Diferrocenyltriphosphines 1: Five-coordinate ruthenium complexes

Article abstract of DOI:10.1016/S0022-328X(98)00505-1

The ligand bis-1-(1′-diphenylphosphinoferrocenyl)phenylphosphine, triphosfer, reacts with dichlorotris-triphenylphosphineruthenium(II) to give the complex [(triphosfer)RuCl₂], (17), in almost quantitative yield. This complex reacts with CO at ambient temperature to afford [(triphosfer)Ru(CO)Cl₂]. Three related complexes, [{(η-R₂PC₅H₄)Fe-μ²-(η-C₅H₄P(R′)η-C₅H₄)Fe(η-C₅H₄R₂)}RuCl₂], R,R′=Ph, iPr (18-20), exhibit identical chemical properties. Reaction of complexes 17-20 with excess PMe₂Ph results in the deligation of the ferrocene ligand. The electrochemistry shows that while the diferrocenyltriphosphine ligands undergo a single two-electron oxidation, irreversible in character, complexes (17-19) exhibit two separated one-electron oxidations having features of chemical reversibility. This data suggests that while within the ligands no electronic communication exists between the two ferrocene subunits, complexation with RuCl₂ allows communication between the two ferrocenes. The single crystal X-ray structure of [(triphosfer)RuCl₂] has been determined.

Full text of DOI:10.1016/S0022-328X(98)00505-1

Journal of Organometallic Chemistry 565 (1998) 243–258
Diferrocenyltriphosphines 1: five-coordinate ruthenium complexes1
Ian R. Butler a,*, Ulrich Griesbach a, Piero Zanello b, Marco Fontani b, David Hibbs c,
Michael B. Hursthouse c, K.L.M. Abdul Malik c
a Department of Chemistry, Uni6ersity of Wales, Bangor, Gwynedd, LL57 2UW, UK
b Dipartimento di Chimica dell’Uni6ersita di Siena, Pian dei Mantellini, 44-53100 Siena, Italy
c EPSRC Crystallographic Centre, Cathedral Park, PO Box 912, Park Place, Cardiff, CF1 3TB, UK
Received 1 January 1998
Abstract
The ligand bis-1-(1%-diphenylphosphinoferrocenyl)phenylphosphine, triphosfer, reacts with dichlorotris-triphenylphos-
phineruthenium(II) to give the complex [(triphosfer)RuCl2], (17), in almost quantitative yield. This complex reacts with CO at
ambient temperature to afford [(triphosfer)Ru(CO)Cl2]. Three related complexes, [{(h-R2PC5H4)Fe–m2-(h-C5H4P(R%)h-
C5H4)Fe(h-C5H4R2)}RuCl2], R,R%=Ph, iPr (18–20), exhibit identical chemical properties. Reaction of complexes 17–20 with
excess PMe2Ph results in the deligation of the ferrocene ligand. The electrochemistry shows that while the diferrocenyltriphosphine
ligands undergo a single two-electron oxidation, irreversible in character, complexes (17–19) exhibit two separated one-electron
oxidations having features of chemical reversibility. This data suggests that while within the ligands no electronic communication
exists between the two ferrocene subunits, complexation with RuCl2 allows communication between the two ferrocenes. The single
crystal X-ray structure of [(triphosfer)RuCl2] has been determined. © 1998 Elsevier Science S.A. All rights reserved.
Keywords: Synthesis; Ferrocene; Phosphine; Ruthenium(II); Electrochemistry; Crystal structure
1. Introduction
been carried out to a significant extent with the prepa-
ration of derivatives of the general form 2. Changing
the phosphorus substituents changes both the steric
and electronic features of the ligand which in turn is
reflected in significant differences in the rates of cataly-
sis of derived complexes (catalyst precursors).
In related chemistry we have previously examined
the coordination chemistry of ferrocenylamine ligands
of the type 3 with ruthenium [5]. The five-coordinate
complexes of the type 4 are obtained directly from the
ligands on treatment with [Ru(PPh3)3Cl2]. Interestingly
these complexes react readily with small molecules ei-
ther quasi-reversibly (H2) or irreversibly (CO) to pro-
duce the appropriate six-coordinate pseudo-octahedral
species 5. In all of these complexes the ferrocene lig-
ands are bidentate and we were interested if the work
could be developed to use tripodal ligands to eliminate
the extraneous triphenylphosphine ligand from such
complexes.
The preparation and use of ferrocene-based ligands
has gained considerable momentum in the past 5 years
or so following on from the early chemistry which was
developed almost a quarter of a century ago. Arguably
the most important ligand of all the ferrocene-based
compounds is the bdppf(dppf) 1 which was initially
prepared by Sollot et al. [1] and was later used by
Davison in 1971 [2]. This ligand has been widely used
in both coordination studies and as a ligand in transi-
tion metal-based catalysts [3,4]. Modification of the
ligand by tailoring the phosphorus substituents has
* Corresponding author.
1 This paper is dedicated to Professor Michael Bruce, a good
friend, on the occasion of his 60th birthday.
0022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(98)00505-1

244
I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
Scheme 1. Tripodal diferrocenyltriphosphine ligand synthetic scheme.
2. Experimental
In recent work we have published the synthesis of
tripodal diferrocenyltriphosphine ligands, such as
triphosfer 6 [6]. The synthetic scheme developed is a
relatively simple one beginning with 1,1%-dibromofer-
rocene as a key precursor (Scheme 1).
We have been able to successfully adapt the method-
ology to obtain a broad range of ferrocenylphosphine
ligands with secondary functional groups. In this work
we present the first of these modifications and we
present a more rigorous treatment of the ligand charac-
terisation with the preparation of some key ruthenium
complexes which derive from the tripodal ligands. Elec-
trochemical and crystallographic structural characteri-
sations have also been carried out.
2.1. General experimental details
All reactions were carried out under an inert (N2 or
Ar) atmosphere. Chromatography was carried out un-
der aerobic conditions rapidly to avoid oxidation of
phosphine ligands. Mass spectra were recorded on ei-
ther Finnegan 4500 or Finnegan 8200 instruments at
the University of Wales, Bangor or University of
Wales, Swansea (EPSRC Central Service), and peaks
are reported as m/z (relative intensity). H-, 13C- and
1
31P-NMR data were obtained using a Bru¨ker WM250
instrument operating at 250, 62.9 and 101.3 MHz,

I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
245
Table 1
Table 2
Crystal data and details of data collection and structure refinement
for [C50H41Cl2Fe2P3Ru]·2CHCl3 (1)
Atomic coordinates (×104) and equivalent isotropic displacement
2
3
˚
parameters (A ×10 ) for [C50H41Cl2Fe2P3Ru]·2CHCl3 (1)
Empirical formula
Formula weight
Temperature (K)
Wavelength (A)
Crystal system
Space group
[C50H41Cl2Fe2P3Ru]·2CHCl3
1257.14
150(2)
0.71069
Monoclinic
P21/c
Atom
x
y
z
Ueq
Ru(1)
Fe(1)
Fe(2)
Cl(1)
Cl(2)
P(1)
P(2)
P(3)
C(1)
C(2)
C(3)
C(4)
C(5)
C(6)
C(7)
C(8)
C(9)
7145.3(6)
7041.1(6)
9913.8(6)
5588.8(10)
8657.2(10)
6000.7(11)
7348.6(10)
8242.3(10)
6235(4)
5536(4)
6093(4)
7112(4)
7208(4)
7517(4)
6850(4)
7389(5)
8373(4)
8448(4)
8549(4)
9536(4)
10143(4)
9563(4)
8589(4)
9647(4)
10130(4)
11154(4)
11340(4)
10418(4)
4527(4)
4149(4)
3051(4)
2329(4)
2684(4)
3794(4)
6066(4)
5778(5)
5738(5)
5986(5)
6322(5)
6339(4)
6251(4)
6395(4)
5599(4)
4616(4)
4478(4)
5294(4)
8373(4)
9218(4)
9250(4)
8417(5)
7551(4)
7517(4)
7747(4)
8239(4)
7959(4)
7183(4)
6658(4)
6950(4)
2986(4)
3751.6(13)
1905.6(13)
1033.1(2)
−170.4(3)
1985.6(3)
1493.7(5)
375.1(5)
214.6(6)
1266.8(5)
1763.8(5)
−319(2)
−472(2)
−900(2)
−1029(2)
−673(2)
673(2)
2673.3(2)
4538.7(4)
4136.5(4)
1814.1(6)
2955.4(6)
2627.8(7)
3912.2(6)
2370.7(7)
3405(3)
3900(3)
4439(3)
4270(3)
3637(3)
4613(2)
5146(2)
5673(3)
5476(3)
4822(3)
4440(2)
5002(2)
5324(3)
4981(3)
4445(3)
2964(2)
3212(2)
3747(3)
3847(3)
3369(2)
2443(3)
3000(3)
2849(3)
2144(3)
1590(3)
1729(3)
1776(3)
1712(3)
1019(4)
407(3)
14(1)
19(1)
17(1)
20(1)
22(1)
17(1)
14(1)
16(1)
22(1)
28(1)
27(1)
26(1)
23(1)
17(1)
23(1)
30(1)
27(1)
22(1)
16(1)
20(1)
22(1)
24(1)
21(1)
16(1)
21(1)
27(1)
23(1)
19(1)
18(1)
23(1)
34(1)
39(2)
47(2)
32(1)
23(1)
36(2)
52(2)
46(2)
41(2)
29(1)
14(1)
19(1)
21(1)
25(1)
28(1)
20(1)
17(1)
27(1)
35(2)
33(1)
30(1)
24(1)
17(1)
21(1)
28(1)
32(2)
34(2)
21(1)
28(1)
52(1)
39(1)
˚
Unit cell dimensions
˚
a (A)
12.555(2)
23.182(4)
17.691(2)
105.180(9)
4969.3(13)
4
1.680
1.441
2528
0.22×0.18×0.14
1.76–25.04
˚
b (A)
˚
c (A)
i (°)
3
˚
Volume (A )
Z
Dcalc. (g cm−3
)
Absorption coefficient (mm−1
F(000)
)
Crystal size (mm)
Theta range for data collection
(°)
556(2)
112(2)
−31(2)
296(2)
Index ranges
−145h59, −265k527,
−215l521
20 163
7505 (Rint=0.0877)
0.839–0.997
C(10)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
C(17)
C(18)
C(19)
C(20)
C(21)
C(22)
C(23)
C(24)
C(25)
C(26)
C(27)
C(28)
C(29)
C(30)
C(31)
C(32)
C(33)
C(34)
C(35)
C(36)
C(37)
C(38)
C(39)
C(40)
C(41)
C(42)
C(43)
C(44)
C(45)
C(46)
C(47)
C(48)
C(49)
C(50)
C(01)
Cl(01)
Cl(02)
1707(2)
1513(2)
1998(2)
2499(2)
2317(2)
1892(2)
2442(2)
2359(3)
1759(2)
1470(2)
349(2)
Reflections collected
Independent reflections
Absorption correction factors
Refinement method
Full-matrix least-squares on
F2
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I\2s(I)]
R indices (all data)
7505/0/595
0.764
aR1=0.0359, bwR2=0.0725
R1=0.0632, wR2=0.0759
Largest difference peak and hole 0.753 and −0.488
−3
˚
(e A
)
659(2)
819(2)
691(2)
382(3)
a R1=S(Fo−Fc)/S(Fo).
b wR2=[S{w(Fo2−Fc2)2}/S{w(F2o)2}]0.5; w=1/[s2(Fo2)].
207(2)
−224(2)
−797(2)
−1093(3)
−825(3)
−262(3)
41(2)
1694(2)
1872(2)
2169(2)
2318(2)
2162(2)
1853(2)
1518(2)
1153(2)
906(2)
1010(2)
1369(2)
1620(2)
2501(2)
2905(2)
3473(2)
3665(2)
3281(2)
2707(2)
2103(2)
2745.6(7)
2209.5(7)
respectively2. 31P-NMR data is reported relative to 85%
H3PO4 as internal standard. All solvents were pre-dried
and were freshly distilled prior to use. Ferrocene was
used as obtained from commercial suppliers, as were
472(3)
1155(3)
4171(2)
4952(3)
5187(3)
4646(3)
3868(3)
3634(3)
1417(2)
1357(3)
659(3)
the
reagents
dichlorophenylphosphine
and
chlorodiphenylphosphine. Other phosphines were pre-
pared according to methods previously used by us
following adaption of literature procedures [6]. Chro-
matography was carried out on silica gel 60–80 mesh
unless reported otherwise. TLC data were similarly
obtained using silica gel coated glass plates. Materials
and apparatus for electrochemistry have been described
elsewhere [7]. All potential values are referenced to the
saturated calomel electrode (SCE).
−14(3)
40(3)
735(3)
2207(2)
1803(2)
1769(3)
2130(3)
2523(3)
2561(3)
1623(3)
1658.4(9)
2071.3(7)
2 In the solvent CDCl3, carbon resonances, in some cases, mask
ferrocene carbon resonances. In addition, because of the proximity of
the ferrocenyl carbon resonances to each other, it is difficult to
distinguish between individual resonances and phosphorus couplings.

