Synthesis and nonlinear optical properties of η5-monocyclopentadienyliron(II) acetylide derivatives. X-ray crystal structures

Article abstract of DOI:10.1021/om0104619

A series of new acetylide complexes of the type [Fe(η⁵-C₅H₅)(P down curve sign P)(p-C≡CC₆H₄R)] (P down curve sign P = DPPE (=l,2-bis(diphenylphosphino)ethane), (P)-PROPHOS (=R)-(+)-l,2-bis(diphenylphosphino)propane), R = NO₂, C₆H₄NO₂, (Z)-C(H)=C(H)C₆H₄NO₂, (E)-C(H)=C(H)C₆H₄NO₂) have been synthesized by halide abstraction from the precursors [Fe(η⁵-C₅H₅)(P down curve sign P)(I)] and fully characterized. Quadratic hyperpolarizabilities (β) for the complexes have been determined by hyper-Rayleigh scattering at 1064 nm. The influence on the nonlinear response of acetylide chain length in proceeding from 4-nitrophenylethynyl to 4-nitrobiphenylethynyl and 4-nitro-(E)-stibenylethynyl has been studied, revealing values of the first hyperpolarizabilities among the highest reported for organometallic molecular materials. Comparisons on the nonlinear efficiencies are drawn with the related well-known families of compounds [Ru(η⁵-C₅H₅)-(PR₃)₂(p -C≡C-(aryl)-NO₂)] and [Fe(η⁵-C₅H₅)(P down curve sign P)(p-N≡C-(aryl)-NO₂)]⁺. Cubic hyperpolarizabilities (γ) determined by Z-scan at 800 nm are consistent with an increase in γupon replacing Ru by the more easily oxidizable Fe and upon chain-lengthening the delocalizable π-bridging unit (proceeding from 4-C₆H₄ to (E)-4,4′-C₆H₄CH=CHC₆H₄). X-ray crystallographic structures of complexes [Fe(η⁵-C₅H₅(DPPE)(p-C≡CC₆H ₄NO₂)] and [Fe(η⁵C₅H₅)(DPPE)((E)-p-C≡CC₆ H₄C(H)=C(H)C₆H₄NO₂)] were studied in order to investigate the existence of π-back-donation suggested by spectroscopic and electrochemical data. Crystal packing was analyzed with the aim of assessing the alignment of the molecules in the lattice and hence suggesting the magnitude of NLO properties at the macroscopic level.

Full text of DOI:10.1021/om0104619

Organometallics 2002, 21, 2107-2118
2107
Syn th esis a n d Non lin ea r Op tica l P r op er ties of
η5-Mon ocyclop en ta d ien ylir on (II) Acetylid e Der iva tives.
X-r a y Cr ysta l Str u ctu r es of
[F e(η5-C5H5)(DP P E)(p-CtCC6H4NO2)] a n d
[F e(η5-C5H5)(DP P E)((E)-p-CtCC6H4C(H)dC(H)C6H4NO2)]
M. Helena Garcia,*,†,‡ M. Paula Robalo,†,§ Alberto R. Dias,† M. Teresa Duarte,†
Wim Wenseleers, G. Aerts, Etienne Goovaerts, Marie P. Cifuentes,|
Steph Hurst,| Mark G. Humphrey,| Marek Samoc,# and Barry Luther-Davies#
Centro de Quı´mica Estrutural, Instituto Superior Te´cnico, Av. Rovisco Pais,
1049-001 Lisboa, Portugal, Departamento de Quı´mica e Bioquı´mica, Faculdade de Cieˆncias da
Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal, Departamento de
Quı´mica, Universidade de EÄ vora, Cole´gio Luı´s Anto´nio Verney, Rua Roma˜o Ramalho no. 59,
7000-671 EÄ vora, Portugal, Physics Department, University of Antwerp (UIA),
Universiteitsplein 1, 2610 Wilrijk-Antwerpen, Belgium, Department of Chemistry,
Australian National University, Canberra, ACT 0200, Australia, and Australian Photonics
Cooperative Research Centre, Laser Physics Centre, Research School of Physical Sciences and
Engineering, Australian National University, Canberra, ACT 0200, Australia
Received J une 1, 2001
A series of new acetylide complexes of the type [Fe(η5-C5H5)(PkP)(p-CtCC6H4R)] (PkP )
DPPE ()1,2-bis(diphenylphosphino)ethane), (R)-PROPHOS ()(R)-(+)-1,2-bis(diphenylphos-
phino)propane), R ) NO2, C6H4NO2, (Z)-C(H)dC(H)C6H4NO2, (E)-C(H)dC(H)C6H4NO2) have
been synthesized by halide abstraction from the precursors [Fe(η5-C5H5)(PkP)(I)] and fully
characterized. Quadratic hyperpolarizabilities (â) for the complexes have been determined
by hyper-Rayleigh scattering at 1064 nm. The influence on the nonlinear response of acetylide
chain length in proceeding from 4-nitrophenylethynyl to 4-nitrobiphenylethynyl and 4-nitro-
(E)-stibenylethynyl has been studied, revealing values of the first hyperpolarizabilities among
the highest reported for organometallic molecular materials. Comparisons on the nonlinear
efficiencies are drawn with the related well-known families of compounds [Ru(η5-C5H5)-
(PR3)2(p-CtC-(aryl)-NO2)] and [Fe(η5-C5H5)(PkP)(p-NtC-(aryl)-NO2)]+. Cubic hyper-
polarizabilities (γ) determined by Z-scan at 800 nm are consistent with an increase in γ
upon replacing Ru by the more easily oxidizable Fe and upon chain-lengthening the
delocalizable π-bridging unit (proceeding from 4-C6H4 to (E)-4,4′-C6H4CHdCHC6H4). X-ray
crystallographic structures of complexes [Fe(η5-C5H5)(DPPE)(p-CtCC6H4NO2)] and [Fe(η5-
C5H5)(DPPE)((E)-p-CtCC6H4C(H)dC(H)C6H4NO2)] were studied in order to investigate the
existence of π-back-donation suggested by spectroscopic and electrochemical data. Crystal
packing was analyzed with the aim of assessing the alignment of the molecules in the lattice
and hence suggesting the magnitude of NLO properties at the macroscopic level.
In tr od u ction
d-electron count, and ligand environment in tuning NLO
performance make organometallics very promising sys-
tems for nonlinear optics.
Over the past 15 years there has been a growing
interest in the design and synthesis of new organic/
organometallic molecules with large quadratic and cubic
molecular nonlinearities. Investigations of the nonlinear
optical (NLO) responses of materials have recently been
focused on organometallic complexes;1-3 a combination
of fast response time, low-lying intense metal-to-ligand
or ligand-to-metal charge transfer (MLCT or LMCT)
transitions, and the potential of variable oxidation state,
After the initial stimulating results achieved with
ferrocene derivatives, design criteria have been sug-
gested to further improve the NLO response, namely,
the incorporation of the metal into the plane of the
π-system of the chromophore and the possible introduc-
tion of metal-carbon multiple-bond character.4
(1) Goovaerts, E.; Wenseleers, W. E.; Garcia, M. H.; Cross, G. H.
Design and Characterisation of Organic and Organometallic Molecules
for Second-Order Nonlinear Optics. In Handbook of Advanced Elec-
tronic and Photonic Materials, Vol. 9, “Nonlinear Optical Materials”;
Nalwa, H. S., Ed.; Academic Press (ISBN 0125137451): San Diego,
2000; Chapter 3, pp 127-191.
(2) Whittall, I. R.; McDonagh, A. M.; Humphrey, M. G.; Samoc, M.
Adv. Organomet. Chem. 1998, 42, 291.
(3) Whittall, I. R.; McDonagh, A. M.; Humphrey, M. G.; Samoc, M.
Adv. Organomet. Chem. 1999, 43, 349.
* Corresponding author. E-mail: lena.garcia@ist.utl.pt (M.H.G.).
† Centro de Qu´ımica Estrutural.
‡ Universidade de Lisboa.
§ Universidade de EÄ vora.
University of Antwerp.
| Australian National University.
#
Australian Photonics Cooperative Research Centre.
10.1021/om0104619 CCC: $22.00 © 2002 American Chemical Society
Publication on Web 04/17/2002

