Mononuclear and dinuclear complexes with a [Ru(tBu 2PCH2CH2PtBu2)(CO)] core

Article abstract of DOI:10.1039/b415812a

Thermolysis of solid [Ru(d^tbpe)(CO)₂Cl₂] (2, d^tbpe = ^tBu₂PCH₂CH ₂P²Bu₂) under vacuum affords the five-coordinate complex [Ru(d^tbpe)(CO)Cl₂] (4), which was shown by X-ray crystallography to contain a weak remote agostic interaction. In solution, 4 can be readily trapped by CO, CH₃CN or water to give [Ru(d^tbpe)(CO)(L)Cl₂] (L = CO, 2; L = CH₃CN, 6; L = H₂O,7). Reaction of 4 with AgOTf/H₂O yields the tris-aqua complex [Ru(d^tbpe)(CO)(H₂O)₃](OTf) ₂ (8), which has been structurally characterised and probed in solution by pulsed-gradient spin echo (PGSE) NMR spectroscopy. The water ligands in 8 are labile and easily substituted to give [Rutd^tbpe)(CO) (NCCH₃)₃](OTf)₂ (10) and [Ru(d ^tbpe)(CO)(DMSO)₃](OTf)₂ (11). In the presence of CO, the tris-aqua complex undergoes water-gas shift chemistry with formation of the cationic hydride species [Ru(d^tbpe)(CO)₃H](OTf) (12) and CO₂. X-Ray crystal structures of complexes 2, 4, 6, 8 and 11-12 are reported along with those for [{Ru(d^tbpe)(CO)} ₂(μ-Cl)₂(μ-OTf)](OTf) (3), [{Ru(d ^tbpe)(CO)}₂(μ-Cl)₃][Ru(d^tbpe)(CO) Cl₃] (5) and [Ru(d^tbpe)(CO)(H₂O) ₂(OTf)](OTf) (9).

Full text of DOI:10.1039/b415812a

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Mononuclear and dinuclear complexes with a
t
t
[
Ru( Bu PCH CH P Bu )(CO)] core
2 2 2 2
a
a
a
b
Jose M. Goicoechea, Mary F. Mahon, Michael K. Whittlesey,* P. G. Anil Kumar and
b
Paul S. Pregosin
a
Department of Chemistry, University of Bath, Claverton Down, Bath, UK BA2 7AY.
E-mail: chsmkw@bath.ac.uk; Fax: 44 (0)1225 386231; Tel: 44 (0)1225 383748
Laboratory of Inorganic Chemistry, ETHZ, H o¨ nggerberg, CH-8093, Zurich, Switzerland
b
Received 13th October 2004, Accepted 25th November 2004
First published as an Advance Article on the web 17th December 2004
t
t
t
t
Thermolysis of solid [Ru(d bpe)(CO)
2
Cl
2
] (2, d bpe = Bu
2
PCH
2
CH
2
P Bu ) under vacuum affords the five-coordinate
2
t
complex [Ru(d bpe)(CO)Cl
2
] (4), which was shown by X-ray crystallography to contain a weak remote agostic
t
interaction. In solution, 4 can be readily trapped by CO, CH
; L = CH CN, 6; L = H O, 7). Reaction of 4 with AgOTf/H
CO)(H O) ](OTf) (8), which has been structurally characterised and probed in solution by pulsed-gradient spin
3
CN or water to give [Ru(d bpe)(CO)(L)Cl
O yields the tris-aqua complex [Ru(d bpe)-
2
] (L = CO,
t
2
3
2
2
(
2
3
2
echo (PGSE) NMR spectroscopy. The water ligands in 8 are labile and easily substituted to give
t
t
[
Ru(d bpe)(CO)(NCCH
3
)
3
](OTf)
2
(10) and [Ru(d bpe)(CO)(DMSO)
3
](OTf) (11). In the presence of CO, the tris-aqua
2
t
complex undergoes water–gas shift chemistry with formation of the cationic hydride species [Ru(d bpe)(CO) H](OTf)
3
(
12) and CO
2
. X-Ray crystal structures of complexes 2, 4, 6, 8 and 11–12 are reported along with those for
t
t
t
[
[
{Ru(d bpe)(CO)}
2
(l-Cl)
2
(l-OTf)](OTf) (3), [{Ru(d bpe)(CO)}
2
(l-Cl)
3
][Ru(d bpe)(CO)Cl ] (5) and
3
t
Ru(d bpe)(CO)(H
2
2
O) (OTf)](OTf) (9).
Introduction
of an inseparable mixture containing all-cis and trans-(CO) -
2
t
isomers upon refluxing [Ru(CO)
2
Cl
2
]
2
with d bpe under a slow
For some time now we have been interested in the properties
of low-valent organometallic ruthenium complexes containing
water as a ligand. Our previous work has concentrated on
5
flow of CO. Assignment of all stereochemistries was based
31
13
upon P/ C NMR spectroscopy and IR data. During one
of our first efforts to prepare this mixture, we were able to
crystallise the more unusual trans-dicarbonyl isomer (unusual
on the grounds of the two p-acceptor ligands occupying trans
1
2
+
[
Ru(dppe)(CO)(H
2
O)
3
]
(1), which can be isolated and crys-
−
−
tallised with a range of counterions, including OTf (1a), BF
4
−
−
(
1b), SbF
6
(1c) and NTf
2
(1d). The nature of the anion affects
6
positions) and confirm the geometry unambiguously through
hydrogen bonding patterns in the solid structures of 1a–d, water
solubility and, as recently shown using pulsed gradient spin-
echo (PGSE) NMR measurements, the diffusion properties of
X-ray crystallography, as shown in Fig. 1. The structure exhibits
some distortion from a regular octahedron, demonstrated most
◦
clearly by the C(2)–Ru(1)–C(1) angle of 160.86(9) (Table 1).
2
the complex as a whole in solution. As expected, considering
This trans-CO–Ru–CO angle is much more obtuse in the
the nature of the anions employed, all four complexes display the
same substitution chemistry at the metal centre; thus, all three
coordinated water ligands undergo facile replacement by co-
◦
7
few reported analogues; all-trans-[Ru(PPh
3
)
2
(CO)
(180.0 ) or trans-[Ru(Ph
P–R–PPh calix[4]arene
2
Cl
2
] (178 ),
◦
8
all-trans-[Ru(PBzl)
R–PPh )(CO) Cl
2
(CO)
(177.2 ; Ph
2
Cl
2
]
2
P–
◦
2
2
2
]
2
2
=
ordinating two-electron donors, including acetonitrile, Me
2
SO
9
t
phosphine). This distorting influence by the bulky Bu groups
of the bidentate phosphine is a common factor in all structurally
characterised compounds reported herein. We note that the
3
and bipyridyl.
As a step towards investigating how reactivity might be
influenced by the other ancillary ligands on the metal, we now
describe our efforts to prepare a bidentate alkylphosphine
analogue of 1a, namely the 1,2-di-tert-butylphosphinoethane
complex, [Ru(d bpe)(CO)(H
◦
phosphine bite angle in 2 (P(1)–Ru–P(2), 86.570(18) ) is similar
t
5,10
to that reported for other complexes bearing d bpe.
t
2
O)
3
][OTf] (8). While examples of
2
ruthenium complexes bearing chelating alkylphosphines are
4
still relatively small in number, of most relevance to our
current work were previous studies by Caulton and co-workers
t
5
who isolated [Ru(d bpe)(CO)
to [Ru(dppe)(CO) Cl
2
Cl
2
] (2). However, in contrast
2
2
], which afforded 1a–d upon silver salt
t
metathesis in the presence of water, the 18-electron d bpe
dichloride (2) first needs to be ‘activated’ by CO loss to form the
t
isolable coordinatively unsaturated species [Ru(d bpe)(CO)Cl
2
]
(
4). Herein, we report the synthesis and characterisation of a
range of complexes including 4 and 8, and their reactions with
donor solvents. While much of the solution chemistry of 8 proves
to be comparable to 1d, the former does show enhanced reaction
towards CO in a water–gas shift pathway.
Results and discussion
Metathesis reactions of 2
t
The dichloride complex, [Ru(d bpe)(CO)
2
Cl
2
] (2), has been
Fig. 1 Molecular structure of trans-(CO) -2 (thermal ellipsoids repre-
2
reported by Caulton and co-workers to be formed as part
sented at 30% probability).