246
I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
Table 2 (Continued)
Table 3
Selected
˚
bond
lengths
(A)
and
angles
(°)
for
Atom
x
y
z
Ueq
[C50H41Cl2Fe2P3Ru]·2CHCl3 (1)
˚
Cl(03)
C(02)
Cl(04)
Cl(05)
Cl(06)
2471.0(12)
9152(4)
10497.8(12)
8992.0(15)
8903.6(16)
1873.9(8)
−667(3)
−380.3(7)
−935.0(9)
−1196.4(7)
652.4(7)
1699(3)
2041.2(9)
748.2(10)
2318.5(12)
49(1)
41(2)
52(1)
78(1)
70(1)
Bond lengths (A)
Ru(1)–P(2)
Ru(1)–P(1)
Ru(1)–Cl(1)
Fe(1)–C(1)
Fe(1)–C(6)
Fe(1)–C(8)
Fe(1)–C(3)
Fe(1)–C(4)
Fe(2)–C(17)
Fe(2)–C(11)
Fe(2)–C(12)
Fe(2)–C(14)
Fe(2)–C(18)
P(1)–C(21)
P(2)–C(6)
2.2067(12) Ru(1)–P(3)
2.3688(14) Ru(1)–Cl(2)
2.3903(12) Fe(1)–C(10)
2.3322(14)
2.3838(13)
2.019(5)
2.030(5)
2.046(5)
2.049(5)
2.054(5)
2.031(5)
2.024(4)
2.032(5)
2.044(4)
2.055(5)
1.814(5)
1.837(5)
1.849(5)
1.816(5)
1.829(5)
1.439(7)
1.420(7)
1.442(6)
1.434(7)
1.405(7)
1.443(6)
1.420(7)
1.431(6)
1.397(6)
1.410(6)
2.026(4)
2.040(5)
2.047(5)
2.049(5)
2.055(5)
2.026(5)
2.031(5)
2.036(5)
2.046(5)
2.052(5)
1.819(5)
1.827(5)
1.856(5)
1.825(4)
1.439(7)
1.426(7)
1.422(7)
1.429(6)
1.409(7)
1.414(7)
1.392(6)
1.402(6)
1.431(7)
1.415(7)
Fe(1)–C(5)
Fe(1)–C(7)
Fe(1)–C(9)
Fe(1)–C(2)
Fe(2)–C(20)
Fe(2)–C(16)
Fe(2)–C(15)
Fe(2)–C(13)
Fe(2)–C(19)
P(1)–C(1)
P(1)–C(27)
P(2)–C(33)
P(3)–C(45)
P(3)–C(39)
C(1)–C(2)
C(3)–C(4)
C(6)–C(7)
C(7)–C(8)
C(9)–C(10)
C(11)–C(12)
C(13)–C(14)
C(16)–C(20)
C(17)–C(18)
C(19)–C(20)
Ueq is defined as one third of the trace of the orthogonalized Uij
tensor.
2.2. Preparation of bis-1-(1%-diphenylphosphinoferro-
cenyl)phenylphosphine (triphosfer), 6
Bis-1-(1%-bromoferrocenyl)phenylphosphine (3.61 g,
5.68 mmol) was dissolved in anhydrous THF (50
cm3). The solution was cooled to −70°C (−78° ex-
ternal acetone/dry ice bath) and the mixture was
treated with n-BuLi (5.0 cm3 of a 2.5 M solution in
hexanes, 12.5 mmol, 2.2 equivalents). After 5 min the
reaction was quenched with chlorodiphenylphos-
phine (2.30 cm3, 12.5 mmol, 2.2 equivalents). After
warming to room temperature (r.t.) the reaction mix-
ture was treated with water (30 cm3) and diethyl
ether (20 cm3). The organic fraction was separated,
dried over anhydrous MgSO4 and reduced to an oil
following filtration. The crude concentrate was
purified by flash chromatography on silica. Initial elu-
tion with petrol (b.p. 40–60°C) removed non-polar
by-products. Elution with a mixture of petrol and
diethyl ether (1:1, v/v) followed by evaporation of the
solvent from the eluted yellow fraction left 3.31 g
(60%) of an amber oil which crystallized to leave a
voluminous yellow orange solid, m.p. 96–100°C (lit.
124–126°).
P(2)–C(11)
P(3)–C(16)
C(1)–C(5)
C(2)–C(3)
C(4)–C(5)
C(6)–C(10)
C(8)–C(9)
C(11)–C(15)
C(12)–C(13)
C(14)–C(15)
C(16)–C(17)
C(18)–C(19)
Bond angles (°)
P(2)–Ru(1)–P(3)
P(3)–Ru(1)–P(1)
P(3)–Ru(1)–Cl(2)
P(2)–Ru(1)–Cl(1)
P(1)–Ru(1)–Cl(1)
C(1)–P(1)–C(21)
C(21)–P(1)–C(27)
C(21)–P(1)–Ru(1)
C(6)–P(2)–C(33)
C(33)–P(2)–C(11)
C(33)–P(2)–Ru(1)
C(45)–P(3)–C(16)
C(16)–P(3)–C(39)
C(16)–P(3)–Ru(1)
C(5)–C(1)–C(2)
C(2)–C(1)–P(1)
C(4)–C(3)–C(2)
C(4)–C(5)–C(1)
C(7)–C(6)–P(2)
C(8)–C(7)–C(6)
C(10)–C(9)–C(8)
97.35(5)
164.47(4)
91.55(5)
112.62(5)
87.05(4)
102.0(2)
102.1(2)
116.6(2)
101.4(2)
98.1(2)
P(2)–Ru(1)–P(1)
P(2)–Ru(1)–Cl(2)
P(1)–Ru(1)–Cl(2)
P(3)–Ru(1)–Cl(1)
98.15(5)
94.04(4)
86.41(5)
87.89(4)
Cl(2)–Ru(1)–Cl(1) 153.19(4)
C(1)–P(1)–C(27)
C(1)–P(1)–Ru(1)
C(27)–P(1)–Ru(1)
C(6)–P(2)–C(11)
C(6)–P(2)–Ru(1)
102.2(2)
122.9(2)
108.3(2)
97.9(2)
116.9(2)
120.1(2)
104.9(2)
121.5(2)
100.7(2)
124.2(4)
107.9(5)
108.0(5)
106.9(4)
125.3(4)
107.6(5)
108.5(5)
125.1(3)
Rf=0.32
(petrol:Et2O=12:1,
v/v).
1H-NMR
118.51(14) C(11)–P(2)–Ru(1)
(CDCl3, 250 MHz): d 3.86 (m, 2H, ferrocenyl H),
(3.99, m, 2H, fc-H), 44.03 (m, 2H, fc-H), 4.07 (m,
2H, fc-H), 4.25 (m, 2H, fc-H), 4.31 (m, 2H, fc-H),
4.34 (m, 2H, fc-H), 4.42 (m, 2H, fc-H), 7.20–7.87 (br,
m, 25H, Ph, C6H5). {1H}31P-NMR (CDCl3, 100
MHz): d −32.42 (s, 1P, fc2P), −17.94 (s, 2P, fcP).
{1H}13C-NMR (CDCl3, MHz): d 71.54, 71.59, 72.25,
72.34, 72.42, 72.51, 72.71, 73.41, 73.63, 73.66, 73.88,
128.05, 128.15, 128.41, 128.48, 133.18, 133.31, 133.48,
133.63, 138.94, 138.79.
100.2(2)
106.0(2)
121.7(2)
107.0(4)
128.8(4)
108.5(5)
108.4(5)
127.0(4)
107.7(5)
109.2(5)
C(45)–P(3)–C(39)
C(45)–P(3)–Ru(1)
C(39)–P(3)–Ru(1)
C(5)–C(1)–P(1)
C(3)–C(2)–C(1)
C(3)–C(4)–C(5)
C(7)–C(6)–C(10)
C(10)–C(6)–P(2)
C(9)–C(8)–C(7)
C(9)–C(10)–C(6)
C(15)–C(11)–P(2)
C(15)–C(11)–C(12) 106.5(4)
C(12)–C(11)–P(2) 127.9(4)
C(12)–C(13)–C(14) 108.7(4)
C(14)–C(15)–C(11) 109.2(4)
C(13)–C(12)–C(11) 108.1(4)
C(15)–C(14)–C(13) 107.5(4)
C(20)–C(16)–C(17) 106.2(4)
The ligands 10, 11 and 12 were prepared us-
ing an identical method. The ligands 11 and 12
were isolated initially as oils and in the latter
C(20)–C(16)–P(3)
127.0(4)
C(17)–C(16)–P(3)
126.1(4)
C(18)–C(17)–C(16) 109.0(5)
C(18)–C(19)–C(20) 108.1(4)
C(17)–C(18)–C(19) 108.2(5)
C(19)–C(20)–C(16) 108.5(4)
case the oil solidified only on standing for
days. Characterisational data is as previously de-
scribed [6].
8