2108 Organometallics, Vol. 21, No. 10, 2002
Garcia et al.
F igu r e 1. Structural formulas of complexes 1-8.
Among the organometallic compounds presenting this
structural feature half-sandwich ruthenium-acetylide
and half-sandwich iron/ruthenium-nitrile complexes
revealed significant second-order and third-order non-
linearities.5-11 Although the existence of M-C or M-N
multiple bonding is uncertain for both acetylide and
nitrile families of compounds, spectroscopic evidence
was found supporting π-back-donation and therefore
some metal ligand double-bond character. In our early
studies, carried out by electric-field-induced second-
harmonic generation (EFISHG), hyper-Rayleigh scat-
tering (HRS), Z-scan, and the Kurtz powder technique5-15
we have been probing how structural modification of
metal acetylide/nitrile complexes modifies the optical
nonlinearities in a systematic fashion.5,10,12,13,15,16 The
electron-rich iron half-sandwich moieties tuned by phos-
phine co-ligands were shown to behave as stronger
donor groups than the ruthenium analogues, when
incorporated in this type of donor-NC(π-conjugated
system)-acceptor structure,15,17 the dppe-containing
fragment being the best donor group.17 On the other
hand compounds derived from ruthenium half-sandwich
moieties possessing acetylide ligands revealed better
NLO properties than the nitrile analogues, possibly due
to more favorable metal-carbon back-bonding. These
results suggest that combination of acetylide ligands
with the best metallic donor fragment [Fe(η5-C5H5)-
(dppe)]+ would maximize the nonlinear response for this
half-sandwich transition metal family of compounds.
We report herein the syntheses of such complexes,
structural characterization, and electrochemical studies,
together with correlations of their spectroscopic
responses and solvatochromic behavior. X-ray struc-
tural studies for compounds [Fe(η5-C5H5)(DPPE)(p-
CtCC6H4NO2)] (1) and [Fe(η5-C5H5)(DPPE)((E)-p-
CtCC6H4C(H)dC(H)C6H4NO2)] (4) were also performed.
Measurements of the molecular quadratic hyperpolar-
izabilities for a selection of the systematically varied (4-
nitroaryl)acetylides were carried out by HRS at the
fundamental wavelength of 1064 nm, and cubic optical
nonlinearities were evaluated by the Z-scan technique
at the fundamental wavelength of 800 nm.
NO2, (E)-C(H)dC(H)C6H4NO2) (1-8) were prepared in
good yields (45-83%) by extension of literature methods
or modifications thereof.5,18 Reaction of [Fe(η5-C5H5)-
(PkP)(I)] (PkP ) DPPE, (R)-PROPHOS) with 4-eth-
ynylnitrobenzene in refluxing methanol, in the presence
of NH4PF6, afforded iron vinylidene complexes, which
are deprotonated in situ using a sodium methoxide
solution to give the σ-acetylide products. The compounds
are fairly stable toward oxidation in air and to moisture
both in the solid state and in solution. Complexes 1-8
1
(Figure 1) were characterized by IR, H, 13C, and 31P
NMR spectroscopies, and satisfactory microanalyses.
X-ray structural characterizations of 1 and 4 confirmed
their identities.
For compounds 1-8, a characteristic ν(CtC) band in
the IR spectrum was found in the range 2040-2060
cm-1. A significant low-energy shift was observed in
ν(CtC) upon coordination of the 4-nitrophenylethynyl
ligand (70 and 60 cm-1 for the DPPE and (R)-PROPHOS
derivatives, respectively), possibly explained by a de-
crease in CtC bond order as electron density is drawn
to the nitro group.
(4) Calabrese, J . C.; Cheng, L.-T.; Green, J . C.; Marder, S. R.; Tam,
W. J . Am. Chem. Soc. 1991, 113, 7227.
(5) Whittall, I. R.; Humphrey, M. G.; Hockless, D. C. R.; Skelton, B.
W.; White, A. H. Organometallics 1995, 14, 3970.
(6) Whittall, I. R.; Humphrey, M. G.; Persoons, A.; Houbrechts, S.
Organometallics 1996, 15, 1935.
(7) Whittall, I. R.; Humphrey, M. G.; Samoc, M.; Swiatkiewicz, J .;
Luther-Davies, B. Organometallics 1995, 14, 5493.
(8) Dias, A. R.; Garcia, M. Helena; Robalo, M. Paula; Green, M. L.
H.; Lai, K. K.; Pulham, A. J .; Kuebler, S. M.; Balavoine, G. J .
Organomet. Chem. 1993, 453, 241.
(9) Dias, A. R.; Garcia, M. H.; Rodrigues, J . C.; Green, M. L. H.;
Kuebler, S. M. J . Organomet. Chem. 1994, 475, 241.
(10) Dias, A. R.; Garcia, M. H.; Rodrigues, J . C.; Bjørnholm, T.;
Geisler, T.; Petersen, J . C. J . Mater. Chem. 1995, 5, 1961.
(11) Dias, A. R.; Garcia, M. H.; Mendes, P. J .; Robalo, M. P.;
Rodrigues, J . C. Trends Organomet. Chem. 1994, 1, 335.
(12) Whittall, I. R.; Cifuentes, M. P.; Humphrey, M. G.; Luther-
Davies, B.; Samoc, M.; Houbrechts, S.; Persoons, A.; Heath, G. A.;
Hockless, D. C. R. J . Organomet. Chem. 1997, 549, 127.
(13) Whittall, I. R.; Humphrey, M. G.; Houbrechts, S.; Persoons, A.;
Hockless, D. C. R. Organometallics 1996, 15, 5738.
(14) McDonagh, A. M.; Cifuentes, M. P.; Whittall, I. R.; Humphrey,
M. G.; Samoc, M.; Luther-Davies, B.; Hockless, D. C. R. J . Organomet.
Chem. 1996, 526, 99.
(15) Wenseleers, W.; Gerbrandij, A. W.; Goovaerts, E.; Garcia, M.
H.; Robalo, M. P.; Mendes, P. J .; Rodrigues, J . C.; Dias, A. R. J . Mater.
Chem. 1998, 8, 925.
(16) McDonagh, A. M.; Whittall, I. R.; Humphrey, M. G.; Skelton,
B. W.; White, A. H. J . Organomet. Chem. 1996, 519, 229.
(17) Garcia, M. H.; Robalo, M. P.; Dias, A.R.; Piedade, M. F. M.;
Galva˜o, A.; Wenseleers, W.; Goovaerts, E. J . Organomet. Chem. 2001,
619, 252.
(18) Houbrechts, S.; Clays, K.; Persoons, A.; Cadierno, V.; Gamasa,
M. P.; Gimeno, J . Organometallics 1996, 15, 5266.
Resu lts a n d Discu ssion
Syn th eses a n d Ch a r a cter iza tion of (Acetylid e)-
ir on Com p lexes. The new acetylide complexes, [Fe-
(η5-C5H5)(PkP)(p-CtCC6H4R)] (PkP ) DPPE, (R)-
PROPHOS, R ) NO2, C6H4NO2, (Z)-C(H)dC(H)C6H4-

η5-Monocyclopentadienyliron(II) Acetylides
Organometallics, Vol. 21, No. 10, 2002 2109
Ta ble 1. Op tica l Sp ectr a l Da ta for Com p lexes
[F e(η5-C5H5)(P kP )(p-CtCR)] in Ch lor ofor m
Solu tion (ca . 2.0 × 10-4 m ol d m -3
)
F igu r e 2. UV-visible spectra of complexes 1, 2, and 4 in
chloroform.
ca. 500 nm, which can be attributed to a MLCT (metal-
to-ligand charge transfer) transition, with ꢀ up to 1.45
× 104 M-1 cm-1 for the 4-nitro-(E)-stilbenylethynyl
derivative (4). In addition, we observe one band at
higher energy (with peak positions in the range 306-
360 nm) probably due to internal transitions occurring
at the acetylide chromophores and another band in the
UV region (λ ∼ 240 nm) characteristic of the cyclopenta-
dienyl and phosphine co-ligands. Figure 2 illustrates the
behavior of the [Fe(η5-C5H5)(DPPE)] derivatives with
the different acetylide chromophores.
Although the chain-lengthening of the chromophore
led to a bathochromic effect on the transition band
attributed to the coordinated chromophore (320 nm (1),
329 nm (2), 355 nm (3), and 360 nm (4)), as would be
expected, this effect was not observed on the MLCT
bands. In fact, the introduction of a second phenyl ring
proceeding from [Fe(η5-C5H5)(DPPE)(p-CtCC6H4NO2)]
(1) to [Fe(η5-C5H5)(DPPE)(p-CtCC6H4C6H4NO2)] (2)
leads to a blue shift; this effect was already reported
for the ruthenium(II) acetylide analogues16 and also for
the related iron(II)nitrile compounds15,17 and was ex-
plained on the basis of a breaking of the conjugation
through the second ring, due probably to a significant
torsion angle in the biphenyl linkage. Introduction of a
stilbene group proceeding from [Fe(η5-C5H5)(DPPE)(p-
CtCC6H4NO2)] (1) to [Fe(η5-C5H5)(DPPE)((Z)-p-
CtCC6H4C(H)dC(H)C6H4NO2)] (3) and [Fe(η5-C5H5)-
(DPPE)((E)-p-CtCC6H4C(H)dC(H)C6H4NO2)] (4) leads
to two different isomers (Z and E) and consequently to
different behaviors, since only the E-isomer (compound
4) retains the planarity of the coordinated chromophore.
As consequence, we observe also a blue shift for com-
pound 3, while compound 4 presents this MLCT band
at the same position of complex 1 with a clear enhance-
ment on the intensity of that band.
The H and 13C NMR spectra contain characteristic
resonances for the cyclopentadienyl group (in the range
4.15-4.29 ppm and 79.25-79.95 ppm). The effect on
NMR resonances of coordination of the acetylides to the
iron(II) fragments is mainly found in the shielding
observed at the ortho (relative to the CtC group)
protons of the benzene ring. This shielding was found
to vary from 0.82 ppm for [Fe(η5-C5H5)(DPPE)((Z)-p-
CdCC6H4C(H)dC(H)C6H4NO2)] (3) up to 1.2 ppm for
[Fe(η5-C5H5)(DPPE)(p-CtCC6H4-NO2)] (1). This effect
was also found for [Fe(η5-C5H5)(PkP)(p-NtCC6H4R)]-
[PF6] ((PkP) ) DPPE, (R)-PROPHOS, etc.; R ) NO2,
C6H4NO2, etc.)17 and might be explained by enhanced
π-back-donation between the metal orbitals and the
CtC acetylide group. The shifts in the resonances are
consistent with the possibility of some contribution of
a vinylidene form in solution:
1
UV/vis absorption spectra were also recorded for
compounds 1, 2, and 4 in other solvents of higher
polarity (acetone and DMF), in the range of wavelengths
200-800 nm, to examine their solvatochromic behavior
(Table 2).
For [Fe(η5-C5H5)(DPPE)(p-CtCC6H4NO2)] (1), a red
shift of the MLCT absorption band of ca. 25 nm occurs
with increasing solvent polarity, while the positions of
the other bands remain almost unchanged. This slight
positive solvatochromic behavior exhibited by this com-
pound is characteristic of ICT transitions with an
increase of the dipole moment upon photoexcitation.
Nevertheless, compound [Fe(η5-C5H5)(DPPE)((E)-p-
CtCC6H4C(H)dC(H)C6H4NO2)] (4), with the most in-
tense MLCT band (ꢀ ) 1.45 × 104 mol-1 dm3 cm-1), does
not reveal any significant solvatochromism in the polar-
ity range used.
[Fe]-CtC-Ar-R T [Fe]+dCdCdArdR-
The metal-bound acetylide R-carbon exhibits a triplet
resonance in the range 136.8-141.9 ppm, which shows
a phosphine dependence (4 ppm upfield shift on replac-
ing DPPE by (R)-PROPHOS), although it is insensitive
to acetylide modification. 31P{1H} NMR data of com-
plexes reported in ppm downfield from the external
standard (85% H3PO4) showed a deshielding for coor-
dinated phosphines, as would be expected.
The electronic spectra of complexes [Fe(η5-C5H5)-
(PkP)(p-CtCC6H4R)] (PkP ) DPPE, (R)-PROPHOS; R
) NO2, C6H4NO2, (Z)-C(H)dC(H)C6H4NO2, (E)-C(H)d
C(H)C6H4NO2) (1-8) were recorded as ca. 2.0 × 10-4
M solutions in chloroform (Table 1). The main feature
of these spectra is the presence of an intense band at

