Activation and transfer of O2 to organic electrophiles by [RuH(dcpe)2]+

Article abstract of DOI:10.1039/a808860h

The crystal structure of the 16-electron hydride [RuH(dcpe)₂]PF₆ (1·PF₆) [dcpe = 1,2-bis(dicyclohexylphosphino)ethane] shows a short contact between Ru and a cyclohexyl CH₂, which was not observed in the disordered tetraphenylborate salt [RuH(dcpe)₂]BPh₄ (1·BPh₄). The comparison of the X-ray structure of 1·PF₆ with those of [RuCl(dcpe)₂]PF₆ (2) and trans- [RuCl₂(dcpe)₂] (3) indicates that electronic effects co-operate with the steric ones in the stabilization of these 16-electron species. The hydrido dioxygen complex [RuH(η²-O₂)(dcpe)₂]⁺ (4), formed by reaction of 1 with O₂, does not react with nucleophiles such as olefins and PPh₃, but does show nucleophilic properties, reacting with activated olefins (TCNE, 2- methyl-1,4-naphthoquinone) and with carbonyl compounds [heptanal, PhC(O)Cl]. The reaction of 4 with heptanal shows an unprecedented reaction mode for a dioxygen complex of a late transition metal, as the O₂-activating complex 1 is partially regenerated upon oxygen transfer from 4 to the substrate. All these reactions show that oxygen transfer from 4 to an organic substrate effectively competes with oxidation of the dcpe ligand.

Full text of DOI:10.1039/a808860h

View Article Online / Journal Homepage / Table of Contents for this issue
Activation and transfer of O to organic electrophiles by
2
[RuH(dcpe) ]‘¤
2
Arianna Martelletti,a Volker Gramlich,b Fabio Zurchera and Antonio Mezzetti*,a
a L aboratorium f ur Anorganische Chemie and b Institut f ur Kristallographie und Petrographie,
ET H-Zentrum, CH-8092 Zurich, Switzerland. E-mail: Mezzetti=inorg.chem.ethz.ch
Received (in Montpellier France) 1st July 1998, Revised m/s received 12th November 1998,
Accepted 3rd December 1998
The crystal structure of the 16-electron hydride [RuH(dcpe) ]PF (1 É PF ) [dcpe \ 1,2-bis(dicyclohexylphosphino)-
2
6
6
ethane] shows a short contact between Ru and a cyclohexyl CH , which was not observed in the disordered
2
tetraphenylborate salt [RuH(dcpe) ]BPh (1 É BPh ). The comparison of the X-ray structure of 1 É PF with those of
[RuCl(dcpe) ]PF (2) and trans-[RuCl (dcpe) ] (3) indicates that electronic e†ects co-operate with the steric ones in
2
4
4
6
2
6
2
2
the stabilization of these 16-electron species. The hydrido dioxygen complex [RuH(g2-O )(dcpe) ]` (4), formed by
2
2
reaction of 1 with O , does not react with nucleophiles such as oleÐns and PPh , but does show nucleophilic
2
3
properties, reacting with activated oleÐns (TCNE, 2-methyl-1,4-naphthoquinone) and with carbonyl compounds
[heptanal, PhC(O)Cl]. The reaction of 4 with heptanal shows an unprecedented reaction mode for a dioxygen
complex of a late transition metal, as the O -activating complex 1 is partially regenerated upon oxygen transfer
2
from 4 to the substrate. All these reactions show that oxygen transfer from 4 to an organic substrate e†ectively
competes with oxidation of the dcpe ligand.
The activation of dioxygen at a transition metal complex is
one of the dream reactions that have been pursued during the
last 30 years.1 In spite of much e†ort spent, progress in the
Ðeld has been very slow, and, despite important achievements,
the nature of the metalÈdioxygen interaction remains gener-
ally elusive. This was also the case in the recent discovery that
ordinate hydride [RuH(dcpe) ]PF (1 É PF ) is possibly stabil-
2
6
6
ized by a weak agostic interaction. This structural feature was
obscured by disorder in a recent X-ray study of the tetra-
phenylborate analogue 1 É BPh .12
4
Although a number of d6 dioxygen complexes have been
recently prepared,3,8 their reactivity has not yet been investi-
gated. Also, the reaction of the Ðve-co-ordinate species
[MX(dcpe) ]` (M \ Ru and Os) with O is a suitable starting
ruthenium-substituted
polyoxometallates
catalyse
the
hydroxylation of adamantane with dioxygenase-like activity.2
2
2
The 30-year-old discovery of O activation by phosphino
point for understanding the role played by the metal centre in
the activation of dioxygen into a number of aerobic oxidation
2
complexes of the late transition metals has led to the prep-
aration of a large number of dioxygen complexes.1 However,
attempts to use these dioxygen complexes in metal-catalysed
oxidation reactions of organic substrates have remained
largely unsuccessful with long periods of oblivion followed by
recurring bursts of interest.
reactions
catalysed
by
phosphine
complexes
of
ruthenium(II).13 Five-co-ordinate ruthenium(II) species such as
[RuCl (PPh ) ] and [RuCl(dppp) ]` [dppp \ 1,3-bis(di-
2
3 3
2
phenylphosphino)propane] have been used in aerobic catalytic
oxidation reactions.13a,b Although these co-ordinatively
unsaturated complexes have been proposed to react with
molecular oxygen, the formation of peroxo adducts of these
complexes under catalysis conditions has not been conclu-
sively proven.1b
The
co-ordinatively
unsaturated
[PwP \ 1,2-bis(dicyclohexylphosphino)-
or 1,2-bis(diisopropylphosphino)ethane
d6
complexes
[RuH(PwP) ]`
ethane
2
(dcpe)
(dippe)]3a and [OsX(PwP) ]` [PwP \ dcpe3b or 1,2-bis(di-
2
phenylphosphino)ethane (dppe)3c] have been recently found to
More generally, an investigation of the reactivity of d6 RuII
and OsII dioxygen complexes is of interest in view of their
intermediate position between the ““electrophilicÏÏ dioxygen
complexes of the early transition elements and the
““nucleophilicÏÏ analogues of the late ones, and is of relevance
to the recent mechanistic debate14 about the catalytic co-
oxidation reactions developed primarily by Mukaiyama and
co-workers using a variety of substrates and reductants, such
as aldehydes, alcohols, and ketones.15 Thus, we have investi-
gated the reactivity of the dioxygen hydride [RuH(g2-
O )(dcpe) ]PF (4) towards electrophilic substrates such as
react with molecular oxygen to give dioxygen complexes4 of
the type [MX(g2-O )(PwP) ]`.3 The 16-electron species
2
2
[MX(PwP) ]` (M \ Ru, Os) are unique in their ability to
2
form both dioxygen and dihydrogen5,6 complexes. Dioxygen
complexes of d6 metals3,7,8 are less common than those of d8
and d10 metal ions, such as IrI and PtO complexes,1 but also
RuO and OsO.9 Parallel to the study of the reactivity of the
Ðve-co-ordinate fragments [MX(dcpe) ]` (M \ Ru,10a Os;10b
2
X \ Cl, H) with H 10c and, more recently, O ,3b theoretical
2
2
studies of 16-electron species have contributed substantially to
the understanding of the steric and electronic factors that
a†ect the stability of these 16-electron complexes.11 We report
herein a new X-ray investigation suggesting that the Ðve-co-
2
2
6
heptanal, PhC(O)Cl, tetracyanoethylene (TCNE), and 2-
methyl-1,4-naphthoquinone (menadione).
Results and discussion
Structural aspects of [RuH(dcpe) ]‘
¤ Supplementary material available: NMR spectra. For direct elec-
tronic access see http://www.rsc.org/suppdata/nj/1999/199/, otherwise
available from BLDSC (No. SUP 57473, 2 pp) or the RSC Library.
See Instructions for Authors, 1999, Issue 1 (http://www.rsc.org/njc).
2
Recently, Winter and Hornung have reported the X-ray struc-
ture of [RuH(dcpe) ]BPh (1 É BPh ), in which both crystallo-
2
4
4
New J. Chem., 1999, 199È206
199

View Article Online
graphically independent cations are located in a special
position and possess a crystallographic two-fold axis of sym-
value of 179.69(4)¡. The P and Ru atoms are within ^0.11 A
from the best RuP plane, and the RuwP(1)wP(2) and
4
metry lying in the RuP plane.12 Since square-pyramidal
RuwP(3)wP(4) planes form an angle of 8.79(2)¡. The dihedral
4
cations are not C -symmetric about any direction in the RuP
angle between these planes plays a major role: although the
conformation of both chelate rings is a nearly ideal twist, the
closing of the P(2)wRuwP(4) angle pushes C(1) and C(3)
2
4
plane due to the presence of the apical hydride, this arrange-
ment results in a disordered situation in which the hydride
ligand must be located with 50% occupancy on both sides
away, by 0.40 and 0.36 A, from the best RuP plane toward
4
of the RuP plane. With the idea that disorder might
the hydride. The bite angles are normal for Ðve-membered
4
conceal some structural feature, we prepared the [PF ]~
rings (average value is 82.6¡).
