Organometallic complexes for nonlinear optics. Part 27. Syntheses and optical properties of some iron, ruthenium and osmium alkynyl complexes

Article abstract of DOI:10.1016/S0020-1693(03)00126-9

The syntheses of the alkynyl complexes M(4-C≡CC₆H₄NO₂)(dppe)(η-C ₅H₅) [M=Fe (1), Ru (2), Os (3)], Os(4-C≡CC₆H₄NO₂)(PPh₃) ₂(η-C₅H₅) (4) and Ru(4-C≡CC₆H₄NO₂)(CO)₂ (η-C₅H₅) (5) are reported. Structural studies reveal a decrease in Ru-C(1) distance on proceeding from 5 to 2, consistent with greater back-donation of electron density to the alkynyl ligand from the more electron-rich metal center in 2. Electrochemical data show that the M^II/III couple for the dicarbonyl complex 5 is at a significantly more positive potential than that of the related diphosphine complex 2, consistent with ligand variation modifying the electron richness and hence donor strength of the metal center. Time-dependent density functional calculations on model complexes M(4-C≡CC₆H₄NO₂)(PH₃) ₂(η-C₅H₅) (M=Fe, Ru, Os) have been employed to assign the intense low-energy optical transition in these complexes as MLCT in character, the higher energy band being phenyl-phenyl* in nature. Molecular quadratic optical nonlinearities have been measured using the hyper-Rayleigh scattering procedure at 1064 nm. β values vary as Fe≤Ru≤Os for metal variation and CO

Full text of DOI:10.1016/S0020-1693(03)00126-9

Inorganica Chimica Acta 352 (2003) 9ꢁ18
/
www.elsevier.com/locate/ica
Organometallic complexes for nonlinear optics.
Part 27. Syntheses and optical properties of some iron, ruthenium
and osmium alkynyl complexesꢀ
Clem E. Powell a, Marie P. Cifuentes a, Andrew M. McDonagh a,
Stephanie K. Hurst a, Nigel T. Lucas a, Christopher D. Delfs a, Robert Stranger a,
Mark G. Humphrey a,*, Stephan Houbrechts b, Inge Asselberghs b, Andre´ Persoons b,
David C.R. Hockless c
a Department of Chemistry, Australian National University, Canberra, ACT 0200, Australia
b Laboratory for Chemical and Biological Dynamics, University of Leuven, Celestijnenlaan 200D, B-3001 Leuven, Belgium
c Research School of Chemistry, Australian National University, Canberra, ACT 0200, Australia
Received 27 September 2001; accepted 30 October 2001
Dedicated to Professor Martin Bennett in recognition of his outstanding contributions to organometallic chemistry
Abstract
The syntheses of the alkynyl complexes M(4-Cꢀ
CC6H4NO2)(PPh3)2(h-C5H5) (4) and Ru(4-CꢀCC6H4NO2)(CO)2(h-C5H5) (5) are reported. Structural studies reveal a decrease in
RuꢁC(1) distance on proceeding from 5 to 2, consistent with greater back-donation of electron density to the alkynyl ligand from
/
CC6H4NO2)(dppe)(h-C5H5) [Mꢀ
/
Fe (1), Ru (2), Os (3)], Os(4-Cꢀ
/
/
/
the more electron-rich metal center in 2. Electrochemical data show that the MII/III couple for the dicarbonyl complex 5 is at a
significantly more positive potential than that of the related diphosphine complex 2, consistent with ligand variation modifying the
electron richness and hence donor strength of the metal center. Time-dependent density functional calculations on model complexes
M(4-Cꢀ
these complexes as MLCT in character, the higher energy band being phenylꢁ
nonlinearities have been measured using the hyper-Rayleigh scattering procedure at 1064 nm. b values vary as Fe5
metal variation and COBphosphines for co-ligand variation, the latter consistent with the variation in donor strength of the metal
/
CC6H4NO2)(PH3)2(h-C5H5) (Mꢀ
/
Fe, Ru, Os) have been employed to assign the intense low-energy optical transition in
phenyl* in nature. Molecular quadratic optical
Ru5Os for
/
/
/
/
center inferred from electrochemical and crystallographic data. The observed trend in b on metal variation follows the trend in
backbonding energies calculated by DFT.
# 2003 Elsevier Science B.V. All rights reserved.
Keywords: Iron; Ruthenium; Osmium; Cyclic voltammetry; X-ray structure; Alkynyl; NLO; Density functional theory
1. Introduction
investigated have the donorꢁ
/
bridgeꢁacceptor composi-
/
tion which has been shown to afford organic materials
with high NLO merit. In this complex composition, the
ligated metal can adopt the donor, bridge, or acceptor
role, and for some bimetallics the metal can function as
both donor and acceptor. A plethora of studies have
focussed on modification of the organic bridging
ligands, and many studies have assessed the effect of
variation of the co-ligands at the ligated metal center on
optical nonlinearity. One less-explored area which is of
fundamental importance for the utility of organometal-
lics in NLO applications is the effect of metal variation
The nonlinear optical (NLO) properties of organo-
metallic complexes have attracted significant attention
over the past decade [2ꢁ6]. The vast majority of
/
organometallics for which NLO properties have been
ꢀ Part 26: see Ref. [1].
* Corresponding author: Tel.: ꢂ
0760.
E-mail address: mark.humphrey@anu.edu.au (M.G. Humphrey).
/
61-2-6125 2927; fax: ꢂ61-2-6125
/
0020-1693/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0020-1693(03)00126-9

10
C.E. Powell et al. / Inorganica Chimica Acta 352 (2003) 9ꢁ18
/
on nonlinearity. Group 8 metal variation in metalloce-
nyl [7ꢁ10] and nitrile [11] complexes revealed the
quadratic NLO efficiency series FeꢀRu, suggesting a
correlation with ease of oxidation; for the nitrile
complexes, correlation with n(CꢀN) values was found,
/
/
/
Fig. 1. NMR atom numbering scheme.
suggestive of the importance of backbonding [11]. We
have previously examined analogous 14 valence electron
gold alkynyl complexes and 18 valence electron nickel
alkynyl complexes, noting an increase in b values on
increasing metal valence electron count [12,13], and 18
valence electron ruthenium complexes, observing a
further increase in b values which may result from
increasing ease of oxidation [14,15]. We report herein
synthesis of a systematically varied series of 4-nitrophe-
nylalkynyl complexes of the Group 8 metals, examina-
tion of their spectroscopic data, electrochemical
properties and selected crystallographic data, their
molecular quadratic nonlinearities at 1064 nm, and
approximate density functional calculations which
have been undertaken in order to understand the
electronic structure and optical properties, and rationa-
lize the observed trend in optical nonlinearities.
assignments follow the numbering scheme shown in Fig.