2110 Organometallics, Vol. 21, No. 10, 2002
Garcia et al.
F igu r e 3. Cyclic voltamogramm of [Fe(η5-C5H5)(DPPE)((E)-p-CtCC6H4C(H)dC(H)C6H4NO2)] in CH2Cl2 containing 0.1
M n-Bu4NPF6 (sweep rate ) 200 mV/s).
Ta ble 2. Solva toch r om ic Beh a vior of th e MLCT Ba n d of [F e(η5-C5H5)(DP P E)(p-CtCR)] Com p lexes
chloroform
acetone
ꢀ (M-1 cm-1
DMF
ꢀ (M-1 cm-1
)
solvent
λexp (nm)
ꢀ (M-1 cm-1
)
λexp (nm)
)
λexp (nm)
C6H4NO2 (1)
C6H4C6H4NO2 (2)
(E)-C6H4C(H)dC(H)C6H4NO2 (4)
504
479
506
0.80
0.50
1.45
510
475
495
1.15
0.59
1.11
527
488
507
0.81
1.18
0.51
Electr och em ica l Stu d ies. To obtain a deeper in-
sight into the electronic richness of the organometallic
moiety, the electrochemical behaviors of compounds 1-6
were studied by cyclic voltammetry in dichloromethane
between the limits imposed by the solvent, i.e. ca. -1.6
and 1.2 V. All measurements were carried out at a scan
rate of 200 mV s-1 in a solution of 0.1 M tetra-n-
butylammonium hexafluorophosphate as the support-
ing electrolyte.
As an example, the cyclic voltammetry response of
[Fe(η5-C5H5)(DPPE)((E)-p-CtCC6H4 C(H)dC(H)C6H4-
NO2)] (4) is shown in Figure 3, and the most relevant
parameters for the redox changes exhibited by all the
complexes are summarized in Table 3. The electrochemi-
cal behavior of the family of compounds [Fe(η5-C5H5)-
(PkP)(p-CtCC6H4R)] is characterized by one quasi-
reversible redox wave attributed to FeII f FeIII oxidation,
in the range 0.06-0.18 V, and an irreversible reduction
wave assigned to reduction of the nitro group, in the
range -1.10 to -1.25 V.
the acetylide ligand, and show a decrease of metal-
ligand back-donation as the chain-lengthening in-
creases.
On the basis of a HOMO-LUMO correlation with
the FeII/III couple and nitro reduction, respectively,
we also found that chain-lengthening leads to a de-
crease in the HOMO-LUMO gap. Therefore, we
expect that compound [Fe(η5-C5H5)(DPPE)((E)-p-
CtCC6H4C(H)dC(H)C6H4NO2)] (4) should give the
highest â value in this family of compounds, in agree-
ment with earlier studies.19,20
Qu a dr a tic Hyper p ola r iza bilities. Our earlier stud-
ies of cyclopentadienyl nitrile complexes,15,17 in which
the metal itself was systematically varied (Co, Ni, Ru,
and Fe) and combined with either electron donor or
acceptor substituents at the other end of a conjugated
ligand, have shown that the ruthenium and iron com-
plexes can form a very effective push-pull system in
combination with an electron-accepting para-substitu-
ent (some of these results are included in Table 4 for
comparison). This observation, both from quadratic
hyperpolarizability measurements by hyper-Rayleigh
scattering and from spectroscopic studies, led us to
conclude that the Ru organometallic moiety and even
more so the Fe organometallic moiety (e.g., [Fe(η5-C5H5)-
(DPPE)]) act as extremely efficient electron donor
groups and are suitable as building blocks for second-
order NLO molecules. Fe complexes were found to
possess about 3 times higher hyperpolarizabilities (â)
than analogous Ru complexes. This led to remarkably
The data are consistent with an FeII/III couple rela-
tively insensitive to the variation of the phosphine co-
ligand and susceptible to the variation of the substitu-
ents attached to the acetylide aromatic ring, i.e., affected
by the chain-lengthening of the acetylide ligand. Replac-
ing the acetylide ligand in proceeding from [Fe(η5-C5H5)-
(DPPE)(p-CtCC6H4NO2)] (1) to the other complexes
results in a decrease of the redox potential at the iron
center, with the most significant variation (0.12 V) being
with [Fe(η5-C5H5)(DPPE)((Z)-p-CtCC6H4C(H)dC(H)C6H4-
NO2)] (3). The trend observed for this family of iron(II)
acetylides is analogous to one observed for the related
ruthenium(II)acetylides.12 The results reveal the mag-
nitude of the metal-ligand back-donation, indicate that
the iron moiety imparts some electron donation through
(19) Whitaker, C. M.; Patterson, E. V.; Kott, K. L.; McMahon, R. J .
J . Am. Chem. Soc. 1996, 118, 9966.
(20) Dias, A. R.; Garcia, M. H.; Mendes, P.; Piedade, M. F. M.;
Duarte, M. T. L.; Calhorda, M. J .; Mealli, C.; Wenseleers, W.;
Gerbrandij, A. W.; Goovaerts, E. J . Organomet. Chem. 1998, 553, 115.

η5-Monocyclopentadienyliron(II) Acetylides
Organometallics, Vol. 21, No. 10, 2002 2111
Ta ble 3. Electr och em ica l Da ta for Com p lexes [F e(η5-C5H5)(P kP )(p-CtCC6H4R)] in CH2Cl2 Solu tion
high hyperpolarizabilities for molecules with only an
extremely short conjugated chain (e.g., [Fe(η5-C5H5)-
(DPPE)(p-NtCC6H4NO2)]+; â ) 395 × 10-30 esu17), or
even higher with a longer chain (e.g., [Fe(η5-C5H5)-
(DPPE)((E)-p-NtCCHdCHC6H4NO2)]+; â ) 570 × 10-30
esu17). Also, this comparison of a complex with a phenyl
ring as the conjugated chain with one with styrene sug-
gested that a further dramatic increase is to be expected
if an even longer conjugated chain would be used.
On the other hand, comparison of the results for the
Ru nitrile complexes (e.g., [Ru(η5-C5H5)(DPPE)(p-
NtCC6H4NO2)]+; â ) 126 × 10-30 esu)15 with analogous
Ru acetylide complexes (e.g., [Ru(η5-C5H5)(PPh3)2(p-
CtCC6H4NO2)]; â ) 468 × 10-30 esu12) reveals that the
acetylide linkage works much more efficiently than the
nitrile linkage. To check that these data, measured in
different experimental conditions and using different
calibration methods and standards, can be directly
compared, we also measured one of the RuCtC com-
pounds in our setup (see Table 4). The values are in
reasonable agreement, especially considering that the
measurements were performed in different solvents (we
use chloroform for internal reference). The much higher
hyperpolarizability of the Ru-CtC compounds com-
pared to the Ru-NtC complexes cannot be attributed
to the phosphine variation, as earlier studies have
demonstrated that varying the phosphine co-ligand only
marginally affects the hyperpolarizability.17
than on the nitrogen atom, while those of the CtC
ligand are distributed more symmetrically over both
carbon atoms. Therefore, a better overlap of the anti-
bonding ligand orbitals with the metal d-orbitals is
expected in the case of the CtC ligands, resulting in a
stronger back-donation and thus in a more efficient
coupling between the metal and the conjugated ligand.
We therefore expected to obtain the highest hyper-
polarizabilities by combining the most effective donor
group Fe(η5-C5H5)(DPPE) with the most effective coor-
dinating link CtC. Our results obtained for Fe-CtC
complexes are given in Table 4 and compared with
related Ru and Fe complexes. Indeed, extremely high
hyperpolarizabilities are obtained for the new com-
pounds. The value of â ) 1160 × 10-30 esu for [Fe(η5-
C5H5)(DPPE)(p-CtCC6H4NO2)] (1) is highly unusual for
a molecule containing only a single phenyl ring in the
conjugated backbone, and the value of 2315 × 10-30 esu,
comparable in magnitude to the best organic chro-
mophores, is one of the highest values for an organo-
metallic compound.1,2,21,22 (Several measurements of â
values in the same order of magnitude have later been
reported to be erroneous due to fluorescence contribu-
tions (e.g., in ref 22).) Nevertheless, even though the
hyperpolarizability of [Fe(η5-C5H5)(DPPE)(p-CtCC6H4-
NO2)] (1) is dramatically increased (2.5 times) compared
to [Ru(η5-C5H5)(PPh3)2(p-CtCC6H4NO2)], this increase
upon replacement of Ru by Fe is not as large as in the
The increase of the hyperpolarizabilities when replac-
ing the NtC linkage with a CtC linkage can be
understood in terms of π-back-bonding, which is known
to occur in these compounds.17,20 Indeed, the antibond-
ing orbitals of the NtC ligand, involved in the π-back-
bonding, are localized more on the nitrile carbon atom
(21) For a recent report of an organometallic with a large nonlin-
earity, see: McDonagh, A. M.; Cifuentes, M. P.; Humphrey, M. G.;
Houbrechts, S.; Maes, J .; Persoons, A.; Samoc, M.; Luther-Davies, B.
J . Organomet. Chem. 2000, 610, 71.
(22) (a) Schma¨lzin, E.; Meerholz, K.; Stadler, S.; Bra¨uchle, Ch.;
Patzelt, H.; Oesterhelt, D. Chem Phys. Lett. 1997, 280, 551. (b) Pauley,
M. A.; Wang, C. H. Chem. Phys. Lett. 1997, 280, 554.