I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
247
fc-H), 7.30 (m, 13H, Ph-H), 7.40–7.60 (brm, 2H,
PPh-o-H). 13C-NMR: d 13.63, 19.97, 68.99, 69.83,
70.47, 71.09, 71.56, 71.80, 71.97, 72.69, 72.83, 73.38,
76.51, 77.02, 127.80, 127.92, 128.02, 128.14, 128.40,
128.53, 128.74, 133.16, 133.31, 133.47, 133.62, 134.00,
134.13, 138.53, 138.72, 138.94, 139.10. {1H}31P-
NMR: −31.02 (s, Pfc2Ph), −16.83 (s, PPh2fc), 0.82 (s,
P(iPr)2fc).
Scheme 2. Production of compound 13 from compound 9.
2.5. [Ru(L-15)Cl2]
A solution of [Ru(PPh3)3Cl2] (958 mg, 1 mmol) and
L-15 (800 mg, 1.03 mmol) in ethanol (10 cm3) was
refluxed for 20 min. On cooling, the solvent was re-
moved under high vacuum and the residue was ex-
tracted and filtered with hot hexane. The insoluble
brown powder was further washed with hot hexane
(2×25 cm3) to give the product complex as a brown/vi-
olet powder.
2.3. 1-(1%-Bromoferrocenyl)-1-(1%-diphenylphosphino-
ferrocenyl)phenylphosphine, snowdonphos, 14
Bis-1-(1%-bromoferrocenyl)phenylphosphine (3.00 g,
4.72 mmol) in anhydrous THF (20 cm3), cooled to ca
−70°C was treated with n-BuLi (1.90 cm3 of a 2.5 M
solution in hexane, 4.72 mmol, one equivalent). The
4.37 (m, 1H), 4.39 (m, 1H), 7.27–7.42 (brm, 13H,
Ph-H), 7.42–7.56 (brm, 2H, PPh-o-H). 31P-NMR
(CDCl3, 100 MHz): d −32.60 (s, 1P, fc2P), −17.83
(s, 1P, fcP). {1H}13C-NMR: 68.36, 68.87, 70.90,
71.54, 73.26, 73.88, 74.52 (overlapping last two re-
sonances) 127.89, 128.27, 128.70, 130.03, 133.18,
133.35, 133.49, 133.60, 133.90 (many overlapped).
Anal. Calc. for C38H31BrFc2P2: C 61.6, H 4.2. Found:
C 63.4, H 4.8%. High resolution mass spectrum (FAB)
m/z calc. for C38H3P2Fc2Br 741.976272. Found
741.976405.
1
For [Ru(L-15)Cl2]: H-NMR (CDCl3): d 0.6–1.0 (2
dd’s, overlapping, 12H), 2.82 (spt, 1H), 3.05 (spt, 1H),
3.38 (m, 1H, fc-H), 3.76 (m, 1H, fc-H), 3.83 (m, 1H,
fc-H), 3.86 (m, 1H, fc-H), 3.93 (m, 1H, fc-H), 3.97 (m,
1H, fc-H), 4.06 (m, 1H, fc-H), 4.09 (m, 1H, fc-H), 4.28
(2m, 2H, fc-H), 4.40 (m, 1H, fc-H), 4.49 (m, 1H,
fc-H), 4.92 (m, 1H, fc-H), 5.18 (m, 1H, fc-H), 5.25
(m, 1H, fc-H), 6.16 (m, 1H, fc-H), 7.09–8.09 (series of
m’s, 15H). {1H}31P-NMR (CDCl3):
fcP(iPr)2, JP–P=12 Hz, 322 Hz), 41.90 (td, fcPPh2,
P–P=322 Hz, 31 Hz), 86.19 (dd, fcPfc2, JP–P=12 Hz,
d 38.49 (td,
J
31 Hz).
2.4. 1-(1%-Diphenylphosphinoferrocenyl)-1-
(1%-diisopropylphosphinoferrocenyl)-phenylphosphine,
dppfdipf, 15
2.6. [Ru(L-15)(CO)Cl2], in situ
A solution of [Ru(L-15)Cl2] in CDCl3 was bubbled
with CO for 2 min. The solution changed colour from
violet to yellow. The product was subsequently charac-
terized in solution.
To 1-(1%-bromoferrocenyl)-1-(1%-diphenylphosphino-
ferrocenyl)phenylphosphine (1.14 g, 1.54 mmol) in an-
hydrous THF (20 cm3) maintained at −70°C, was
added a n-BuLi (0.74 cm3 of a 2.5 M solution in
hexane, 1.84 mmol, 1.20 equivalents). After 10 min the
reaction mixture was quenched with chlorodiisopropy-
lphosphine (0.28 g, 1.84 mmol, 1.20 equivalents). Di-
ethyl ether (18 cm3) was then added to facilitate layer
separation, followed by water (10 cm3). Following the
separation of the organic layer and drying over anhy-
drous MgSO4 the fraction was reduced to a crude oil.
The product fraction was eluted using a mixture of
petroleum spirit (b.p. 40–60°C) and diethylether(9:2,
v/v). Evaporation in vacuo gave the target compound
as an orange oil (0.35 g, 29%). Residual 1-(1%-bromofer-
rocenyl) - 1 - (1% - diphenylphosphino - ferrocenyl)
phenylphosphine (0.57 g, 50%) was also recovered.
Rf 0.50 (petrol:Et2O, 1:2, v/v). 1H-NMR (CDCl3,
250 MHz): d 1.10 (2d, 12H, CH3’s), 1.90 (brm,
2H, CH-H), 3.85 (m, 1H, fc-H), 3.92 (m, 2H,
fc-H), 4.10 (m, 5H, fc-H), 4.13–4.35 (brm, 8H,
1
For [Ru(L-15)(*CO)Cl2]: H-NMR (CDCl3): d 1.10
(dd, 3H), 1.40 (dd, 3H), 1.60 (dd, 6H), (CH3’s), 2.37 (br
spt, 1H), 2.42 (br spt, 1H), 3.63 (m, 1H, fc-H), 3.85
(2m, 2H, fc-H), 3.96 (m, 1H, fc-H), 4.02 (m, 1H, fc-H),
4.06 (m, 1H, fc-H), 4.12 (m, 1H, fc-H), 4.27 (m, 1H,
fc-H), 4.34 (2m, 2H, fc-H), 4.43 (m, 1H, fc-H), 4.62 (m,
1H, fc-H), 4.89 (m, 1H, fc-H), 5.19 (m, 1H, fc-H), 5.36
(m, 1H, fc-H), 5.44 (m, 1H, fc-H), 6.94 (m, 4H),
7.10–8.00 (m’s, 9H), 8.50 (pseudo t, 2H). {1H}31P-
NMR (CDCl3): d −11.24 (pseudo t, Pfc2), 19.75 (td,
PPh2, JP–P=33 Hz, 324 Hz), 24.87 (td, P(iPr)2, JP–P
=
25 Hz, 324 Hz). IR: nCO=1974 cm−1).
2.7. 1-(1%-Diphenylphosphinoferrocenyl)-1-(1%-carboxa-
latoferro-cenyl)phenylphosphine, erasmusphos, 16
To 1-(1%-bromoferrocenyl)-1-(1%-diphenylphosphino-