6
analogue [RuH(dcpe) ]PF (1 É PF ) by reaction of trans-
[RuHCl(dcpe) ] with TlPF . Selected interatomic distances
and angles are given in Table 1, and an ORTEP view is
shown in Fig. 1. In order to assess the role of steric and elec-
tronic e†ects in a series of RuIIÈdcpe complexes, we have also
determined the X-ray structure of trans-[RuCl (dcpe) ] (3)
The axial hydride ligand was not located: an electron
density peak about 1.0 A below Ru was found in the Ðnal *F
map, but could not be reÐned to a chemically reasonable
RuwH distance. However, the hydride ligand is most prob-
2
6
6
6
2
ably located below the RuP plane (Fig. 1), since the apical
4
position above this plane is involved in two short contacts
2
2
(Fig. 2).
between Ru and two cyclohexyl groups. The shortest RuÉ É ÉH
contact [2.59(5) A], which involves the cyclohexyl equatorial
b-hydrogen on C(34), suggests a weak c-agostic interaction
(see below).16 The next shortest RuÉ É ÉH contact [2.84(5) A]
involves C(6), which can replace C(34) in the interaction to Ru
with a minor rearrangement. The shortest RuÉ É ÉH distances
[RuH(dcpe) ]PF , 1 Æ PF . The unit cell of 1 É PF contains
2
6
6
6
the complex cation (in general positions) and two crystallo-
graphically independent [PF ]~ anions at 50% occupancy in
6
special positions. All non-hydride H atoms have been located
on the opposite side of the RuP plane involve the equatorial
H atoms H(22A) and H(42A) and are much longer (3.30 and
and reÐned. The major distortion from the ideal square-
pyramidal geometry is the displacement of two of the four P
4
3.32 A, respectively), in agreement with the presence of the
hydride ligand.
atoms away from the RuP plane. The P(2)wRuwP(4) angle
4
is closed down to 171.3¡, whereas P(1)wRuwP(3) has an ideal
Comparison to 1 É PF clearly indicates that the tetrahedral
6
distortion observed in 1 É BPh is an artifact of disorder. In
4
fact, the P(1)wRuwP(3) and P(2)wRuwP(4) angles of 1 É PF
6
are markedly di†erent [179.69(4)¡ and 171.26(4)¡,
respectively], whereas in 1 É BPh the crystallographic two-
4
fold
axis
bisecting
the
P(11)wRuwP(11a)
and
P(12)wRuwP(12a) angles forces the angles between the trans
P atoms to be equivalent. Accordingly, the average of the
trans PwRuwP angles in 1 É PF (175.5¡) is close to the value
6
of 174.4¡ observed in 1 É BPh . Also, the thermal ellipsoids of
4
P(11) and P(21) in 1 É BPh are elongated perpendicular to the
RuP plane, which is not physically reasonable and indicates
that the real atomic positions are split between two minima.
4
4
The same is true for C(11) and C(21). Again, the comparison
with 1 É PF is revealing. Since both C(1) and C(3) lie on the
6
same side of the RuP plane in 1 É PF , the conformation of
4
6
the two chelate rings is not compatible with any binary axis
lying in the RuP plane: thus, C(11) and C(12) in 1 É BPh are
4
4
disordered.
The disorder in 1 É BPh masks the most interesting struc-
4
tural feature of 1, namely the short RuÉ É ÉH non-bonded dis-
tances. In 1 É PF , the closest RuÉ É ÉH contacts are 2.59(5) and
6
2.84(5) A on one side of the RuP plane, and 3.30(5) and
4
Fig. 1 ORTEP view of the cation 1 in [RuH(dcpe) ]PF .
3.32(5) A on the opposite one. Also, the orientation of the
2
6
cyclohexyl groups is signiÐcantly di†erent on either side of the
RuP plane. In 1 É BPh , the pseudo-symmetry imposes aver-
4
4
Table
1 Selected interatomic distances (A) and angles (¡) in
aged distances (B3.0 A) between Ru and H atoms lying on
[RuH(dcpe) ]PF , 1 É PF
2
6
6
opposite sides of the RuP plane, which obscures the
4
RuÉ É ÉHwC close contact. However, the agostic interaction in
RuwP(1)
RuwP(2)
RuwP(3)
RuwP(4)
2.3484(10) C(34)wH(34B)
2.3710(11) RuÉ É ÉC(6)
2.3454(11) RuÉ É ÉC(34)
2.3746(11) RuÉ É ÉH(6B)
1.14(6)
3.511(5)
3.332(5)
2.84(5)
2.59(5)
1 É PF is, in the best of cases, very weak. Applying CrabtreeÏs
6
parametrization to the shortest contact gives an e†ective cova-
lent radius of the CwH bonding electrons r of 1.46 A, an
bp
C(6)wH(6B)
0.99(6)
RuÉ É ÉH(34B)
MwHwC angle H of 120¡, and a corrected RuÉ É ÉH(34B) of
P(1)wRuwP(2)
P(1)wRuwP(3)
P(1)wRuwP(4)
82.90(4)
179.69(4)
97.91(4)
RuwP(2)wC(2)
RuwP(2)wC(17)
RuwP(2)wC(23)
RuwP(3)wC(3)
RuwP(3)wC(29)
RuwP(3)wC(35)
RuwP(4)wC(4)
RuwP(4)wC(41)
RuwP(4)wC(47)
P(1)wC(5)wC(6)
P(1)wC(5)wC(10)
108.6(2)
2.54 A. This is suggestive of a weak c-agostic interaction
rather than a simple van der Waals contact.16b The presence
116.2(1)
125.2(2)
111.3(2)
110.0(1)
120.2(2)
108.9(1)
115.6(1)
125.9(1)
110.5(3)
118.6(3)
of an attractive interaction in 1 É PF is further indicated
P(1)wRuwH(34B) 110 (1)
6
by (i) the closing of the RuwP(3)wC(29) angle, (ii) the
P(2)wRuwP(3)
P(2)wRuwP(4)
P(2)wRuwH(34B)
P(3)wRuwP(4)
P(3)wRuwH(34B)
P(4)wRuwH(34B)
RuwP(1)wC(1)
RuwP(1)wC(5)
RuwP(1)wC(11)
96.81(4)
171.26(4)
89 (1)
82.40(4)
69 (1)
99 (1)
110.0(1)
114.1(1)
119.3(1)
shortening of the RuwP(3) distance and (iii) the tilting of
the C(29)wC(34) cyclohexyl group towards the Ru
atom [P(3)wC(29)wC(34) \ 108.8(3)¡, P(3)wC(29)wC(30) \
119.4(3)¡]. The X-ray structures of the 16-electron species
[W(CO) (PCy ) ] and [Re(CO) (PCy ) ]` reveal distortions
3
3 2
3
3 2
that are qualitatively similar, albeit much larger in magni-
tude.17 Being attracted toward the metal, the C(5)wC(10) and
C(29)wC(34) cyclohexyl groups push the C(23)wC(28) and
P(3)wC(29)wC(30) 119.4(3)
P(3)wC(29)wC(34) 108.8(3)
200
New J. Chem., 1999, 199È206

View Article Online
Table 2 Average bond lengths in 1, 2 and 3 (A)
[RuH(dcpe) ]`
(1)
[RuCl(dcpe) ]`
(2)
[RuCl (dcpe) ]
(3)
2
2
2
2
RuwP
2.360(1)
2.424(4) (ax)
2.294(3) (eq)
2.386(3)
2.424 (2)
2.435 (2)
RuwCl
cm~1 (CH Cl solution) can be tentatively attributed to the
2
2
stretching vibration of the agostic CwH bond. At room tem-
perature, the 31PM1HN NMR spectrum is a sharp singlet at d
73.6, which broadens upon cooling, indicating that a dynamic
process slows down on the NMR time scale. At [80 ¡C, two
broad signals are observed (Fig. 3), but no resolved 31PM1HN
spectrum is achieved down to the lowest available tem-
perature (ca. [90 ¡C). The 1H NMR spectra recorded
between 20 and [60 ¡C show a quintet for the hydride and
broad signals without unusual features for the methylene
protons. The quintet structure is lost below this temperature,
and a broad signal integrating as 1 H decoalesces from the
Fig. 2 ORTEP view of [RuCl (dcpe) ], 3. Selected bond lengths (A)
2
2
and angles (¡) are: RuwP(1) 2.400(2), RuwP(2) 2.447(2), RuwCl
2.435(2), ClwRuwP(1) 91.8(1), ClwRuwP(2) 96.0(1), P(1)wRuwCl@
high frequency side of the featureless CH hump. However,
2
the assignment of the ““agosticÏÏ CwH is not conclusive. In
88.2(1),
P(2)wRuwCl@ 84.0(1).