2112 Organometallics, Vol. 21, No. 10, 2002
Garcia et al.
Ta ble 4. Secon d -Or d er NLO P ola r iza bilities, â, of a Selection of th e F e(II) Acetylid e Com p ou n d s Mea su r ed
a t 1.064 µm by HRS
a
Corrected value, see ref 17; *In THF; †In methanol. Comparison is made with related ruthenium complexes and with analogous iron
and ruthenium nitrile compounds. All measurements are in chloroform solution except when otherwise indicated.
case of the nitrile complexes, and in the acetylide
complexes with longer (stilbene-based) conjugated
ligands, no significant further increase is observed upon
replacement of Ru by Fe. This apparent “saturation”
partly results from the red shift of the MLCT bands in
proceeding from an Ru-containing complex to an Fe-
containing homologue and the concomitant increase in
resonant enhancement of â, which is more important
for the shorter than for the longer ligands (see Table
4). Remarkably, there is hardly any shift of the MLCT
band with ligand substitution in our Fe-acetylide com-
pounds.
indicate that the biphenyl group has a smaller torsional
distortion in the CtC compared to the NtC compounds,
but even so the results do not correspond to expectations
for a doubling of the conjugation length.
Cu bic Hyp er p ola r iza bilities. The third-order opti-
cal nonlinearities of organometallics have been of
considerable interest.3 Recently, attention has turned
to metal acetylide complexes,7,12,14,23 as the presence of
the metal in the plane of the alkynyl ligand may result
in enhanced nonlinearities compared to those observed
with the extensively investigated ferrocenyl class of
compounds (in which the MLCT interaction is perpen-
For the Ru as well as the Fe compounds and for the
nitriles as well as the acetylides, the first hyperpolar-
izability increases strongly with a styrene segment
inserted instead of a phenyl ring (see Table 4), showing
that lengthening the conjugation path has indeed the
potential to improve the NLO properties. However, in
the nitrile compounds a marked decrease was found
going from phenyl to biphenyl, which we attributed to
breaking of the conjugation path by a nonplanar con-
figuration of the biphenyl ligand.15,17 This is less the
case in the acetylides (see Table 4), for which substitu-
tion of phenyl by biphenyl yields a small increase or a
steady value of â, for Ru and Fe, respectively. This may
(23) (a) Thompson, M. E.; Chiang, W.; Myers, L. K.; Langhoff, C.
Proc. SPIE-Int. Soc. Opt. Eng. 1991, 1497, 423. (b) Myers, L. K.; Ho,
D. M.; Thompson, M. E.; Langhoff, C. Polyhedron 1995, 14, 57. (c)
Myers, L. K.; Langhoff, C.; Thompson, M. E. J . Am. Chem. Soc. 1992,
114, 7560. (d) Whittall, I. R.; Humphrey, M. G.; Samoc, M.; Luther-
Davies, B. Angew. Chem., Int. Ed. Engl. 1997, 36, 370. (e) Whittall, I.
R.; Humphrey, M. G.; Cifuentes, M. P.; Samoc, M.; Luther-Davies, B.;
Persoons, A.; Houbrechts, S. Organometallics 1997, 16, 2631. (f)
Frazier, C. C.; Chauchard, E. A.; Cockerham, M. P.; Porter, P. L. Mater.
Res. Soc. Symp. Proc. 1988, 109, 323. (g) Porter, P. L.; Guha, S.; Kang,
K.; Frazier, C. C. Polymer 1991, 32, 1756. (h) Naulty, R. H.; Cifuentes,
M. P.; Humphrey, M. G.; Houbrechts, S.; Boutton, C.; Persoons, A.;
Heath, G. A.; Hockless, D. C. R.; Luther-Davies, B.; Samoc, M. J . Chem.
Soc., Dalton Trans 1997, 4167. (i) McDonagh, A. M.; Humphrey, M.
G.; Samoc, M.; Luther-Davies, B.; Houbrechts, S.; Wada, T.; Sasabe,
H.; Persoons, A. J . Am. Chem. Soc., in press.

η5-Monocyclopentadienyliron(II) Acetylides
Organometallics, Vol. 21, No. 10, 2002 2113
Ta ble 5. Lin ea r Op tica l a n d Cu bic Non lin ea r Op tica l Resp on se P a r a m eter s for Selected
(Cyclop en ta d ien yl)(p h osp h in e)ir on a n d -r u th en iu m Alk yn yl Com p lexesa
γ800 (10-36 esu)
λ (nm)
complex
[ꢀ (104 M-1 cm-1)]
real
imaginary
|γ|
ref
[Fe(CtC-4-C6H4NO2)(dppe)(η5-C5H5)]
[Fe(CtC-4-C6H4-(E)-CHdCH-4-C6H4NO2)-
(dppe)(η5-C5H5)]
497 [0.9]
499 [0.9]
-410 ( 200
580 ( 200
710 ( 280
this work
this work
-2200 ( 600
1200 ( 300
2500 ( 670
[Ru(CtC-4-C6H4NO2)(PPh3)2(η5-C5H5)]
[Ru(CtC-4-C6H4-(E)-CHdCH-4-C6H4NO2)-
(PPh3)2(η5-C5H5)]
460 [1.1]
476 [2.6]
-210 ( 50
-450 ( 100
e10
210 ( 60
210 ( 50
500 ( 120
7, 12
7, 12
a
All complexes are optically transparent at 800 nm. All measurements in THF solvent.
dicular to the cyclopentadienyl plane and its polarizable
substituents). We previously reported the cubic molec-
ular nonlinearities of (cyclopentadienyl)bis(triphenyl-
phosphine)ruthenium acetylide complexes9,12 and sub-
sequently examined the effect of ligated metal variation
for metals with differing d-electron counts (nickel, gold).
The series of (cyclopentadienyl){bis(diphenylphosphino)-
ethane}iron complexes from the present work provides
the opportunity to assess the impact of ligated metal
variation on cubic NLO merit for metals with the same
d-electron count, for which factors such as relative ease
of oxidation may be expected to be important.
Third-order nonlinearities of compounds 1 and 4 were
evaluated by the Z-scan technique;24 the results of these
measurements are given in Table 5, together with the
previously reported data for related (cyclopentadienyl)-
bis(triphenylphosphine)ruthenium acetylide complexes.
The real components of the hyperpolarizabilities for the
new complexes are negative, and the imaginary com-
ponents are significant, consistent with two-photon
absorption contributing to the observed responses; com-
ment on the effect of structural variation on NLO merit
must therefore be cautious, particularly with the sig-
nificant error margins.
Replacing the 4-nitrophenylethynyl ligand by the
4-nitro-(E)-stilbenylethynyl ligand (which has a longer
π-system) for the iron complexes results in no significant
variation in λmax or ꢀ, but a significant increase in |γ|; a
similar increase in molecular cubic hyperpolarizabilities
on this ligand replacement has been noted previously
with the ruthenium,7,12 gold,23d and nickel23e complexes.
Replacing Ru(PPh3)2 by Fe(DPPE) for both the 4-nitro-
phenylethynyl and 4-nitro-(E)-stilbenylethynyl com-
plexes results in a significant red-shift in λmax, a small
decrease in ꢀ, and a significant increase in |γ|. It is
important to try and deconvolute the contributions from
the two simultaneous molecular variations (Ru f Fe;
2PPh3 f DPPE). We have previously examined the
effect of replacing PPh3 by PMe3 for the ruthenium
complexes, for which a significant decrease in cubic NLO
merit was noted.7,12 There are two important factors
with the phosphine co-ligands that may influence the
magnitude of the NLO response, namely, their donor
strength (which should make the metal more electron
rich) and the presence of phenyl rings (which increases
the polarizable π-system). The donor strength increases
in the order PPh3 < 0.5 DPPE < PMe3, while the phenyl
ring count increases in the order PMe3 < 0.5 DPPE <
PPh3. The significant decrease noted on replacing PPh3
by PMe3 suggests that the number of polarizable phenyl
rings is more important than ligand donor strength for
cubic hyperpolarizability in complexes of this type. One
would therefore expect that the (conceptual) ligand
substitution of 2PPh3 by DPPE would decrease the cubic
hyperpolarizabilities, in the absence of other variations.
The significant increase in progressing from the ruthe-
nium complexes to the present series of iron complexes
is consistent with the metal variation more than com-
pensating for the ligand replacement and with the
replacement Ru f Fe providing a means of significantly
increasing cubic NLO merit in alkynyl complexes. Iron
complexes are in general more easily oxidized than their
ruthenium homologues, and this may be a relevant
factor in the significant increase in cubic nonlinearities
noted in the present work; further data are required to
confirm this, as the contribution of two-photon states
to the observed nonlinearities is not expected to be
uniform across this series.
Cr ysta llogr a p h ic Stu d ies of [F e(η5-C5H5)(DP P E)-
(p-CtCC6H4NO2)] (1) a n d [F e(η5-C5H5)(DP P E)((E)-
p-CtCC6H4C(H)dC(H)C6H4NO2)] (4). With the goal
of finding structural evidence for a vinylidine con-
tribution to this family of compounds, expected from
experimental data presented before, single-crystal
X-ray diffraction studies of [Fe(η5-C5H5)(DPPE)(p-
CtCC6H4NO2)] (1) and [Fe(η5-C5H5)(DPPE)((E)-p-
CtCC6H4C(H)dC(H)C6H4NO2)] (4) were carried out to
afford bond length and angle data about the metal-
acetylide linkage and the degree of planarity and bond
alternation of the π-system of the chromophores.
Crystals of [Fe(η5-C5H5)(DPPE)(p-CtCC6H4NO2)] (1)
and [Fe(η5-C5H5)(DPPE)((E)-p-CtCC6H4 C(H)dC(H)-
C6H4NO2)] (4) suitable for X-ray analysis were grown
by slow evaporation from methanol solutions. The
molecular structures of both compounds are shown in
Figures 4 and 5, along with the atom-numbering
scheme. Selected bond lengths and angles for both
complexes are given in Table 6.
The structures of both complexes have a pseudo-
octahedral three-legged piano stool arrangement of the
terminal acetylide and dppe ligands around the iron
atom (on the assumptation that the cyclopentadienyl
group takes up three coordination sites), with bite
angles of P1-Fe-P2 (85.15(8)° for (1) and 86.88(9)° for
(4)), characteristic of the chelated dppe complexes.
Since a large number of iron acetylide complexes have
been crystallographically characterized, important bond
length and angle data are collected in Table 7 in order
to allow some structural comparisons with data obtained
in this work. The Fe-Cp distances [2.060-2.088 Å
range (1); 2.072-2.094 Å range (4)] are entirely unex-
ceptional for an iron(II) compound of this type. The
phosphine structural data, i.e., Fe-P bond lengths
(24) Sheikh-bahae, M.; Said, A. A.; Wei, T.; Hagan, D. J .; van
Stryland, E. W. IEEE J . Quantum Electron. 1990, 26, 760.