248
I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
Scheme 3. Production of compounds 14 [1,-(1%-bromoferrocenyl)-1-(1%-diphenylphosphinoferrocene)phenyl phosphine, Snowdonphos], 15 [1-(1%-
diphenylphosphinoferrocenyl)-1-(1%-diisopropylphosphinoferrocenyl)-phenylphosphine, dppfdipf] and 16 [1-(1%-diphenylphosphinoferrocenyl)-1-(1%-
carboxalatoferrocenyl)phenylphosphine, Erasmusphos] via the intermediate compound 9.
Scheme 4. Production of the palladium complex 24 using [Pd(1,5-COD)Cl2].
ferrocenyl)phenylphosphine (0.62 g, 0.84 mmol) dis-
solved in anhydrous THF (8 cm3) and cooled to ca.
−70°C was added n-BuLi (0.42 cm3 of a 2.5 M solu-
tion in hexane, 1.09 mmol, 1.30 equivalents). After
stirring for 15 min at this temperature, a slow stream of
CO2 was passed through the reaction mixture for 5 min
which resulted in the formation of a yellow precipitate
from the dark solution. The mixture was allowed to
warm to r.t. and was subsequently quenched with a
10% HCl solution (5 cm3). Diethyl ether was added and
the layers were separated. The organic layer was
washed with water (2×5 cm3) and brine (5 cm3). The
aqueous layer was extracted with diethyl ether and the
combined organic fractions were dried over MgSO4
before being reduced to an oil under vacuum.
lowing flash chromatography on silica using a mixture
of petroleum spirit (b.p. 40–60°C) and diethyl ether
(1:1, v/v). Evaporation in vacuo gave (0.10 g, 17%) of
an amber oil which solidified on standing.
1
Rf 0.77 (petrol:Et2O, 1:1, v/v). H-NMR (CDCl3, 250
MHz): d 3.80 (m, 1H, fc-H), 3.87 (m, 1H, fc-H), 4.00
(m, 1H, fc-H), 4.04 (m, 1H, fc-H), 4.06 (m, 1H, fc-H),
4.09 (m, 1H, fc-H), 4.11-4.15 (m, 2H, fc-H), 4.20 (m,
2H, fc-H), 4.24–4.29 (m, 2H, fc-H), 4.29–4.34 (m, 2H,
fc-H), 4.68 (m, 1H, fc-H), 4.76 (m, 1H, fc-H), 7.28 (br
m, 13H, Ph-H), 7.45 (br m, 2H, PPh-o-H). 13C-NMR:
d 69.03, 70.26, 71.28, 71.29, 71.77, 72.40, 73.08, 73.13,
73.25, 73.37, 73.45, 73.56, 73.67, 73.87, 73.97, 74.01,
74.19, 127.95, 128.06, 128.16, 128.42, 128.50, 129.09.
{1H}31P-NMR (CDCl3, 100 MHz): d −31.91 (s,
PPh2fc), −16.91 (s, Pfc2Ph).
The product was isolated from this concentrate fol-

I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
249
Table 4
31P chemical shift data on ligation
Ligand
Unbound
[RuLCl2]
Coordination shift
[RuL(CO)Cl2]
ꢀ"
Pa
Pb
Pa
Pb
Pa
Pb
Pa
Pb
Pa
Pb
Pb
Triphosfer, fc2PPh
Triphosfer, fcPPh2
L-10, fc2PPh
L-10, fcP(iPr)2
L-11, fc2PiPr
L-11, fcP(iPr)2
L-12, fc2PiPr
L-12, fcPPh2
−32.10
−17.63
−31.59
+0.08
−18.54
−0.07
−19.03
−17.59
−31.05
−16.83
+0.82
91.74
38.33
82.40
37.75
77.90
34.06
95.21
34.31
+123.84
+55.96
+113.99
+37.67
+96.44
+34.13
+114.24
+51.90
+117.24
+58.75
+37.67
−8.78
+22.22
−20.59
+21.13
+2.77
+22.16
+8.40
+18.05
−11.24
+24.87
+19.75
−99.78 *
−16.11 ꢁ
−102.99 *
−16.62 ꢂ
−75.13 ꢀ
−11.90 ꢂ
−86.81 ꢀ
−16.26 ꢁ
−97.43 *
−17.12 ꢁ
−18.74 ꢂ
L-15, fc2PPh
L-15, fcPPh2
L-15, fcP(iPr)2
+86.19
+41.90
+38.49
2.8. Preparation of palladium complex of ligand 14, 24
(m, 2H, fc-H), 4.09 (m, 2H, fc-H), 4.28 (m, 2H, fc-H),
4.49 (m, 2H, fc-H), 5.22 (m, 2H, fc-H), 5.95 (m, 2H,
fc-H), 7.10–7.40 (br, m, 14H, Ph-H), 7.40–7.52 (br, m,
5H, Ph-H), 7.52–7.62 (br, m, 4H, Ph-H), 7.75 (dd,
J=10.7, 7.1 Hz, 2H, PPh-o-H). {1H}31P-NMR
(CDCl2, 100 MHz): d +38.33 (d, J=36 Hz, 2P, fcP),
91.74 (t, J=36 Hz, 1P, fc2P). {1H}13C-NMR: 68.11,
68.65, 69.34, 74.02, 74.17, 74.32, 78.17 (other reso-
nances masked by solvent), 126.11, 126.28, 126.75,
127.80, 128.35, 129.20, 131.88, 133.94, 135.30, 136.80,
137.01. MS (FAB): clusters at m/z=948 ([Ru(triphos-
fer)]), 984 ([Ru(triphosfer)Cl]).
To a solution of [Pd(COD)Cl2] (200 mg, 0.75 mmol)
in the minimum quantity of dichloromethane and
ethanol (1:1 v/v) required for dissolution 1:1 equivalent
of the ligand 14 was added. The deep red solution
obtained was stirred for 1 h before hexane was added
which resulted in the precipitation of the product as a
deep violet powder which was isolated by filtration and
1
washed with diethyl ether, m.p.\200°C (decomp). H-
NMR (CDCl3, 250 MHz): d 3.65 (m, 1H, fc-H), 3.90
(m, 1H, fc-H), 4.12 (m, 1H, fc-H), 4.29 (brm, 3H, fc-H),
4.48 (br m, 3H, fc-H), 4.50 (br m, 5H, fc-H), 4.72 (m,
1H, fc-H), 4.90 (m, 1H, fc-H), 7.40 (m, 4H, Ph-H),
7.48–7.70 (br m, 6H, Ph-H), 7.82 (m, 1H, Ph-H), 8.05
(m, 1H, Ph-H), 8.25 (m, 3H, Ph-H). MS clusters at m/z
883 (LPdCl)+. {1H}31P-NMR: 30.41 (bs), 38.26 (bm).
An identical methodology was used in the prepara-
tion of the complexes [Ru(L-10)Cl2], [Ru (L-11)Cl2] and
[Ru(L-12)Cl2].
2.10. [Ru(L-10)Cl2], 18
M.p. 295–296°C. 1H-NMR (CDCl3, 250 MHz): d
1.13 (pq(dd), 6H), 1.30 (pq(dd), 6H), 1.49 (2dd, 12H),
2.98 (m, 4h), 3.38 (m, 4H), 3.73 (m, 2H), 3.92 (m, 2H),
4.03 (t, 2H), 4.16 (t, 2H), 4.34 (m, 2H), 4.48 (t, 2H),
5.10 (t, 2H), 5.45 (t, 2H), 7.13 (td, 2H), 7.26 (m, 2H),
7.92 (dd, 1H). {1H}13C-NMR (CDCl3, 75 MHz): d
190.22, 19.57, 21.57, 22.63, 30.1; 67.20, 68.34, 69.21,
72.06, 72.98, 75.22, 75.34, 125.36, 125.55, 129.62,
136.88, 137.06. {1H}31P-NMR (CDCl3, 101 MHz): d
37.75 (bd), fcPiPr2, 82.40 (bt), fc2PPh. MS (FAB) calc.
2.9. Ruthenium(II). Ru(triphosfer)Cl2, 17
Tris-(triphenylphosphine)dichlororuthenium(II) (270
mg, 0.59 mmol) was dissolved in a mixture of ethanol
and dichloromethane (40 cm3, 8:1, v/v) and refluxed for
5 min under N2. Bis-1-(1%-diphenylphosphinoferro-
cenyl)phenylphosphine dissolved in
a mixture of
ethanol and dichloromethane (20 cm3, 1:1, v/v) was
added and the mixture was refluxed for a further 2 h
upon which the brown colour changed to black and
then red, forming ultimately a brown/red precipitate.
Following cooling and solvent reduction to a few cm3,
the precipitate was collected by filtration, was washed
with dry diethyl ether and finally dried under vacuum
to give 153 mg (25%) of red microcrystals. A low yield
caused by subsequent finding that ruthenium sample
contained excess triphenylphosphine, after preparation
using literature methods [8]. Subsequent preparations
using highly purified reagent ruthenium–phosphine
complex gave yields in excess of 95%.
for
C38H50Cl526Fe102RuP3
883.024474.
Found
883.020008 (D=5.1 ppm). Anal. Calc. for
C38H50Cl2Fe2P3Ru: C 52.20, H 4.75. Found: C 52.43, H
5.55%.
2.11. [Ru(L-11)Cl2], 19
M.p. 282–283°C. 1H-NMR (CDCl3, 250 MHz): d
1.06 (pq(dd), 6H), 1.20 (two overlapping dd sets, 12H),
1.48 (m, 12H), 2.60 (spt, 1H), 3.10 (m, 4H), 3.18 (m,
4H), 4.20 (2m, 4H), 4.36 (m, 2H), 4.41 (m, 2H), 4.50
(m, 2H), 4.71 (m, 2H), 5.22 (m, 2H). {1H}13C-NMR
(CDCl3, 75 MHz): d 19.98, 19.29, 20.28, 21.67, 21.96,
M.p.\240°C (decomp). 1H-NMR (CDCl3, 250
MHz): d 3.88 (m, 2H, fc-H), 3.99 (m, 2H, fc-H), 4.03