5
8 8
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5

View Article Online
◦
◦
t
Table 1 Selected bond lengths ( A˚ ) and angles ( ) for trans-(CO)
-
Table 2 Selected bond lengths ( A˚ ) and angles ( ) for [{Ru(d bpe)-
2
t
[
Ru(d bpe)(CO)
2
Cl
2
]·CH
2
Cl
2
(2)
(CO)}
2
(l-OTf)(l-Cl
2
](OTf)·3CHCl
3
(3)
Ru1–C2
Ru1–P2
Ru1–Cl1
1.950(2)
2.3759(6)
2.4475(6)
Ru1–C1
Ru1–P1
Ru1–Cl2
1.956(2)
2.3763(6)
2.4560(6)
Ru1–C2
Ru1–P1
Ru1–Cl2
Ru2–C3
Ru2–P3
Ru2–Cl2
1.806(4)
Ru1–O2
2.225(3)
2.3711(9)
2.4721(8)
1.808(4)
2.3714(10)
2.4742(9)
Ru1–P2
Ru1–Cl1
Ru2–O1
Ru2–P4
Ru2–Cl1
2.3803(10)
2.4823(9)
2.219(3)
2.3883(10)
2.4842(8)
C2–Ru1–C1
C1–Ru1–P2
C1–Ru1–P1
C2–Ru1–Cl1
P2–Ru1–Cl1
C2–Ru1–Cl2
P2–Ru1–Cl2
Cl1–Ru1–Cl2
160.86(9)
C2–Ru1–P2
C2–Ru1–P1
P2–Ru1–P1
C1–Ru1–Cl1
P1–Ru1–Cl1
C1–Ru1–Cl2
P1–Ru1–Cl2
100.11(7)
94.03(7)
100.23(7)
83.76(7)
176.04(2)
83.23(7)
92.59(2)
88.67(2)
93.48(7)
86.570(18)
82.38(7)
92.40(2)
83.31(7)
176.41(2)
C2–Ru1–O2
O2–Ru1–P1
O2–Ru1–P2
C2–Ru1–Cl2
P1–Ru1–Cl2
C2–Ru1–Cl1
P1–Ru1–Cl1
Cl2–Ru1–Cl1
C3–Ru2–P3
C3–Ru2–P4
P3–Ru2–P4
O1–Ru2–Cl2
P4–Ru2–Cl2
O1–Ru2–Cl1
P4–Ru2–Cl1
Ru1–Cl1–Ru2
162.78(13)
95.53(7)
101.45(7)
85.35(11)
178.40(3)
84.29(12)
95.46(3)
83.26(3)
94.89(12)
91.72(13)
86.33(3)
83.08(7)
95.91(3)
80.91(7)
178.51(3)
95.89(3)
C2–Ru1–P1
C2–Ru1–P2
P1–Ru1–P2
O2–Ru1–Cl2
P2–Ru1–Cl2
O2–Ru1–Cl1
P2–Ru1–Cl1
C3–Ru2–O1
O1–Ru2–P3
O1–Ru2–P4
C3–Ru2–Cl2
P3–Ru2–Cl2
C3–Ru2–Cl1
P3–Ru2–Cl1
Cl2–Ru2–Cl1
Ru1–Cl2–Ru2
95.48(12)
92.44(12)
86.27(4)
83.35(7)
95.07(3)
81.52(7)
176.43(3)
164.06(13)
96.42(7)
100.16(7)
85.18(12)
177.75(3)
87.04(13)
94.59(3)
83.17(3)
96.41(3)
Attempts to react 2 with AgOTf (OTf = OSO
2
CF
3
) in CH
2
Cl
2
either in the absence or presence of water failed to cleanly
t
yield either the bis-triflate complex [Ru(d bpe)(CO)
the aqua complex 8. When the solvent was changed to dry
toluene, partial metathesis of the ruthenium chloride bonds
2
(OTf) ] or
2
in 2 occurred yielding the cationic triflate bridged dimer
t
[
{Ru(d bpe)(CO)}
2
(l-OTf)(l-Cl) ](OTf) (3), which was isolated
2
as a yellow microcrystalline solid in 64% yield (Scheme 1).
3
1
1
P{ H} NMR spectroscopy of 3 in CDCl
3
revealed a singlet
1
9
at 99.8 ppm, while the F NMR spectrum exhibited a sharp,
coordinated triflate resonance at −77.84 ppm, along with a
broader signal from non-coordinated triflate at −78.63 ppm.
Crystals of the complex suitable for X-ray crystallography were
from an octahedral geometry at the ruthenium centres; the
angles between the carbonyl ligands and the coordinated
oxygen atoms of the bridging triflate are 162.78(13) and
◦
164.06(13) for Ru(1) and Ru(2), respectively (Table 2). The
grown from CHCl
Fig. 2.
3
–hexane and revealed the structure shown in
bridging triflate group is a rare structural motif in late tran-
sition metal complexes and has only been identified in the
1
1
solid state. The Ru–O bond lengths in 3 (2.225(3), 2.219(3)
˚
A) are in between those found in the bis-bridging triflate
11d
species [{Ru(CO)
2
(CH(OCOC
2
H
5
)CH
3
)(OTf)}
2
]
(2.171(3),
˚
2
.293(3) A), but significantly longer than the terminal Ru–O
˚
distances of 2.182(4) and 2.197(4) A reported by us in all-cis
Ru(dppe)(CO)
1
[
2
(OTf)
2
].
t
Solid-state thermolysis of [Ru(d bpe)(CO)
2
2
Cl ]
◦
Upon heating a solid sample of 2 at 120 C under vacuum
for 36 h (Scheme 2), a distinct colour change from yellow
to orange-red occurred due to formation of the 16-electron
t
complex [Ru(d bpe)(CO)Cl
2
] (4). The appearance of a deep
colouration has previously been reported as being consistent
with formation of such coordinatively unsaturated ruthenium
systems.
Scheme 1 Synthesis of 3 by silver salt metathesis.
12
Scheme 2 Formation of the coordinatively unsaturated complex 4.
The low-temperature (150 K) X-ray structure (Fig. 3) con-
firmed the unsaturated and monomeric nature of 4 in addition
to showing that it adopts a distorted square-pyramidal structure
with the chelating phosphine and the two chloride ligands
forming the base plane and the apical CO trans to the vacant
13
site. The ruthenium is lifted a little out of the basal plane with
P(1)–Ru(1)–Cl(2) and P(2)–Ru(1)–Cl(1) angles of 167.902(15)
◦
and 172.512(14) , respectively (Table 3). The coordinative
t
unsaturation is stabilised to some extent by the presence of a Bu
C–H · · · Ru interaction. Thus, the Ru(1)–P(1)–C(8) angle is con-
Fig. 2 X-Ray structure of the cation in 3 (thermal ellipsoids represented
◦
at 30% probability).
siderably more acute (105.78(5) ) than expected (cf. Ru(1)–P(2)–
◦
C(16), 116.94(5) ) while the Ru(1)–C(11) distance (3.1205(17)
t
˚
˚
The two [Ru(d bpe)(CO)] fragments are bridged by two
A) is over 0.5 A shorter than the comparative distance Ru(1)–
˚
chloride substituents located in the equatorial plane along with
the phosphines. Particularly striking is the distortion away
C(19) of 3.6722(18) A. This provides clear evidence for a
contact between the ruthenium metal centre and C(11). The
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7
5 8 9

View Article Online
◦
t
Table 3 Selected bond lengths ( A˚ ) and angles ( ) for [Ru(d bpe)-
CO)Cl ] (4)
CO (1 atm) in CH
of [Ru(d bpe)(CO) Cl ] in a thermodynamic ratio. Complex 4
Cl
2
at room temperature to give both isomers
2
2
2
t
(
2
16
exhibited no reaction with H
Over a period of several hours at room temperature, CD
solutions of 4 began to exhibit two new, broad resonances in
2
.
Ru1–C1
Ru1–P2
Ru1–Cl1
Ru1–H19B
Ru1–H19A
1.7838(15) Ru1–P1
2.3267(4)
2.4031(4)
2.55(2)
2
Cl
2
2.3660(4)
2.4103(4)
3.12(2)
Ru1–Cl2
Ru1–H11A
Ru1–H11B
3
1
1
the P{ H} NMR spectrum at 101.3 and 101.9 ppm. These
3.20(2)
signals arise from what have been identified by Caulton as the
3.83(2)
t
5,17
cis-(CO)
2
and trans-(CO)
2
isomers of [Ru(d bpe)(CO)Cl ] .