2114 Organometallics, Vol. 21, No. 10, 2002
Garcia et al.
Ta ble 6. Selected Bon d Len gth s (Å) a n d An gles
(d eg) for [F e(η5-C5H5)(DP P E)(p-CtCC6H4NO2)] (1)
a n d [F e(η5-C5H5)(DP P E)((E)-
p-CtCC6H4C(H)dC(H)C6H4NO2)] (4)
j
1
4
Fe-C(11)
2.084(9)
2.060(9)
2.088(10)
2.086(10)
2.070(10)
2.158(2)
2.157(2)
1.837(9)
1.848(8)
1.508(11)
1.822(8)
1.819(9)
1.829(8)
1.822(8)
1.856(8)
1.216(10)
1.442(11)
1.399(12)
1.379(12)
1.352(12)
1.351(14)
1.383(13)
1.395(12)
1.450(13)
2.082(7)
2.094(8)
2.072(8)
2.090(8)
2.089(7)
2.159(2)
2.158(2)
1.849(6)
1.838(7)
1.495(9)
1.821(7)
1.827(7)
1.823(7)
1.829(8)
1.897(7)
1.202(8)
1.414(9)
1.387(9)
1.377(9)
1.376(10)
1.388(10)
1.396(9)
1.391(10)
Fe-C(12)
Fe-C(13)
Fe-C(14)
Fe-C(15)
Fe-P(1)
Fe-P(2)
P(1)-C(131)
P(2)-C(231)
C(131)-C(231)
P(1)-C(111)
P(1)-C(121)
P(2)-C(211)
P(2)-C(221)
Fe-C(1)
C(1)-C(2)
C(2)-C(3)
F igu r e 4. ORTEP diagram for[Fe(η5-C5H5)(DPPE)(p-
CdCC6H4NO2)] (1), with 40% thermal ellipsoids, showing
the labeling scheme.
C(3)-C(4)
C(4)-C(5)
C(5)-C(6)
C(6)-C(7)
C(7)-C(8)
C(3)-C(8)
C(6)-N(1)
C(6)-C(9)
1.508(11)
1.227(10)
1.494(11)
1.416(11)
1.386(10)
1.362(10)
1.360(10)
1.354(10)
1.370(12)
1.456(10)
1.171(10)
1.208(10)
178.0(7)
172.5(9)
85.1(2)
C(9)-C(10)
C(10)-C(11)
C(11)-C(12)
C(12)-C(13)
C(13)-C(14)
C(14)-C(15)
C(15)-C(16)
C(11)-C(16)
C(14)-N(1)
N(1)-O(2)
1.198(15)
1.240(14)
178.7(8)
173.2(9)
85.5(3)
N(1)-O(1)
Fe-C(1)-C(2)
C(1)-C(2)-C(3)
C(1)-Fe-P(1)
P(2)-Fe-P(1)
C(1)-Fe-P(2)
C(121)-P(1)-Fe
C(111)-P(1)-Fe
C(131)-P(1)-Fe
C(221)-P(2)-Fe
C(211)-P(2)-Fe
C(231)-P(2)-Fe
O(2)-N(1)-O(1)
F igu r e 5. ORTEP diagram for [Fe(η5-C5H5)(DPPE)((E)-
p-CdCC6H4CHdCHC6H4NO2)] (4), with 40% thermal el-
lipsoids, showing the labeling scheme.
85.15(8)
85.2(3)
86.88(9)
87.8(2)
120.5(3)
115.5(3)
109.1(3)
117.3(3)
120.9(3)
108.3(3)
124.6(13)
121.9(3)
118.2(2)
107.0(2)
116.7(3)
119.4(3)
110.1(3)
120.8(10)
[2.157 Å (1), 2.159 Å (4), and 2.158 Å (1, 4)] and P1-
Fe-P2 angles reflecting the presence of a five-mem-
bered chelate ring are similar to the Fe-P distances and
angles in the related iron(II) acetylide complexes (Table
7).
The Fe-C bond length in 1 (1.856(8) Å) is one of the
shortest Fe-C(1) interactions in Fe(η5-C5H5)(phosphine)
complexes containing acetylides. This shortening is
associated with a marginally longer CtC bond length
(1.21(1) Å) and seems to confirm the existence of metal
f acetylide back-donation, which is suggested by spec-
troscopic data. It is well established that an effective
transfer of electrons from the metal to the alkynyl ligand
increases the M-C bond order and decreases the C-C
bond order. This feature is also supported in this case,
by the almost linear geometry of the acetylide group
Fe-C(1)-C(2) and C(1)-C(2)-C(3) with the angles of
178.7(8)° and 173.2(9)°, respectively. However, distances
and angles within the benzene ring in the acetylide
ligand are consistent with the retention of the benzene
aromaticity. In particular, there is no obvious bond
length alternation that would be expected in case of an
appreciable quinoidal contribution, as was suggested by
the spectroscopic IR and NMR data obtained for the
complex. On the contrary, the Fe-C(1) bond length in
4 (1.897(7) Å), within the values found for other Fe(η5-
C5H5)(DPPE) complexes containing acetylides (Table 7),
suggests that metal to acetylide back-donation does not
contribute to describe the Fe-CtC bond in the solid
state, although this effect could also in this case be
supported by the almost linear geometry of the acetylide
ligand (Fe-C(1)-C(2) and C(1)-C(2)-C(3) angles of
178.0(7)° and 172.5(9)°, respectively).
While the focus of this work is on the molecular NLO
properties of the acetylide complexes, it was also of
interest to examine crystal packing as an indicator of
bulk material SHG, since a noncentrosymmetric ar-
rangement in the lattice is required in order to obtain
a nonzero bulk response. With this in mind, the cell
packing of compounds 1 and 4 has been investigated.
The crystal packing of 4 is centrosymmetric as expected,
due to the crystallization in the monoclinic space group
P21/a, which makes this compound unsuitable for
macroscopic second-order NLO purposes in this crystal-
line form. In contrast, compound 1 has a noncentro-

η5-Monocyclopentadienyliron(II) Acetylides
Organometallics, Vol. 21, No. 10, 2002 2115
Ta ble 7. Str u ctu r a l Da ta for F e(η5-C5H5) Der iva tives Con ta in in g Acetylid es Liga n d s
Fe-Cp
(Å)
Fe-P
(Å)
P1-Fe-P2
Fe-C
(Å)
CdC
(Å)
CC-C
(Å)
Fe-C-C
(deg)
compound
(deg)
ref
[FeCp(CO)2(CdCC6H5)]
2.090(7)
2.113(7)
1.920(6) 1.201(9) 1.444(9)
174.4(4) 25a
[FeCPkPh2PCH2PPh2)(CdCC6H5)]
Fe(Cl)(Me2PCH2CH2PMe2)(CdCC6H5)]
[Fe(PMe2CH2CH2PMe2)2(p-CdCC6H4CdCH)2]
{[FeCp*(DPPE)]2(CdC-CdC)}
2.070(7) 2.167(2)
2.086(8) 2.162(2)
2.213(2)
74.8(2)
1.900(7) 1.206(10) 1.438(10) 177.4(7) 25b
86.0(1)
86.2(1)
85.64
1.880(5) 1.216(8) 1.421(8)
1.933(3) 1.193(4) 1.438(4)
1.830(8) 1.236(9)
177.8(5) 25c
178.41 25d
167.0(6) 25e
2.216(2)
2.209(1)
2.216(1)
2.118
2.170
2.206(2)
2.238(2)
84.65(8)
86.70(2)
85.15(8)
86.88(9)
[FeCp(DPPE)(CdC-CpFeCp)]
2.094(12) 2.159(3)
2.122(12) 2.155(3)
2.060(9) 2.157(2)
2.088(10) 2.158(2)
2.072(8) 2.158(2)
1.910(11) 1.196(16) 1.455(17) 177.0(9) 25f
[FeCp(DPPE)(p-CdCC6H4NO2)]
1.856(8) 1.216(10) 1.442(11) 178.7(8) this work
[FeCp(DPPE)((E)-p-CdCC6H4-
C(H)dC(H)C6H4NO2)]
1.897(7) 1.202(8) 1.414(9)
178.0(7) this work
2.094(8) 2.159(2)
F igu r e 6. Crystal packing for [Fe(η5-C5H5)(DPPE)(p-CdCC6H4NO2)] (1). Hydrogen atoms (not involved in intermolecular
hydrogen bonds) have been omitted for clarity.
symmetric crystal packing in the orthorhombic space
group P212121.
along the a axis. In this way, an infinite bidimensional
hydrogen bond network in the ac plane (Figure 6) is
formed.
In an effort to search for other features that could be
relevant to the crystal packing of the compounds and
to their NLO properties, we have examined the inter-
molecular contacts. A common feature emerges from
inspection of the two crystal packings: the interactions
between the neutral molecules in the lattice are based
on several C-H‚‚‚O hydrogen bonds between the oxy-
gens of the terminal NO2 group and the C-H aromatic
systems of two different molecules. In Figure 6 we can
observe the hydrogen bond between the oxygen atom
(O1) of compound 1 and the phosphine aromatic hydro-
gen (H122) belonging to a neighboring molecule, along
the c crystallographic direction. The second hydrogen
interaction, between O2 and another phosphine aro-
matic hydrogen (H213) belonging to a neighbor molecule
related by a 21 symmetry axis, extends this 1D array
Con clu sion
Spectroscopic data and molecular quadratic and cubic
optical nonlinearities for the present series of complexes
have shown that coordination of acetylide ligands to the
[Fe(η5-Cp)PkP] fragment affords complexes that are
more NLO efficient than their ruthenium analogues and
significantly more efficient than the corresponding iron
and ruthenium nitrile derivatives. Chain-lengthening
by inclusion of one ene-linkage between phenyl rings
results in a value of the molecular first hyperpolariz-
ability for the iron acetylide derivative that is among
the highest reported for organometallic molecular ma-
terials.
(25) (a) Goddard, R.; Howard, J .; Woodward, P. J . Chem. Soc., Dalton
Trans. 1974, 2025. (b) Gamasa, M. P.; Gimeno, J .; Lastra, E.;
Lanfranchi, M.; Tipiricchio, A. J . Organomet. Chem. 1991, 405, 333.
(c) Field, L. D.; George, A. V., Hambley, T. W. Inorg. Chem. 1990, 29,
4565. (d) Field, L. D.; George, A. V.; Malouf, E. Y.; Slip, I. H. M.;
Hambley, T. W. Organometallics 1991, 10, 3842. (e) Le Narvor, N.;
Toupet, L.; Lapinte, C. J . Am. Chem. Soc. 1995, 117, 7129. (f) Sato,
M.; Hayashi, Y.; Kumakura, S.; Shimizu, N.; Katada, M.; Kawata, S.
Organometallics 1996, 15, 721.
Exp er im en ta l Section
Gen er a l Com m en ts. All experiments were carried out
under a nitrogen atmosphere by use of standard Schlenk
techniques. Solvents were dried according to the usual pub-
lished methods.26 Column chromatography was performed
using Merck aluminum oxide 90 active neutral (activity stage