1. UVꢁVis spectra were recorded using a Cary 5
/
spectrophotometer as thf solutions in 1 cm cells and
are reported as lmax, cmꢃ1 [o, 104 Mꢃ1 cmꢃ1]. Mass
spectra were recorded using a VG ZAB 2SEQ instru-
ment (30 kV Csꢂ ions, current 1 mA, accelerating
potential 8 kV, 3-nitrobenzyl alcohol matrix) at the
Australian National University; peaks are reported as
m/z (assignment, relative intensity).
Electrochemical measurements were recorded using a
MacLab 400 interface and MacLab potentiostat from
ADInstruments (using a glassy carbon disc working, Pt
auxiliary and Ag/AgCl reference mini-electrodes from
Cypress Systems). Scan rates were typically 100 mV sꢃ1
.
Electrochemical solutions contained 0.1 M [NBun4 ]PF6
and approximately 10ꢃ3 M complex in dichloro-
methane. Solutions were purged and maintained under
an atmosphere of nitrogen. All values are referenced to
an internal sample of ferrocene (0.56 V).
2. Experimental
Calculations were performed on a Linux-based Pen-
tium III computer (600 MHz) using the Amsterdam
2.1. General remarks
Density Functional Theory (ADF) Release 1999 [22ꢁ
/
Reactions were carried out using Schlenk techniques
under an atmosphere of nitrogen, although subsequent
work up was carried out without any precautions to
exclude air. The following were prepared by literature
24]. The local exchange correlation approximation of
Vosko et al. [25] was used with the corrections of Becke
[26] and Perdew et al. [27,28]. Triple-j basis sets (Type
IV) were used for all atoms. Core orbitals were frozen
through 1s (C), 2p (P), 3p (Fe), 4p (Ru), and 4f (Os).
Relativistic corrections were incorporated using the
ZORA functionality [29]. Geometries were optimized
using the algorithm of Versluis and Ziegler [30]. Optical
spectra were calculated using the time-dependent density
functional theory (TD-DFT) functionality available in
ADF [31].
methods: 4-HCꢀ
[17], FeCl(dppe)(h-C5H5) [18], RuCl(dppe)(h-C5H5)
[19], OsBr(dppe)(h-C5H5) [19], and Fe(4-CꢀCC6-
/
CC6H4NO2 [16], RuCl(CO)2(h-C5H5)
/
H4NO2)(CO)2(h-C5H5) (6) [20]. The literature proce-
dure for the preparation of OsBr(PPh3)2(h-C5H5) [21]
was modified in the following way: MeOH was sub-
stituted for EtOH as solvent, reaction time was reduced
from 12 h to 30 min, Et2O was substituted for C6H6 as
the eluant for column chromatography. NH4PF6,
NaBH4, NEt3, CuI and HC2Ph were obtained commer-
cially and used as received. Petrol refers to a mixture
2.2. Fe(4-Cꢀ
/
CC6H4NO2)(dppe)(h-C5H5) (1)
with boiling point range 60ꢁ
/
80 8C. Solvents CH2Cl2,
A mixture of FeCl(dppe)(h-C5H5) (200 mg, 0.36
mmol), 4-HCꢀCC6H4NO2 and NH4PF6 was stirred in
Et2O, MeOH, and n-hexane were dried using standard
procedures. Thin layer chromatography (tlc) was on
Merck GF254 silica gel (0.5 mm); column chromatogra-
phy used ungraded basic alumina.
Elemental microanalyses were performed by the
Microanalysis Service Unit at the Australian National
University. Infrared spectra were recorded as CH2Cl2
/
MeOH (40 ml) for 90 min at reflux. The mixture was
allowed to cool to room temperature and then 2 ml of
NaOMe solution (0.33 M in MeOH) was added with
stirring. The solvent was removed in vacuo and the
residue was taken up in CH2Cl2 and placed onto an
alumina column. Excess acetylene was eluted using 3:7
solutions in CaF2 cells using a Perkinꢁ/Elmer System
CH2Cl2ꢁ/petrol and the product was eluted using 3:5
2000 FT-IR. NMR spectra were recorded in CDCl3
using a Varian Gemini-300 FT NMR spectrometer and
are referenced to residual solvent (1H, 300 MHz; 13C, 75
MHz) or external 85% H3PO4 (31P, 121 MHz). NMR
CH2Cl2ꢁ
/
petrol. The solvent was reduced to Â10 ml
/
and the resultant solid collected by filtration and washed
with petrol to afford a red solid identified as 1 (75 mg,
31%). MS: 665 ([M]ꢂ, 65), 649 ([MꢃO]ꢂ, 7), 519
/

C.E. Powell et al. / Inorganica Chimica Acta 352 (2003) 9ꢁ
/
18
11
([Fe(dppe)(C5H5)]ꢂ, 100). Anal. Calc. for C39H33Fe-
NO2P2: C, 70.39; H, 5.00; N, 2.10. Found: C, 70.37; H,
2.5. Os(4-Cꢀ
/
CC6H4NO2)(PPh3)2(h-C5H5) (4)
5.10; N, 2.30%. IR (cmꢃ1): 2044w n(Cꢀ
000 [1.3], 31 500 [1.0]. H NMR: d 2.23 (m, 2H, CH2),
/
C). UVꢁ/Vis: 20
A mixture of OsBr(PPh3)2(h-C5H5) (250 mg, 0.29
mmol), 4-HCꢀCC6H4NO2 (130 mg, 0.88 mmol) and
1
/
2.55 (m, 2H, CH2), 4.29 (s, 5H, C5H5), 6.25 (m, 2H, H4),
7.84 (m, 22H, H5ꢂ
Ph). 31P NMR: d 106.2 (s (br),
PPh2).
NH4PF6 (150 mg, 0.92 mmol) was stirred in MeOH (40
ml) for 3 h at reflux. The mixture was allowed to cool to
room temperature and then 2 ml of NaOMe solution
(0.3 M in MeOH) was added with stirring. The solvent
was removed and the product taken up in CH2Cl2 and
purified by column chromatography on alumina, eluting
7.20ꢁ
/
/
2.3. Ru(4-Cꢀ
/
CC6H4NO2)(dppe)(h-C5H5) (2)
firstly with 3:7 CH2Cl2ꢁ
acetylene, and then with 6:4 CH2Cl2ꢁpetrol to remove
the product. The solvent was removed and the product
/
petrol to remove any excess
A mixture of RuCl(dppe)(h-C5H5) (250 mg, 0.35
mmol), 4-HCꢀCC6H4NO2 (130 mg, 0.88 mmol) and
NH4PF6 (150 mg, 0.92 mmol) was stirred in MeOH (40
ml) for 1 h at reflux. On cooling to room temperature,
CH2Cl2 (40 ml) was added and the mixture filtered.