P(1)wRuwP(2)
81.9(1),
P(1)wRuwP(2@)
98.1(1),
fact, the signal of the single H is still broad at [90 ¡C: its
chemical shift of d [0.36 is indicative of the RuÉ É ÉHwC
contact, but the width of this 1H NMR signal (w \ 268 Hz
C(47)wC(52) ones away, as shown by the P(2)wRuwP(4)
angle, which is closed toward the hydride ligand. As a result,
the angles between P(2) and P(4) and the axial hydride must
be less than 90¡. Thus, the conformation of the cyclohexyl
groups probably derives from the weak RuÉ É ÉHwC inter-
action rather than from steric factors, as claimed for
1@2
at 250 MHz) prevents the measurement of the J(C,H) coupling
constant. Even at 500 MHz the half-line width w is as large
1@2
as 150 MHz. This suggests that the chemical exchange
between di†erent CwH groups is not frozen out even at
[90 ¡C. At the same temperature, the occurrence of a
dynamic process is observed also in the 31P NMR spectrum
[Fig. 3(b)]. In a further attempt to freeze out the dynamic
1 É BPh .12 In fact, the steric crowding due to the cyclohexyl
4
rings cannot be invoked for 1, since, even in the much more
process, the 1H NMR spectra were recorded in
a
crowded trans-[RuCl (dcpe) ] (3), the rotational freedom of
2
2
CD Cl È(CD ) CO mixture (4 : 1) down to the freezing point
the cyclohexyls releases the steric strain (Fig. 2).
2
2
3 2
of the solution (ca. [120 ¡C). Unexpectedly, these spectra
show that a further exchange process slows down. Instead of
sharpening, the signals of the methylene protons broaden
further, and new signals appear in the hydridic region about d
[RuCl (dcpe) ], 3. The unit cell contains discrete
2
2
[RuCl (dcpe) ] and two CH Cl molecules with normal non-
2
2
2 2
bonded distances. Since the Ru atom lies on an inversion
centre, only one-half of the molecule is crystallographically
independent. However, this is compatible with the complex
symmetry. The co-ordination geometry is a slightly distorted
octahedron, apparently due to the constraint of the
P(1)wRuwP(2) bite angle of 81.9(1)¡, a value similar to that
found in 1. The crystal symmetry imposes angles of exactly
180¡ for all mutually trans ligands. The average RuwP bond
length of 2.424(2) A is up to 0.06 A longer than in the Ðve-co-
ordinate hydride 1 [2.360(1) A] and falls at the upper end of
the range found for a number of ruthenium(II) complexes.18
Interestingly, the concerted opening of the ClwRuwP(2)
angle to 96.0(1)¡ [due to the C(7)wC(12) cyclohexyl ring
pointing toward Cl] and the rotation of the cyclohexyl groups
release the steric congestion.
Besides conÐrming the theoretical studies of the MXL
4
complexes of d6 ions that predict a square-pyramidal struc-
ture for X \ H and a Y-shaped one for X \ Cl,11 the com-
parison of the structures of 1 É PF , 2 and 3 is extremely
6
informative about the secondary interactions that stabilize the
Ðve-co-ordinate species [RuX(dcpe) ]`, together with the
2
bulkiness of the diphospine ligands. Thus, in 1 É PF it is the
6
weak MÉ É ÉHwC interaction that stabilizes the 16-electron
complex,19 whereas in the chloro analogue 2, it is the Cl ] Ru
p-donation, as indicated by the short RuwCl bond length
[2.386(3) A, (Table 2)].20 The latter value falls in the lower
quartile of the distribution found for 115 RuII complexes
(average RuwCl distance of 2.416 A), and is dramatically
shorter than in the dichloro analogue 3 [2.435 (2) A].
Solution studies of 1 Æ PF . The existence of some degree of
6
RuÉ É ÉHwC interaction in 1 É PF in solution was investigated
Fig. 3 (Left) 31PM1HN and (right) 1H NMR spectra of 1 (250 MHz,
LB \ 4 Hz, CD Cl ) at (a) room temperature and (b) 183 K.
6
by IR and NMR spectroscopies. A weak IR band at 2709
2
2
New J. Chem., 1999, 199È206
201

View Article Online
[30 (see Fig. S1 in the Supplementary Material). In view of
this further complication, attempts to record spectra in
FreonÈCD Cl mixtures were not undertaken.
stretching vibrations.28,29 The band at 2209 cm~1 probably
results from the overlap of the CN bands of tcva with
m (NxCxO). As recrystallization attempts only yielded
2
2
as
Although the 1H NMR data possibly suggest that some
sort of RuÉ É ÉHwC interaction is retained in solution, at least
at low temperature, this interaction must be very weak. Also,
the hydride chemical shift observed for 1 (d [32.0) is in the
range observed for Ðve-co-ordinate complexes, whereas
related complexes displaying strong agostic interactions to the
vacant site have hydride resonances at considerably lower
Ðeld (around d [12).6a Although agostic interactions are a
common feature of 16-electron species, they have been rarely
observed in Ðve-co-ordinate ruthenium(II) complexes either by
1H NMR spectroscopy or X-ray di†raction. An unusual one is
that involving a methylene group of the diphosphine back-
intractable oils, no analytically pure product was isolated. The
nature of this product is currently under investigation.
The reaction of
4 with 2-methyl-1,4-naphthoquinone
(menadione), another electron-poor, highly activated oleÐn,
was monitored by 1H and 31PM1HN NMR in an NMR tube.
After 24 h, 1 (20% of starting 4) and dcpe oxide 5 (30%) are
observed, but no trace of menadione oxide is visible in the 1H
NMR spectrum. After 5 days in the dark at room tem-
perature, 4 is consumed and the acetato complex [Ru(g2-
O CMe)(dcpe) ]` (6a) is formed, together with dcpe oxide
2
2
(Table 3, entry 1). Column chromatography of the reaction
solution yielded pure 6a and unreacted menadione (about
70% of starting amount). The identity of 6a is supported by
comparison of the FAB` MS and NMR data with those of an
authentic sample. The acetato ligand is probably formed by
oxidative degradation of menadione. A control reaction
showed that 4 reacts slowly with menadione oxide (4 is con-
sumed after 5 days), giving unidentiÐed products, but only a
small amount of 6a (5%).
bone in [RuH(PwP) ]` [PwP \ 1,4-bis(diphenylphosphino)-
2
butane (dppb) or 2,3-isopropylidene-2,3-dihydroxy-1,4-bis(di-
phenylphosphino)butane (diop)].6a
Reactivity of [RuH(g2-O )(dcpe) ]‘, 4
2
2
Reactivity with nucleophiles. The reactivity of 4 with nucleo-
philes and electrophiles was investigated.21 No reaction
occurs with electron-rich oleÐns, such as 1-hexene and styrene,
Complex 4 reacts with other organic electrophiles such as
PhC(O)Cl and aldehydes. The reaction of 4 with PhC(O)Cl
gives the carboxylato derivative [Ru(g2-O CPh)(dcpe) ]` (6b)
or with PPh . The 31PM1HN and 1H NMR spectra of 4 remain
3
2
2
unchanged in the presence of these substrates (either in excess
(by comparison with an authentic sample) (Table 3, entry 2).
After 24 h, 80% of starting 4 has reacted and some benzoato
complex 6b is formed, together with 2, dcpe oxide 5, and as yet
unidentiÐed products. The reaction of 4 with aldehydes,
which are weaker electrophiles than acyl chlorides, is most
interesting. Acid-free heptanal and 4 (2:1 molar ratio) react
quantitatively within 24 h giving a mixture of products: 1,
[RuCl(g2-H )(dcpe) ]` (7),10c and the carboxylato complex
or in a 1 : 1 molar ratio) in CD Cl solution after several days.
2
2
The failure of 4 to react with electron-rich oleÐns and PPh
3
parallels the behaviour of the recently reported [M(g2-
O )(CO)(CH CN)(PPh ) ]` (M \ Rh, Ir),22 and suggests that
the dioxygen ligand in these complexes has nucleophilic char-
2
3
3 2
acter (see below).