2116 Organometallics, Vol. 21, No. 10, 2002
Garcia et al.
C6H5); 127.92 (C-meta, C6H5); 128.80 (C-para, C6H5); 129.18
(C-para, C6H5); 131.72 (C-ortho, C6H5); 133.6 (C-ortho, C6H5);
137.02 (d, J CP ) 23.5 Hz, C-ipso).
Syn th esis of [F e(η5-C5H5)(P kP )(p-CtCC6H4R)]. All the
complexes were prepared by the process described below. To
a suspension of [Fe(η5-C5H5)(PkP)(I)] (PkP ) DPPE or (R)-
PROLOPHOS) (1 mmol) and the appropriate acetylene p-HCt
CC6H4R (R) NO2, C6H4NO2, (Z)-C(H)dC(H)C6H4NO2, (E)-
C(H)dC(H)C6H4NO2) (1.1 mmol) in methanol (15 mL) was
added NH4PF6 (1.1 mmol) at room temperature with stirring.
The mixture was refluxed for 1-2 h and then allowed to cool.
A change was observed from dark violet to orange-red. A
solution of sodium methoxide in methanol (1.2 mmol, 0.1 M)
was added, the mixture was stirred, and then concentration
of the solvent under reduced pressure and filtration gave a
dark red-violet solid.
[F e(η5-C5H5)(DP P E)(p-CtCC6H4NO2)] (1). The product
was recrystallized from dichloromethane/methanol giving a
violet crystalline powder; 60% yield, mp 236-238 °C. Anal.
Calcd for C39H33FeNO2P2: C, 70.39; H, 5.00; N, 2.10. Found:
C, 70.37; H, 5.15; N, 2.05. IR (KBr): ν(CtC) 2040 cm-1, ν(NO2)
F igu r e 7. Numbering scheme for NMR spectral assign-
ments.
1500, 1320 cm-1 1H NMR (CDCl3): δ 4.29 (s, 5H, η5-C5H5);
.
6.40 (d, 2H, J HH ) 8.1 Hz, H4,H8); 7.78 (d, 2H, J HH ) 8.1 Hz,
I, 70-230 mesh ASTM). Thin-layer chromatography was
performed using Merck aluminum oxide 60254. Petroleum ether
refers to a fraction of boiling point range 60-80 °C. Solid state
IR spectra were recorded on a Perkin-Elmer 683 spectro-
photometer in KBr pellets; only significant bands are cited in
the text. 1H, 13C{1H}, and 31P{1H} NMR spectra were recorded
on a Varian Unity 300 spectrometer at probe temperature. The
electronic spectra and solvatochromic behavior of the acetylide
compounds were measured in CHCl3, (CH3)2CO, and di-
methylformamide (DMF) solutions of aproximately 2.0 × 10-4
M concentration in quartz cells using a Shimadzu 1202
spectrophotometer over the range 200-800 nm. Microanalyses
were performed in our laboratories using a Fisons Instruments
EA1108 system. Data acquisitions, integration, and handling
were performed using a PC with the software package Eager-
200 (Carlo Erba Instruments). Melting points were obtained
on a Reichert Thermovar melting point apparatus.
H5,H7). 13C{1H} NMR (CDCl3): δ 79.67 (η5-C5H5); 123.31
2
(C5,C7); 129.96 (C4,C8); 137.49 (C3); 141.39 (t, J CP ) 15.2
Hz, C1); 142.14 (C6); (C2).a 31P{1H} NMR (CDCl3): δ 101.65.
aNot located due to overlapping with other signals.
[Fe(η5-C5H5)(DP P E)(p-CtCC6H4C6H4NO2)] (2). The prod-
uct was recrystallized from dichloromethane/n-hexane, giving
a purple crystalline powder; 56% yield, mp 198 °C (dec). Anal.
Calcd for C45H37FeNO2P2: C, 72.88; H, 5.03; N, 1.89. Found:
C, 72.56; H, 4.84; N, 1.69. IR (KBr): ν(CtC) 2060 cm-1, ν(NO2)
1510, 1340 cm-1 1H NMR (CDCl3): δ 4.28 (s, 5H, η5-C5H5);
.
6.53 (d, 2H, J HH ) 8.1 Hz, H4,H8); 7.30 (d, 2H, J HH ) 8.1 Hz,
H5,H7); 7.60 (d, 2H, J HH ) 8.9 Hz, H10,H14); 8.20 (d, 2H, J HH
)
8.9 Hz, H11,H13). 13C{1H} NMR (CDCl3): δ 79.27 (η5-C5H5);
123.97 (C11,C13); 126.24, 126.70 (C5,C7,C10,C14); 130,97 (C3);
2
131.70 (C4,C8); 132.22 (C6); 141.88 (t, J CP ) 17.8 Hz, C1);
146.13 (C12); 147.62 (C9); (C2).a 31P{1H} NMR (CDCl3):
106.02. Not located due to overlapping with other signals.
δ
a
The 1H and 13C{1H} (chloroform-d) chemical shifts are
reported in parts per million downfield from internal Me4Si,
and the 31P{1H} NMR spectra are reported in parts per million
downfield from external 85% H3PO4. Spectral assignments
follow the numbering scheme shown in Figure 7.
[F e(η5-C5H 5)(DP P E )((Z)-p -CtCC6H 4C(H )dC(H )C6H 4-
NO2)] (3). The product was recrystallized from dichloro-
methane/petroleum ether giving a purple crystalline solid; 75%
yield, mp 139-141 °C. Anal. Calcd for C47H39FeNO2P2: C,
73.54; H, 5.12; N, 1.82. Found: C, 73.21; H, 5.17; N, 1.74. IR
DPPE, (R)-PROPHOS, and NH4PF6 were used as purchased
from Aldrich Chemical Co. The acetylenes 4-HCtCC6H4NO2,27
4,4′-HCtCC6H4C6H4NO2,12 (Z)-4,4′-HCtCC6H4C(H)dC(H)C6H4-
1
(KBr): ν(CtC) 2060 cm-1, ν(NO2) 1515, 1340 cm-1. H NMR
(CDCl3): δ 4.25 (s, 5H, η5-C5H5); 6.34 (d, 2H, J HH ) 8.1 Hz,
28
H4,H8); 6.35 (d, 1H, J HH ) 12.6 Hz, H15); 6.61 (d, 1H, J HH
)
NO2, and (E)-4,4′-HCtCC6H4C(H)dC(H)C6H4NO2 were pre-
pared by literature procedures. [Fe(η5-C5H5)(DPPE)(I)] and
[Fe(η5-C5H5)((R)-PROPHOS)(I)] were prepared following a
procedure described previously.17
12.3 Hz, H16); 6.79 (d, 2H, J HH ) 7.8 Hz, H5,H7); 7.35 (d, 2H,
J HH ) 8.4 Hz, H10,H14); 8.03 (d, 2H, J HH ) 9.0 Hz, H11,H13).
13C{1H} NMR (CDCl3): δ 79.35 (η5-C5H5); 123.41 (C11,C13);
125.93 (C15); 129.60 (C5,C7,C10,C14); 130.70 (C3); (C16);a
131.1 (C6); (C4,C8);a 141.5 (t, 2J CP ) 15.0 Hz, C1); 144.77 (C9);
1
The phosphines’ H and 13C{1H} NMR data are similar
for all the complexes containing the same phosphine. (R)-
a
146.16 (C12); (C2).b 31P{1H} NMR (CDCl3): δ 102.27. Signal
PROPHOS: 1H NMR (CDCl3): δ 1.05 and 1.07 (3H, dd, J HH
8.1 Hz, CH3); 1.91 (m, 1H, CH2), 2.55 (1H, br s, CH); 2.70 and
)
b
obscured by the aromatic carbons of phosphine. Not located
due to overlapping with other signals.
2.86 (1H, dm, CH2); 7.31-7.48 (18H, m, C6H5); 8.01 (2H, t,
[F e(η5-C5H 5)(DP P E )((E)-p -CtCC6H 4C(H )dC(H )C6H 4-
NO2)] (4). The crude product was isolated by thin-layer
chromatography (25% dichloromethane/75% n-hexane eluant)
and was recrystallized from dichloromethane/methanol, giving
a violet crystalline solid; 63% yield, mp 292 °C (dec). Anal.
Calcd for C47H39FeNO2P2: C, 73.54; H, 5.12; N, 1.82. Found:
C, 73.20; H, 5.27; N, 1.80. IR (KBr): ν(CtC) 2050 cm-1, ν(NO2)
C6H5). 13C{1H} NMR (CDCl3): δ 16.17 and 16.36 (dd, J CP
)
6.8 Hz, CH3); 32.58 and 33.03 (dd, J CP ) 10.2 Hz, CH2); 34.53
(m, CH); 128.14 (d, J CP ) 8.3 Hz, C6H5); 128.52 (C6H5); 128.86
(C6H5); 130.45 (d, J CP ) 8.3 Hz, C6H5); 131.54 (d, J CP ) 8.3
Hz, C6H5); 134.47 (d, J CP ) 9.6 Hz, C6H5); 135.77 (d, J CP ) 9.6
Hz, C6H5); 136.71 (d, J CP ) 32.1 Hz, C-ipso, C6H5).
DPPE: 1H NMR (CDCl3): δ 2.28 (m, 2H, CH2); 2.57 (m, 2H,
CH2); 7.23-7.41 (m, 16H, C6H5); 7.86 (m, 4H, C6H5). 13C{1H}
NMR (CDCl3): δ 28.40 (t, CH2, J CP ) 22.1 Hz); 127.56 (C-meta,
1500, 1335 cm-1
6.46 (d, 2H, J HH ) 8.4 Hz, H4,H8); 6.88 (d, 1H, J HH ) 16.2 Hz,
15); 7.09 (d, 1H, J HH ) 16.2 Hz, H16); 7.12 (d, 2H, J HH ) 8.1
.
1H NMR (CDCl3): δ 4.26 (s, 5H, η5-C5H5);
H
(26) Perrin, D. D.; Amarego, W. L. F.; Perrin, D. R. Purification of
Laboratory Techniques, 2nd ed.; Pergamon: New York, 1980.
(27) Takahashi, S.; Kuroyama, Y.; Sonogashira, K.; Hagihara, N.
Synthesis 1980, 627.
(28) Hockless, D. C. R.; Whittall, I. R.; Humphrey, M. G. Acta
Crystallogr. 1996, C52, 3222.
Hz, H5,H7); 7.50 (d, 2H, J HH ) 9.0 Hz, H10,H14); 8.15 (d, 2H,
J HH ) 9.0 Hz, H11,H13). 13C{1H} NMR (CDCl3): δ 79.25 (η5-
C5H5); 123.42 (C15); 124.08 (C11,C13); 126.28, 126.37
(C5,C7,C10,C14); 130.47 (C3); 130.58 (C16); 130.70 (C6);
133.61 (C4,C8); 141.94 (t, J CP ) 15.8 Hz, C1); 144.54 (C9);

