, W. Wu, J. Yamamoto,
Organometallics 13 (1994) 2357. (b) R.D. Adams, T.S. Barnard,
Z. Li, W. Wu, J.H. Yamamoto, J. Am. Chem. Soc. 116 (1994)
9103.
edged.
.

2
60
M. Castiglioni et al. / Journal of Organometallic Chemistry 571 (1998) 251–260
[
[
3] M. Castiglioni, R. Giordano, J. Organomet. Chem. 491 (1995)
11.
4] (a) D. Cauzzi, R. Giordano, E. Sappa, A. Tiripicchio, M.
Tiripicchio Camellini, J. Cluster Sci. 4 (1993) 279. (b) R. Gior-
dano, E. Sappa, S.A.R. Knox, J. Cluster Sci. 7 (1996) 179.
5] (a) M. Castiglioni, R. Giordano, E. Sappa, J. Organomet. Chem.
Chem. Soc., Chem. Commun. (1991) 1393.
1
[15] Hydrogenation of benzene at 120°C and 1 atm of H2 was
attempted with clusters 1 (S/C 579.9) and 2 (S/C 1347.3) for
reaction times ranging from 30 to 90 min. No reaction was
observed. These data are therefore not in Table 4.
16] Together with metal fragment condensation.
17] A survey of clusters with ‘parallel’ alkynes is in: S. Deabate, R.
Giordano, E. Sappa, J. Cluster Sci. 8 (1997) 407.
[
[
[
[
[
[
3
62 (1989) 399. (b) Idem, Ibidem, 369 (1989) 419. (c) Idem,
Ibidem, 407 (1991) 377.
6] S.A.R. Knox, B.R. Lloyd, D.A.V. Morton, S.M. Nicholls, A.G.
Orpen, J.M. Vinas, M. Weber, K.G. Williams, J. Organomet.
Chem. 394 (1990) 385.
7] F. Van Gastel, S.A. MacLaughlin, M. Lynch, A.J. Carty, E.
Sappa, A. Tiripicchio, M. Tiripicchio Camellini, J. Organomet.
[
[
18] See for example: (a) G. Huttner, K. Knoll, Angew. Chem. Int.
Ed. Engl.26 (1987) 743. (b) A.J. Carty, Pure and Appl. Chem. 54
(
1982) 133.
19] J.P.H. Charmant, H.A.A. Dickson, N.J. Grist, S.A.R. Knox,
A.G. Orpen, K. Saynor, J.M. Vinas, J. Organomet. Chem. 565
Chem. 326 (1987) C65.
8] Cluster 4 is obtained by reacting the electron-poor (62 e
(
1998) 141.
−
)
[
[
20] R.M. Laine, J. Mol. Catal. 14 (1982) 137.
cluster Ru (CO) (v-PPh) with C Ph under either stoichiometric
4
13
2
2
21] See for example: (a) P. Briard, J.A. Cabeza, A. Llamazares, L.
Ouahab, V. Riera, Organometallics 12 (1993) 1006. (b) J.A.
Cabeza, S. Garcia-Granda, A. Llamazares, V. Riera, J.F. Van
der Maelen, Organometallics 12 (1993) 2973.
22] D. Campagnola, M. Castiglioni, W. Dastr u´ , S. Deabate, R.
Giordano, P.J. King, Inorg. Chim. Acta 262 (1997) 157.
23] After submission of this paper, hydrosilylation of diarylalkynes
was reported and a reaction mechamism comparable with that of
or homogeneous catalytic conditions. Stoichiometric conditions:
a) J. Lunniss, S.A. MacLaughlin, N.J. Taylor, A.J. Carty, E.
(
Sappa, Organometallics 4 (1985) 2066. (b) J.F. Corrigan, S.
Doherty, A.J. Carty, Organometallics 12 (1993) 1365. Catalytic
conditions, see Ref. [5]b.
[
[
[
9] Servizio di Spettrometria di Massa, Dipartimento di Scienza e
Tecnologia del Farmaco, Universit a´ di Torino.
[
[
10] J.P.H. Charmant, S.A.R. Knox, unpublished results.
11] 1H-NMR: 7.46–7.18 mm (15H, Ph), 6.88 s (1H), 6.27 s (1H,
Scheme
1
was proposed; R.D. Adams, T.S. Barnard,
Organometallics 17 (1998) 2567.
C–H) ppm. Mass spectrum: sharp organic signal at 237 m/z.
12] A. Basu, S. Bhaduri, H. Khwaja, P.G. Jones, T. Shroeder, G.M.
Sheldrick, J. Organomet. Chem. 290 (1985) C19.
13] M.I. Bruce, J.M. Guss, R. Mason, B.W. Skelton, A.H. White, J.
Organomet. Chem. 251 (1983) 261.
[
24] (a) R. Gobetto, E. Sappa, A. Tiripicchio, M. Tiripicchio
Camellini, M.J. Mays, J. Chem. Soc., Dalton Trans (1990) 807.
(b) D. Belletti, A. Tiripicchio, M. Tiripicchio Camellini, E.
Sappa, J. Chem. Soc., Dalton Trans. (1990) 703. (c) R. Gior-
dano, E. Sappa, G. Predieri, A. Tiripicchio, J. Organomet.
Chem. 547 (1997) 49.
[
[
[
14] J.P.H. Charmant, H.A.A. Dickson, N.J. Grist, J.B. Keister,
S.A.R. Knox, D.A.V. Morton, A.G. Orpen and J.M. Vinas, J.
.
.