2
2
C1–Ru1–P1
P1–Ru1–P2
P1–Ru1–Cl2
C1–Ru1–Cl1
P2–Ru1–Cl1
C1–Ru1–H11A
P2–Ru1–H11A
Cl1–Ru1–H11A
P1–Ru1–H19B
Cl2–Ru1–H19B
P1–Ru1–H11B
Cl2–Ru1–H11B
C1–Ru1–H19A
P2–Ru1–H19A
Cl1–Ru1–H19A
C4–P1–Ru1
92.53(5)
86.702(13)
167.902(15)
90.98(5)
172.512(14)
154.9(5)
101.6(5)
71.9(5)
110.8(4)
58.4(4)
61.5(4)
106.6(4)
148.2(3)
51.8(3)
120.8(3)
122.85(5)
115.54(5)
173.16(13)
C1–Ru1–P2
C1–Ru1–Cl2
P2–Ru1–Cl2
P1–Ru1–Cl1
96.38(5)
99.57(5)
91.915(14)
94.384(14)
85.471(15)
71.3(5)
97.3(5)
146.2(4)
62.4(4)
110.4(4)
152.0(4)
73.2(4)
100.8(4)
87.5(3)
In many cases monomer/dimer interconversion (via bridging
halides) is detected, although this largely depends upon the steric
bulk of the phosphine; bulky PR
Lewis acidity of a [Ru(PR (CO)Cl
of a monomer, whereas small PR ligands lead to dimers, that
3
groups that outweigh the
3
)
2
2
] fragment allow isolation
Cl2–Ru1–Cl1
P1–Ru1–H11A
Cl2–Ru1–H11A
C1–Ru1–H19B
P2–Ru1–H19B
Cl1–Ru1–H19B
C1–Ru1–H11B
P2–Ru1–H11B
Cl1–Ru1–H11B
P1–Ru1–H19A
Cl2–Ru1–H19A
C2–P1–Ru1
3
18
reduces the overall Lewis acidity. Attempted crystallisation of
one or both of the Caulton dimers afforded instead a mixture of
t
4
Cl)
and the cationic trichloride bridged salt [{Ru(d bpe)(CO)}
2
(l-
t
3
][Ru(d bpe)(CO)Cl
3
] (5) (Scheme 3).
82.2(3)
107.76(5)
105.78(5)
116.94(5)
C12–P2–Ru1
O1–C1–Ru1
C8–P1–Ru1
C16–P2–Ru1
Scheme 3 Trichloride bridged salt 5.
t
t
Structure of [{Ru(d bpe)(CO)}
2
(l-Cl)
3
][Ru(d bpe)(CO)Cl
3
] (5)
The structure of 5 (Fig. 4, Table 4) revealed a cationic moiety
consisting of two face-sharing distorted octahedra, with a non-
19
˚
bonding Ru · · · Ru separation of 3.39 A. Accompanying the
t
− 20
cation is an anion of formula [Ru(d bpe)(CO)Cl ] . Unfortu-
3
nately, the data were not of the highest quality, but the structure
determination is unequivocal. While we do not make detailed
claims with respect to the metric data for 5, some trends are
obvious. In particular, the anion exhibits the same distortion of
t
the octahedral geometry with d bpe; this is most observable in
◦
◦
the Cl(4)–Ru(3)–C(3) angle (160.3(8) ), which is ca. 20 more
acute than would be expected of an ideal octahedral geometry.
The two Ru–CO distances in the cationic moiety are comparable
Fig. 3 Molecular structure of 4 (thermal ellipsoids represented at 30%
probability).
˚
1.76(3) and 1.83(2) A) to the Ru–CO distance exhibited by the
(
hydrogen atoms are included at calculated positions in the final
refinement of 4, except for those attached to C(9) and C(11).
Hence we can see that the Ru(1)–H(11a) distance (2.55 A) is
◦
t
Table 4 Selected bond lengths ( A˚ ) and angles ( ) for [{Ru(d bpe)-
˚
t
(
CO)}
2
(l-Cl
3
][Ru(d bpe)(CO)Cl
3
] (5)
also considerably shorter than any other distance between a
hydrogen atom and the metal centre. The distances involved
imply that this interaction could only be classed as a very
Ru1–P2
Ru1–Cl2
Ru2–P3
Ru2–Cl2
Ru1–Cl1
Ru2–Cl1
Ru3–P5
Ru3–Cl4
2.382(6)
2.479(6)
2.401(7)
2.476(6)
2.548(6)
2.543(6)
2.369(7)
2.473(6)
Ru1–P1
2.379(6)
2.477(6)
2.369(6)
2.475(6)
2.368(6)
2.452(6)
2.478(6)
Ru1–Cl3
Ru2–P4
Ru2–Cl3
Ru3–P6
Ru3–Cl5
Ru3–Cl6
14
weak ‘remote’ agostic interaction. Interestingly, there are also
short Ru–Cl · · · H–C contacts in this structure (Cl(1) · · · H(7A),
˚
˚
˚
2
.89 A; Cl(1) · · · H(10B) 2.70 A; Cl(2) · · · H(19B) 2.77 A;
15
˚
Cl(2) · · · H(17A) 2.66 A). Unsurprisingly, the Ru–CO bond
˚
length (1.7838(15) A) is substantially shorter than the com-
˚
parable distances in 2 (1.948, 1.954 A); a similar constriction
C1–Ru1–P1
P1–Ru1–P2
93.0(7)
86.3(2)
95.4(2)
177.9(2)
94.7(2)
83.64(19)
104.2(2)
75.8(2)
95.0(2)
95.1(2)
83.75(19)
103.3(2)
76.8(2)
Ru2–Cl2–Ru1
P1–Ru1–Cl2
P1–Ru1–Cl2
P1–Ru1–Cl1
Cl3–Ru1–Cl1
P4–Ru2–P3
P4–Ru2–Cl3
P3–Ru2–Cl2
P3–Ru2–Cl1
Cl2–Ru2–Cl1
P5–Ru3–Cl5
P6–Ru3–Cl4
Cl5–Ru3–Cl4
P6–Ru3–Cl6
Cl5–Ru3–Cl6
Ru2–Cl1–Ru1
Ru2–Cl3–Ru1
86.4(2)
178.1(2)
178.1(2)
105.6(2)
76.7(2)
86.2(2)
178.8(2)
177.5(2)
105.9(2)
76.0(2)
91.4(2)
96.6(2)
84.2(2)
90.1(2)
91.4(2)
83.52(18)
86.4(2)
occurs in the Ru–Cl distances (2.4031(4), 2.4103(4) cf . 2.4475(6),
˚
.4560(6) A). Consistent with the expected enhanced back-
1
2
P1–Ru1–Cl3
P2–Ru1–Cl3
P2–Ru1–Cl2
Cl3–Ru1–Cl2
P2–Ru1–Cl1
Cl2–Ru1–Cl1
P3–Ru2–Cl3
P4–Ru2–Cl2
Cl3–Ru2–Cl2
P4–Ru2–Cl1
Cl3–Ru2–Cl1
P6–Ru3–P5
P6–Ru3–Cl5
P5–Ru3–Cl4
P5–Ru3–Cl6
Cl4–Ru3–Cl6
−
bonding from the metal, m(CO) appeared at 1934 cm in the
IR spectrum, significantly lower in frequency than the bands
−
1
arising from 2 (all-cis isomer: 2064, 1984 cm ; trans-CO isomer,
−
1
2
013 cm ).
The solution behaviour of 4 seems to be somewhat
contradictory–this is dicussed in more detail later with respect
to diffusion properties. What is evident upon dissolving a
microcrystalline sample of 4 in CD
2
2
Cl is that there was no
evidence for any classical agostic interaction in solution by either
1
13
1
◦
H or C{ H} NMR spectroscopy down to −80 C. As expected
for the geometry shown in Fig. 3, one CO triplet (d 205.9, JCP
87.1(2)
178.4(2)
102.1(3)
175.9(2)
81.2(2)
=
1
3
1
1
6.5 Hz) was observed in the C{ H} NMR spectrum and a
31 1
singlet seen by P{ H} spectroscopy at d 96.2. Additional proof
for the unsaturated nature of 4 was shown by its reaction with
5
9 0
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7

View Article Online
Fig. 4 Molecular structure of 5 (thermal ellipsoids represented at 50% probability).
˚
Table 5). The Ru–CO distance (1.825(3) A) is longer than that
(
˚
seen in 4 (1.7838(15) A) due to the presence of the good r-
donating acetonitrile ligand in the trans position.
t
In wet dichloromethane, the mono-aqua complex [Ru(d bpe)-
(
CO)(H
2
O)Cl ] (7) was formed, although it proved to be only
2
sparingly soluble in conventional NMR solvents, preventing full
characterisation. However, the appearance of a single band in
−
1
the carbonyl region of the IR spectrum at 1950 cm compares
−
1
favourably with the values of 1934 and 1961 cm found for
complexes 4 and 6, respectively.
Scheme 4 Reactions of the coordinatively unsaturated complex 4.
˚
anion (Ru(3)–C(3) 1.88(3) A). The Ru–Cl bond distances for the
˚
cation (2.479(6), 2.477(6), 2.475(6), 2.476(6) A) are of similar
values and slightly shorter than the bond distances between
the metal centre and the axial chloride substituent (2.548(6),
˚
2
.543(6) A). The three Ru–Cl distances in the anion are similar.