/
/
was collected to yield 210 mg (78%) of dark red
H]ꢂ, 100),
microcrystals identified as 4. MS: 927 ([Mꢂ
/
781 ([Os(PPh3)2(C5H5)]ꢂ, 10), 664 ([MꢃPPh3]ꢂ, 35).
/
Addition of petrol (Â20 ml) and volume reduction of
/
Anal. Calc. for C49H39NO2OsP2: C, 63.56; H, 4.25; N,
1.51. Found: C, 63.55; H, 4.42; N, 1.61%. IR (cmꢃ1):
the solvent in vacuo afforded a solid which was collected
by filtration in air and then redissolved in CH2Cl2 (10
ml). To this solution was added NaOMe solution (10 ml
of a 0.1 M solution in MeOH) with stirring. The solvent
was removed and the product purified by column
chromatography on alumina, eluting with 65:35
2054w n(Cꢀ
1H NMR: d 4.43 (s, 5H, C5H5), 6.98 (d, JHH
2H, H4), 7.05ꢁ7.35 (m, 30H, Ph), 8.00 (d, JHH
2H, H5). 31P NMR: d 2.8 (s, PPh3).
/
C). UVꢁ
/
Vis: 21 000 [2.2], 36 400 (sh) [1.6].
9 Hz,
9 Hz,
ꢀ
/
/
ꢀ
/
CH2Cl2ꢁpetrol. Reduction of the solvent volume af-
/
2.6. Ru(4-Cꢀ
/
CC6H4NO2)(CO)2(h-C5H5) (5)
forded a red solid identified as 2 (75 mg, 30%). MS: 711
([M]ꢂ, 100), 565 ([Ru(dppe)(C5H5)]ꢂ. Anal. Calc. for
C39H33NO2P2Ru: C, 65.91; H, 4.68; N, 1.97. Found: C,
RuCl(CO)2(h-C5H5) (230 mg, 0.89 mmol), 4-HCꢀ
/
CC6H4NO2 (157 mg, 1.07 mmol) and CuI (5 mg) were
stirred in NEt3 (30 ml) for 2 h. The mixture was filtered
and the filtrate taken to dryness. The resulting residue
was taken up in CH2Cl2 and subjected to tlc, eluting
65.91; H, 4.60; N, 1.79%. IR (cmꢃ1): 2056w n(Cꢀ
/
C).
Vis: 22 200 [1.8]. H NMR: d 2.30 (m, 2H, CH2),
2.57 (m, 2H, CH2), 4.79 (s, 5H, C5H5), 6.31 (d, JHH
Hz, 2H, H4), 7.26ꢁ7.87 (m, 22H, H5ꢂ
Ph). 31P NMR: d
86.2 (s, PPh2).
1
UVꢁ
/
ꢀ7
/
/
/
with 6:4 CH2Cl2ꢁpetrol. The major band was collected,
taken to dryness, and the residue extracted with acetone.
The solvent was removed to afford a red solid which was
/
2.4. Os(4-Cꢀ
/
CC6H4NO2)(dppe)(h-C5H5) (3)
crystallized from Et2Oꢁ
/
hexane to afford reddishꢁbrown
/
crystals of 5 (110 mg, 34%). IR (cmꢃ1): 2108w n(Cꢀ
/C),
1
2056s, 2005s n(CO). H NMR: d 8.03 (d, JHH
A mixture of OsBr(dppe)(h-C5H5) (160 mg, 0.22
mmol), 4-HCꢀCC6H4NO2 (50 mg, 0.34 mmol) and
ꢀ/9 Hz,
/
2H, H5), 7.35 (d, JHHꢀ9 Hz, 2H, H4), 5.45 (s, 5H,
/
NH4PF6 (75 mg, 0.46 mmol) was stirred in MeOH (30
ml) for 65 h at reflux. The solvent was removed in vacuo
and the residue extracted into CH2Cl2 (10 ml) and
filtered through a sintered glass funnel. Et2O (50 ml) was
added to the filtrate and 100 mg of yellow powder was
collected. This was dissolved in CH2Cl2 (10 ml) and 2 ml
of NaOMe solution (0.3 M in methanol) was added with
stirring. The solvent was removed and the product taken
up in CH2Cl2 and purified by column chromatography
on alumina, eluting with CH2Cl2. The solvent was
removed and the resulting dark red solid collected and
identified as 3 (30 mg, 17%). MS: 802 ([M]ꢂ, 100), 655
C5H5). 13C NMR: d 196.2 (CO), 144.7 (C6), 134.4 (C1),
131.9 (C4), 123.3 (C5), 110.4 (C3), 96.7 (C2), 88.0 (C5H5).
Anal. Calc. for C15H9NO4Ru: C, 48.92; H, 2.46; N,
3.80. Found: C, 49.23; H, 2.23; N, 3.79%.
2.7. Spectroscopic data for Fe(4-Cꢀ
CC6H4NO2)(CO)2(h-C5H5) (6)
/
IR (cmꢃ1): 2104m n(Cꢀ
/
C), 2044s, 1999s n(CO)
cmꢃ1. H NMR: d 8.03 (d, JHH
9 Hz, 2H, H5), 7.33
(d, JHH
9 Hz, 2H, H4), 5.06 (s, 5H, C5H5). 13C NMR:
1
ꢀ
/
ꢀ
/
d 211.7 (CO), 144.7 (C6), 134.7 (C1), 131.5 (C4), 123.4
(C5), 116.2 (C3), 104.7 (C2), 85.5 (C5H5).
([Os(dppe)(C5H5)]ꢂ,
C39H33NO2OsP2: C, 58.57; H, 4.16; N, 1.75. Found:
C, 58.40; H, 4.11; N, 1.57%. IR (cmꢃ1): 2055w n(Cꢀ
C).
UVꢁ
Vis: 21 600 [1.7], 36 800 (sh) [1.0]. 1H NMR: d 2.42
(m, 4H, CH2), 4.78 (s, 5H, C5H5), 6.25 (d, JHH
2H, H4), 7.17ꢁ7.38 (m, 20H, Ph), 7.72 (d, JHH
2H, H5). 31P NMR: d 46.7 (s, PPh2).
25).
Anal.