η5-Monocyclopentadienyliron(II) Acetylides
Organometallics, Vol. 21, No. 10, 2002 2117
146.06 (C12); (C2).a 31P{1H} NMR (CDCl3): δ 107.28. aNot
located due to overlapping with other signals.
used as counter electrode and the working electrode was a 2
mm piece of Pt wire. Potentials are referenced to an aqueous
calomel electrode (SCE) containing a saturated solution of
potassium chloride. The reference electrode was calibrated
using a 1.0 × 10-3 M solution of ferrocene in acetonitrile
containing 0.10 M LiClO4, for which the ferricinium/ferrocene
potential was in agreement with the literature value.29 Solu-
tions studied were 1 mM in solute and 0.1 M in the supporting
electrolyte, tetra-n-butylammonium hexafluorophosphate (pur-
chased from Sigma Chemical, Co.).
[F e(η5-C5H5)((R)-P ROP HOS)(p-CtCC6H4NO2)] (5). The
crude product was purified from a neutral aluminum oxide
chromatography column by elution with 25% acetone/75%
petroleum ether, affording a dark red crystalline solid; 68%
yield, mp 88-90 °C. Anal. Calcd for C40H35FeNO2P2: C, 70.70;
H, 5.19; N, 2.06. Found: C, 70.30; H, 5.26; N, 1.94. IR (KBr):
ν(CtC) 2050 cm-1, ν(NO2) 1500, 1320 cm-1. 1H NMR (CDCl3):
δ 4.17 (s, 5H, η5-C5H5); 6.51 (d, 2H, J HH )7.8 Hz, H4,H8); 7.86
(d, 2H, J HH )8.1 Hz, H5,H7). 13C{1H} NMR (CDCl3): δ 79.95
(η5-C5H5); 123.39 (C5,C7); 130.20 (C4,C8); 136.84 (t, 2J CP ) 15.1
Hz, C1); 142.14 (C3); 142.64 (C6); (C2).a 31P{1H} NMR
(CDCl3): δ 104.01 (d, J CP ) 23.0 Hz); 124.69 (d, J CP ) 24.8
The dichloromethane (reagent grade material) was dried
over CaH2 and distilled under a nitrogen atmosphere just
before use. Solutions were degassed and kept under an argon
atmosphere during each experiment.
a
Hz). Not located due to overlapping with other signals.
â Mea su r em en ts. Hyper-Rayleigh scattering (HRS) meas-
urements were performed using 70 ps pulses from a Nd:YAG
regenerative amplifier (λ ) 1064 nm) operating at a repetition
rate of 2 kHz. A detailed description of the experimental
procedure has been given in refs 1 and 17. In this setup, any
erroneous signals due to two-photon luminescence are ad-
equately eliminated by systematically scanning a narrow
region around the second-harmonic wavelength using a mono-
chromator and to some extent also by using a nanosecond
electronic time gating. Hyperpolarizabilities are determined
from the HRS intensity by internal reference relative to
chloroform. The intensity of the unpolarized HRS light is
proportional to the concentration and to the sum of the
[F e(η5-C5H5)((R)-P ROP HOS)(p-CtCC6H4C6H4NO2)] (6).
The product was recrystallized from dichloromethane/petro-
leum ether, giving a purple powder; 60% yield, mp 125 °C (dec).
Anal. Calcd for C46H39FeNO2P2: C, 73.12; H, 5.20; N, 1.85.
Found: C, 73.52; H, 5.37; N, 2.09. IR (KBr): ν(CtC) 2060
cm-1, ν(NO2) 1510, 1345 cm-1 1H NMR (CDCl3): δ 4.15 (s,
.
5H, η5-C5H5); 6.67 (d, 2H, J HH ) 8.4 Hz, H4,H8); 7.56 (d, 2H,
J HH ) 8.6 Hz, H5,H7); 7.74 (d, 2H, J HH ) 8.4 Hz, H10,H14); 8.23
(d, 2H, J HH ) 8.7 Hz, H11,H13). 13C{1H} NMR (CDCl3): δ 79.51
(η5-C5H5); 124.01 (C11,C13); 126.32, 126.71 (C5,C7,C10,C14);
130.02 (C4,C8); 130.83 (C3); 132.44 (C6); 137.86 (t, 2J CP ) 18.7
Hz, C1); 146.13 (C9); 147.68 (C12); (C2).a 31P{1H} NMR
(CDCl3): δ 104.79 (d, J CP )25.3 Hz); 125.32 d, J CP )18.7 Hz).
aNot located due to overlapping with other signals.
orientational averages âZ2 ZZ
+
â2XZZ (upper case indices:
laboratory coordinates) of the molecular hyperpolarizability.
For simplicity we assume that the â tensors of both solvent
and solute molecules are dominated by a single (diagonal)
tensor component âzzz (lower case indices: molecular coordi-
nates). For the present NLO compounds, this is expected to
be a very good approximation, as previous measurements15 on
[F e (η5-C5H 5)((R )-P R OP H OS)((Z)-p -CtCC6H 4C(H )d
C(H)C6H4NO2)] (7). The product was recrystallized from
dichloromethane/petroleum ether, affording a violet solid; 45%
yield, mp 128 °C (dec). Anal. Calcd for C48H41FeNO2P2: C,
73.76; H, 5.29; N, 1.79. Found: C, 73.41; H, 5.44; N, 1.66. IR
similar complexes all showed depolarization ratios âX2 ZZ
/
1
(KBr): ν(CtC) 2060 cm-1, ν(NO2) 1520, 1340 cm-1. H NMR
(CDCl3): δ 4.15 (s, 5H, η5-C5H5); 6.11 (d, 2H, J HH ) 8.7 Hz,
â2ZZZ very close to the theoretically predicted value of 1/5,
which is also reproduced by calculations on related Fe(II)
model compounds.17 With this assumption, the total HRS
H4,H8); 6.51 (d, 1H, J HH ) 13.2 Hz, H15); 6.69 (d, 1H, J HH
)
12.8 Hz, H16); 6.73 (d, 2H, J HH ) 7.8 Hz, H5,H7);. 7.43 (d, 2H,
10,H14);a 8.07 (d, 2H, J HH ) 8.4 Hz, H11,H13). 13C{1H} NMR
intensity is proportional to â2HRS ) â2ZZZ + â2XZZ ) (6/35)âz2zz
.
H
(CDCl3): δ 79.46 (η5-C5H5); 123.45 (C11,C13); 125.64 (C15);
128.15 and 129.55 (C5,C7,C10,C14); 130.27 (C4,C8); 133.10
(C6); 134.41 (C16); 142.96 (C1); 144.97 (C9); 146.16 (C12);
(C2);a (C3).a 31P{1H} NMR (CDCl3): δ 87.01 (d, J CP ) 24.3 Hz);
107.60 (d, J CP ) 24.2 Hz). aSignal obscured by the aromatic
protons and carbons of phosphine.
In this way, the ratio âHRS (solute)/âHRS(CHCl3) obtained from
the experiments can be directly interpreted as the ratio of the
molecular zzz-components. The value âHRS(CHCl3) ) 0.49 ×
10-30 esu from EFISHG measurements30 is adopted for the
solvent. Typical concentrations in the range (1-5) × 10-5 mol/
L-1 were used in the case of the Fe-acetylide compounds.
Although for CHCl3 significant off-diagonal components may
be expected, additional assumptions on the tensor components
would be needed to improve significantly on this analysis.
Experimental errors on â values are estimated to be on the
order of 5%, not including the uncertainty on the CHCl3
reference value.
γ Mea su r em en ts. Measurements were performed at 800
nm using a system consisting of a Coherent Mira Ar-pumped
Ti-sapphire laser generating a mode-locked train of ap-
proximately 100 fs pulses and a Ti-sapphire regenerative
amplifier pumped with a frequency-doubled Q-switched pulsed
YAG laser (Spectra Physics GCR) at 30 Hz and employing
chirped pulse amplification. THF solutions were examined in
a glass cell with a 0.1 cm path length. The Z-scans were
recorded at two concentrations for each compound, and the
real and imaginary part of the nonlinear phase change
determined by numerical fitting. The real and imaginary parts
of the hyperpolarizability of the solute were then calculated
assuming linear concentration dependencies of the nonlinear
phase change. The nonlinearities and light intensities were
calibrated using Z-scan measurements of a 1 mm thick silica
plate for which the nonlinear refractive index n2 ) 3 × 10-16
cm2 W-1 was assumed.
[F e (η5-C5H 5)((R )-P R OP H OS)((E )-p -CtCC6H 4C(H )d
C(H)C6H4NO2)] (8). The crude product was isolated by thin-
layer chromatography (25% dichloromethane/75% n-hexane
eluant), affording a violet crystalline solid; 83% yield. Anal.
Calcd for C48H41FeNO2P2: C, 73.76; H, 5.29; N, 1.79. Found:
C, 73.37; H, 5.42; N, 1.64. IR (KBr): ν(CtC) 2060 cm-1, ν(NO2)
1510, 1340 cm-1
6.58 (d, 2H, J HH ) 8.1 Hz, H4,H8); 6.94 (d, 1H, J HH ) 15.6 Hz,
15); 7.16 (d, 1H, J HH ) 15.8 Hz, H16); 7.20 (d, 2H, J HH ) 8.1
.
1H NMR (CDCl3): δ 4.14 (s, 5H, η5-C5H5);
H
Hz, H5,H7); 7.55 (d, 2H, J HH ) 8.1 Hz, H10,H14); 8.18 (d, 2H,
J HH ) 8.1 Hz, H11,H13). 13C{1H} NMR (CDCl3): δ 79.51 (η5-
C5H5); 123.44 (C15); 124.12 (C11,C13); 126.32, 127.45
(C5,C7,C10,C14); 132.77 (C3); 130.60 (C16); (C6);a 133.69
2
(C4,C8); 140.92 (t, C1, J CP ) 15.6 Hz); 144.62 (C9); 146.10
(C12); (C2).b 31P{1H} NMR (CDCl3): δ 95.19 (d, J CP ) 25.8 Hz);
107.87 (d, J CP ) 24.2 Hz). aSignal obscured by the aromatic
carbons of phosphine. bNot located due to overlapping with
other signals.
Electr och em ica l Ap p a r a tu s. The electrochemical instru-
mentation for cyclic voltammetry consisted of a EG&G Prin-
ceton Applied Research Potentiostat Model 273A, connected
to the data acquisition software (EG&G PAR Electrochemical
Analysis Model 273 Version 3.0). The voltammetric experi-
ments were performed at room temperature, in an argon
atmosphere, in a standard single-compartment three-electrode
design (PAR polarographic cell). A platinum spiral wire was
(29) Nelson, I. V.; Iwamoto, R. T. Anal. Chem. 1990, 35, 867.
(30) Kajzar, F.; Ledoux, I.; Zyss, J . Phys. Rev. A 1987, 36, 2210.