The IR spectrum of 5 in Nujol displayed two intense CO bands
−
1
at 1979 and 1932 cm , the former also showing a shoulder at
−
1
1
m
[
963 cm . Useful comparisons are somewhat limited, although
−
1
CO at 1977 cm has been reported for the arylphosphine cation
+
{Ru((S)-biphemp)(CO)}
2
(l-Cl)
3
] , while the IR spectrum of
Fig. 5 X-Ray structure of one molecule in the asymmetric unit of 6
−
the more distantly related anion [Ru(AsPh
3
)
2
(CO)Cl
3
] contains
(
thermal ellipsoids represented at 30% probability).
−
1 20,21
a band at 1928 cm .
In solution, the salt proved to be
unstable, cleaving to yield resonances for 4 and the Caulton
t
31
1
◦
t
dimer, [Ru(d bpe)(CO)Cl
2
]
2
, by P{ H} NMR spectroscopy.
Table 5 Selected bond lengths ( A˚ ) and angles ( ) for [Ru(d bpe)-
(
3 2
CO)(NCCH )Cl ] (6)
Reaction of the agostic complex with coordinating solvents
Ru1–C1
Ru1–P1
Ru1–C1
1.825(3)
2.3744(8)
2.4675(8)
Ru1–N1
Ru1–P2
Ru1–Cl2
2.128(3)
2.3824(8)
2.4807(8)
In addition to reacting with CO, complex 4 was readily
trapped by coordinating solvents as illustrated in Scheme 4.
t
[
Ru(d bpe)(CO)(NCCH
3
)Cl
2
] (6) was isolated upon addition
Cl solution of 4.
C1–Ru1–N1
N1–Ru1–P1
N1–Ru1–P2
C1–Ru1–Cl1
P1–Ru1–Cl1
C1–Ru1–Cl2
P1–Ru1–Cl2
Cl1–Ru1–Cl2
164.91(11)
99.27(6)
93.82(7)
84.82(9)
90.71(3)
86.72(10)
177.43(3)
91.24(3)
C1–Ru1–P1
C1–Ru1–P2
P1–Ru1–P2
N1–Ru1–Cl1
P2–Ru1–Cl1
N1–Ru1–Cl2
P2–Ru1–Cl2
91.79(10)
97.02(9)
86.86(3)
84.81(7)
176.99(3)
82.57(7)
91.24(3)
of 10 equivalents of acetonitrile to a CD
2
2
3
1
1
The single resonance revealed by P{ H} NMR spectroscopy
indicates that the two chloride ligands must be coplanar with the
t
5
d bpe; this was proven unambiguously by a single-crystal X-ray
diffraction study on crystals grown from acetonitrile solution
at low temperature. The structure (Fig. 5) shows a distorted
◦
octahedral geometry with a C(1)–Ru(1)–N(1) angle of 164.91
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7
5 9 1

View Article Online
t
◦
t
Characterisation and isolation of [Ru(d bpe)(CO)(H
(
2
O)
3
](OTf)
2
Table 6 Selected bond lengths ( A˚ ) and angles ( ) for [Ru(d bpe)-
8)
Addition of just over two equivalents of AgOTf to a solution
of 4 in CH Cl in the presence of water afforded the tris-aqua
complex [Ru(d bpe)(CO)(H
(CO)(H
2
O)
2
(OTf)](OTf) (9)
Ru1–C1
Ru1–O2
Ru1–P1
1.808(5)
2.176(3)
2.3652(12)
Ru1–O3
Ru1–O4
Ru1–P2
2.176(4)
2.209(4)
2.3490(11)
2
2
t
2
O)
3
](OTf)
2
(8) as a colourless crys-
talline solid in 68% isolated yield (Scheme 4). In contrast to the
analogous dppe complex 1a, which is essentially insoluble in wa-
ter, 8 displayed remarkably high water-solubility (120 mg/mL).
C1–Ru1–O3
O3–Ru1–O2
O3–Ru1–O4
C1–Ru1–P1
O2–Ru1–P1
C1–Ru1–P2
O2–Ru1–P2
P1–Ru1–P2
162.77(19)
79.86(14)
79.23(19)
92.65(16)
178.32(10)
92.92(16)
91.52(10)
86.81(4)
C1–Ru1–O2
C1–Ru1–O4
O2–Ru1–O4
O3–Ru1–P1
O4–Ru1–P1
O3–Ru1–P2
O4–Ru1–P2
87.40(19)
89.6(2)
91.43(14)
100.43(11)
90.25(11)
98.91(12)
176.18(14)
In D O solution, 8 appeared as a singlet at d 100.5 in the
2
3
1
1
P{ H} NMR spectrum and showed the expected triplet in the
1
3
1
carbonyl region of the C{ H} NMR spectrum at 205.0 ppm.
1
9
The F NMR spectrum displayed a single uncoordinated triflate
resonance at −79.5 ppm, while in the IR spectrum, a single
−
1
carbonyl band was found at 1971 cm . This shift to lower
−
1
frequency compared with mCO for 1a (1990 cm ) is as expected for
as it located trans to the electron-donating phosphine (Table 6).
The distortion of the molecular geometry because of the d bpe
ligand is best illustrated by the acute C(1)–Ru(1)–O(3) bond
angle of 163.8(2) .
t
replacement of dppe by the more electron-donating d bpe. The
t
structure of 8, elucidated by X-ray crystallography, is shown in
Fig. 6. However, as this is a poor quality refinement for reasons
outlined later, we make no further claims herein based on the
associated metric data for 8.
◦
Diffusion properties of 8 in solution
Pulsed gradient spin-echo (PGSE) diffusion studies performed
on the tris-aqua complex 8 show that it behaves very similarly
2,22
to the dppe analogue 1a in solution (Table 7).
Diffusion
constants of 8.78 × 10 and 8.88 × 10 m s were recorded
for the cation and anion, respectively, in CD Cl /water solution;
−
10
−10
2
−1
2
2
these almost identical values can be rationalised by either ion-
pairing or hydrogen bonding of the anion to the coordinated
water ligands. As the experiments were necessarily performed
in the presence of excess water to help stabilise 8, ion pairing
would be broken up leaving us to favour the hydrogen bonding
explanation.
We now return to the solution behaviour of the structurally
t
characterised unsaturated species [Ru(d bpe)(CO)Cl
2
]
(4).
Table 7 shows the observed D and r values for the mixture
H
Fig. 6 X-ray structure of one of the cations in the asymmetric unit of
of cis and trans isomers of 2, at both ambient temperature and
223 K; as expected, both sets of data are consistent with the
properties of a mononuclear complex. Interestingly, 2 was found
to have a larger D value than 8, in keeping with a larger volume
for 8 due to hydrogen bonding. The low-temperature diffusion
measurements were repeated with different mixing times to
exclude the probability of convection. However, upon comparing
the D values (223 K) for the cis/trans mixture of 2 with that for 4,
shows that the latter is diffusing at a much slower rate. The ratio
of the D values (2.84/2.23) = 1.27 is in excellent agreement with
a structure which has ca. double the molecular volume of 2 i.e.
this indicates that upon dissolution, 4 exists as a dimeric species
and not the monomeric coordinatively unsaturated species that
we have structurally characterised.
8
(thermal ellipsoids represented at 30% probability).
During repeated attempts to grow higher quality crystals
t
of 8, the bis aqua complex [Ru(d bpe)(CO)(H
2
2
O) (OTf)](OTf)
23
(
9), was continually detected as an ‘impurity’ in crystalline
batches of 8. The asymmetric unit in 9 consists of two cations,
two triflate anions and two heavily disordered molecules of
dichloromethane (Fig. 7). The coordinated triflate substituent
is occupying one of the equatorial positions trans to P(1) and cis
to P(2) with a coordinated water molecule (O(2)) occupying the
remaining equatorial position. The water molecule in the equa-
torial position exhibits a slightly longer Ru–OH bond distance
2
˚ ˚
2.171(4) A) to that of the axial water substituent (2.165(4) A),
(
Substitution reactions of 8
The tris-aqua complex displayed the same rapid react-
ivity towards CH
3
CN or DMSO as seen with 1a leading to
replacement of all three water ligands by solvent to afford
a,b
c
Table 7 Diffusion constants and radii for 2, 4 and 8
b
H
r /A
˚
Compound
Fragment
T/K
D
2
300
10.55
2.84
2.85
2.23
8.78
8.88
5.1d
5.4
5.4
6.8
6.1
6.0
2
2
23
23
e
4
8
223
300
300
Cation
Anion
a
1
c
All measurements are for 2 mM solutions. Cations were measured by
19
b
−10
2
−1
H and anions using F. D values are given in units of 10 m s
.
The viscosities used in the calculations are as follows: CH Cl , 0.405
2
2
d
e
Fig. 7 X-Ray structure of the ordered cation in the asymmetric unit of
(300 K); 1.072 (223 K). d3 = 75 ms. d3 = 50 ms.
9
(thermal ellipsoids represented at 30% probability).