250
I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
Fig. 1. Comparison of the 1H-NMR spectra of (a) [Ru(L-6)Cl2], (b) [Ru(L-10)Cl2] and (c) [Ru(L-11)Cl2].
2.12. [Ru(L-12)Cl2], 20
23.88, 24.05, 24.21, 29.79, 29.94, 30.09, 31.96, 32.54,
68.61, 68.80, 69.73, 71.92, 72.17, 77.70, 79.70. {1H}31P-
NMR (CDCl3, 101 MHz): d 34.06 (bm)iPr2Pfc, 77.90
(bm) fc2PiPr. MS calc. for C35H51Cl2Fe2P3Ru
848.03233. Found 848.072306 (D=10 ppm). Anal.
Calc. for C35H51Cl2Fe2P3Ru: C 49.55, H 6.06. Found: C
49.06, H 5.85%.
1
M.p. 302–303°C (with decomp). H-NMR (CDCl3,
250 MHz): d 1.32 (dd, 6H, PCHCH3), 2.71 (pspt, 1H,
PCH), 3.85 (m, 2H, fc-H), 4.06 (m, 2H, fc-H), 4.13 (m,
2H, fc-H), 4.20 (m, 2H, fc-H), 4.44 (m, 2H, fc-H), 4.56
(m, 2H, fc-H), 4.66 (m, 2H, fc-H), 5.81 (m, 2H, fc-H),

I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
251
Fig. 2. The change in the proton NMR spectra of (a) [Ru(L-6)Cl2] on transformation to (b) [Ru(L-6)(CO)Cl2]. The spectrum of (c) [Rh(L-6)Cl]
is shown for comparison.
7.10 (m’s, 6H), 7.33 (m’s, 6H), 7.56 (m, 4H), 7.80 (m,
2.13. Reactions of complexes 17–20 with CO
4H). {1H}31P-NMR: d 34.31 (d, PPh2fc), 95.21 (bt,
Pfc2iPr,
JP–P=29.1
Hz).
Anal.
Calc.
for
The ruthenium complexes in CDCl3 were bubbled
with dry CO for 4 min in standard 5 mm NMR tubes.
The colour of the solutions, in each case changed to
C47H43Cl2Fe2P3Ru: C 57.34, H 4.40. Found: C 57.54, H
4.25%.

252
I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
Table 5
yellow whereupon the product solutions were charac-
terised by NMR and infrared spectroscopy. In the case
of complex 20, the poorly soluble precursor dissolved
readily on CO addition to form the carbonylated
product.
Formal electrode potentials (in V, vs. SCE) and peak-to-peak separa-
tion (in mV) for the redox processes exhibited by complexes 17–19
and related molecules in dichloromethane solution with ([NBu4][PF6]
(0.2 M) supporting electrolyte)
Complex/Ligand
Eo%
DEap Eo%
Deap EbpRuII/0
0/+
+/2+
2.14. [Ru(triphosfer)(CO)Cl2], 21
fc2PPh
6
10
17
18
19
+0.60b
+0.56c
+0.49c
+0.59
+0.40
+0.30
+0.39
72
—
—
69
76
74
80
+0.75b
+0.56c
+0.48c
+0.86
+0.91
+0.86
—
76
—
—
—
—
−1.6
−1.6
−1.6
—
1H-NMR (CDCl3, 250 MHz): d 3.91 (m, 4H), 4.07
(m, 2H), 4.12 (m, 2H), 4.45 (m, 4H), 5.24 (m, 2H), 5.52
(m, 2H), 7.11 (pspt, 4H), 7.21–7.88 (m’s, 19H), 8.57
(pspt, 2H). {1H}13C-NMR (CDCl3, 75 MHz): d 68.55,
70.61, 72.12, 72.48, 74.37, 75.60, 77.06, 77.67, 78.17,
78.58, 78.76, 126.84, 127.04, 127.36, 128.15, 128.41,
128.60, 129.40, 124.55, 130.62, 132.01, 132.17, 133.25,
74
74
76
—
Fe(C5H5)2
a
Measured at 0.2 V s−1 b [NBu4][ClO4] (0.2 M) supporting elec-
.
trolyte, see text. c Peak potential value for irreversible processes.
135.65, 138.15, 138.32 (NB; no 13C
{1H}31P-NMR (CDCl3, 101 MHz): d −8.78 (t, fc2PPh)
P–P=30.5 Hz, 22.22 (d, fcPPh2). IR: nCO=1988 cm−1
6 O observed).
Table 6
Separations between the two redox potentials (in V) of the bis-fer-
rocene derivatives reported in Table 5 and Robin–Day classification
of their mixed-valent monocations
J
.
Complex
DEo%
Kcom
Class
[(C5H5)Fe(C5H4)]2PPh
0.15
0.27
0.51
0.56
3.4×102
3.7×104
4.3×108
3.0×109
II
II
III
III
17
18
19
2.15. [Ru(L-10)(CO)Cl2], 22
1H-NMR (CDCl3, 250 MHz): d 1.46 (2 dd’s, overlap-
ping resonances, 12H), 1.65 (2 dd’s, overlapping, 12h),
2.64 (m, 4H, 3.03 (psp, 4H), 3.82 (m, 2H), 4.13 (m, 2H),
4.19 (m, 2H), 4.29 (2m, 4H), 4.50 (m, 2H), 4.97 (m, 2H),
5.22 (m, 2H), 7.35–7.50 (m’s, 3H), 8.17 (pst, 2H).
{1H}13C-NMR (CDCl3, 75 MHz): d 18.68, 19.54, 20.35,
20.59, 27.77, 29.38, 68.42, 69.70, 71.90, 73.60, 73.87,
75.96, 126.31, 126.45, 130.03, 138.20, 138.36 (no 13CO
observed). {1H}31P-NMR (CDCl3, 101 MHz): d −
20.59 (t) fc2PPh, 21.13 (d), fcP(iPr)2. IR: nCO=1962.2
Fig. 3. Cyclic voltammetric responses recorded at a platinum electrode
cm−1
.
in CH2Cl2 solutions containing (a) [(C5H5)Fe(C5H4)]2PPh (7×10−4
M) and [NBu4][ClO4] (0.2 M), and (b)
6
(3×10−4 M) and
2.16. [Ru(L-11)(CO)Cl2], 22
[NBu4][PF6] (0.2 M). Scan rate 0.2 V s−1
.
1H-NMR: d 1.08 (pq(dd), 6H), 1.43 (m’s, 15H), 1.82
(pq(dd), 6H), 2.75 (m, 4H), 3.05 (2m’s, 5H), 4.20 (2m,
4H), 4.23 (m, 2H), 4.36 (m, 2H), 4.45 (m, 2H), 4.46 (m,
2H), 4.90 (m, 2H), 5.31 (m, 2H). {1H}13C-NMR
(CDCl3, 75 MHz): 18.96, 20.37, 20.63, 21.60, 21.63,
23.24 (t), 28.10 (several lines), 68.77, 70.00, 71.39, 72.44,
72.50 (d), 74.21, 74.22 (several overlapping) (N.B. no
13CO observed). {1H}31P-NMR (CDCl3, 101 MHz): d
2.77 (t) fc2PiPr, 22.16 fcP(iPr)2. IR: nCO=1954.3 cm−1
.
Fig. 4. Cyclic voltammetric response recorded at a platinum electrode
2.17. [Ru(L-12)(CO)Cl2], 23
in a CH2Cl2 solution containing [Ru(L-6)Cl2] (5×10−4 M) and
[NBu3][PF6] (0.2 M). Scan rate 0.2 V s−1
.
1H-NMR (CDCl3): d 1.34 (dd, 6H, CHCH3), 3.25

I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
253
Fig. 5. Structure of [RuCl2(triphosfer)] showing the atom labelling scheme. Thermal ellipsoids are drawn at 50% probability. All hydrogen atoms,
except H(32) which blocks the vacant ‘octahedral’ coordination site around ruthenium, are omitted for clarity.
(bm, 1H), 4.00 (m, 2H, fc-H), 4.15 (m, 2H, fc-H), 4.17
(m, 2H, fc-H), 4.41 (m, 2H, fc-H), 4.51 (m, 2H, fc-H),
4.67 (m, 2H, fc-H), 5.68 (m, 2H, fc-H), 7.18 (m, 4H),
7.37 (overlapping m’s, 8H), 7.62 (bm, 4H), 8.05 (bm,
4H). 13C-NMR: d 20.32, 27.89 (d), 69.01, 70.20, 70.69,
72.00, 73.01, 75.09, 76.52, 78.48, 126.82, 126.44, 127.99,
127.13, 124.83, 130.08, 133.27, 133.34, 133.40, 136.75,
136.88, 136.95 (no 13CO observed). {1H}31P-NMR: d
8.40 (t, PiPrfc), 18.05 (d, PPh2fc, JP–P=30.5 Hz). IR:
anisotropic. The H atoms were included in calculated
positions (riding model) with Uiso=1.2×Ueq of the
parent carbon. The diagram was drawn using SNOOPI
[13]. Sources of scattering factor data are given in ref.
[12]. The calculations were done on a 200 MHz personal
computer.
Fractional coordinates of the non-H atoms, and se-
lected bond lengths and angles are given in Tables 2 and
3, respectively. Anisotropic displacement coefficients, H
atom parameters and complete lists of bond lengths and
angles have been deposited as supplementary material.
n
CO=1974 cm−1
.
2.18. X-ray crystallography
The crystal used for X-ray work was obtained as
described above. All measurements were made on a Delft
Instruments FAST area detector diffractometer posi-
tioned at the window of a rotating anode generator with
3. Results and discussion
In the ligand synthetic scheme presented in Scheme 1,
it is clear to see that the intermediate compound 9, by
analogy with 1,1%-dibromoferrocene should be amenable
to monolithiation in addition to the dilithiation already
described in Scheme 2.
The first reaction of this type was carried out at low
temperature in THF to avoid Wurtz coupling. The key
compound, 1,-(1%-bromoferrocenyl)-1-(1%-diphenylphos-
phinoferrocene)phenyl phosphine, snowdonphos3 14
(Scheme 3), was isolated initially as a yellow/orange oil
˚
Mo–Ka radiation (l=0.71069 A) by following proce-
dures described earlier [9]. The crystal data, details of
data collection and structure refinement are summarised
in Table 1. The cell parameters were determined by
least-squares refinement of diffractometer angles for 250
reflections. The data were corrected for Lorentz and
polarisation factors and also for absorption effects (DI-
FABS) [10].
The structure was solved by direct methods
(SHELXS86) [11] and refined by full-matrix least-squares
on F2 using all unique data with intensities greater than
zero (SHELXL93) [12]. The non-H atoms were all
3 The ligand was named due to the similarity of the 1H-NMR
spectrum of its palladium complex to the profile of a nearby topolog-
ical feature.