2118 Organometallics, Vol. 21, No. 10, 2002
Garcia et al.
Ta ble 8. Cr ysta l Da ta a n d Deta ils of th e Str u ctu r e Deter m in a tion of Com p lexes
[F e(η5-C5H5)(DP P E)(p-CtCC6H4NO2)] (1) a n d [F e(η5-C5H5)(DP P E)((E)-p-CtCC6H4C(H)dC(H)C6H4NO2)] (4)
1
4
empirical formula
fw
C
39H33FeNO2P2
C47H39FeNO2P2
767.66
665.52
temperature
wavelength
293(2) K
1.54052 Å
293(2) K
0.71069 Å
cryst syst, space group
unit cell dimensions
orthorhombic, P212121
a ) 16.230(1) Å
monoclinic, P21/a
a ) 15.389(1) Å
b ) 24.053(1) Å
b ) 16.555(1) Å, â ) 108.26(1)°
c ) 15.923(1) Å
c ) 8.509(2) Å
volume
Z, calcd density
abs coeff
3321.8(5) Å3
3852.5(4) Å3
4, 1.331 Mg/m3
4, 1.324 Mg/m3
4.709 mm-1
0.465 mm-1
F(000)
1384
1600
θ range for data collection
3.28-59.90°
1.82-24.97°
index ranges
0 < h < 18, -27 < k < 27, 0 < l < 8
5397/4844 [R(int) ) 0.0739]
full-matrix least-squares on F2
4844/0/406
-18 < h < 17, -2 < k <19, -13 < l < 18
8594/6771 [R(int) ) 0.2120]
full-matrix least-squares on F2
6771/0/478
no. of reflns collected/unique
refinement method
no. of data/restraints/params
goodness-of-fit on F2
final R indices [I>2σ(I)]
R indices (all data)
absolute struct param
largest diff peak and hole
1.118
1.103
R1 ) 0.0688, wR2 ) 0.1455
R1 ) 0.1067, wR2 ) 0.1858
-0.056(11)
R1 ) 0.0803, wR2 ) 0.0873
R1 ) 0.2167, wR2 ) 0.1336
0.273 and -0.514 e Å-3
0.322 and -0.297 e Å-3
ORTEP33 and SCHAKAL.34 The atomic scattering factors and
anomalous scattering terms were taken from International
Tables.35
X-r a y Str u ctu r e Deter m in a tion s. Diffraction measure-
ments of compounds 1 and 4 were made on an Enraf-Nonius
TURBOCAD4 (using Cu rotating anode) and MACH3 (using
conventional Mo tube) diffractometers, respectively, at room
temperature. The unit cell dimensions and orientation matrix
were obtained by least-squares refinement of 25 centered
reflections with 13° < θ < 16° for 1 and with 15° < θ < 17° for
4. Data were collected in the range 3.28° < θ < 59.90° for 1
and 1.82° < θ < 24.97° for 4. As a general procedure, the
intensity of three standard reflections was measured periodi-
cally every 2 h. This procedure did not reveal any appreciable
decay. Using the CAD4 software, data were corrected for
Lorentz and polarization effects and empirically for absorption
(from Ψ-scan measurements). Details for data collection and
structure determination are included in Table 8.
For both compounds 1 and 4, the position of the Fe atom
was obtained by a tridimensional Patterson synthesis, while
all the other non-hydrogen atoms were located in subsequent
difference Fourier maps. All non-hydrogen atoms were refined
by full-matrix least squares on F2 with anisotropic thermal
motion parameters. The hydrogen atoms were inserted in
calculated positions and refined isotropically riding on the
parent carbon atom. The largest peak in the final difference
Fourier synthesis was 0.273 and 0.322 e Å-3, respectively.
Lists of observed and calculated structure factors, tables of
final atomic coordinates, anisotropic thermal parameters for
all non-hydrogen atoms, hydrogen atomic coordinates, bond
lengths and angles, and inter- and intramolecular contact
distances for both compounds are available from the authors
and have been deposited as Supporting Information. Structure
solution and refinement were done with SHELXS-8631 and
with SHELX97.32 The illustrations were drawn with programs
Ack n ow led gm en t. We are grateful to the J unta
Nacional de Investigac¸a˜o Cient´ıfica e Tecnolo´gica (JNICT)
for financial support (PRAXIS XXI/2/2.1/QUI/143/94 and
PRAXIS XXI/PCEX/C/QUI/96). The work in Antwerp is
partly funded by the Flanders Government in its action
for the promotion of participation in EU-research pro-
grams and also by the Fund for Scientific Research of
Flanders (FWO). E.G. is a Research Director of the
FWO. We thank the Australian Research Council for
financial support (M.G.H.), an ARC Australian Post-
doctoral Research Fellowship (M.P.C.), an ARC Aus-
tralian Research Fellowship (M.G.H.), and an ARC
Australian Senior Research Fellowship (M.G.H.).
A listing of crystal data and structural refinement,
atomic coordinates, anisotropic displacement param-
eters, and a complete list of bond lengths and angles
for both compounds is available free of charge via the
Internet at http://pubs.acs.org.
Su p p or tin g In for m a tion Ava ila ble: This material is
available free of charge via the Internet at http://pubs.acs.org.
OM0104619
(32) Scheldrick, G. M. SHELX97, Crystallographic Calculation
Program; University of Cambridge, 1997.
(33) J ohnson, C. K.; Burnett, M. N. ORTEPIII, Report ORNL-6895;
Oak Ridge National Laboratory: TN, 1996. McArdle, P. PC Windows
version. J . Appl. Crystallogr. 1995, 28, 65.
(34) SCHAKAL97: Keller, E. Graphical Representation of Molecular
Models, University of Freiburg: Germany, 1997.
(35) International Tables for X-ray Crystallography, Vol. IV; Kynoch
Press: Birmingham, UK, 1974.
(31) Scheldrick, G. M. In Crystallographic Computing 3; Scheldrick,
G. M., Kruger, C., Goddard, R., Eds.; Oxford University Press: Oxford,
1985; pp 175-189.