2
2
Reactions with electrophiles. The classiÐcation scheme of the
relative nucleophilic reactivity of transition metal peroxo com-
plexes, devised by Valentine and co-workers, prompted us to
investigate the reactivity of 4 toward carbonyl compounds
and electron-poor oleÐns.21 The results of this study allow
comparison to early reports of the reactions between
[M(g2-O )L ] (M \ Ni, Pd, Pt; L \ RNC or PPh ) and
[Ru(g2-O CC H )(dcpe)]` (6c), together with dcpe oxide
2
6 13
5 (Table 3, entry 3). The occurrence of oxygen transfer to the
aldehyde is indicated by the presence in the reaction mixture of
1, 6c and C H C(O)OH, as detected by GC-MS and 1H and
6
13
31PM1HN NMR spectroscopies. Low-intensity broad signals in
the 1H NMR spectrum (d region 0 to [4) suggest that minor
amounts of as yet unidentiÐed paramagnetic products are also
formed. When a stoichiometric amount of TEMPO (2,2,6,6,-
tetramethyl-1-piperidinyloxyl radical) is added to the reaction
solution, the formation of dcpe oxide is inhibited, and the car-
boxylato species 6c is the main product after 24 h (Table 3,
entry 4). By contrast, when the radical initiator azaisobutyron-
itrile, AIBN, is added, dcpe oxide is the major product (Table
3, entry 5). Attempts to detect the corresponding peroxy car-
boxylic acid as an intermediate were not successful.30
2
2
3
PhC(O)Cl, RCHO, or TCNE.23h26
When 4 and TCNE (1 equiv.) are dissolved in anhydrous
CD Cl under argon, a green solution is formed instantane-
2
2
ously, which turns dark brown within 6 h. Work-up yields a
dark brown crystalline compound that is not a hydride, as
shown by 1H NMR spectroscopy. The IR spectrum suggests
oxidative fragmentation of TCNE: bands at 2209 and 1611
cm~1 are typical of a tricyanovinylethanol (tcva) ligand.27 The
presence of an N-bound isocyanato ligand is suggested by the
peaks in the FAB` mass spectrum at m/z 1005 and 973, corre-
sponding to [Ru(NCO)(O)(dcpe) ]` and [Ru(CN)(dcpe) ]`,
The reactions with PhC(O)Cl, RCHO, and TCNE indicate
that 4 is a nucleophile of strength comparable to that of
[M(g2-O )L ] (M \ Ni, Pd, Pt; L \ RNC or PPh ).23h26 The
2
2
2
2
3
and is in agreement with the IR bands at 2209 and 1327
reaction of 4 with aldehydes is particularly interesting, as it is
cm~1, attributed to the asymmetric and symmetric NxCxO
an example of an ““atom-transfer redox reactionÏÏ according to
Table 3 Oxygen-transfer reactions of 4a
Run
Reagent(s)
Ratiob
t/days
Conv/%
Products (yield/%)c
1
2
3
4
5
6
7
8
Menadione
PhC(O)Cl
C H CHO
1 : 1
1 : 1
1 : 2
1 : 2
5
1
1
1
1
5
7
6
[99
80
6a (30), 5 (50)
6b (10), 2 (10), 5 (20)
1 (29), 6c (7), 7 (17), 5 (40)
6c (50), 5 (10)
[99
[90
[90
[90
[90
[99
6
13
C H CHO d
13
C H CHO e
13
6
1 : 2
5 (80)
6
PhC(O)ClÈPPh
1 : 1 : 1
1 : 1 : 1
1 : 1 : 1
Ph PxO (25),f 6b (25), 2 (23), 1 (15), 5 (12)
3
3
PhC(O)ClÈstyrene
2 (30), 6b (28), 1 (3), 5 (10)
C H CHOÈstyrene
5 (25), 1 (20), 6c (5), 7 (13)
6
13
aReactions monitored by 1H and 31PM1HN NMR in an NMR tube (CDCl , under puriÐed N or sealed under vacuum). Products are
3
2
[RuH(dcpe) ]` (1), [RuCl(dcpe) ]` (2), dcpe oxide (5), [Ru(g2-O CR)(dcpe) ]` (R \ Me, 6a; Ph, 6b; C H , 6c), and [RuCl(g2-H )(dcpe) ]` (7).
2
2
2
2
6
13
2
2
bMole ratio of complex 4 to reagent. cBased on starting 4 (see also Experimental). dTEMPO added (1 : 1 mole ratio) to the reaction solution.
eAIBN (1 : 1 mole ratio) as additive.fBased on starting PPh .
3
202
New J. Chem., 1999, 199È206

View Article Online
the deÐnition by Collman and Otsuka (that is, the formal oxi-
dation state of the metal does not change in the net oxygen-
transfer reaction).23 In the present case, the overall
oxygen-transfer reaction from 4 to the substrate regenerates
the O -activating species 1 without oxidation of the metal. To
2
the best of our knowledge, such a reactivity is unprecedented
for dioxygen complexes of the late transition metals, which
generally react with aldehydes to give a carboxylato complex
with the metal in an oxidized form.1h,31,32 Thus, [Ru(g2-
O )(CO)(CNR)(PPh ) ], formally a d8 dioxygen complex,
2
3 2
reacts with CH CHO to give the bis(carboxylato) d6 complex
3
[Ru(g1-O CMe) (CO)(CNR)(PPh ) ],31a whereas [Pt(g2-
2
3 2
2
3 2
Scheme 2
O )(PPh ) ] reacts with aldehydes forming the metallacycle
2
[Pt(O1OC(R)(H)O3-O1,O3)(PPh ) ]30b or, in the presence of
3 2
air, [Pt(g1-O CR) (PPh ) .31b
2
2
3 2
In view of its nucleophilic character, 4 was also reacted with
acids to check whether the dioxygen ligand is protonated. The
reaction of 4 with CH CO H (1 : 1 molar ratio) gives the
co-oxidation of PhC(O)Cl and PPh by 4 can be explained
3
with the generally accepted mechanism of electrophilic peroxi-
dation (Scheme 1).1f,g The cyclic peroxymetallation reaction1f
3
2
acetato complex [Ru(g2-O CCH )(dcpe) ]` (6a): after 1 h,
involves opening of the M(g2-O ) ring, followed by nucleo-
2
3
2
2
70% of 4 is converted to 6a with 40% yield. The remaining 4
is probably converted to undetectable paramagnetic species
(see Experimental). Protonation of 4 by [Et OH]BF gives a
philic attack of M(g1-O ) onto the carbonyl C atom.21b Elimi-
2
nation of Cl~ from the peroxy metallacycle 8a gives the
ruthenium(IV) hydride 9, which reductively eliminates H` as it
must be very acidic.33 Reaction of the peroxy intermediate 10
2
4
complex mixture of non-hydridic RuII complexes. Both reac-
tions indicate that the g2-O ligand is eliminated as O , not
as H O . Thus, protonation apparently occurs at the hydride
with PPh before dissociation of RC(O)OOH takes place
2
2
3
would explain why free RC(O)OOH is not detected. The plati-
2
2
ligand rather than at g2-O , and is followed by elimination of
num congener of 10 has been suggested as an intermediate in
2
H .
the analogous reaction involving [Pt(g2-O )(PPh ) ].25b
2
2
3 2
The reaction of 4 with heptanal di†ers from the previous
one in that the formyl H atom can be reductively eliminated
from 8b instead of the hydride, which eventually gives the Ðve-
co-ordinate hydride 1 (Scheme 2). This would explain why
Co-oxidation reactions. The Ðrst attempts to exploit the
oxygen-transfer reaction from 4 in the co-oxidation of a
nucleophile (Nu) and an electrophile (El) (eqn. 1) were only
partially successful.
oxygen transfer from 4 regenerates the O binding species.
The reaction of the peroxy acid with the second equivalent of
2
RCHO can occur either directly or, more probably, with the
mediation of the metal, for instance via formation of a highly
reactive oxometal species from 11. Support for this possibility
comes from our recent observation that the related dioxygen
complex [OsCl(g2-O )(dcpe) ]` forms the OsIV oxo complex
Oxygen transfer from 4 to PPh and PhC(O)Cl (1 : 1 : 1 molar
2
2
[OsCl(O)(dcpe) ]`.34 A peculiar feature of the reactions of 4
3
ratio) is moderately efficient, with 25% of starting PPh con-
2
with heptanal is the formation of considerable amounts of
3
verted into Ph PxO after 5 days (Table 3, entry 6). The other
3
[RuCl(g2-H )(dcpe) ]` (7) (Table 3, entries 3, 8) along with 1.
reaction products are the carboxylato derivative [Ru(g2-
2
2
Independent experiments show that 7 is not produced by reac-
tion of 1 with CD Cl . We suggest that 7 is formed by proto-
O CPh)(dcpe) ]` (6b), 2, and 1. Partial oxidation of the di-
2
2
phosphine ligand occurs: some dcpe oxide 5 is observed.