5
9 2
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7

View Article Online
t
t
[
Ru(d bpe)(CO)(NCCH
3
)
3
](OTf)
2
(10) and [Ru(d bpe)(CO)-
(
DMSO) ](OTf) (11) (Scheme 5).
3
2
Scheme 6 Water–gas shift chemistry to afford 12.
2
4
13
13
by loss of CO
2
1
.
When CO was employed, CO
2
was readily
Scheme 5 Substitution reactions of tris-aqua complex 8.
13
detected ( C{ H} NMR: d 126.4 in acetone-d
6
) in the reaction.
Exchange of CO with the isotopic label was also seen in the
In addition to characterisation by NMR spectroscopy, 11 was
25
starting tris-aqua complex.
also structurally characterised by X-ray diffraction. The struc-
ture is shown in Fig. 8, with selected bond lengths and angles
given in Table 8. As seen previously in [Ru(dppe)(CO)-
The hydride complex 12 showed a characteristic triplet (d
1
31
1
−
7.97) and singlet (d 110.7) in the H and P{ H} NMR
5
spectra respectively. As expected, there was no evidence for
(
DMSO)
3
](OTf)
2
, all three DMSO ligands in 11 are oxygen
19
bound triflate ( F NMR). The molecular geometry of 12
◦
bound to the metal centre with the Ru–O–S angle (123.37 ) of
the axial DMSO substituent intermediate between those found
in the equatorial positions (119.19 and 132.39 ; cf. 121.8 (eq),
was established unequivocally by X-ray crystallography. The
asymmetric unit is shown in Fig. 9 with selected bond lengths
and angles given in Table 9. The complex displays a regular
octahedral geometry with two carbonyl groups occupying
◦
◦
◦
◦
131.4 (eq) and 118.5 (ax) found in the dppe analogue).
equatorial positions along with the bidentate phosphine. The
Water–gas shift chemistry of 8 to afford
Ru–COeq distances are very similar (1.917(3) and 1.923(3) A˚ for
t
[
Ru(d bpe)(CO)
Treatment of a solution of [Ru(d bpe)(CO)(H
acetone–H O with CO gave the cationic monohydride complex
Ru(d bpe)(CO) H](OTf) (12) as the sole ruthenium containing
product (Scheme 6). However, in contrast to the three weeks
needed for formation of [Ru(dppe)(CO) H](OTf), the synthesis
3
H](OTf) (12)
Ru(1)–C(3) and Ru(1)–C(1) respectively) while, as expected, the
˚
axial Ru–CO distance is elongated (1.999(3) A) due to the trans
t
t
6
2
O)
3
](OTf) in d -
2
influence of the hydride ligand. [Ru(d bpe)(CO) H](OTf) slowly
13
3
2
1
3
t
catalyses the transformation of CO (1 atm) to CO
2
, over a
[
3
◦
period of one month at 50 C. This contrasts with the behaviour
of [Ru(dppe)(CO) H](OTf) which, although it undergoes sub-
3
3
1
3
13
stitution with CO, yields no C-labelled carbon dioxide.
of 12 in good yield was complete after only 2 h at room
temperature (1 atm CO). Formation of a hydride product is
consistent with a water–gas shift (WGS) mechanism involving
attack of H O on one of the coordinated CO ligands, followed
2
Fig. 9 X-Ray structure of the cation in 12 (thermal ellipsoids repre-
sented at 30% probability).
Conclusions
A series of mononuclear and dinuclear complexes containing a
t
core [Ru(d bpe)(CO)] fragment have been synthesised during
Fig. 8 X-Ray structure of the cation in 11 (thermal ellipsoids repre-
sented at 30% probability).
efforts directed at the synthesis of the tris-aqua complex
t
2+
[
Ru(d bpe)(CO)(H
2
O)
3
]
(8). In contrast to its dppe analogue
◦
t
1d, formation of 8 via silver salt metathesis of the coordinatively
Table 8 Selected bond lengths ( A˚ ) and angles ( ) for [Ru(d bpe)-
(
CO)(DMSO)
3
](OTf)
2
(11)
◦
t
Table 9 Selected bond lengths ( A˚ ) and angles ( ) for [Ru(d bpe)-
CO) H](OTf) (12)
(
3
Ru1–C1
Ru1–O2
Ru1–P1
1.8170(17)
2.1612(12)
2.3831(4)
Ru1–O3
Ru1–O1
Ru1–P2
2.1297(12)
2.1858(12)
2.3958(4)
Ru1–C3
Ru1–C2
Ru1–P3
O2–C2
1.917(3)
1.999(3)
2.4470(7)
1.129(3)
Ru1–C1
Ru1–P2
O1–C1
O3–C3
1.923(3)
2.4456(6)
1.133(3)
1.131(3)
C1–Ru1–O3
O3–Ru1–O2
O3–Ru1–O1
C1–Ru1–P1
O2–Ru1–P1
C1–Ru1–P2
O2–Ru1–P2
P1–Ru1–P2
167.07(6)
78.92(5)
79.33(5)
93.73(6)
171.11(4)
94.08(5)
101.79(4)
85.750(15)
C1–Ru1–O2
C1–Ru1–O1
O2–Ru1–O1
O3–Ru1–P1
O1–Ru1–P1
O3–Ru1–P2
O1–Ru1–P2
90.44(6)
91.65(6)
79.76(5)
95.85(4)
92.26(3)
95.27(3)
174.04(3)
C3–Ru1–C1
C1–Ru1–C2
C1–Ru1–P2
C3–Ru1–P3
C2–Ru1–P3
91.53(11)
91.35(11)
90.83(8)
90.20(8)
97.96(7)
C3–Ru1–C2
C3–Ru1–P2
C2–Ru1–P2
C1–Ru1–P3
P2–Ru1–P3
91.76(10)
166.69(8)
101.28(7)
170.48(8)
85.40(2)
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7
5 9 3

View Article Online
+
−1
saturated precursor [Ru(P–P)(CO)
reaction via the 16-electron monocarbonyl [Ru(d bpe)(CO)Cl
2
Cl
2
] proved unsuccessful, and
H, 7.24%. FAB -mass spectrum: m/z 1003.2. m˜ /cm (CO):
t
2
]
1979, 1963sh, 1932 (Nujol).
4
was required. The solution chemistry of 8 proves in some
t
[
Ru(d bpe)(CO)(NCCH
3
3
)Cl ] (6). 10 equivalents (40 lL) of
2
respects (substitution reactions with CH CN/DMSO, diffusion
properties) to be comparable to the dppe analogue 1d, although,
importantly, 8 shows higher reactivity towards CO in affording
3
t
CH
CN were added to a solution of [Ru(d bpe)(CO)Cl
.096 mmol) in CH
The solvent was filtered off and the solid sample dried in
vacuo. Crystals of this complex were obtained from acetonitrile
2
] (50 mg,
0
2
2
Cl (0.5 mL) to give a yellow precipitate.
t
+
[
Ru(d bpe)(CO) H] . In water–gas shift chemistry, detection of
3
intermediate metal hydride species is often difficult, although
there are examples where hydride complexes prove to be
characterisable or even isolable.
◦
solution at −20 C (yield 35 mg, 65%) (Found: C, 45.4; H, 7.58;
26
N, 4.64. RuC21
H
43
P
2
ONCl
2
·CH
3
CN requires C, 46.00; H, 7.58;
1
N, 4.66%). H NMR (CD
3
CN): d 2.21 (br, 4H, CH
), 1.38 (d, JPH = 14.4 Hz, 18H, CH
CN): d 209.1 (t,
2
) 1.43 (d,
J
PH = 13.2 Hz, 18H, CH
3
3
).
Experimental
31
1
13
1
P{ H} (CD
3
CN): d 94.8 (s). C{ H} (CD
3
−
1
J
CP = 13.9 Hz, Ru–CO). m˜ /cm (CO) 1961 (Nujol).
General information
t
All reactions were carried out using standard Schlenk tech-
niques. Solvents (unless otherwise stated) were distilled from
purple solutions of sodium benzophenone ketyl (hexane,
[Ru(d bpe)(CO)(H O)Cl ] (7). 20 equivalents (35 lL) of
2
2
t
H O were added to a solution of [Ru(d bpe)(CO)Cl ] (50 mg,
0.96 mmol) in CH Cl (0.5 mL) to afford a yellow precipitate.
This was isolated by filtration and dried under vacuum. m˜ /cm
(CO): 1950 (Nujol). Due to the lack of solubility of this
compound in common solvents, NMR characterisation could
not be achieved.