Calc.
for
/
2.8. X-ray structural determinations
/
ꢀ
/
9 Hz,
9 Hz,
Crystals suitable for the X-ray structural analyses
were grown by liquid diffusion techniques from hexaneꢁ
CH2Cl2 (5) or methanolꢁCH2Cl2 (2) at 276 K. Selected
/
ꢀ
/
/
/

12
C.E. Powell et al. / Inorganica Chimica Acta 352 (2003) 9ꢁ18
/
Table 1
Crystal data and structure refinement parameters for 2 and 5
least-squares refinement were based on Nobs observed
reflections (I ꢀ2s(I)) and converged to R and Rw.
Selected bond lengths and angles for 2, 5 and some
/
2
5
related ruthenium alkynyl complexes are given in Table
2.
Empirical formula
Formula weight
Crystal size (mm)
T (K)
C
710.71
39H33NO2P2Ru
C15H9NO4Ru
368.31
0.16ꢄ
200
0.71069
/0.16ꢄ/0.06
0.36ꢄ
/
0.24ꢄ0.18
/
2.9. Hyper-Rayleigh scattering measurements
296
1.5418
˚
Wavelength (A)
An injection seeded Nd:YAG laser (Q-switched
Nd:YAG Quanta Ray GCR, 1064 nm, 8 ns pulses, 10
Hz) was focussed into a cylindrical cell (7 ml) containing
the sample. The intensity of the incident beam was
varied by rotation of a half-wave plate placed between
crossed polarizers. Part of the laser pulse was sampled
by a photodiode to measure the vertically polarized
incident light intensity. The frequency doubled light was
collected by an efficient condenser system and detected
by a photomultiplier. The harmonic scattering and
linear scattering were distinguished by appropriate
filters; gated integrators were used to obtain intensities
of the incident and harmonic scattered light. All
measurements were performed in thf using p-nitroani-
Crystal system
Space group
Unit cell dimensions
monoclinic
C2/c (no. 15)
monoclinic
P21/n (no. 14)
˚
a (A)
29.1649(7)
11.5689(3)
19.7324(5)
97.389()2)
6602.5(3)
8
15.987(3)
5.775(4)
16.054(2)
104.88(1)
1432.5(9)
4
˚
b (A)
˚
c (A)
b (8)
3
˚
V (A )
Z
Range for data collection, 3.04ꢁ
u (8)
Index ranges
/
25.04
2.85ꢁ60.06
/
ꢃ
ꢃ/
/
345
235
05k 5
/
h 5
/
34,
ꢃ
/
125
175
175
/
k 5
h 5
l 5
/
13,
/
l 523,
/
ꢃ
/
/
/0,
/
/6
ꢃ/
/
/
18
m (mmꢃ1
)
6.08 (Mo Ka)
0.694, 0.656
46 245
9.015 (Cu Ka)
0.197, 0.097
2473
Max./min. transmission
Reflections collected,
Ncollected
line (bꢀ
/
21.4ꢄ
10ꢃ30 esu [50]) as a reference.
/
Independent reflections,
Nunique
Observed reflections
5834
(Rint
2130
(Rint
ꢀ
/
0.060)
ꢀ0.081)
/
3705
1667
3. Results and discussion
(I ꢀ
Goodness of fit
(I ꢀ2s(I))
Final R indices
/2s(I)), Nobs
0.70
1.66
3.1. Synthesis and characterization of metal alkynyl
complexes
/
Rꢀ
Rwꢀ
Rꢀ0.0547,
Rwꢀ0.0411
[s2(Fo)ꢂ
0.00024jFoj ]
0.26 and ꢃ0.36
/0.0271,
Rꢀ
Rwꢀ
Rꢀ0.0539,
Rwꢀ0.0436
[s2(Fo)ꢂ
/0.0377,
(I ꢀ
/
2s(I)) a
/
0.0323
/
0.0411
R indices (all data) a
/
/
We have previously reported synthesis and NLO
studies of (cyclopentadienyl)bis(phosphine) ruthenium
alkynyl complexes [14,15,38ꢁ43], examples of which
have some of the largest quadratic nonlinearities re-
ported for organometallics; we therefore decided to
pursue complexes with this complex composition to
investigate the effect of metal variation on b values. The
/
/
w
/
/
2 ꢃ1
2 ꢃ1
/
0.00027jFoj ]
Largest difference peak
ꢃ3
/
0.57 and ꢃ
/0.52
˚
and hole (e A
)
1=2
/
/
Rꢀa(½½Fo½ꢃ½Fc½½)=a½Fo½;
Rw ꢀ½aw(½Fo½ꢃ½Fc½)2=awFo2ꢀ
:
a
syntheses of the new alkynyl metal complexes M(4-Cꢀ
CC6H4NO2)(dppe)(h-C5H5) [MꢀFe (1), Ru (2), Os (3)]
and Os(4-CꢀCC6H4NO2)(PPh3)2(h-C5H5) (4) were
based on extending published procedures for the related
phenylalkynyl complexes [44], and involve the formation
of an intermediate vinylidene complex by reaction of a
terminal alkyne with a ligated transition metal halide,
followed by deprotonation of the b-carbon of the
vinylidene complex upon addition of methoxide ion
/
/
/
crystal data and structure refinement parameters are
collected in Table 1. For each study, a single orange
crystal was mounted on a glass fibre, and data were
collected at 296 K on a Rigaku AFC6R diffractometer
using graphite-monochromated Cu Ka radiation (5), or
at 200 K on a Nonius KappaCCD diffractometer using
graphite-monochromated Mo Ka radiation (2). The
reduced data [32,33] were corrected for absorption using
empirical c-scan (5) or numerical (2) [34] methods,
implemented from within TEXSAN [33] and MAXUS [35],
respectively. The structures were solved by direct
methods [36], and expanded using difference Fourier
techniques [37] within TEXSAN [33]. Non-hydrogen
atoms were refined anisotropically; hydrogen atoms
were included in idealized positions which were fre-
quently recalculated. The final cycles of full-matrix
(Scheme 1). For the dicarbonyl complex Ru(4-Cꢀ
/
CC6H4NO2)(CO)2(h-C5H5) (5), this route afforded
only unreacted starting material. An alternate route,
involving reaction of the chloro precursor and the
acetylene in NEt3 in the presence of catalytic amounts
of CuI, was found to give the desired product (Scheme
1). The analogous iron complex, 6, was prepared
following the published route [20,45]. The new alkynyl
complexes were characterized by secondary ion (SI)
1
Vis, and IR spectroscopy, H,
mass spectrometry, UVꢁ
/

C.E. Powell et al. / Inorganica Chimica Acta 352 (2003) 9ꢁ
/
18
13
Table 2
˚
Selected bond lengths (A) and angles (8) for 2, 5 and some related ruthenium alkynyl complexes
Bond lengths
Bond angles
Ru(1)ꢁ/C(1)
C(1)ꢁ
/
C(2)
C(2)ꢁ/C(201)
Ruꢁ
/
C(1)ꢁ/C(2)
C(1)ꢁ
/
C(2)ꢁC(201)
/
Ru(4-Cꢀ
Ru(4-Cꢀ
Ru(4-Cꢀ
Ru(4-Cꢀ
Ru(Cꢀ
/
CC6H4NO2)(CO)2(h-C5H5) (5)
CC6H4NO2)(PMe3)2(h-C5H5) a
CC6H4NO2)(PPh3)2(h-C5H5) a
CC6H4NO2)(dppe)(h-C5H5) (2)
2.023(6)
1.99(2)
1.197(9)
1.23(2)
1.427(9)
1.43(3)
178.9(6)
178.0(2)
175.9(4)
177.4(3)
177.7(4)
177.5(8)
177.0(3)
175.0(9)
176.7(3)
170.6(5)
/
/
1.994(5)
1.993(3)
2.017(5)
1.202(8)
1.214(4)
1.214(7)
1.432(7)
1.424(4)
1.462(8)
/
/
CPh)(PPh3)2(h-C5H5) b
a
Ref. [38].