However, styrene is not epoxidized in the presence of 4 and an
electrophile. The reaction of 4 with PhC(O)Cl in the presence
of styrene (1 : 1 : 1 molar ratio) gives 2 and 6b as the main
products, together with dcpe oxide and minor amounts of 1
(Table 3, entry 7). No styrene oxide is detected either by 1H
NMR or by GC-MS after completion of the reaction. In the
reaction of 4 with heptanal and styrene (1:1:1 molar ratio)
under the same conditions, 4 is consumed after 6 days, and
dcpe oxide is the main product, together with 1 and 7 (Table
3, entry 8). The carboxylato complex 6c is slowly formed (5
and 15% after 6 and 12 days, respectively).
2
2
nation of the hydride in 1, followed by abstraction of Cl~
from the solvent. The acidic species could be the RC(O)OOH
moiety in 11, since co-ordination to the metal enhances the
acidity of the peroxy acid. Alternatively, protonation of unre-
acted 4, followed by loss of O , can also produce 7. The car-
2
boxylato derivative 6c is eventually formed by reaction
between heptanoic acid and 1, as conÐrmed by independent
experiments showing that C H CO H reacts with 1 to form
6
13
2
6c, probably via protonation of the hydride ligand and loss of
H from an undetected dihydrogen complex.
Finally, the presence of minor amounts of paramagnetic
2
products, as well as the experiments with radical scavengers
(TEMPO) and initiators (AIBN), leave the possibility open
that more than one mechanism is operative. We can not
Mechanistic considerations. A possible mechanistic interpre-
tation of the reactivity of 4 is shown in Schemes 1 and 2. The
exclude that the g2-O ligand abstracts the formylic H atom
2
to give a MIIIOOH species.35 A mechanism based on the
elimination of HOO~ from 4, analogous to that suggested for
the catalytic oxidation of PPh by [Pt(g2-O )(PPh ) ],36 does
3
2
3 2
not seem to apply here, since the hydride ligand is retained to
give 1.
Only a few M(g2-O ) complexes are known to react rapidly
2
with menadione.21 This has been proposed as a diagnostic
tool for the nucleophilicity of the g2-O moiety. In the case of
2
4, the reaction is very slow (5 days), and complicated by the
oxidative degradation of menadione and of its epoxide. Both
reactions suggest that the CxC bond is oxidatively cleaved in
the presence of 4, a reaction already observed with
Scheme 1
New J. Chem., 1999, 199È206
203

View Article Online
rutheniumÈO systems.13b On the basis of the reactivity
from CH Cl Èethanol. 1H and 31P NMR data as for the
2
2 2
results, we suggest that 4 and [Pt(g2-O )(PPh ) ] are nucleo-
[BPh ]~ analogue.3a UV/VIS (CH Cl ) j/nm (e/mol L~1
2
3 2
4
2 2
philes of similar strength, since 4 reacts with RCHO and
PhC(O)Cl, and only very slowly with menadione.21a
cm~1): 213(sh), 236 (750), 260 (1060), 280 (1052), 353 (530).
Anal. Found: C, 55.29; H, 9.02. Calcd. for C O F P Ru:
H
52 97 2 6 5
C, 55.55; H, 8.69%. FAB` MS (NBA matrix): m/z 980 [M]`,
12%; 964 [M [ O]`, 7%; 948 [M [ 2 O]`, 100%; 864 [M
[ O [ C H ]`, 6%.
Concluding remarks
A weak agostic interaction between a cyclohexyl methylene
group and ruthenium seems to contribute to the stabilization
2
6 11
Reaction of 4 with TCNE. Complex 4 (100 mg, 0.088 mmol)
of the Ðve-co-ordinate hydride [RuH(dcpe) ]`. Thus, steric
and TCNE (12 mg, 0.088 mmol) were dissolved in CH Cl
2
2 2
factors alone do not explain the stability of the 16-electron
and stirred for 2 h. The solution turned Ðrst green-blue, then
brown-black. Pumping the solvent o† resulted in a brown-
black powder. Recrystallization from di†erent solvent mix-
tures gave intractable oils. 1H NMR (300 MHz, CD Cl ,
complexes [MX(PwP) ]` (M \ Ru, Os) and their reactivity
2
toward O (and H ). Electronic factors, such as weak agostic
2
2
interactions and the r- and p-donor/acceptor properties of the
ligands, must be also considered in this context. In particular,
the good donor properties of dcpe and hydride ligands are
2
2
25 ¡C): d 2.4È0.4 (CH ). 31PM1HN NMR (121 MHz, CD Cl ,
2
2 2
25 ¡C): d 70.4 (s), [144.3 (spt, PF , 1J(PF) \ 713 Hz). IR
(KBr): 2209s, 2120m, 2098m, 1611s, 1327m, m(PÈF) 841vs
6
apparently responsible for the reactivity of 1 toward O . The
2
resulting peroxo species 4 is a class II nucleophile, according
cm~1. FAB` MS (NBA matrix): m/z 1005, 9%; 973, 21%.
to ValentineÏs deÐnition, as it reacts with organic electrophiles
such as TCNE, PhC(O)Cl, and RCHO. Of particular interest,
4 oxidizes RCHO to RC(O)OH, restoring the oxygen-binding
species 1. This scarcely documented ““reductiveÏÏ oxygen trans-
fer to organic substrates opens the way to developing catalytic
transformations. Our e†orts are presently directed to getting a
better insight into the mechanistic aspects of these reactions,
and to enhancing the stability of the complex toward oxida-
tive degradation of the diphosphine ligand.
Reactions of 4 in an NMR tube. Complex 4 (20 mg, 18
lmol) and the appropriate reagent were dissolved in CD Cl
(0.5 mL) in an NMR tube under argon: (a) menadione (3 mg,
2
2
18 lmol); (b) acid-free C H CHO (5 ll, 36 lmol), (c) PPh (5
6
13
3
mg, 18 lmol) and acid-free PhC(O)Cl (2 ll, 18 lmol); (d)
styrene (2 ll, 18 mmol) and acid-free PhC(O)Cl (2 ll, 18 lmol),
(e) styrene (2 ll, 18 lmol) and acid-free C H CHO (2 ll, 18
6
13
lmol). The reaction solution was degassed by three freeze-
thaw-pump cycles, and then the NMR tube was sealed under
vacuum. Alternatively, the solutions were prepared in glove
Experimental
General
box, and the tube closed under N . The progress of the reac-
2
tions was monitored by 31PM1HN and 1H NMR spectro-
scopies, using the 31P NMR resonance of [PF ]~ as internal
Manipulations involving solutions of the complexes were per-
formed either under argon with use of Schlenk-line techniques
or in a glove box (Braun) under puriÐed nitrogen. Solvents
were puriÐed by standard methods. All chemicals used were of
reagent grade or comparable purity. The ligand dcpe was pur-
chased from Strem and RuCl É H O from Aldrich; TCNE
6
reference. This is justiÐed by the similarity of the relaxation
times and by the chemical stability of the anion under these
conditions. The total amount of the Ru containing products is
generally below 100% since most reactions yield variable
amounts of paramagnetic by-products, depending on the sub-
strate. Most of these yet unidentiÐed species give a character-
istic high-Ðeld 1H NMR pattern in the d region between 0 and
[4, but cannot be quantiÐed as they are 31P NMR silent. In
some cases, the presence of several products with complicated
31P spectral patterns in small amounts hinders accurate inte-
gration of the spectra. The percentage of dcpe oxide is not
normalized to starting 4. After the reaction, the volatile pro-
ducts were analysed by GC-MS.
3
2
was sublimed twice before use. Complexes 1 É BPh ,3a 3,10a
4
and [RuH(Cl)(dcpe) ]10a were prepared as previously report-
2
ed. Unless otherwise stated, yields are based on the metal.