254
I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
following column chromatography. The oil originally
obtained solidified on standing. The spectroscopic fea-
tures are unremarkable e.g. 31P d −32.60, −17.83 which
is consistent with diferrocenylphenylphosphines (nor-
mally observed at chemical shifts between −24 and −34
d) and diphenylferrocenylphosphines (−16 to −19 d).
The 1H-NMR spectrum recorded at 250 MHz is difficult
to interpret because of overlap of resonances particularly
in the ferrocenyl region, because there are 16 distinct
ferrocenyl protons in the racemate. This compound
forms the expected palladium complex 24 on treatment
with [Pd(COD)Cl2] (Scheme 4).
Lithiation of snowdonphos followed by a carbon
dioxide quench resulted in the formation of the related
carboxylic acid derivative 16, erasmusphos which is
obtained as a pale yellow powder following the synthetic
procedure and standard work up. Finally a mixed
triphosphine ligand was prepared using a chlorodiiso-
propylphosphine quench sequence to give 15.
also predictable within a few ppm, in most cases.
In the case of triphosfer, on coordination to the
‘RuCl2’ fragment the ligand phosphorus resonances are
shifted downfield by ca. 124 and 56 ppm respectively
which in comparison with similar coordination with
Rh(I) are large [6]. The larger of the shifts is for the
unique phosphorus atom, which is as anticipated and the
observed JP–P coupling constant is 30 Hz, typical of
coupling constants in similar complexes. The larger trans
coupling, observed in the unsymmetric ligand complex
[Ru(L-15)Cl2] is ca. 322 Hz. Typical 1H-NMR spectra for
the complexes [Ru(triphosfer)Cl2], [Ru(L-10)Cl2] and
[Ru(L-11)Cl2] are shown in Fig. 1, illustrating the range
of chemical shifts observed in the ferrocenyl region of the
spectrum which are dependent on the geometry and
donor ability of the phosphorus. By examination of the
X-ray structure (following) it is easy to observe the
vacant coordination site. On treatment with CO over
short periods this free site is rapidly occupied. When this
occurs there is again a large shift for the unique phospho-
rus atom resonance (trans to CO, ca. 100 ppm for
L=triphosfer), this time upfield, reflecting the increase
of electron density on the phosphorus atom. There is also
a more reserved effect on the other phosphorus atoms
with a modest upfield shift of ca. 16 ppm. The JP–P value
is essentially unaffected as the mutual P–P coordination
geometry is essentially maintained. The colour of the
complex changes from deep red/violet to yellow on CO
ligation reflecting the increased crystal field stabilization
3.1. Tripodal phosphine ligands
As previously mentioned we have described the synthe-
sis of the ligands 6, 10, 11, 12 in a recent paper but we
have not discussed the detailed spectral conclusions. To
augment the series we have already described the synthe-
sis of 15 in this paper and we have also obtained
derivatives of the type 25 by the quench with
chlorophenylisopropylphosphine [6].
1
in the new pseudo octahedral geometry. The H-NMR
spectrum before and following CO binding for L=
triphosfer are shown in Fig. 2. There are clearly signifi-
cant chemical shifts of the ferrocenyl resonance on
coordination of the ligand reflecting the proximity of the
ferrocenyl protons to the metal centre. The carbonyl
stretching frequency for the complexes [Ru(L-
6)(CO)Cl2], [Ru(L-10)(CO)Cl2], [Ru(L-11)(CO)Cl2],
[Ru(L-12)(CO)Cl2] and [Ru(L-15)(CO)Cl2] are observed
at 1988, 1962, 1954, 1974 and 1974 cm−1, respectively.
Clearly the stretching frequency may be used as a rough
guide to ligand basicity with the stronger metal carbonyl
interaction observed in the more highly alkyl-substituted
phosphines.
A large volume of work has been compiled on the
reaction chemistry of complexes of the type [Ru(P3)Cl2],
P3=tripodal phosphine e.g. [CH{(CH2)2PPh2}3], A, or
[PhP{(CH2)2PPh2}2], B [14]. For example, the complexes
of the general type [HRu(P3)H2]+P3=A are more stable
than their diphosphine analogues, and there are many
examples of other metal hydride complexes of other
transition elements. A series of papers has documented
the reaction chemistry of rhenium and rhodium trispho-
sphine complexes using the unique geometry of the
tripodal ligands [15]. The complexes [LRu(P3)Cl2], P3=
B, on treatment with hydrides readily afford the dihy-
In
compound:
6
R1,R2,R3,R4,R5=Ph:
10
R1,R2,R4,R5=iPr, R3=Ph: 11 R1,R2,R3,R4,R5=iPr:
12 R1R2R4R5=Ph, R3=iPr: 15 R1,R2,R3=Ph,
R4,R5=iPr and 25 R1,R3,R4=Ph, R2,R5=iPr
(racemate).
As a measure of ligand basicity we are able to use the
{1H}31P spectra to give some relative information. There
are three distinct variables dependent on the the phos-
phine substituents which can be classified in tabular
form.
The 31P-NMR chemical shift data for five complexes
is presented in Table 4. There is an internal consistency
in the data allowing for the prediction of chemical shifts
with a few ppm according to the nature of the phospho-
rus substituted. On coordination to ruthenium in the
five-coordinated complexes the typical shift for bridge-
head (Pa) and terminal (Pb) phosphorus is shown in
column 4. The shifts on coordination of CO shown in
column 6 show that for similar groups the shift data is

I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
255
dride complexes which on protonation form molecular
hydrogen complexes [16]. The six-coordinate complexes
themselves are obtained on direct treatment of the parent
five-coordinate species with appropriate ligands e.g.
L=CO, PMe2Ph, CH3CN, DMSO, etc.
tion process, which in controlled potential coulometry
(Ew= +0.7 V) consumes two electrons per molecule.
Only at the scan rate of 10 V s−1 does the appearance
of a directly associated reduction peak take place. We
assign such a two-electron step to the concomitant
oxidation of the two ferrocene units. Identical behaviour
is displayed by ligand 10. The fact that the addition of
the phosphino substituents in the bis-(ferro-
cenyl)phenylphosphine backbone causes instability of
their ferricinium congeners is not unexpected in that for
It is of immediate and relevant interest to find out if
a similar reaction chemistry can be developed for the
ferrocene-based ligand series. The advantage that such
ligands have is that a more rigid ligand backbone should
result in highly stabilized complexes with the added
ability for such ligands to move electron density from one
phosphorus atom to another through the ferrocene
backbone to give secondary stabilization. The simple
synthesis we have developed for these ligands allows for
ease of substitution of the phosphorus with different
alkyl or aryl groups allowing facile steric and electronic
tuning of the ligand template. Interestingly, treatment of
each of the complexes 17–20 with an excess of
dimethylphenylphosphine in hot ethanol resulted in
complete deligation with the formation of the appropri-
ate ruthenium monodentate phosphine complex. Clearly
further chemical studies are warranted in this area. Initial
results on the reactions of complexes 17–20 with hydride
are inconclusive and are ongoing. The reaction chemistry
with pyridinyl bases will be the subject of the next paper
in this series.
instance 1-diphenylphosphinoferrocene undergoes
a
chemically reversible one-electron oxidation whereas the
oxidation of 1,1%-bis-diphenylphosphinoferrocene is
more complex being coupled to chemical complications
[19]. The effect of the two PPh2 substituents is more
complex if one considers that a single two-electron step
precludes no electronic interaction between the two
ferrocene subunits in sharp contrast with what occurs for
bis-(ferrocenyl)phenylphosphine.
Fig. 4, which illustrates the anodic portion of the
electrochemical response of the ruthenium(II) complex of
triphosfer, exemplifies the redox aptitude of the present
ruthenium complexes.
As seen, complexation with RuCl2 seems to restore
either the electronic communication between the two
ferrocene subunits or the stability of the oxidation
congeners, in that two separated one-electron oxidations
(coulometrically measured) appear again, both having
features of chemical reversibility. Nevertheless, the oc-
currence of the sequential ruthenium centre oxidation
Ru(II)/Ru(III)/Ru(IV) cannot be, in principle, ruled out,
even if it were accompanied by the disappearance of
ferrocene oxidations. In all ruthenium complexes 17–19,
step by step controlled potential electrolysis show that
only the corresponding monocations are fairly stable,
whereas the dications tend to decompose. An irreversible
two-electron reduction is present at very negative poten-
tial value, which we confidently assign to the Ru(II)/
Ru(0) step.
Table 5 compiles the redox potentials of the present
complexes together with those of related derivatives.
The differences between the formal electrode poten-
tials of the subsequent oxidation processes in the present
bis-ferrocene molecules are reported in Table 6. Such
differences, allowing computation of the relevant com-
proportionation constants, allow the electrogenerable
mixed-valent monocations to be classified according to
the Robin–Day classification [20].
3.2. Electrochemistry
In order to account for the electrochemical behaviour
of the present ruthenium(II) complexes 17–19, it is useful
to illustrate preliminarily the electrochemical responses
given by their diferrocenyl ligands themselves. In this
connection Fig. 3 compares the cyclic voltametric be-
haviour of unsubstituted bis-(ferrocenyl)phenyl-phos-
phine
with
that
of
its
diphenylphosphino-
substituted triphosfer 6 in dichloromethane solution.
As shown, bis-(ferrocenyl)phenylphosphine undergoes
two separated oxidations with features of chemical
reversibility, which are preceded by a minor adsorption
wave. It must be taken into account that such response
has been obtained on using [NBu4][ClO4] as the support-
ing electrolyte. As a matter of fact, the use of [NBu4][PF6]
affords a response displaying a higher prewave (Ep= +
00.43 V) followed by a broad reversible peak-system.
These data could account for some uncertainties which
still exist about the electrochemical behaviour of bis-(fer-
rocenyl)phenylphosphine: a simple sequential oxidation
of the two ferrocenyl subunits in CH2Cl2/[NBu4][ClO4]
[17] versus a sequential oxidation of the two ferrocenes
preceded by a PPh centred anodic process in DMSO/
[NBu4][BF4] solution [18]. Accordingly, we attribute the
prewave to adsorption phenomena more or less pro-
nounced depending upon the solution composition. By
contrast, free triphosfer, independent from the nature of
supporting electrolyte, undergoes an irreversible oxida-
As can be seen in Table 6, increasing the electron-do-
nating ability of the peripheral biferrocene–phosphine
substituents makes the electronic communication be-
tween the two ferrocenes increase progressively from the
slightly charge-delocalized class II to the completely
charge-delocalized class III.
There are interesting implications of these observa-
tions in that it should be possible to adapt the ligands