Ref. [51].
b
31P and, in selected cases, 13C NMR spectroscopy; we
have also summarized previously unreported spectro-
scopic data for 6.
The SI mass spectra of the new complexes show
signals due to the molecular ion, fragmentation occur-
ring by the loss of the alkynyl ligand. The UVꢁVis
/
Scheme 1. Syntheses of alkynylꢁmetal complexes.
/

14
C.E. Powell et al. / Inorganica Chimica Acta 352 (2003) 9ꢁ18
/
spectra contain a band in the range 20 000ꢁ
/
27 500
cmꢃ1, tentatively assigned to the MLCT transition from
the metal to the alkynyl ligand; frequencies for optical
The structural data in Table 2 provide the opportu-
nity to assess the effect of systematic structural mod-
ifications on important metrical parameters. However,
most differences which exist are not statistically sig-
absorption maxima follow the trend FeB
/
OsBRu. The
/
IR spectra show characteristic n(Cꢀ
/
C) bands at 2044ꢁ
/
nificant. Thus, the Ru(1)ꢁ
/
C(1) distance decreases on
CC6H4NO2)-
2056 cmꢃ1, with the two iron complexes having the
lowest energy bands.
proceeding from 5 to 2 and Ru(4-Cꢀ
/
(PPh3)2(h-C5H5), consistent with greater p-backbonding
to the alkynyl ligand from the more electron-rich metal
centers in the latter complexes, but other variations in
bond distances are not meaningful. The small deviations
3.2. X-ray structural studies of Ru(4-Cꢀ
/
in bond angles Ruꢁ
/
C(1)ꢁ
/
C(2) and C(1)ꢁ
/
C(2)ꢁ
/
C(201)
CC6H4NO2)L2(h-C5H5) [Lꢀ1/2 dppe (2), CO (5)]
/
from the idealized 1808 probably reflect crystal packing
effects.
Structural studies of 2 and 5 were undertaken. Crystal
data and structure refinement parameters are given in
Table 1, and selected bond lengths and angles collected
in Table 2; the latter also includes data for some related
ruthenium alkynyl complexes. Figs. 2 (2) and 3 (5)
contain ORTEP plots showing the molecular geometries
and atomic labelling schemes.
3.3. Electrochemical studies
Cyclic voltammetric data are summarized in Table 3.
Complexes 1, 2, 4, 5, and 7 exhibit anodic and cathodic
processes assigned to metal-centered oxidation and
Fig. 2. Molecular structure and atom numbering scheme for Ru(4-Cꢀ
/
CC6H4NO2)(dppe)(h-C5H5) (2). 30% displacement ellipsoids are shown for
non-hydrogen atoms. Hydrogen atoms have arbitrary radii.
Fig. 3. Molecular structure and atom numbering scheme for Ru(4-Cꢀ
/
CC6H4NO2)(CO)2(h-C5H5) (5). 30% displacement ellipsoids are shown for
non-hydrogen atoms. Hydrogen atoms have arbitrary radii.

C.E. Powell et al. / Inorganica Chimica Acta 352 (2003) 9ꢁ
/
18
15
first hyperpolarizabilities (bHRS) are shown together
with static first hyperpolarizabilities (b0) calculated
from the experimental values using the two-level
approximation [the shortcomings of the two-level model
have been discussed elsewhere (e.g., Ref. [2]). This model
was developed for a restricted class of organic com-
pounds where structural modifications are directed at
the charge-transfer band thought to contribute to the
hyperpolarizability, and may not be useful where there
are several dominant transitions close to 2v].
Table 3
Cyclic voltammetric data
Compound
E1/2 (V) [ipc/ipa
MII/III (V)
]
E1/2 [ipa/ipc
]
NO20/ꢃI (V)
a
Fe(4-Cꢀ
(h-C5H5) (6)
Ru(4-CꢀCC6H4NO2)(CO)2-
(h-C5H5) (5)
Fe(4-CꢀCC6H4NO2)(dppe)-
(h-C5H5) (1)
Ru(4-CꢀCC6H4NO2)(dppe)-
(h-C5H5) (2)
Ru(4-CꢀCC6H4NO2)(PPh3)2-
(h-C5H5) (7) b
Os(4-CꢀCC6H4NO2)(PPh3)2-
(h-C5H5) (4)
/
CC6H4NO2)(CO)2-
ꢃ
/
1.04 [0.7]
1.07 [0.8]
1.13 [0.7]
0.93 [1]
/
0.86 [0.7]
0.29 [0.6]
0.67 [0.9]
0.73 [1]
ꢃ/
/
ꢃ/
/
ꢃ/
A comparison of the bHRS values indicates that the
iron-containing complexes have similar or lower re-
sponses than ruthenium-containing analogues, in con-
/
ꢃ
/
1.08 [1]
/
0.59 [1]
ꢃ
/
1.17 [0.5]
trast to the trend reported for donorꢁ
[11] and metallocenyl complexes [7ꢁ10] for which the
relative efficiency FeꢀRu was found. In the current
/
acceptor nitrile
/
Ag/AgCl reference electrode (FcH/FcHꢂ 0.56 V). See Ref. [46].
/
a
No oxidation process observed to switching potential of 1.2 V.