Infrared spectra were recorded on a Perkin Elmer Paragon
1000 FT-IR spectrometer. Microanalyses were performed by
the Laboratory of Microelemental Analysis at the Organic
Laboratory at ETH. 1H, 13CM1HN, and 31PM1HN NMR spectra
were taken on a Bruker DPX 300 spectrometer. 1H and
31PM1HN NMR chemical shifts (d) are relative to TMS and
85% H PO , respectively. 13CM1HN NMR chemical shifts
[Ru(g2-O CCH )(dcpe) ]PF , 6a. Complex 3 (100 mg,
2
3
2
6
0.098 mmol) was dissolved in benzene (5 mL) and Tl[PF ] (69
3
4
were internally referenced to solvent CDCl (d 77.0). Mass
6
mg, 0.20 mmol) was added to the yellow-orange solution.
3
spectroscopy was carried out by the analytical service of the
After stirring the resulting violet solution for 45 min, a MeOH
solution (1 mL) of NEt (O CCH ) (15 mg, 0.20 mmol) was
Organic Laboratories at the ETH Zurich.
4
2
3
added. The solution turned yellow within 20 min, after which
the solvent was evaporated in vacuum and the residue was
dissolved in CH Cl . After Ðltration over Celite, addition of
Preparations and reactions
[RuH(dcpe) ]PF , 1 Æ PF . [RuH(Cl)(dcpe) ] (200 mg, 0.20
2
2
2
6
6
3
2
MeOH (6 mL) and evaporation of CH Cl , followed by addi-
mmol) was dissolved in CH OH (15 mL) and a CH OH solu-
2
2
3
tion of ether (8 mL), gave 6a (51 mg, 45%) as a yellow precipi-
tate, which was Ðltered o† and dried in vacuum. Anal. Found:
tion (3 mL) of TlPF (75 mg, 0.41 mmol) was added to the
deep orange solution, from which orange microcrystals pre-
6
C, 51.40; H, 7.90. Calcd. for C
H
F O P Ru É 2 CH Cl : C,
cipitated during 2 h. Cooling in ice, decanting the supernatant
solution and drying in vacuum gave the crude product. Re-
crystallization from CH Cl ÈPriOH yielded 1 É PF (140 mg,
54 99 6 2 5
2
2
50.95; H, 7.86%. IR (KBr): l (OCO) 1559s, l (OCO)
asym
sym
1456s, g(PwF) 848vs cm~1. 1H NMR (300 MHz, CDCl ,
3
2
2
6
25 ¡C): d 2.4È0.4 (CH ] CH ). 31PM1HN NMR (121 MHz,
63%). Anal. Found: C, 56.94; H, 8.72. Calcd. for
2
3
CDCl , 25 ¡C): AA@XX@, d 53.2 (t, 2J(PP@) \ 16.4 Hz), 71.4 (t),
C
H
F P Ru: C, 57.18; H, 8.95%. 1H NMR (300 MHz,
52 97 6 5
CD Cl , 25 ¡C): d [32.0 (RuwH, qt, 2J(PH) \ 19.2 Hz).
3
[144.3
(spt,
PF ,
1J(PF) \ 713
Hz).
13CM1HN
6
2
2
NMR (75 MHz, CDCl , 25 ¡C): d 185.3 (s, CH CO ),
31PM1HN NMR (121 MHz, CD Cl , 25 ¡C): d 73.6 (s, 4 P),
[144.3 (spt, PF , 1J(PF) \ 713 Hz).
3
3
2
2
2
53.1 (s, CH CO ), 45È17 (aliphatic CH ). FAB` MS (NBA
3
2
2
6
matrix): m/z 1006 [M]`, 100%; 922 [M [ Cy]`, 10%.
[RuH(g2-O )(dcpe) ]PF , 4. Complex 1 (200 mg, 0.091
2
2
6
mmol) was dissolved in CH Cl (4 mL) and stirred in air for
[Ru(g2-O CC H )(dcpe) ]PF , 6b. Complex 6b (73 mg,
2
2
2
6
5
2
6
6
45 min. Evaporation of the solvent in vacuum gave brown
63%) was prepared from C H CO H (24 mg, 0.20 mmol) and
5
2
microcrystals of 4 (124 mg, 60%), which were recrystallized
NEt (27 lL, 0.20 mmol) as described for 6a. Anal. Found: C,
3
204
New J. Chem., 1999, 199È206

View Article Online
58.28; H, 8.66. Calcd. for C
H
O F P Ru: C, 58.45; H,
standard reÑection was measured every 120 reÑections; no sig-
niÐcant variation was detected. A face-indexed numerical
absorption correction was applied. The structure was solved
using Patterson methods. Of the 2824 independent reÑections
with index ranges 0 O h O 12, 0 O k O 15, [13 O l O 13, 2178
with F [ 4.0r(F) were used in the reÐnement. A total of 296
parameters were reÐned by full-matrix least squares using
SHELXTL PLUS38 [data-to-parameter ratio 7.4 : 1, quantity
minimized &w(F [ F )2] with anisotropic displacement
59 109 2 6 5
8.40%. IR (KBr): l (OCO) 1522s, l (OCO) 1497s, l(PwF)
asym
sym
848vs cm~1. 1H NMR (300 MHz, CDCl , 25 ¡C): d 7.4È7.9
3
(C H ), 2.4È0.4 (CH ). 31PM1HN NMR (121 MHz, CDCl ,
6
5
2
3
25 ¡C): AA@XX@, d 54.8 (t, 2J(PP@) \ 16.4 Hz), 73.4 (t), [144.3
(spt, PF , 1J(PF) \ 713 Hz). 13CM1HN NMR (75 MHz,
6
CDCl , 25 ¡C): d 180.0 (s, RCOO), 132.1, 131.7, 127.9, 127.3 (s,
3
C H ), 48È18 (aliphatic CH and CH ). FAB` MS (NBA
6
5
2
3
matrix): m/z 1068 [M]`, 100%; 984 [M [ Cy]`, 17%; 644
o
c
[M [ dcpe [ H]`, 6%.
parameters for all non-H atoms. The contribution of the H
atoms in idealized positions (riding model with Ðxed isotropic
[Ru(g2-O CC H )(dcpe) ]PF , 6c. Complex 6c was pre-
U \ 0.080 A
Ž 2) was taken into account but not reÐned. Two
2
6
13
2
6
pared from heptanoic acid (28 lL, 0.20 mmol) and NEt (27
independent CH Cl molecules were located in the Ðnal
3
2
2
lL, 0.20 mmol) as described for 6a. Recrystallization from
Fourier map. Final residuals were R \ 0.039 and wR \ 0.055
(obs. data), and R \ 0.050 and wR \ 0.058 (all data), GOF
0.98 (weighting scheme w~1 \ r2(F) ] 0.0021F2). Maximum
benzeneÈether gives 6c (81.3 mg, 67%) as a yellow crystalline
solid. IR (KBr): l (OCO) 1526s, l (OCO) 1453s, l(PwF)
asym
sym
837vs cm~1. 1H NMR (300 MHz, CDCl , 25 ¡C): d 0.4È2.4
and minimum di†erence peaks were ]0.54 and [0.61 e AŽ ~3,
3
(CH ). 31PM1HN NMR (121 MHz, CDCl , 25 ¡C): d AA@XX@,
largest and mean D/p \ 0.040 and 0.002. Selected interatomic
distances and angles are reported on Fig 2.
2
3
52.4 (t, 2J(PP@) \ 16.4 Hz), 70.5 (t), [144.3 (spt, PF ,
6
1J(PF) \ 713 Hz). 13CM1HN NMR (75 MHz, CDCl , 25 ¡C): d
3
CCDC reference number 440/088. See http://www.rsc.org/
suppdata/nj/1999/199/ for crystallographic Ðles in .cif format
188.1 (s, RCOO), 46È13.9 (aliphatic CH and CH ). FAB`
2
3
MS (NBA matrix): m/z 1076 [M]`, 100%; 993 [M [ Cy]`,
23%; 945 [M [ C H O [ 2 H]`, 6%.
7
13 2
Acknowledgements
A.M. thanks the ETH Zurich for Ðnancial support
Crystallography
[RuH(dcpe) ]PF , 1 Æ PF . Dark red crystals of 1 É PF ,
2
6
6
6
suitable for X-ray analysis, were grown from CH Cl ÈPriOH
2
2
in a glove box under puriÐed N . A prism (0.30 ] 0.25 ] 0.20
mm) was mounted in a 0.5 mm glass capillary in glove box.
