Reactions of Fe3(CO)12 and Fe(CO)5 with functionalized alkynes in CH3OH-KOH solution - The crystal structure of Fe2(CO)6(PPh3)[μ-η3- (H2CCCH2)]

Article abstract of DOI:10.1139/V05-252

The complex Fe₂(CO)₇(H₂CCCH₂) (1) is obtained starting from a variety of functionalized alkynes that can release C₃ fragments during the reactions with Fe₃(CO) ₁₂ or Fe(CO)₅ in CH₃OH-KOH solutions. Owing to the oily nature of 1, it was reacted with triphenylphosphine, and the structure of the resulting title compound (1a) has been determined by X-ray analysis. Complex 1a shows a new coordination mode of a H₂C=C=CH₂ ligand to a diiron centre and is, to our knowledge, unprecedented. We have also found that 1 (and 1a) are by-products, and not intermediates in the formation of complex 3. A discussion of the possible reaction pathways is given.

Full text of DOI:10.1139/V05-252

337
Reactions of Fe3(CO)12 and Fe(CO)5 with
functionalized alkynes in CH3OH–KOH solution —
The crystal structure of Fe2(CO)6(PPh3)[␮-␩3-
(H2CCCH2)]1
Giuliana Gervasio, Domenica Marabello, Enrico Sappa, and Andrea Secco
Abstract: The complex Fe2(CO)7(H2CCCH2) (1) is obtained starting from a variety of functionalized alkynes that can
release C3 fragments during the reactions with Fe3(CO)12 or Fe(CO)5 in CH3OH–KOH solutions. Owing to the oily na-
ture of 1, it was reacted with triphenylphosphine, and the structure of the resulting title compound (1a) has been deter-
mined by X-ray analysis. Complex 1a shows a new coordination mode of a H2C=C=CH2 ligand to a diiron centre and
is, to our knowledge, unprecedented. We have also found that 1 (and 1a) are by-products, and not intermediates in the
formation of complex 3. A discussion of the possible reaction pathways is given.
Key words: iron carbonyls, alkynols, CO and methanol activation, acetyl and acetate complexes, allenyl iron complexes,
crystal structure.
Résumé : Le complexe Fe2(CO)7(H2CCCH2) (1) a été préparé à partir d’une variété d’alcynes fonctionnalisés qui peu-
vent perdre des fragments en C3 au cours de réactions avec du Fe3(CO)12 ou du Fe(CO)5 dans des solutions de
CH3OH–KOH. En raison de la nature huileuse du composé 1, on l’a fait réagir avec de la triphénylphosphine et on a
fait appel à la diffraction des rayons X pour déterminer la structure cristalline du produit qui en résulte et qui est men-
tionné dans le titre (1a). Le complexe 1a présente un nouveau mode de coordination du ligand H2C=C=CH2 au centre
difer qui, à notre connaissance était inconnu jusqu’à maintenant. On a aussi trouvé que les composés 1 (et 1a) sont des
sous-produits et non pas des intermédiaires dans la formation du complexe 3. On présente une discussion des voies
réactionnelles potentielles.
Mots clés : fer carbonyles, alkynols, activation du CO et du méthanol, complexes acétyles et acétates, complexes
d’allényl fer, structure cristalline.
[Traduit par la Rédaction] Gervasio et al. 344
Introduction
surfaces (3), or silica and alumina slurries (4). Other
processes observed are deoxygenation (5) and loss of
functionalities (6).
Propargyl alcohols react with M3(CO)12 carbonyls (M =
Fe, Ru, or Os) both under thermal (hydrocarbons, reflux)
and basic conditions (CH3OH–KOH solution, followed by
acidification with HCl) undergoing different processes,
among which the more important is dehydration. Two main
dehydration pathways leading, respectively, to coordinated
ene-yne (or vinylacetylene) or to allenylidene ligands have
been evidenced (1). The latter is apparently preferred when
M = Fe, as discussed in the following. Most of the dehydra-
tion–hydration reactions are promoted by acids (2), silica
We have recently reported the reactions of Fe3(CO)12 with
propargyl alcohol (HCϵCCH2OH) and propargyl chloride
(HCϵCCH2Cl), both under thermal and basic methanolic
conditions. The products obtained were: Fe2(CO)7(C3H4)
(1), the allenylidenic Fe3(CO)9(µ-CO)(C=C=CH2) (2) in very
small yields, Fe2(CO)6[H2CCC(H)C(=O)OCH3] (3), and
Fe3(CO)10[H2CCC(H)C(=O)C{CH2(O)CH3}CCH2] (4) (7).