Ref. [15].
work, the iron acetylide complexes have absorption
bands closer to the second-harmonic wavelength of 532
nm than either the ruthenium or osmium analogues,
suggesting that the bHRS values for the iron complexes
contain a larger resonance contribution. The bHRS
values for the osmium acetylide complexes are in each
case greater than the values for the ruthenium-contain-
ing complexes. Absorption bands for the osmium
complexes are closer to the second harmonic than are
those of their ruthenium homologues. For the cyclo-
pentadienyl complexes, calculating static b values from
the experimental bHRS values using a two-level model
affords the same trend as for the experimental values,
b
nitro-centered reduction, respectively. Reversibility of
the metal-centered oxidation increases on replacing Fe
by Ru (proceeding from 1 to 2) and replacing co-ligand
CO by dppe (proceeding from 5 to 2); complexes 4 and 7
are the only examples in this system with reversible
metal-centered oxidation. Ease of oxidation increases on
replacing Ru by Os (on proceeding from 7 to 4) or Fe
(on proceeding from 2 to 1), consistent with the trend in
optical absorption maxima, for which lmax increases on
replacing Ru by Os or Fe.
except that b0 (1)B/b0 (2). The ruthenium complex has
the highest calculated static value for complexes 8ꢁ
/10.
3.4. Nonlinear optical studies
While this suggests some ambiguity, the two-level model
may have limited applicability with organometallic
complexes of this type (see above). Despite experimental
The results of hyper-Rayleigh scattering (HRS) ex-
periments are presented in Table 4, together with the
data we recently obtained for a series of related
uncertainties (9
cient data to suggest that bHRS values for alkynyl
complexes of this type follow the ordering: iron5
/10% in reported values) there is suffi-
complexes 8ꢁ10 containing 1,2-bis(methylphenylpho-
/
sphino)benzene (diph) ligands [46]. The experimental
/
Table 4
Linear optical and quadratic NLO response parameters a
lmax (nm) [o (104 Mꢃ1 cmꢃ1)]
bHRS (10ꢃ30 esu)
b0 (10ꢃ30 esu)
c
b
Compound
Fe(4-Cꢀ
Ru(4-Cꢀ
Fe(4-CꢀCC6H4NO2)(dppe)(h-C5H5) (1)
Ru(4-CꢀCC6H4NO2)(dppe)(h-C5H5) (2)
Os(4-CꢀCC6H4NO2)(dppe)(h-C5H5) (3)
(ꢃ 436-trans-Fe(4-Cꢀ
CC6H4NO2)Cl{(R,R)-diph}2 (8) d
(ꢃ 589-trans-Ru(4-Cꢀ
CC6H4NO2)Cl{(R,R)-diph}2 (9) d
(ꢃ 365-trans-Os(Cꢀ
CC6H4NO2)Cl{(R,R)-diph}2 (10) d
Ru(4-CꢀCC6H4NO2)(PPh3)2(h-C5H5) (7)
Os(4-CꢀCC6H4NO2)(PPh3)2(h-C5H5) (4)
/
CC6H4NO2)(CO)2(h-C5H5) (6)
370 [1.3]
364 [1.6]
498 [1.3]
447 [1.8]
461 [1.7]
543 [1.7]
467 [2.1]
490 [1.8]
460 [1.1]
474 [2.2]
49
58
22
27
/
CC6H4NO2)(CO)2(h-C5H5) (5)
/
665
664
929
440
528
620
468
1051
64
/
161
188
/
/
)
/
ꢃ
/
14
97
74
96
/)
/
/)
/
/
/
174
a
b
c
All compounds are optically transparent at the fundamental frequencies. All measurements in solvent thf.
HRS at 1064 nm; values910%, using p-nitroaniline (bꢀ21.4ꢄ
10ꢃ30 esu) as a reference.
Data corrected for resonance enhancement at 532 nm using the two-level model with b0ꢀb[1ꢃ
/
/
/
/
/
(2lmax/1064)2][1ꢃ(lmax/1064)2]; damping factors
/
not included.
d
Ref [46]; diphꢀ1,2-bis(methylphenylphosphino)benzene.
/

16
C.E. Powell et al. / Inorganica Chimica Acta 352 (2003) 9ꢁ18
/
Ru(CꢀCH)(PH3)2(h-C5H5) [47]; the present study ex-
/
tends this investigation to other Group 8 metals and a
strongly electron-withdrawing alkynyl ligand. The
square-pyramidal M(PH3)2(h-C5H5) fragment (in which
the h-C5H5 ligand occupies three facially disposed
coordination sites) and the planar 4-Cꢀ
/CC6H4NO2
Fig. 4. Molecular geometry used to calculate optical transitions.
ligand interact strongly via overlap of the singly
2
occupied 15a? orbital of the former (essentially dz in
ruthenium5osmium, contrary to previously reported
/
character) with the singly occupied 18a? orbital of the
latter (an sp-hybridized s-orbital). For the M(PH3)2(h-
C5H5) unit, 13a? and 11a?? (essentially dxy and dx2ꢃy2 in
character, respectively) are non-bonding, transforming
to 34a? and 18a?? in the alkynyl complex. Orbitals 14a?
and 10a?? of the ligated metal unit (comprised, princi-
pally, of dxz and dyz contributions) have p symmetry
trends for other organometallics.
Comparison of the bHRS values for 1, 2, 5, and 6
permits comment on the effect of varying co-ligands.
Replacement of the electron-donating diphosphine li-
gand with the relatively strongly electron-withdrawing
carbonyls results in a significant reduction of the
second-order NLO response. This is readily rationalized
as the amount of electron density available to the
donating metal center (and hence its donor strength)
being reduced on replacing diphosphine by two carbonyl
groups. Ligated metals are, therefore, efficient tunable
donors in alkynyl complexes.
with respect to the alkynylꢁmetal axis, and interact with
/
filled arylalkynyl p orbitals (3a??, 19a?, 6a??) in a
destabilizing manner and with vacant p orbitals (7a??,
9a??, 22a?, 10a??) in a stabilizing manner. The transitions
in the optical spectra summarized in Table 5 correspond
to 16a??0
/
17a?? (MLCT) and 15a??0
/
17a?? (phenyl0
/
phenyl*).