Notes and references
2
1 (a) G. A. Barf and R. A. Sheldon, J. Mol. Catal. A, 1995, 102, 23;
(b) T he Activation of Dioxygen and Homogeneous Catalytic Oxida-
tion, eds. D. H. R. Barton, A. E. Martell and D. T. Sawyer,
Plenum, New York, 1993; (c) R. S. Drago, Coord. Chem. Rev.,
1992, 117, 185; (d) B. R. James, in Dioxygen Activation and Homo-
geneous Catalytic Oxidation, ed. L. I. Simandi, Elsevier, Amster-
dam, 1991, p. 195; (e) H. A. Hill and D. G. Tew, in Comprehensive
Coordination Chemistry, eds. G. Wilkinson and J. A. McCleverty,
Pergamon, Oxford, 1987, vol. 2, p. 318; ( f ) H. Mimoun, Angew.
Chem., Int. Ed. Engl., 1982, 21, 734; (g) B. R. James, Adv. Chem.
Ser., 1980, 191, 251; (h) J. E. Lyons, in Aspects of Homogeneous
Catalysis, ed. R. Ugo, Reidel, Dordrecht, 1977, p. 1; (i) J. S. Valen-
tine, Chem. Rev., 1973, 73, 235.
Crystal data for C
P2/c (No 13), a \ 19.692(1), b \ 12.494(1), c \ 22.413(2) A,
b \ 95.960(7)¡, U \ 5467.0(7) A
H
F P Ru, 1: M \ 1092.22, monoclinic,
52 97 6 5
Ž
3, Z \ 4, D \ 1.327 Mg m~3,
c
k(Mo-Ka) \ 0.487 mm~1, F(000) \ 2328, T \ 298 K, j(Mo-
Ka) \ 0.71073 A. The data were collected on a STOE IPDS
(image plate detector system) in the 2h range 5.0È50.0¡
(oscillation, 6 min exposure time, 181 images). Unit cell
dimensions determination and data reduction were performed
by standard procedures,37 and an additional absorption cor-
rection using ABSCOR, including all non-hydrogen atoms,
was applied. The structure was solved with SHELXTL
PLUS38 using direct methods. Of the 30 551 measured reÑec-
tions with index ranges [22 O h O 20, [14 O k O 14,
2 R. Neumann and M. Dahan, Nature, 1997, 388, 353.
3 (a) M. JimenezÈTenorio, M. C. Puerta and P. Valerga, Inorg.
Chem., 1994, 33, 3515; (b) A. Mezzetti, E. Zangrando, A. Del Zotto
and P. Rigo, J. Chem. Soc., Chem. Commun., 1994, 1597; (c) T. Y.
Bartucz, A. Golombek, A. J. Lough, P. A. Maltby, R. H. Morris,
R. Ramachandran and M. Schlaf, Inorg. Chem., 1998, 37, 1555.
4 The term ““dioxygen complexÏÏ is used throughout the paper, and
the oxidation state of the metal in 4 is deÐned accordingly as
ruthenium(II). Although the di†erence is largely formal, we feel that
[25 O l O 25, 8503 were independent (R \ 0.0581), and
int
5983 with o F o2 [ 2r(o F o2) were used in the reÐnement. A total
of 867 parameters were reÐned by full-matrix least squares
using SHELXS-9639 [data-to-parameter ratio 1 : 7, quantity
minimized &w(o F o2 [ o F o2)2, w~1 \ r2(F2) ] (aP)2 ] bP
o
c
c
o
where 3P \ F2(P 0) ] 2F2] with anisotropic displacement
the alternative description of
ruthenium(IV) is less realistic.
4 as a peroxo complex of
o
parameters for all non-H atoms. All hydrogen atoms other
than the hydride were located in the Fourier peak list and
5 For reviews see: (a) P. J. Jessop and R. H. Morris, Coord. Chem.
Rev., 1992, 121, 155; (b) D. M. Heinekey and W. J. Oldham, Chem.
Rev., 1993, 93, 913.
6 For recent references, see: (a) M. Ogasawara and M. Saburi,
Organometallics, 1994, 13, 1911; (b) M. Ogasawara and M. Saburi,
J. Organomet. Chem., 1994, 482, 7; (c) E. Rocchini, A. Mezzetti, H.
Ruegger, U. Burckhardt, V. Gramlich, A. Del Zotto, P. Martinuzzi
and P. Rigo, Inorg. Chem., 1997, 36, 711; (d) M. Schlaf, A. J. Lough
and R. H. Morris, Organometallics, 1997, 16, 1253.
reÐned with Ðxed isotropic parameters (U \ 0.080 AŽ 2).
iso
Molecular graphics were drawn with ORTEP II38 with 50%
ellipsoids. Final residuals were R \ 0.0457 and wR \ 0.0993
(obs. data), and R \ 0.0787 and wR \ 0.1088 (all data), GOF
0.995. (N.B.: wR values are calculated on o F o2, see Supplemen-
tary material). Maximum and minimum di†erence peaks were
]0.96 and [0.29 e A
0.000.
Ž ~3, largest and mean D/p \ [0.001 and
7 (a) M. M. Taqui Khan, R. K. Andal and P. T. Manoharan, Chem.
Commun., 1971, 561; (b) F. G. Moers, R. W. M. ten Hoedt and J.
P. Langhout, J. Inorg. Nucl. Chem., 1974, 36, 2279.
trans-[RuCl (dcpe) ], 3. A pale orange cube of 3, grown
2
2
8 (a) M. A. Esteruelas, E. Sola, L. A. Oro, U. Meyer and H. Werner,
Angew. Chem., Int. Ed. Engl., 1988, 27, 1563; (b) K. Kirchner, K.
Mauthner, K. Mereiter and R. Schmid, J. Chem. Soc., Chem.
Commun., 1993, 892; (c) S. M. Maddock, C. E. F. Rickard, W. R.
Roper and L. J. Wright, J. Organomet. Chem., 1996, 510, 267;
(d) I. De los Rios, M. J. Tenorio, J. Padilla, M. C. Puerta and
P. Valerga, J. Chem. Soc., Dalton T rans., 1996, 377; (e) M. Sato
and M. Asai, J. Organomet. Chem., 1996, 508, 121.
9 (a) B. E. Cavit, K. R. Grundy and W. R. Roper, J. Chem. Soc.,
Chem. Commun., 1972, 60; (b) D. F. Christian and W. R. Roper,
Chem. Commun., 1971, 1271; (c) K. R. Laing and W. R. Roper,
Chem. Commun., 1968, 1568; (d) M. Ogasawara, S. A. Macgregor,
from CH Cl Èacetone, was mounted on a glass capillary.
2
2
Crystal data for C
H P Ru É 2CH Cl : M \ 1187.0, mono-
52 96 4
2 2
clinic, space group P2 /n, cell dimensions (293 K)
a \ 13.256(2), b \ 16.259(4), c \ 14.189(4) A, b \ 100.05(2)¡,
1
U \ 3011.2(12)
A
Ž
3, Z \ 2, D \ 1.309 Mg m~3, k(Mo-
c
Ka) \ 6.66 cm~1 (graphite monochromated), j \ 0.710 73 A,
F(000) \ 1260. The data were collected on a Syntex P2 dif-
1
fractometer using the u scan mode (2h range \ 3.0È40.0¡)
with variable scan speeds (1.0È4.0¡ min~1 in u) to ensure con-
stant statistical precision on the collected intensities. One
New J. Chem., 1999, 199È206
205

View Article Online
W. E. Streib, K. Folting, O. Eisenstein and K. G. Caulton, J. Am.
Chem. Soc., 1996, 118, 10189; (e) K. Hiraki, S. Kira and H.
Kawano, Bull. Chem. Soc. Jpn., 1997, 70, 1583; ( f ) J. Y. Shen,
E. V. D. Stevens and S. P. Nolan, Organometallics, 1998, 17, 3875.
10 (a) A. Mezzetti, A. Del Zotto, P. Rigo and N. Bresciani Pahor, J.
Chem. Soc., Dalton T rans., 1989, 1045; (b) A. Mezzetti, A. Del
Zotto and P. Rigo, J. Chem. Soc., Dalton T rans., 1990, 2515;
(c) A. Mezzetti, A. Del Zotto, P. Rigo and E. Farnetti, J. Chem.
Soc., Dalton T rans., 1991, 1525.
11 (a) J.-F. Riehl, Y. Jean, O. Eisenstein and M. Pelissier,
Organometallics, 1992, 11, 729; (b) T. J. Johnson, K. Folting, W. E.
Streib, J. D. Martin, J. C. Hu†man, S. A. Jackson, O. Eisenstein
and K. G. Caulton, Inorg. Chem., 1995, 34, 488 and references
therein.
12 R. F. Winter and F. M. Hornung, Inorg. Chem., 1997, 36, 6197.
13 (a) J. E. Lyons and J. O. Turner, T etrahedron L ett., 1972, 29, 2903;
(b) M. Bonaldo, F. Borin, M. Bressan and A. Morvillo, J.