2
2
2
2
−
1
toluene) or CaH
2
(chloroform, dichloromethane, acetonitrile)
and stored under argon. NMR solvents were purchased from
Aldrich and were dried, degassed and stored under argon
in gastight solvent bulbs; CD
uum transferred from CaH , d
thaw degassed and used without any further purification
and D O was degassed by bubbling with argon. AgOTf and
,2-bis(dichlorophosphino)ethane (Aldrich) were used without
2
Cl
2
, CDCl
3
, CD
3
CN were vac-
2
6
-acetone was freeze-pump-
t
[Ru(d bpe)(CO)(H O) ](OTf) (8). Silver triflate (109 mg,
2
3
2
2
0.42 mmol) was added to a Schlenk tube containing a solution
of [Ru(d bpe)(CO)Cl ] (100 mg, 0.19 mmol) in CH Cl (30 mL).
Immediately afterwards 10 equivalents of H O (35 lL, 1.9 mmol)
was added to the solution. The reaction mixture was allowed
to stir at room temperature in the absence of light for 2 h.
The reaction mixture was then filtered and the filtrate reduced
to a solid in vacuo. This solid was washed with hexane and
dried yielding a pale yellow powder which was characterised as
[Ru(d bpe)(CO)(H O) ](OTf) . Analytically pure crystals were
obtained from CH Cl in the presence of trace amounts of
t
1
2
2
2
1
3
further purification. CO (Cambridge Isotopes, 99%) and
CO (BOC, 99.9%) were used as received. Literature routes
2
t
27
t
5
were used to prepare d bpe, [Ru(d bpe)(CO)
2
Cl
2
] (2) and
t
5
[
Ru(d bpe)(CO)Cl ] . NMR spectra were recorded on Bruker
2
2
AVANCE 300, 400, Varian MERCURY 400 MHz (Bath) and
Bruker AVANCE 400 and 500 MHz NMR (ETHZ) spectrom-
eters and referenced to the chemical shifts of residual protio
solvent resonances (acetone d 2.05, dichloromethane d 5.35,
chloroform d 7.27, acetonitrile d 1.94, HOD d 4.79, DMSO
t
2
3
2
2
2
water upon cooling (yield 68 mg, 41%) (Found: C, 30.9; H,
1
3
1
1
d 2.50). C{ H} NMR spectra were referenced to d
d 29.9), CD Cl (d 53.8), CDCl (d 77.2) and CD CN (d 1.3).
P{ H} and F NMR chemical shifts were referenced externally
to 85% H PO (d 0.0) and CFCl (d 0.00), respectively. IR
spectra were typically recorded as Nujol mulls or occasionally
in aqueous solution using CaF
plates on a Nicolet Prot e´ g e´
6
-acetone
5.70. C H P O S F Ru·H O requires C, 30.84; H, 5.92%). H
2
1
46
2
10
2
6
2
(
2
19
2
3
3
NMR (D O): d 2.12 (br m, 2H, CH ), 1.91 (br m, 2H, CH ), 1.38
31 1 19
2
2
2
3
1
1
(br, 36H, CH ). P{ H} (D O): d 100.5 (s). F (D O): d −79.48
3
2
2
1
3
1
−1
3
4
3
(s). C{ H} (D O): 205.0 (t, J = 17.4 Hz, Ru–CO). m˜ /cm
2
CP
(CO) 1971 (H O).
2
2
t
FTIR spectrometer. Mass spectra were recorded on a Fisons VG
AutoSpec. Elemental analyses were performed at the University
of Bath.
[Ru(d bpe)(CO)(NCCH
3
)
3
](OTf)
2
2
(10). A solid sample of
t
[Ru(d bpe)(CO)(H
2
O)
3
](OTf)
(50 mg, 0.063 mmol) was placed
CN
20 mL). The solution was filtered and the filtrate taken
to dryness to give yellow powder, which was charac-
in a Schlenk tube under argon and dissolved in CH
3
(
t
[
{Ru(d bpe)(CO)}
2
(l-Cl) (l-OTf)](OTf) (3). AgOTf (103 mg,
2
0
[
.40 mmol) was added to a 30 mL toluene suspension of
t
terised as [Ru(d bpe)(CO)(NCCH
Found: RuC27
3
)
3
](OTf)
2
(yield 42 mg, 77%)
t
Ru(d bpe)(CO)
in the absence of light. The mixture was filtered by cannula
and the solid residue dried in vacuo. Dissolution in CHCl and
2
Cl
2
] (100 mg, 0.18 mmol) and stirred for 2 h
(
H
49
P
F
2
O
6
7
N
3
S
2
F
6
: C, 37.0; H, 5.42; N, 5.14.
1
RuC27
H
49
P
2
O
3
7
N
3
S
2
requires C, 37.33; H, 5.68; N, 4.84). H
), 1.50 (br, 36H, CH ).
CN): d 98.3 (s). F (CD
CN): d −79.75 (s).
CN): d 201.4 (t, JCP = 13.8 Hz, CO). m˜ /cm (CO):
3
NMR (CD
CN): d 2.11 (br, 4H, CH
2
3
filtration afforded a yellow solution from which analytically pure
material was produced on removal of the solvent (yield 68 mg,
31
1
19
P{ H} (CD
3
3
13
1
−1
C{ H} (CD
3
6
4%) (Found: C, 33.8; H, 5.54. RuC40
H
80
P
4
O
8
S
2
F
6
Cl
2
.2CHCl
3
2
019 (Nujol).
1
requires C, 33.57; H, 5.50%) H NMR (CDCl
H, CH ), 1.77 (br, 2H, CH ), 1.42 (br m, 36H, CH
CDCl
3
): d 2.02 (br,
3
1
1
2
(
2
2
3
). P{ H}
t
[
Ru(d bpe)(CO)(DMSO)
3
](OTf)
2
(11). A solid sample of
1
3
1
): d 99.8 (s). C{ H} (CDCl
3
): d 205.5 (t, JCP = 16.1 Hz,
t
3
[
Ru(d bpe)(CO)(H
O)
](OTf) (50 mg, 0.063 mmol) was placed
2
3
2
−
1
Ru–CO). m˜ /cm (CO): 1984 (Nujol).
in a Schlenk tube under argon and dissolved in DMSO (2 mL).
The solution was filtered, and reduced in volume to give 11
as a yellow powder (yield 37 mg, 60%). X-Ray quality crystals
were formed upon layering a CHCl
hexane. (Found: C, 32.3; H, 5.78. RuC H P O S F requires
t
t
[
Ru(d bpe)(CO)Cl
2
] (4). A sample of [Ru(d bpe)(CO)
2
2
Cl ]
(
50 mg, 0.09 mmol) in a J. Youngs resealable NMR tube was
−
3
3
–DMSO solution of 11 with
placed under vacuum (10 bar) and heated in an oil-bath at
1
◦
2
7
58
2
10
5
6
20 C for 36 h. The sample changed from a yellow powder to
1
C, 33.09; H, 5.96%). H NMR (d
.77 (br, 2H, CH
6
-DMSO): d 1.99 (br, 2H, CH ),
2
an orange solid; this was dissolved in CH
hexane to afford crystals overnight (yield 20 mg, 42%). (Found:
C, 43.7; H, 7.51. RuC19 OCl requires C, 44.02; H, 7.78%)
): d 2.11 (br m, 4H, CH ), 1.49 (d, JCP
2
Cl
2
and layered with
3
1
1
1
9
2
), 1.34 (br, 36H, CH
-DMSO): d −78.33 (s). C{ H} (d
3
). P{ H} (d
6
-DMSO): d
-DMSO): d
1
9
13
1
8.54 (s). F (d
6
6
H
40
P
2
2
206.7 (t, JCP = 14.5 Hz, Ru–CO). m˜ /cm− (CO): 2031 (Nujol).
1
1
H NMR (CD
2
Cl
2
2
=
1
3
1
1
3.6 Hz, 18H, CH
3
3
), 1.44 (d, JCP = 14.0, 18H, CH
3
). P{ H}
t
t
1
1
[Ru(d bpe)(CO)
3
H](OTf) (12). To a solution of [Ru(d bpe)-
(
CD
2
Cl
2
): d 96.2. C{ H} (CD
2
Cl
2
): d 204.4 (t, JCP = 16.5 Hz,
−
1
(CO)(H
Youngs NMR tube, 10 equivalents of H
0.63 mmol). This solution was degassed and placed under one
2
O)
3
](OTf)
2
(50 mg, 0.06 mmol) in d
6
-acetone in a
Ru–CO). m˜ /cm (CO): 1934 (Nujol).