3.5. Calculations
TD-DFT was employed to investigate the optical
spectra, and the bonding between the acetylide ligand
and the metal fragment, in order to shed light on the
observed trend in optical nonlinearities. The model
4. Discussion
Correlating readily accessible spectroscopic or elec-
trochemical parameters with NLO properties would, if
successful, afford information about the NLO responses
of complexes without recourse to less readily available
NLO measurements. While connections between linear
optical properties and NLO properties have been
probed significantly [48], the present studies suggest
that relationships between other physical data (e.g.,
redox potentials and NMR resonances) and quadratic
hyperpolarizability should be made with caution. Garcia
and co-workers [11] established a correlation between
complexes M(4-Cꢀ
were used in calculations to approximate the complexes
M(4-CꢀCC6H4NO2)(dppe)(h-C5H5) (for which the
complete set MꢀFe, Ru, Os exists, and which were
/
CC6H4NO2)(PH3)2(h-C5H5) (Fig. 4)
/
/
examined experimentally in the present study). The
calculated optical transitions show good correlation
with the experimental results (Table 5); differences
between the calculated and observed MLCT and
phenyl0
/
phenyl* transitions are small, considering the
approximations used to generate the calculated spectra
(phosphine replacement, isolated molecules in the gas
phase c.f. experimental data obtained in solution). A
molecular orbital diagram of one example, namely Fe(4-
n(Nꢀ
/
C) and bHRS for the nitrile complex cations [M(Nꢀ
/
CC6H4-4-R)(dppe)(h-C5H5)]ꢂ (Mꢀ
/
Fe, Ru; RꢀPh,
/
NMe2, NO2, C6H4NO2), for which lower energy IR
bands correspond to larger optical nonlinearities. This
CꢀCC6H4NO2)(PH3)2(h-C5H5), is shown in Fig. 5. We
/
have previously described the electronic structure of
correspondence can be rationalized if n(Nꢀ
/
C) is indica-
Table 5
Observed and calculated optical transitions
nmax (cmꢃ1) (o, Mꢃ1 cmꢃ1) (Expt.) a
20 000 (13 000)
nmax (cmꢃ1) (f) (Calc.) b
Transition
MLCT
Fe(4-Cꢀ
Ru(4-Cꢀ
Os(4-CꢀCC6H4NO2)(PH3)2(h-C5H5)
/
CC6H4NO2)(PH3)2(h-C5H5)
18 000 (0.36)
28 700 (0.25)
31 490 (9800)
phenyl0
/
phenyl*
phenyl*
phenyl*
/
CC6H4NO2)(PH3)2(h-C5H5)
22 200 (18 000)
33 700 (8800)
19 000 (0.45)
32 400 (0.18)
MLCT
phenyl0
/
/
21 600 (17 000)
36 800 (10 400)
20 100 (0.49)
30 800 (0.18)
MLCT
phenyl0
/
a
Experimental results are taken from M(4-Cꢀ
/
CC6H4NO2)(dppe)(h-C5H5).
b
All other calculated transitions are at least two orders of magnitude lower in oscillator strength.

C.E. Powell et al. / Inorganica Chimica Acta 352 (2003) 9ꢁ
/
18
17
Fig. 5. Molecular orbital diagram of Fe(4-CꢀCC6H4NO2)(PH3)2(h-C5H5). Noninteracting orbitals have been removed for clarity.
/
tive of the extent of p-backbonding, and if p-back-
bonding is an important determinant of NLO merit. No
orbitals of osmium should be significantly higher in
energy than those of iron, resulting in a stronger
interaction with the alkynyl ligand p* orbitals, and
hence stronger p-backbonding. The present work con-
firms this ordering in the cyclopentadienyl system,
establishes the transition which is the most important
contributor to optical nonlinearity as clearly MLCT in
character, and reproduces the trend in absorption
maxima. The calculations suggest that p-backbonding
is important for NLO merit, but that correlations with
easily accessible spectroscopic or electrochemical para-
meters should be made with caution.
similar correspondence between n(Cꢀ
nonlinearity is seen with the present series of complexes.
It is probable, though, that n(CꢀC) is not a useful
/C) and quadratic
/
determinant of backbonding, because stretching vibra-
tions are frequently coupled. We have recently assessed
p-backbonding in the complexes trans-M(Cꢀ
CR)Cl(PH3)4 (MꢀFe, Ru, Os; RꢀH, Ph, 4-
C6H4NO2) using DFT, for which MꢁC bond energy
varies as FeBRuBOs when relativistic terms are
included [49]. This is not unexpected; the relevant dp
/
/
/
/
/
/

18
C.E. Powell et al. / Inorganica Chimica Acta 352 (2003) 9ꢁ18
/
[17] A. Eisenstadt, R. Tannenbaum, A. Efraty, J. Organomet. Chem.
221 (1981) 317.
5. Supplementary material
[18] R.B. King, L.W. Houk, K.H. Pannell, Inorg. Chem. 8 (1969)
1042.
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Centre, CCDC Nos. 168645 (5), 169737 (2).
Copies of this information may be obtained free of
charge from The Director, CCDC, 12 Union Road,
[19] G.S. Ashby, M.I. Bruce, I.B. Tompkins, R.C. Wallis, Aust. J.
Chem. 32 (1979) 1003.
[20] H. Fischer, F. Leroux, G. Roth, R. Stumpf, Organometallics 15
(1996) 3723.
[21] G. Jia, W.S. Ng, J. Yao, C.P. Lau, Y. Chen, Organometallics 15
(1996) 5039.
Cambridge, CB2 1E2, UK (fax: ꢂ
email: deposit@ccdc.cam.ac.uk,
www.ccdc.cam.ac.uk).
/
44-1223-336-033;
www: http://
[22] E.J. Baerends, D.E. Ellis, P. Ros, Chem. Phys. 2 (1973) 42.
[23] E.J. Baerends, P. Ros, Chem. Phys. 2 (1973) 52.
[24] E.J. Baerends, P. Ros, Int. J. Quantum Chem. S12 (1978) 169.
[25] S.H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 58 (1980) 1200.
[26] A.D.J. Becke, Chem. Phys. 84 (1986) 4524.
Acknowledgements
[27] J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R.
Pederson, D.J. Singh, C. Fiolhais, Phys. Rev. A 46 (1992) 6671.
[28] J.P. Perdew, Phys. Rev. B 33 (1986) 8822.
We thank the Australian Research Council (MGH),
the Fund for Scientific Research, Flanders (G.0338.98,
G.0407.98) (AP), the Belgian Government (IUAP-IV/
11) (AP) and the K.U. Leuven (GOA/2000/03) (AP) for
[29] E. van Lenthe, E.J. Baerends, J.G.J. Snijders, Chem. Phys. 99
(1993) 4597.
[30] L. Versluis, T. Ziegler, J. Chem. Phys. 88 (1988) 322.
[31] S.J.A. van Gisbergen, J.G. Snijders, E.J. Baerends, Comput. Phys.
Commun. 118 (1999) 119.
support of this work, and JohnsonꢁMatthey Technol-
/
ogy Centre (MGH) for the generous loan of ruthenium
and osmium salts. MGH holds an ARC Australian
Senior Research Fellowship.
[32] Z. Otwinowski, W. Minor (Eds.), Methods in Enzymology,
Academic Press, New York, 1997, p. 307.