The structures of complexes 2–4 are shown in Scheme 1.
On the basis of previous results (8), the formation of com-
Received 31 May 2005. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 25 March 2006.
Dr. Carty is now at the top of a brilliant career. This paper is dedicated to the chemist who agreed to collaborate with us, the
teacher who told me how to decently write a paper in English, and to the friend with whom we shared nice ski and walking runs
in the Alps. In a word: Arthur Carty.
G. Gervasio and D. Marabello. Dipartimento di Chimica Inorganica, Chimica Fisica e Chimica dei Materiali, Universita’ di
Torino, Via Pietro Giuria 7, I-10125 Torino, Italy.
E. Sappa2 and A. Secco. Dipartimento di Scienze e Tecnologie Avanzate, Universita’ del Piemonte Orientale, Via Bellini 25/g,
I-15100, Alessandria, Italy.
1This article is part of a Special Issue dedicated to Professor Arthur Carty.
2Corresponding author (e-mail: enrico.sappa@unipmn.it).
Can. J. Chem. 84: 337–344 (2006)
doi:10.1139/V05-252
© 2006 NRC Canada

338
Can. J. Chem. Vol. 84, 2006
Scheme 1.
Table 1. These alkynes, except for the last two, are characterized
by CϵCCH2OH or HCϵCCH2 moieties, which could form
cumulene systems.
H
C
H
Ligand
Name
HCϵCCH2OH
Propargylic alcohol
C
ClCϵCCH2OH
PhCϵCCH2OH
Propargyl chloride
C
3-Phenyl-2-propyn-1-ol
Trimethylsilyl propargyl alcohol
Propargyloxytrimethylsilane
Monopropargylamine
2
(CH3)3SiCϵCCH2OH
(CH3)3SiOCH2CϵCH
H2NCH2CϵCH
(CH3COO)CH2CϵCH
(Ph)2(OH)CCϵCH
(CH3)3SiCϵCH
Fe
Fe
C
(CO)
(OC)
3
3
Fe
(CO)
O
Propargyl acetate
3
1,1-Diphenyl-2-propyn-1-ol
Trimethylsilyl acetylene
OCH
CH
3
C
O
Experimental
Fe(CO)5, Fe3(CO)12 (Strem Chemicals), and the alkynes
(Lancaster Syntheses) were used as received. Methanol,
KOH (pellets), and HCl (37%) were laboratory grade chemi-
cals. Solvents (hexane, heptane, toluene, and dietyl ether)
were dehydrated (over sodium when possible). The reactions
were performed under dry nitrogen in conventional three-
necked flasks, equipped with a gas inlet, cooler, mercury
check valve, and magnetic stirring.
3
C
Fe
(OC)
3
CH
2
Fe
(CO)
3
All reactions of Fe3(CO)12 were performed in CH3OH–
KOH solution and the following “standard” conditions were
adopted: dissolution of KOH (20 pellets, ca. 1.5 g) in
200 cm3 of CH3OH, then addition and dissolution of 5.0 g
(ca. 10 mmol) of Fe3(CO)12, and warming at 40 °C for
10 min; addition of 4.0 cm3 of the liquid alkynes (or of 1.0 g
for the solid ones) and warming at 40 °C for a further 10–
15 min. After cooling, the substrate was acidified with HCl
(37%) to pH 1. After 1 h, the solution was extracted with
three 75 cm3 portions of n-heptane. The solution was then
reduced to a small volume under reduced pressure and then
purified using TLC (Kieselgel PF Merck; eluants were mix-
tures of light petroleum (40–70 °C) and diethyl ether in a
9:1 v/v ratio). Finally, the TLC bands were extracted with di-
ethyl ether.