256
I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
into conducting polymeric assemblies which could, in
theory, be used to sense metal ions because an elec-
tronic communication between the iron centres will be
established only when the metal ion binds into the
coordination pocket.
and [RuCl2(CMTP)(PPh3)] [CMTP=PhP(C8H11)P
(Ph)(C8H11)PPh] [27], whilst the second group includes
the [Ru2Cl3(ETP)]+ cation in the Ph4B− [28] and
F3CSO3− [22] salts, and the [Ru2Cl3[TRIPHOS)2]+
cation [TRIPHOS=(Ph2PCH2O)3C(Me)] in the Ph4B−
salt [29]. Of all these complexes, the ETP compound
[24] is quite unusual since it contains a monodentate
ETP bonded to ruthenium in addition to the usual
tridentate ligand. Simpler phosphine ligands are
monodentate (PR3, R=Me, Et, Ph, etc.) or bidentate
(R2PCH2PR2, R=Me, Et, Ph, etc.) and tend to yield
both monomeric and dimeric species depending on the
relative size, shape and orientation of the ligands
around the metal. An important point to note here is
that the five-coordinate ruthenium complexes are ex-
tremely useful, since the associated ‘coordinative unsat-
uration’ plays a vital role in catalytic property of such
materials [30]. In fact two very important catalysts used
in homogeneous hydrogenation i.e. RhCl(PPh3)3 [31]
and RuCl2(PPh3)3 [32–34] are both ‘coordinatively un-
saturated’ 16-electron species.
3.3. X-ray structure of Ru(triphosfer)Cl2, 17
The molecular structure of 17 is shown in Fig. 5,
which also indicates the crystallographic atom number-
ing scheme used. Selected molecular geometry parame-
ters are presented in Table 2.
Although there are several reports on the structures
of metal complexes containing tridentate PPP donor
ligands, the present compound appears to be unique in
that it is the first example of a metal compound bound
to a tridentate PPP ligand containing two ferrocenyl
units (triphosfer). The ruthenium atom is five-coordi-
nate and has a distorted square pyramidal geometry
with the bridging phosphorus [P(2)] at the apical posi-
tion. The main deviations from ideal geometry are the
angles (a) P(2)–Ru(1)–Cl(1) which, instead of being ca.
90°, has a value of 112.62(5)°, and (b) P(1)–Ru(1)–P(3)
and Cl(1)–Ru(1)–Cl(2) which, instead of 180°, are
164.47(4) and 153.19(4)°, respectively. The much wider
P(2)–Ru(1)–Cl(1) angle may be explained by the short
The Ru–P distances are different, those to the termi-
nal phosphorus atoms [Ru(1)–P(1) 2.3688(14), Ru(1)–
˚
P(3) 2.3322(14) A] being significantly longer that to the
˚
central atom [Ru(1)–P(2) 2.2067(12) A]. These values
are comparable with the corresponding values [2.280(2),
˚
˚
˚
Cl(1)···H(38) 2.547 A steric interaction. The P–Ru–P
2.261(2), 2.198(2) A] in [RuCl2(TTP)] [22] and [2.306(1),
chelate angles 97.35(5) and 98.15(5)° are comparable
with that [99.82(8)°] in the palladium complex
[PdBr(C14H8Cl){Fe(C5H4–PPh2)2}] [21] and are dic-
tated by the ‘bite’ associated with the ferrocenyl bispho-
sphine moiety. Other Cl–Ru–P angles lie in the range
86.41(5)–94.04(4)° and closer to 90°.
Besides the present compound, there are only two
other five-coordinate ruthenium structures reported to
date, which contain the ‘RuCl2PPP’ moiety, i.e.
2.276(2), 2.212(2) A] in [RuCl2(CyTTP)]·2Me2SO [23].
These distances are also close to the respective values
observed in [RuCl2(ETP)(MeCN)3]·2CF3SO3 [2.314(6),
˚
2.309(6), 2.261(6) A] [22], [RuCl2(ETP)2]·CHCl3 ·Et2
˚
[2.389, 2.358, 2.295 A] [24], [RuCl2(ETP)(Me2
˚
SO)]·C6H5Me5 [2.366, 2.289, 2.257 A] [25],
˚
[RuCl2(CMTP)(PPh3)] [2.350(1), 2.320(1), 2.308(1) A]
[27], [R2Cl3(ETP)]BPh4 ·CHCl3 ·0.5H2O [2.308, 2.292,
˚
2.252 A] [28] and [Ru2Cl3(ETP)]·CF3SO3 ·CH2Cl2
˚
[RuCl2(TTP)]
[TTP=Ph2P(CH2)3P(Ph)(CH2)3PPh2]
[2.287(2), 2.266(2), 2.251(2) A] [22]. Examination of the
[22], and [RuCl2(CyTTP)]·2Me2SO [CyTTP=(cyclo-
hexyl)2P(CH2)3P(Ph)(CH2)3P(cyclohexyl)2] [23]. In the
TTP complex, the coordination geometry was described
as square pyramidal, whereas a trigonal bipyramidal
arrangement was reported for the CyTTP complex,
with considerable distortions being observed in both
cases. In our compound the two chloride ligands are
mutually trans, in contrast with their cis arrangement in
the above two compounds. It may be pointed out that
‘RuCl(PPP)’ complexes generally prefer six-coordinate
octahedral geometry, which may be achieved by the
five-coordinate monomeric species taking a ‘sixth’ lig-
and or forming a trichloro-bridged dimer in a ‘face
sharing bioctahedral’ arrangement. The first category of
these complexes includes [RuCl2 (ETP)2]·CHCl3 ·Et2O
Ru–P distances in all the above compounds suggests
that the distances involving the two terminal (basal)
atoms are always longer than that involving the central
atom. Although the small differences in the two basal
bonds could be attributed to steric factors, the shorten-
ing of the apical bonds are most probably due to
electronic effects and certainly consistent with the sug-
gestion that there should be strong p-bonding in square
pyramidal type metal complexes [35]. It should be
noted, however, that in the [Ru2Cl3(TRIPHOS)2]+
cation, which contains sterically rigid tripodal PPP
ligands, all the Ru–P distances are virtually identical
˚
[2.296(3)–2.310(3), average 2.305 A] [29].
The Ru–Cl two distances in the present complex
˚
[2.3838(13), 2.3903(11) A] are nearly the same, but
[ETP=Ph2P(CH2)2P(Ph)(CH2)2PPh2]
(ETP)(Me2SO)]·C6H5Me [25],
(OMe)3}] [MTP=Me2P(CH2)3P(Me)(CH2)3PMe2] [26],
[RuCl2(STP)(PMe3)] [STP=(Me2PCH2)3Si(Me)] [26],
[24],
[RuCl2(MTP){P
[RuCl2
those in the related complexes RuCl2(TTP)][2.394(2),
˚
2.439(2) A] [22] and [RuCl2(CyTTP)]·2Me2SO
˚
[2.406(2), 2.455(1) A] [23] are distinctly different, the
longer bond in each case being trans to a phosphorus