[33] TEXSAN, Single Crystal Structure Analysis Software, Version 1.8,
Molecular Structure Corporation, The Woodlands, TX, 1996.
[34] P. Coppens (Ed.), Crystallographic Computing, Munksgaard,
Copenhagen, 1970, p. 255.
References
[35] S. Mackay, C.J. Gilmore, C. Edwards, M. Tremayne, N. Stuart,
K. Shankland, MAXUS Computer Program for the Solution,
Refinement of Crystal Structures, Nonius, Delft, MacScience,
Japan, The University of Glasgow, 2000.
[1] S.K. Hurst, N.T. Lucas, M.G. Humphrey, I. Asselberghs, A.
Persoons, Aust. J. Chem., 54 (2001) 447.
[36] A. Altomare, M. Cascarano, C. Giacovazzo, A. Guagliardi, M.C.
Burla, G. Polidori, M. Camalli, J. Appl. Cryst. 27 (1994) 425.
[37] P.T. Beurskens, G. Admiraal, W.P. Bosman, S. Garcia-Granda,
R.O. Gould, J.M.M. Smits, C. Smykalla, The DIRDIF Program
System. Technical Report of the Crystallographic Laboratory,
University of Nijmegen, Nijmegen, The Netherlands, 1994.
[38] I.R. Whittall, M.G. Humphrey, D.C.R. Hockless, B.W. Skelton,
A.H. White, Organometallics 14 (1995) 3970.
[2] I.R. Whittall, A.M. McDonagh, M.G. Humphrey, M. Samoc,
Adv. Organomet. Chem. 42 (1998) 291.
[3] I.R. Whittall, A.M. McDonagh, M.G. Humphrey, M. Samoc,
Adv. Organomet. Chem. 43 (1999) 349.
[4] N.J. Long, Angew. Chem., Int. Ed. Engl. 34 (1995) 21.
[5] T. Verbiest, S. Houbrechts, M. Kauranen, K. Clays, A. Persoons,
J. Mater. Chem. 7 (1997) 2175.
[6] S.R. Marder, in: D.W. Bruce, D. O’Hare (Eds.), Inorganic
Materials, Wiley, New York, 1992.
[39] I.R. Whittall, M.G. Humphrey, M. Samoc, J. Swiatkiewicz, B.
Luther-Davies, Organometallics 14 (1995) 5493.
[7] S.R. Marder, B.G. Tiemann, J.W. Perry, L.T. Cheng, W. Tam,
W.P. Schaefer, R.E. Marsh, in: S.R. Marder, G.D. Stucky, J.E.
Simon, Materials for Nonlinear Optics, ACS Symp. Ser.455, 1991
ACS, New York, p. 187.
[40] R.H. Naulty, M.P. Cifuentes, M.G. Humphrey, S. Houbrechts, C.
Boutton, A. Persoons, G.A. Heath, D.C.R. Hockless, B. Luther-
Davies, M. Samoc, J. Chem. Soc., Dalton Trans. (1997) 4167.
[41] A.M. McDonagh, M.P. Cifuentes, N.T. Lucas, M.G. Humphrey,
S. Houbrechts, A. Persoons, J. Organomet. Chem. 605 (2000) 193.
[42] A.M. McDonagh, N.T. Lucas, M.P. Cifuentes, M.G. Humphrey,
S. Houbrechts, A. Persoons, J. Organomet. Chem. 605 (2000) 184.
[43] S. Hurst, N.T. Lucas, M.P. Cifuentes, M.G. Humphrey, M.
Samoc, B. Luther-Davies, I. Asselberghs, R. Van Boxel, A.
Persoons, J. Organomet. Chem. 633 (2001) 114.
[8] J.C. Calabrese, L.-T. Cheng, J.C. Green, S.R. Marder, W. Tam, J.
Am. Chem. Soc. 113 (1991) 7227.
[9] L.T. Cheng, W. Tam, G.R. Meredith, S.R. Marder, Mol. Cryst.
Liq. Cryst. 189 (1990) 137.
[10] S.R. Marder, D.N. Beratan, B.G. Tiemann, L.T. Cheng, W. Tam,
Org.Mater.NonlinearOpt.,Spec.Publ.R.Soc.Chem.91(1991)165.
[11] W. Wenseleers, A.W. Gerbrandij, E. Goovaerts, M.H. Garcia,
M.P. Robalo, P.J. Mendes, J.C. Rodrigues, A.R. Dias, J. Mater.
Chem. 8 (1998) 925.
[44] M.I. Bruce, R.C. Wallis, Aust. J. Chem. 32 (1979) 1471.
[45] M.P. Gamasa, J. Gimeno, E. Lastra, M. Lanfranchi, A.
Tiripicchio, J. Organomet. Chem. 405 (1991) 333.
[12] I.R. Whittall, M.G. Humphrey, S. Houbrechts, A. Persoons,
D.C.R. Hockless, Organometallics 15 (1996) 5738.
[13] I.R. Whittall, M.P. Cifuentes, M.G. Humphrey, B. Luther-
Davies, M. Samoc, S. Houbrechts, A. Persoons, G.A. Heath, D.
Bogsanyi, Organometallics 16 (1997) 2631.
[46] A.M. McDonagh, M.P. Cifuentes, M.G. Humphrey, S. Hou-
brechts, A. Persoons, J. Organomet. Chem. 610 (2000) 71.
[47] J.E. McGrady, T. Lovell, R. Stranger, M.G. Humphrey, Orga-
nometallics 16 (1997) 4004.
[48] D.R. Kanis, M.A. Ratner, T.J. Marks, Chem. Rev. 94 (1994) 195.
[49] C.D. Delfs, R. Stranger, M.G. Humphrey, A.M. McDonagh, J.
Organomet. Chem. 607 (2000) 208.
[14] I.R. Whittall, M.G. Humphrey, A. Persoons, S. Houbrechts,
Organometallics 15 (1996) 1935.
[15] I.R. Whittall, M.P. Cifuentes, M.G. Humphrey, B. Luther-
Davies, M. Samoc, S. Houbrechts, A. Persoons, G.A. Heath,
D.C.R. Hockless, J. Organomet. Chem. 549 (1997) 127.
[16] S. Takahashi, Y. Kuroyama, K. Sonogashira, N. Hagihara,
Synthesis (1980) 627.
[50] M. Stahelin, D.M. Burland, J.E. Rice, Chem. Phys. Lett. 191
¨
(1992) 245.
[51] M.I. Bruce, M.G. Humphrey, M.R. Snow, E.R.T. Tiekink, J.
Organomet. Chem. 314 (1986) 213.