Fe(CO)
4
H COCH
3
2
CH
2
C
C
O
CH
CH
4
C
Fe
(OC)
3
2
Fe
(CO)
3
The reactions of Fe(CO)5 with propargylic alcohol and
propargyl chloride were performed under the following con-
ditions: to a solution of 30 pellets of KOH in 200 cm3 of
methanol, 30 cm3 (ca. 65 mmol) of Fe(CO)5 was added. The
resulting solution was stirred and warmed at 40 °C for
30 min after which time 6 cm3 of the liquid alkynes were
added. Stirring and warming were continued for a further
30 min. The resulting solution was allowed to cool and acid-
ified (to pH 1) with HCl (32%). A violent evolution of gases
and vapours was observed. After filtering, the solutions were
extracted with three 75 cm3 portions of heptane. After reduc-
tion to a small volume under vacuum, the solutions obtained
were chromatographed as described previously.
plexes such as 3 had been explained by invoking the
intermediacy of complexes of type 2, with no evidence for
the role of complex 1 in the reaction sequence. In this work
the attention has been paid to the synthetic pathways leading
to complexes 1 and 3 and on some aspects of their reactivity
with the aim of obtaining more details on the reaction mech-
anisms leading to methoxycarbonyl derivatives (3, 4). The
reactions of Fe(CO)5 and Fe3(CO)12 with the alkynes
collected in Table 1, under basic methanolic conditions,
have been therefore considered. These alkynes, except for
the last two, are characterized by CϵCCH2OH or HCϵCCH2
moieties, which could form cumulene systems.
Elemental analyses were performed in the laboratories of
the DISTA (Università del Piemonte Orientale). The IR
spectra were obtained on a Bruker Equinox 55 (KBr cells,
Reacting the oily complex 1 with triphenylphosphine, the
title complex (1a) was obtained as the major product, and its
structure was determined by X-ray analysis. The structural
features for this unprecedented derivative are discussed.
1
path length 0.5 mm). The H and 13C NMR were obtained
on a JEOL GX 270 spectrometer. The EI (and CI) mass
spectra were obtained on a Finnigan Mat TSQ-700 mass
© 2006 NRC Canada

Gervasio et al.
339
spectrometer (Servizio di Spettrometria di Massa, Diparti-
mento di Scienza e Tecnologia del Farmaco, Università di
Torino).
68.2 (s), 72.3 (s), 203.3 (s), 206.6 (s), 220.0 (s). EI-MS:
M+ = 586 m/z, release of 12 fragments with m/z = 28. Anal.
calcd. (%): C 38.1, H 1.7, Fe 29.1; found: C 38.2, H 1.8, Fe
28.0.
Reactions of propargyl alcohol with Fe(CO)5
After extraction with heptane, the following products were
obtained: complex 1 (20%), Fe3(CO)12 (15%), and orange
complex 3 (20%).
Reaction of Fe3(CO)12 with propargyloxytrimethylsilane
After extraction, the red-yellow heptane solution was
chromatographed on TLC plates and the following bands
were obtained: dark yellow (15%, complex 1), Fe3(CO)12
(15%), orange (20%, complex 3), red (5, 10%), trace
amounts of a red and a yellow compound (not collected),
and decomposition.
Complex 1
IR (heptane, cm–1) νCO: 2098 (m), 2040 (vs), 2020 (vs),
2012 (s sh), 1989 (m), 1980 (m sh). 1H NMR (CDCl3, r.t.) δ:
3.77 (s, 2H), 2.42 (s, 2H). 13C NMR δ: 70.87 (s), 75.86 (s,
CH2), 186.0 (s, b), 206.1 (s, b, terminal CO), 212.0 (s, CO).
EI-MS: M+ = 348 m/z, release of a fragment with 16 m/z,
then release of 7CO; intense signal at 152 m/z [Fe2(C3H4)].
Anal. calcd. (%): C 34.5, H 1.15, Fe 32.2; found: C 34.6, H
1.2, Fe 32.1.
Reaction of Fe3(CO)12 with propargyl acetate
After extraction with heptane, TLC showed the following
bands: yellow (20%, complex 1), Fe3(CO)12 (20%), orange
(20%, complex 3), traces of a red compound, and decompo-
sition.
Complex 3
IR (cm–1): 2078 (m-s), 2033 (vs), 2008 (s, sh), 1996 (vs),
Reaction of Fe3(CO)12 with monopropargylamine
After acidification, the dark yellow heptane extract was
chromatographed and the following bands were observed:
yellow (complex 1, 13%), orange yellow (complex 3, 30%)
and some decomposition.