2
O were added (11 lL,
t
t
[
{Ru(d bpe)(CO)}
2
(l-Cl)
3
][Ru(d bpe)(CO)Cl
3
]
(5). Found:
3
1
1
C, 41.2; H, 7.39. Ru
3
C
57
H
120
P
6
O
3
Cl
6
·2CH
2
Cl
2
requires C, 41.06;
atmosphere of CO. The reaction was monitored via P{ H} and
5
9 4
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7

View Article Online
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7
5 9 5

View Article Online
1
H NMR. After 2 h at room temperature, all of the starting
confirmed disorder rather than twinning. Consequently, it can
be readily seen that, despite exhaustive attempts, this is a very
poor quality refinement on many levels, and we therefore make
no detailed claims from these data save to say that they afford
definitive proof for the presence of compound 8. Refinement
in lower symmetry space groups did not assist in negating any
of the refinement problems. Our residual assessment is that the
crystal may have been undergoing a phase transition. (This was
our logic for collecting data at different temperatures, in which
case we have not successfully frozen out either phase).
material had converted to 12; layering with hexane afforded X-
ray quality crystals of 12 (yield 19 mg, 46%) (Found: C, 39.3;
H, 6.27. RuC22
H
41
P
2
O
6
SF
-acetone): d 2.31 (m, 4H, CH
4.0 Hz, 18H, CH
), 1.38 (d, JHP = 13.6 Hz, 18H, CH
t, JHP = 21.6 Hz, Ru–H; added splitting with CO trans to
RuH, JHC = 27.6 Hz). P{ H} (d -acetone): d 110.4 (s). F (d
-acetone): d 194.0 (t, JCP
3
·H
2
O requires C, 39.34; H, 6.45).
), 1.41 (d, JHP
), −7.96
1
H NMR (d
6
2
=
1
(
3
3
1
3
3
1
1
19
6
6
-
=
1
3
1
acetone): d −79.51 (s). C{ H} (d
6
.2 Hz, Ru–COax), 192.7 (m, Ru–COeq). m˜ /cm− (CO) 2092, 2038,
1
8
2
028 (Nujol).
Thus, as presented, the asymmetric unit in 8 consists of the
following cornucopia: on one full cation (based on Ru1), one
half of an anion, three triflate anions, and two half-occupancy
water molecules. As already stated, this structure is riddled
with disorder. In particular, in the full cation, the carbonyl
ligand is disordered in a 1 : 1 ratio with the trans water moiety.
Similar disorder is present in the half cation (wherein the central
metal, Ru2, is located on a two-fold rotation axis). Only one
of the triflate anions (based on S101) is not disordered. The
remaining two triflates are disordered in a 40 : 60 ratio between
the fragments based on sulfur pairs S201–S301 and S401–S501,
respectively. During the final least squares, the S–C, C–F, S–O, S–
F and C–O distances of atoms or partial atoms within individual
triflates were restrained, the reason being that the disorder
is more widespread than will allow for decent modelling.
The ADPs in regions for which disorder was modeled were
also restrained. Hydrogen atoms were included at calculated
positions on carbon atoms, but omitted in the case of the waters
as they could not be reliably located. The data completeness
of 77.8% in this structure is a consequence of the crystal
orientation, its unit cell parameters and restricted goniometer
movement during HELIX use. The gross structure is heavily
dominated by hydrogen bonding, based on O · · · O distances.
Compound 9 also exhibited a rather profuse asymmetric unit
which contained two cations (each containing one coordinated
triflate moiety), two triflate anions, one full water molecule
(based on O19), 0.25 of a water molecule (O20) and 2 very diffuse
solvent regions. The phosphine ligand in the cation based on Ru1
exhibited 45:55 disorder over two sites. With respect to solvent
regions, partial atom occupancies were initially freely refined
using isotropic thermal parameters, and subsequently fixed to
the arising values prior to anisotropic refinement. Indeed, some
four dozen attempts were made to model the solvent in this
structure before obtaining the most satisfactory convergence. In
essence, these two regions each represent one very disordered
dichloromethane molecule, which was ‘mopped-up’ using par-
tial isotropic carbon and chlorine atoms. Some of the partial
chlorine atoms were also positionally constrained to aid the
final convergence of the model. Once modelling of the lattice
solvent had been optimized, it was readily possible to locate
and refine the hydrogen atoms on the ligated water molecules.
Pulsed gradient spin-echo measurements
PGSE measurements were recorded as previously described for
2
complexes 1a–d. Compound 8 was dissolved on 0.6 mL of
CD
2
Cl
2
(10 lL of water was added to stabilise the complex)
1
to give a concentration of 2 mM. The H diffusion constant for
the cation [Ru(d bpe)(CO)(H
t
2+
2
O)
3
] was obtained from the tert-
butyl signals of the phosphine while data for the triflate anion
1
9
was obtained by F NMR spectroscopy.
X-Ray experimental data
Single crystals of compounds 2, 3, 4, 5, 6, 8, 9, 11 and 12 were
analysed using a Nonius Kappa CCD diffractometer. Details
of the data collections, solutions and refinements are given in
Table 10. The structures were universally solved using SHELXS-
28
9
7 and refined using full-matrix least squares in SHELXL-97.
Convergence was uneventful, with the following exceptions and
points of note.
The asymmetric unit in 2 was seen to contain one molecule
of CH
2
2
Cl in addition to the ruthenium complex, while in 3,
the motif was composed of one cationic dimer, one triflate
anion (disordered over two sites in a 1 : 1 ratio) plus three
molecules of chloroform, one of which was also disordered in a
4
: 1 ratio over two sites. Refinement of 4 was routine and the
hydrogen atoms attached to C11 and C19 therein were located
and refined without any restraints, because of the potential for
agostic interactions involving these two methyl groups. In 5,
the asymmetric unit was seen to consist of two molecules of
dichloromethane in addition to one molecule of the salt. Despite
copious re-crystallisation efforts, the quality of the sample could
not be improved, and the crystal used in the diffraction study
was, at best, of mediocre quality. Diffraction intensities were
somewhat eroded at higher Bragg angles and this is reflected in
both Rint and R(sigma) as well as in the final residual values. Data
quality necessitated inclusion of separate restraints on the ADPs
of the carbons in each ion during the final stages of refinement.
Nonetheless, the structure determination provides unambiguous
identification of the species present.
Solvent also featured in the structure of 6, which contains
three molecules of acetonitrile per molecule of the complex.
Moreover, there is evidence for intramolecular C–Hsolvent · · · Cl
interactions in this structure.
˚
These hydrogens were uniformly refined at a distance of 0.89 A
from the relevant parent oxygen. In 12, the asymmetric unit
was seen to consist of one molecule of water in addition to one
molecule of the salt. The water and hydride hydrogen atoms
Obtaining the structure of compound 8 certainly kept us
exercised several times over a period of approximately one year.
Several batches of beautiful single crystals were grown (no
indication of twinning using a polarising microscope), and
approximately seven datasets were very carefully collected at
varying temperatures between 30 and 298 K. The room-
temperature data set was useless, as the crystal decayed. The
issues were consistently the same: beautiful spots, routine
indexation, early fall-off in diffraction intensity, a promising
structure solution, a horrible refinement, lots of disorder, and a
suspect weighting scheme. Indeed, there seemed more evidence
at the refinement stage to indicate that the data sets arose from a
twinned sample, and consequently, a myriad of twin laws (using
both PLATON and ROTAX) were examined, all to no avail. Still
convinced that we may have overlooked something, we asked two
leading experts look at our data objectively, and the responses
˚
were located and refined at distance of 0.89 A from O7, and
˚
1.6 A from the ruthenium centre respectively. Unsurprisingly,
the water hydrogens are implicated in hydrogen bonding to
the oxygen atoms from proximate lattice anions. The absolute
structure parameters for 2 and 3 refined to a values of 0.04(2)
and 0.00(2), respectively.
CCDC reference numbers 250924–250932.
See http://www.rsc.org/suppdata/dt/b4/b415812a/ for cry-
stallographic data in CIF or other electronic format.
Acknowledgements
Financial support was provided by the EPSRC and the Uni-
versity of Bath. Johnson Matthey plc generously provided
ruthenium trichloride under their loan scheme. We thank
5
9 6
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7

View Article Online
Professor Ken Caulton (Indiana University) and Dr Andrew
Weller (University of Bath) for fruitful discussions.
11, 1655; (b) D. Huang, W. E. Streib, J. C. Bollinger, K. G. Caulton,
R. F. Winter and T. Scheiring, J. Am. Chem. Soc., 1999, 121, 8087.
4 R. Cini and A. Cavaglioni, Inorg. Chem., 1999, 38, 3751.
5 Comparable Ru · · · C and Ru · · · H distances have been seen else-
where: R. M. Stoop, C. Bauer, P. Setz, M. W o¨ rle, T. Y. H. Wong
and A. Mezzetti, Organometallics, 1999, 18, 5691; M. Sanford,
L. M. Henling, M. W. Day and R. H. Grubbs, Angew. Chem.,
Int. Ed., 2000, 39, 3451; J. N. Coalter III, J. C. Huffman, W. E.
Streib and K. G. Caulton, Inorg. Chem., 2000, 39, 3757; for a
recent discussion of stronger agostic interactions in coordinatively
unsaturated ruthenium complexes, see: W. Baratta, C. Mealli, E.
Herdtweck, A. Ienco, S. A. Mason and P. Rigo, J. Am. Chem. Soc.,
2004, 126, 5549.