Organomet. Chem., 1989, 363, 175; (c) S.-I. Murahashi, Pure Appl.
Chem., 1992, 64, 403; (d) S.-I. Murahashi, Y. Oda and T. Naota, J.
Am. Chem. Soc., 1992, 114, 7913; (e) S.-I. Murahashi, Y. Oda, T.
Naota and T. Kuwabara, T etrahedron L ett., 1993, 34, 1299.
14 (a) Y. Nishida, N. Tanaka and M. Okazaki, Polyhedron, 1994, 13,
2245; (b) M. Suzuki, T. Ishikawa, A. Harada, S. Ohba, M. Saka-
moto and Y. Nishida, Polyhedron, 1997, 16, 2553; (c) W. Nam,
H. J. Kim, S. H. Kim, R. Y. N. Ho and J. S. Valentine, Inorg. Chem.,
1996, 35, 1045; (d) S. S. Bennett, M. Brown, G. Conole, M. Kessler,
S. Rowling, E. Sinn and S. Woodward, J. Chem. Soc., Dalton.
T rans., 1995, 367; (e) P. Mastrorilli, C. F. Nobile, G. P. Suranna
and L. Lopez, T etrahedron, 1995, 51, 7943; ( f ) N. Mizuno, H.
Weiner and R. G. Finke, J. Mol. Catal. A, 1996, 114, 15.
15 (a) T. Mukaiyama, in T he Activation of Dioxygen and Homoge-
neous Catalytic Oxidation, eds. D. H. R. Barton, A. E. Martell and
D. T. Sawyer, Plenum, New York, 1993, p. 133; (b) K. Yorozu, T.
Takai, T. Yamada and T. Mukaiyama, Bull. Chem. Soc. Jpn., 1994,
67, 2195; (c) R. I. Kureshy, N. H. Khan, S. H. R. Abdi and P. Iyer,
J. Mol. Catal. A, 1997, 124, 91.
23 S. Otsuka, A. Nakamura, Y. Tatsuno and M. Miki, J. Am. Chem.
Soc., 1972, 94, 3761.
24 R. A. Sheldon and J. A. van Doorn, J. Organomet. Chem., 1975, 94,
115, and references therein.
25 (a) The reaction of [Pt(g2-O )(PPh ) ] with PhC(O)Cl is fast even
2
3 2
at [78 ¡C. The resulting highly reactive peroxy carboxylato
complex [PtCl(OOC(O)Ph)(PPh ) ] reacts at low temperature
3 2
with Ph P giving Ph PxO and [PtCl(O CPh)(PPh ) ], but
3
3
2
3 2
decomposes by oxene transfer onto co-ordinated PPh when the
3
temperature is raised;25b (b) M. J. Y. Chen and J. K. Kochi, J.
Chem. Soc., Chem. Commun., 1977, 204.
26 TCNE has been found to react with g2-O complexes in
2
a
number of ways: peroxymetallacycles are formed with
[Pt(g2-O )(PPh ) ],24 whereas its reaction with [Fe(g2-
2
3 2
O )(porphyrin)]~ gives malononitrile as the main product.21
2
27 (a) E. J. Brandon, G. P. A. Yap, A. L. Rheingold, A. Arif and J. S.
Miller, Inorg. Chim. Acta, 1995, 240, 515; (b) A. J. Blake, L. E.
Horsburgh, M. Schroder and L. J. Yellowlees, J. Chem. Soc.,
Dalton T rans., 1997, 1973; (c) B. W. Sullivan and B. M. Foxman,
Organometallics, 1983, 2, 187.
28 K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coor-
dination Compounds, Wiley, New York, 1997.
29 F. Ragaini, T. Longo, S. Cenini and F. Demartin, J. Chem. Soc.,
Dalton T rans., 1996, 3221.
30 (a) Iodometric titrations of solutions containing a number of
peroxy compounds have been found to give low results (30 to 60%
of theoretical), and palladium peroxycarbonato complexes decom-
posed so rapidly that no liberated iodine was detected;30b (b) P. J.
Hayward, D. M. Blake, G. Wilkinson and C. J. Nyman, J. Am.
Chem. Soc., 1970, 92, 5873.
31 (a) D. F. Christian, G. R. Clark, W. R. Roper, J. M. Waters and
K. R. Whittle, J. Chem. Soc., Chem. Commun., 1972, 458; (b) R. Ugo,
S. Conti, S. Cenini, R. Mason and G. B. Robertson, Chem. Commun.,
1968, 1498.
32 C. Bianchini, A. Meli, M. Peruzzini and M. Vizza, J. Am. Chem.
Soc., 1990, 112, 6726.
16 (a) M. Brookhart, M. L. H. Green and L.-L. Wong, Prog. Inorg.
Chem., 1988, 36, 1; (b) R. H. Crabtree, E. M. Holt, M. Lavin and
S. M. Morehouse, Inorg. Chem., 1985, 24, 1986; (c) R. H. Crabtree,
Angew. Chem., Int. Ed. Engl., 1993, 32, 789.
17 (a) H. J. Wasserman, G. J. Kubas and R. R. Ryan, J. Am. Chem.
Soc., 1986, 108, 2294; (b) D. M. Heinekey, B. M. Schomber and
C. E. Radzewich, J. Am. Chem. Soc., 1994, 116, 4515; (c) M. D. Butts,
J. C. Bryan, X. L. Luo and G. J. Kubas, Inorg. Chem., 1997, 36,
3341.
33 (a) The irreversible electrochemical oxidation of [RuH(Cl)(dppp) ]
2
gives
a
RuIV hydride that rapidly dissociates H` to give
[RuCl(dppp) ]` with formal RuIV ] RuII reduction.33b (b) A. A.
2
Batista, L. A. Centeno Cordeiro and G. Oliva, Inorg. Chim. Acta,
1993, 203, 185.
34 P. Barthazy, M. Worle and A. Mezzetti, J. Am. Chem. Soc., in
press.
35 (a) An analogous mechanism has been suggested for RhIII and IrIII
peroxo complexes,1g and both terminal and bridged metalÈoxo
oxidants with non-radical character abstract H~ from hydrocar-
bons.35b,c However, this is not probable here, since 4 does not
18 A. G. Orpen, L. Brammer, F. H. Allen, O. Kennard, D. G. Watson
and R. Taylor, J. Chem. Soc., Dalton T rans., 1989, S1.
19 Recently, reports of 14-electron ruthenium(II) complexes stabilized
by agostic interactions have appeared: (a) D. Huang, J. C.
Hu†man, J. C. Bollinger, O. Eisenstein and K. G. Caulton, J. Am.
Chem. Soc., 1997, 119, 7398; (b) D. Huang, W. E. Streib, O. Eisen-
stein and K. G. Caulton, Angew. Chem., Int. Ed. Engl., 1997, 36,
2004.
react with cyclohexene [BDE(CwH
) B 86 kcal mol~1,
formyl
BDE(CwH ) B 87 kcal mol~1]; (b) K. A. Gardner, L. L. Kuhe-
allyl
nert and J. M. Mayer, Inorg. Chem., 1997, 36, 2069. (c) K. Wang
and J. M. Mayer, J. Am. Chem. Soc., 1997, 119, 1470.
36 A. Sen and J. Halpern, J. Am. Chem. Soc., 1977, 99, 8337.
37 Stoe & Cie Scanner Stoe IPDS Di†ractometer Software, version
2.75, Stoe & Cie, Darmstadt, Germany, 1996.
38 G. M. Sheldrick, SHEL XT L PL US. Program for the Solution of
Crystal Structures, University of Gottingen, Germany, 1985.
39 Bruker AXS, Madison, WI, USA, SHEL XT L Program Package V .
5.1, 1997
20 For a discussion of p-donor e†ects, see: K. G. Caulton, New J.
Chem., 1994, 18, 25.
21 (a) M. F. Sisemore, M. Selke, J. N. Burstyn and J. S. Valentine,
Inorg. Chem., 1997, 36, 979; (b) M. Selke, M. F. Sisemore and J. S.
Valentine, J. Am. Chem. Soc., 1996, 118, 2008. (c) M. F. Sisemore,
J. N. Burstyn and J. Valentine, Angew. Chem., 1996, 35, 195; (d) M.
Selke, M. F. Sisemore, R. Y. N. Ho, D. L. Wertz and J. S. Valen-
tine, J. Mol. Catal. A, 1997, 117, 71.
22 M. Selke, L. Rosenberg, J. M. Salvo and C. S. Foote, Inorg. Chem.,
1996, 35, 4519.
Paper 8/08860H
206
New J. Chem., 1999, 199È206