1
1979 (m-s). H NMR δ: 4.08 (s, 1H), 3.68 (s, 3H, Me), 3.56
(s, 1H), 2.97 (s, 1H). 13C NMR δ: 54.4 (s, Me), 58.8 (d,
CH2), 66.3 (d, CH), 189.1 (d, C ring?), 197.1 (s, CO), 212.0
(vb, CO acetate). EI-MS: M+ = 378 m/z, release of 6CO,
then complex fragmentation. Anal. calcd. (%) C 34.9, H 1.6,
Fe 29.6; found: C 35.0, H 1.7, Fe 29.7. The analytical results
for complexes 1 and 3 have already been reported in ref. 7.
These are given here for comparison with the other com-
plexes described.
Reaction of Fe3(CO)12 with 3-phenyl-2-propyn-1-ol in
methanol
After acidification and extraction, the greenish solution
gave the following TLC bands: light yellow (10%, complex
6), Fe3(CO)12 (15%), orange (15%, complex 7), and orange,
purple, purple (tr, not collected).
Reaction of propargyl chloride with Fe(CO)5
After filtration of the acidified solution a considerable
amount of solid residual was observed; this was partially
soluble in methanol. The methanolic solution was added to
the filtrate. After extraction with heptane, TLC gave the fol-
lowing bands: yellow (ca. 10%, complex 1), Fe3(CO)12 (30%),
orange (ca. 10%, complex 3) and red (tr, not collected).
Complex 6
IR (cm–1): 2074 (m), 2033 (vs), 2005 (s), 1996 (s, sh),
1
1978 (m, sh). H NMR δ: 7.26–7.19 (m, 5H, Ph), 4.28 (s,
1H), 3.88 (s, 1H), 3.69 (s, 1H). EI-MS: M+ = 424 m/z, loss
of seven fragments with m/z = 28. Proposed formula:
Fe2(CO)7C3H3(Ph). Anal. calcd. (%): C 45.3, H 1.9, Fe 26.4;
found: C 45.2, H 1.8, Fe 26.5.
Reaction of Fe3(CO)12 with trimethylsilyl propargyl
alcohol
After acidification and extraction, the red, clear heptane
solution was chromatographed on TLC plates and the fol-
lowing bands were obtained: yellow (10%, complex 1),
triiron dodecacarbonyl (10%), orange (25%, complex 3),
orange-red (20%, complex 5), and some decomposition.
Complex 7
IR (cm–1): 2074 (m-s), 2033 (vs), 2005 (s), 1996 (s, sh),
1978 (m-s). 1H NMR δ: 7.45–7.26 (m, 5H, Ph), 6.30 (s, 1H),
3.88 (s, 1H). EI-MS: M+ = m/z 454, loss of 6CO (see ref. 7).
Proposed formula: Fe2(CO)6[RHC3R′C(=O)OCH3] (R, R′ =
Ph, H). Anal. calcd. (%): C 45.0, H 2.2, Fe 24.6; found: C
45.1, H 2.3, Fe 24.5.
Complex 5
IR (heptane, cm–1): 2100 (m), 2074 (m-s), 2032 (vs), 2006
1
(s), 1992 (s, sh), 1974 (m). H NMR δ: 4.18 (d), 3.95 (d),
Reactions of complexes 1 and 3
3.86 (s, 1H), 3.75 (d), 3.67 (d), 3.53 (s, 1H), 3.28 (d), 3.04
(q, 2H), 2.94 (s, 1H), 2.77 (s, 1H), 2.55 (s, 1H). EI-MS:
M+ = 586 m/z, release of 12 fragments with m/z = 28. Anal.
calcd. (%) C 38.1, H 1.7, Fe 29.1; found: C 38.3, H 1.7, Fe
28.8.