1
1
References
1
2
3
4
M. F. Mahon, M. K. Whittlesey and P. T. Wood, Organometallics,
999, 18, 4068.
1
P. G. A. Kumar, P. S. Pregosin, J. M. Goicoechea and M. K.
Whittlesey, Organometallics, 2003, 22, 2956.
M. D. Hargreaves, M. F. Mahon and M. K. Whittlesey, Inorg. Chem.,
2
002, 41, 3137.
(a) A. Mezzetti, A. Del Zotto, P. Rigo and N. Bresciani Pahor,
J. Chem. Soc., Dalton Trans., 1989, 1045; (b) J. F. Hartwig, R. A.
Andersen and R. G. Bergman, Organometallics, 1991, 10, 1710;
16 D. Huang, K. Folting and K. G. Caulton, Inorg. Chem., 1996, 35,
(
1
c) G. C. Hsu, W. P. Kosar and W. D. Jones, Organometallics, 1994,
3, 385; (d) D. E. Fogg and B. R. James, Inorg. Chem., 1995, 34, 2557;
7035.
17 Heating the mixture of the Caulton dimers in the solid state under
◦
(
(
1
e) R. F. Winter and F. M. Hornung, Inorg. Chem., 1997, 36, 6197;
f) M. C. Nicasio, R. N. Perutz and P. H. Walton, Organometallics,
vacuum (120 C, 30 h) resulted in formation of 2.
18 C. F. J. Barnard, J. A. Daniela, J. Jeffery and R. J. Mawby, J. Chem.
Soc., Dalton Trans., 1976, 953; S. C. Grocott and S. B. Wild, Inorg.
Chem., 1982, 31, 3535; E. Bustelo, M. Jim e´ nez-Tenorio, M. C. Puerta
and P. Valerga, J. Chem. Soc., Dalton Trans., 1999, 31, 2399; A. V.
Marchenko, J. C. Huffman, P. Valerga, M. Jim e´ nez-Tenorio, M. C.
Puerta and K. G. Caulton, Inorg. Chem., 2001, 40, 6444; V. Cadierno,
P. Crochet, J. D ´ı ez, S. E. Garc ´ı a-Garrido, J. Gimeno and S. Garc ´ı a-
Granda, Organometallics, 2003, 22, 5226.
997, 16, 1410; (g) S. M. Hansen, F. Rominger, M. Metz and
P. Hofmann, Chem. Eur. J., 1999, 5, 557; (h) A. Martelletti, V.
Gramlich, F. Z u¨ rcher and A. Mezzetti, New. J. Chem., 1999, 199; (i) T.
Gottschalk-Gaudig, J. C. Huffman, K. G. Caulton, H. G e´ rard and
O. Eisenstein, J. Am. Chem. Soc., 1999, 121, 3242; (j) T. Gottschalk-
Gaudig, J. C. Huffman, H. G e´ rard, O. Eisenstein and K. G. Caulton,
Inorg. Chem., 2000, 39, 3957; (k) M. S. Kirkham, M. F. Mahon and
M. K. Whittlesey, Chem. Commun., 2001, 813; (l) S. D. Drouin, D.
Amoroso, G. P. A. Yap and D. E. Fogg, Organometallics, 2002, 21,
19 A. M. Joshi, I. S. Thorburn, S. J. Rettig and B. R. James, Inorg. Chim.
Acta, 1992, 198–200, 283; M. Faure, L. Maurette, B. Donnadieu and
G. Lavigne, Angew. Chem., Int. Ed., 1999, 38, 518; A. C. da Silva,
H. Piotrowski, P. Mayer, K. Polborn and K. Severin, Eur. J. Inorg.
1
042; (m) S. Gauthier, L. Quebatte, R. Scopelliti and K. Severin,
Chem. Eur. J., 2004, 10, 2811.
5
6
7
T. Gottschalk-Gaudig, K. Folting and K. G. Caulton, Inorg. Chem.,
Chem., 2001, 38, 685.
−
1
999, 38, 5241.
20 The analogous anions [RuL
2
(CO)Cl
3
]
(L = Me
2 2 3
S, Me SO, AsPh ,
D. W. Krassowski, J. H. Nelson, K. R. Brower, D. Hauenstein and
R. A. Jacobson, Inorg. Chem., 1988, 27, 4294.
3 5 5
SbPh , C H N) have been isolated as their phosphonium salts and
studied spectroscopically.20a The data suggest an axial CO ligand,
M. Peres de Araujo, O. M. Porcu, A. A. Batista, G. Oliva, D.-H. F.
Souza, M. Bonfadini and O. R. Nascimento, J. Coord. Chem., 2001,
trans to p-donor chloride. (a) L. Ruiz-Ramirez and T. A. Stephenson,
J. Chem. Soc., Dalton Trans., 1974, 1640.
21 A. Mezzetti, L. Costella, A. Del Zotto, P. Rigo and G. Consiglio,
Gazz. Chim. Ital., 1993, 123, 155.
22 For a general overview on the use of PGSE in coordination chemistry,
see: P. S. Pregosin, E. Martinez-Viviente and P. G. A. Kumar, Dalton
Trans., 2003, 4007.
23 E. Martinez-Viviente and P. S. Pregosin, Helv. Chim. Acta, 2003, 86,
2364.
24 P. C. Ford and A. Rokicki, Adv. Organomet. Chem., 1988, 28,
5
4, 87.
8
L. H. Wilkes, J. H. Nelson, J. P. Mitchener, M. W. Babich, W. C.
Riley, B. J. Helland, R. A. Jacobsen, M. Y. Cheng, K. Seff and L. B.
McCusker, Inorg. Chem., 1982, 21, 1376.
9
C. Wieser-Jeunesse, D. Matt and A. deCian, Angew. Chem., Int. Ed.,
1
998, 37, 2861.
1
1
0 T. Gottschalk-Gaudig, J. C. Huffman, H. G e´ rard, O. Eisenstein and
K. G. Caulton, Organometallics, 2000, 19, 3957; M. A. Volland,
B. F. Straub, I. Gruber, F. Rominger and P. Hofmann, J. Organomet.
Chem., 2001, 617–618, 288; T. Pechmann, C. D. Brandt and H.
Werner, Organometallics, 2003, 22, 3004.
1 (a) G. A. Lawrance, Chem. Rev., 1986, 86, 17; (b) for examples
of structures with bridging triflates, see: R. C. Schnabel and
D. M. Roddick, Organometallics, 1993, 12, 704; (c) P. Leoni, M.
Pasquali, M. Sommovigo, A. Albinati, F. Lianza, P. S. Pregosin
and H. R u¨ egger, Organometallics, 1993, 12, 4503; (d) T. Funaioli, F.
Marchetti and G. Fachinetti, Angew. Chem., Int. Ed., 1998, 37, 2861;
139.
13
25 During the formation of 12 from 6 with CO, labelled carbon
monoxide is incorporated into all three ligand sites. In the reaction
of isolated [Ru(d bpe)( CO) H](OTf) with CO, substitution of the
3
t
12
13
axial CO (trans to RuH) can be seen after ca. 2 h at 298 K while
substitution into the equatorial positions is apparent only after about
one week.
26 H. Ishida, K. Tanaka, M. Morimoto and T. Tanaka, Organometallics,
1986, 5, 724; J. G. Haasnoot, W. Hinrichs, O. Weir and J. G. Vos, Inorg.
Chem., 1986, 25, 4140; T. Funaioli, C. Cavazza, F. Marchetti and G.
Fachinetti, Inorg. Chem., 1999, 38, 3361; K. L. Breno, M. D. Pluth,
C. W. Landorf and D. R. Tyler, Organometallics, 2004, 23, 1738.
27 K.-R. P o¨ rschke, C. Pluta, B. Proft, F. Lutz and C. Z. Kr u¨ ger,
Z. Naturforsch., Teil B, 1993, 48, 608.
28 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 467–473, A46;
G. M. Sheldrick, SHELXL-97, a computer program for crystal
structure refinement, University of G o¨ ttingen, 1997.
(
e) C. R. Baar, L. P. Carbray, M. C. Jennings and R. J. Puddephatt,
Organometallics, 2000, 19, 2482.
1
1
2 M. Ogasawara, S. A. Macgregor, W. Streib, K. Folting, O. Eisenstein
and K. G. Caulton, J. Am. Chem. Soc., 1996, 118, 10189.
3 Similar geometries have been reported in both [Ru(PCy
3
)
2
(CO)Cl
] 13a
2
i
13b
and [Ru(P Pr
3
)
2
(CO)Cl
2
], although both of these show a trans,trans
configuration of the phosphine and chloride groups. (a) F. G. Moers,
P. T. Beurskens and J. H. Noordijk, Cryst. Struct. Commun., 1982,
D a l t o n T r a n s . , 2 0 0 5 , 5 8 8 – 5 9 7
5 9 7