I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
257
donor atom. The average value in the present complex
Acknowledgements
˚
[2.3871 A] is shorter the average values in these two
˚
complexes, 2.416 and 2.430 A, respectively, but com-
The NMR spectra were recorded by Eric Lewis of the
University of Wales, Bangor whom I.R. Butler gratefully
thanks. Elemental analyses were performed in the same
department by G. Connolly. He also thanks the EU for
funding the Erasmus placement of U. Griesbach. The
high resolution mass spectra were obtained at the EP-
SRC central research facility at the University of Wales,
Swansea and he gratefully thanks the staff who per-
formed the analyses. P. Zanello gratefully acknowledges
the financial support of the University of Siena (ex quota
60%). Thanks are due to G. Montomoli for her technical
assistance. Finally I.R. Butler acknowledges the patience
of his research group allowing him to carry out much of
this synthetic work, particularly S. Quayle, G. Kelly and
J. Szewczyk. A loan of RuCl3 ·xH2O from Johnson-
Mathey p.l.c. is gratefully acknowledged.
parable with those in the monomeric five-coordinate
˚
complexes [RuCl2(PPh3)3][2.387 A] [36] and [RuCl2
˚
{PPh2(CH2SiMe3)}3] [2.392 A] [37].
The geometry parameters involving the triphosfer
ligand are basically as expected. However, the relative
orientations of the two ferrocenyl moieties with respect
to the RuP2 chelate planes are significantly different, as
indicated by the different values for the equivalent
torsion angles, Ru(1)–P(1)–C(1)–Fe(1) 25.0(4) and
Ru(1)–P(3)–C(16)–Fe(2) −51.7(3)°, and Ru(1)–P(2)–
C(11)–Fe(1) −26.5(4)° and Ru(1)–P(2)–C(11)–Fe(2)
−7.1(3)°. The Fe–C(Cp) distances vary from 2.019(5)
˚
to 2.055(5), average 2.040 A, and are comparable with
those in the related ferrocenyl–bis(phosphine) palladium
complex [PdBr(C14H8Cl){Fe(C5H4–PPh2)2}] [1.976(9)–
˚
2.048(8), average 2.014 A] [21] and ferrocenyl–an-
thracene complex [Fe(C5H4–C14H9)2] [2.023(4)–
2.083(4), average 2.047 A] [21]. The two Cp rings around
References
˚
[1] G.P. Sollot, J.L. Snead, S. Portnoy, W.R. Peterson Jr., H.E.
Mertwoy, Chem. Abstr. 63 (1965) 18147b.
[2] J.J. Bishop, A. Davison, M.L. Katcher, D.W. Lichtenberg, R.E.
Merrill, J.C. Smart, J. Organomet. Chem. 27 (1971) 241.
[3] A. Togni, T. Hayashi, editors. Ferrocenes. Weinheim: VCH
Press, 1995.
both iron atoms in the present complex are almost
parallel as indicated by the dihedral angles 4.1(4) [for
Fe(1)] and 2.7(3)° [for Fe(2)]. However, the Cp rings
around Fe(1) are almost eclipsed, whilst those around
Fe(2) are ca. 35° rotated from a totally eclipsed arrange-
ment. The bond lengths and angles in all the Cp rings
are as expected with minor deviations from ideal pen-
tagon due to the strain suffered by bonding with the
phosphorus atoms. In particular the bonds involving the
phosphorus bonded carbon atoms are slightly but defin-
itely lengthened compared with the other C–C bonds as
[4] (a) W.R. Cullen, J.D. Woolins, Coord. Chem. Rev. 39 (1981) 1.
(b) W.R. Cullen, S.J. Rettig, T.-C. Zheng, Organometallics 11
(1992) 277. (c) A. Togni, G. Rihs, R.E. Blumer, Organometallics
11 (1992) 613. (d) T.S.A. Hor, L.-T. Phang, Polyhedron 9 (1990)
2305. (e) T.S.A. Hor, S.P. Neo, C.S. Tan, T.C.W. Mak, K.W.P.
Leung, R.-J. Wang, Inorg. Chem. 31 (1992) 4510. (f) M.I. Bruce,
P.A. Humphrey, O. Shawkataly, M.R. Snow, E.R.T. Tiekint,
W.R. Cullen, Organometallics 9 (1990) 2910. (g) S.T. Chacon,
W.R. Cullen, M.I. Bruce, O. Shawkataly, F.W.B. Einstein, R.H.
Jones, A.C. Wilkes, Can. J. Chem. 68 (1990) 2001. (h) M. Sato,
M. Sekino, J. Organomet. Chem. 444 (1993) 185. (i) T. Hayashi,
A. Yamamoto, Y. Ito, E. Nishioka, H. Miura, K. Yanagi, J.
Am. Chem. Soc. 14 (1989) 6301. (j) I.R. Butler, W.R. Cullen,
T.-J. Kim, S.J. Rettig, J. Trotter, Organometallics 4 (1985) 972.
[5] C.R.S.M. Hampton, I.R. Butler, W.R. Cullen, B.R. James, J.P.
Charland, J. Simpson, Inorg. Chem. 31 (1992) 5509.
[6] (a) I.R. Butler, R.L. Davies, Synthesis (1996) 1350. (b) I.R.
Butler, S. Sellarajah, Undergraduate Dissertation, University of
Wales, 1996.
˚
indicated by their average values 1.434 versus 1.413 A.
The C–C–C angles at these atoms are also slightly
smaller than those at other carbon atoms [106.2(4)–
107.0(4), average 106.7° vs. 107.5(4)–109.2(4), average
108.3°]. Similar distortions of the Cp rings due to
anthracenyl substituents were also observed in the above
mentioned [Fe(C5H4–C14H9)2] complex [21]. The phenyl
rings also suffer some distortions from ideal hexagon,
mainly at the carbon bonded to phosphorus; the C–C–
C internal angles for these atoms are all slightly nar-
rower [117.4(4)–119.6(5), average 118.2°] than the ideal
value 120°.
[7] A. Togni, M. Hobi, G. Rihs, G. Rist, A. Albinati, D. Zech, H.
Keller, Organometallics 13 (1994) 1224.
[8] R. Holm, Inorg. Synth. 12 (1971) 238.
It has been mentioned earlier that the ruthenium metal
in the present compound has a five-coordinate square
pyramidal coordination; however, it is also of interest to
note that one of the phenyl hydrogens [H(32)] makes a
close approach to the metal, from the opposite side of
P(2), within 2.89 and this gives it a pseudo-octahedral
environment. This Ru(1)···H(32) contact is significant in
that it blocks the unused octahedral site of the metal
centre. Similar situation, also observed in other cases,
e.g. [RuCl2(PPh3)3] [36] and [RhBr(1-naphthyl)2(PPh3)2]
[38], has been suggested to give extra stability to the
electron deficient metal centres in such species [38].
[9] J.A. Darr, S.R. Drake, M.B. Hursthouse, K.M.A. Malik, Inorg.
Chem. 32 (1993) 5704.
[10] N.P.C. Walker, D. Stuart. DIFABS Program for Absorption
Correction, Acta Crystallogr. Sect. A 39 (1983) 158; adapted for
FAST geometry by A.I. Karaulov, University of Wales Cardiff,
1991.
[11] G.M. Sheldrick, SHELXS86 program for crystal structure solu-
tion, Acta Crystallogr. Sect. A 46 (1990) 467.
[12] G.M. Sheldrick, SHELXL93 program for crystal structure refin-
ement, University of Go¨ttingen, Germany, 1993.
[13] K. Davies, SNOOPI program for crystal structure drawing,
University of Oxford, UK, 1983.
[14] M.M. Taquikhan, R. Mohiuddina J. Coord. Chem. 6 (1977) 171
and references therein.

258
I.R. Butler et al. / Journal of Organometallic Chemistry 565 (1998) 243–258
[15] (a) X.-L. Luo, R.H. Crabtree, J. Chem. Soc. Dalton Trans.
(1991) 587. (b) M.T. Costello, P.E. Fanwick, M.A. Green, R.A.
Walton, Inorg. Chem. 31 (1992) 2359. (c) S.A. Westcott, G.
Stringer, S. Anderson, N.J. Taylor, T.B. Marder, Inorg. Chem.
33 (1994) 4589.
[16] (a) D. Michos, X.-L. Luo, R.H. Crabtree, Inorg. Chem. 31
(1992) 4245. (b) C. Bianchini, P.J. Perez, M. Peruzzini, F.
Zanobini, A. Vacca, Inorg. Chem. 30 (1991) 279.
[17] J.C. Kotz, C.L. Nivert, J.M. Lieber, R.C. Reed, J. Organomet.
Chem. 91 (1975) 87.
[24] R.R. Guimerans, E.C. Hernandez, M.M. Olmstead, A.L. Balch,
Inorg. Chim. Acta 165 (1989) 45.
[25] G. Delgado, A.V. Rivera, T. Suarez, B. Fontal, Inorg. Chim.
Acta 233 (1995) 145.
[26] I. Dahlenburg, S. Kerstan, D. Werner, J. Organomet. Chem. 411
(1991) 457.
[27] L. Solujic, E.B. Milosavijevic, J.H. Nelson, N.W. Alcock, J.
Fischer, Inorg. Chem. 28 (1989) 3453.
[28] W.S. Sheldrick, K. Brandt, Inorg. Chim. Acta 217 (1994) 51.
[29] R.F. Rhodes, C. Sorato, L.M. Venanzi, F. Bachechi, Inorg.
Chem. 26 (1987) 604.
[30] D.W. Krassowski, J.H. Nelson, K.R. Brower, D. Hauenstein,
R.A. Jacobson, Inorg. Chem. 27 (1988) 4294.
[31] J. Halpern, C.S. Wong, J. Chem. Soc. Chem. Comm. (1973) 629.
[32] B.R. James, L.O. Markham, Inorg. Chem. 13 (1974) 97.
[33] P.S. Halmann, B.R. McGarvey, G. Wilkinson, J. Chem. Soc. A
(1968) 3143.
[34] T.A. Stephenson, G. Wilkinson, Inorg. Nucl. Chem. 28 (1966)
945.
[18] A. Houlton, P.T. Bishop, R.M.G. Roberts, J. Silver, J.
Organomet. Chem. 364 (1989) 381.
[19] P. Zanello, G. Opromolla, G. Giorgi, G. Sasso, A. Togni, J.
Organomet. Chem. 506 (1996) 61 and references therein.
[20] (a) M.D. Ward, Chem. Soc. Rev. (1995) 121. (b) R.R. Gagne´,
C.A. Kowal, T.J. Smith, M.C. Cimolind, J. Am. Chem. Soc. 101
(1979) 4571. (c) A.W. Bott, Current Separations, Bioanalyt. Syst.
16 (1997) 61.
[21] I.R. Butler, L.J. Hobson, S.J. Coles, M.B. Hursthouse, K.M.A.
Malik, J. Organomet. Chem. 540 (1997) 27.
[22] A. Albianati, Qiongzhong Jiang, H. Ruegger, L.M. Venanzi,
Inorg. Chem. 32 (1993) 4940.
[35] M.R.M. Fontes, G. Oliva, L.A.C. Cordeiro, A.A. Batista, J.
Coord. Chem. 30 (1993) 125.
[36] S.J.L. Placa, J.A. Ibers, Inorg. Chem. 4 (1965) 778.
[37] F.A. Cotton, M. Matusz, Inorg. Chim. Acta 131 (1987) 213.
[38] J. Chatt, A.E, Underhill, J. Chem. Soc. (1963) 2088.
[23] G. Jia, I. Lee, D.W. Meek, J.C. Gallucci, Inorg. Chim. Acta 177
(1990) 81.
.