Reaction of complex 1 with PPh3 under thermal
conditions
About 0.5 g of complex 1 (ca. 1.5 mmol) were dissolved
in heptane; PPh3 (0.5 g, ca. 1.9–2.0 mmol) and Me3NO were
added and the solution was warmed to reflux (10 min) under
nitrogen. After 4 min reflux (after which time the colour
turned from yellow to dark red), the suspension was allowed
to cool, was reduced to a small volume under vacuum, and
chromatographed on TLC plates. The following bands were
obtained: orange (50%, complex 1a) and red (50%, complex
Complex 4
IR (cm–1): 2100 (m), 2075 (m-s), 2031 (vs), 2026 (vs),
1
1992 (s), 1973 (m). H NMR δ: 4.16–4.12 (d, 2H), 3.89–
3.78 (d, 2H), 3.68–3.66 (d, 1H), 3.29 (s, 3H, Me), 3.07–3.04
(d, 2H, CH2). 13C NMR ␦: 37.3 (s), 58.4 (s), 66.4 (s),
© 2006 NRC Canada

340
Can. J. Chem. Vol. 84, 2006
Scheme 2.
1b). Complex 1b is nearly insoluble in solvents commonly
used for obtaining X-ray grade crystals.
-
.-
[HFe3 (CO)11] and (or) [Fe3 (CO)11]
Complex 1a
IR (cm–1): 2061 (m-s), 2017 (s), 1988 (vs), 1978 (s, sh),
Propargylic alcohol
Propargyl chloride
1
1946 (w). H NMR δ: 7.56 (m), 7.45 (m), 7.32 (m, 15H,
Ph), 3.55 (s, 2H), 2.86 (s, 2H). 31P NMR δ: –4.50 (s). Anal.
calcd. (%): C 55.67, H 3.26, Fe 19.24; found: C 55.8, H 3.3,
Fe 19.1. (MW = 582).
2
1
Complex 1b
IR (cm–1): 2050 (m b), 1970 (vs), 1900 (m b). 1H NMR δ:
7.54 (m), 7.40 (m, 30H, Ph), 3.50 (s, 2H), 2.16 (s, 2H). 31P
NMR δ: 59.78 (s), 79.20 (s). CI-MS: M+ = 842 m/z (exp.
844), loss of 5CO. Anal. calcd. (%): C 55.67, H 3.26, Fe
19.24; found: C 55.7, H 3.3, Fe 19.3. (MW = 844).
3
X-ray structure analysis for complex 1a
Reaction of complex 1 with CH3OH under thermal
conditions
The reflection data were collected on a Siemens P4
diffractometer equipped with a Bruker APEX CCD detector
using graphite-monochromated Mo Kα radiation (λ =
0.710 73 Å). The complex 1a (C27H19Fe2O6P) crystallized
from n-heptane in a monoclinic P21/n space group, with a =
12.7888(15) Å, b = 9.3735(10) Å, c = 21.4652(26) Å, β =
92.163(2)°, V = 2571.3(5) Å3, MW = 582.09, Z = 4,
Dcalcd. = 1.504 g cm–3, ␮ = 1.230 mm–1. The red crystal
used was prismatic and had dimensions of 0.08 × 0.10 ×
0.20 mm3. The θ range for measurement was 1.90°–28.32°;
23 563 reflections were measured at 293 K and 5444 were
unique (Rint = 0.077). The intensities were corrected semi-
empirically for absorption, based on symmetry equivalent re-
flections. The refinement of 325 parameters was made using
full-matrix least-squares on F2. All non-hydrogen atoms
were refined anisotropically. The hydrogen atoms were
calculated and refined with Uiso set at 1.2 times Ueq of the
corresponding C atom. The final parameters were: R =
Σ||Fo| – |Fc||/ΣFo| = 0.0732 for 2975 “observed” reflections
Approximately 0.2 g of complex 1 (ca. 0.6 mmol) were
dissolved in CH3OH (10 cm3), brought to reflux (3 min),
and allowed to react for 15 min. The dark yellow suspension
was brought to dryness under N2 flow; after TLC purifica-
tion, only parent 1 could be observed together with some de-
composition.
Approximately 0.25 g of complex 1 (ca. 0.72 mmol) were
dissolved in 10 cm3 of CH3OH (and 0.2 cm3 of HCl were
added), brought to reflux (5 min), and allowed to react for
13 min. No change in colour was observed. The suspension
was dried under a nitrogen stream; TLC purification and
spectroscopic analyses showed the presence of complex 1 only.
Reactions of 1a with methanol under thermal conditions
Approximately 25 mg of 1a (crystals used for the X-ray
study) were dissolved in 10 cm3 of CH3OH under N2 and al-
lowed to reflux for 5 min. No change in colour was ob-
served. The mixture was brought to dryness under reduced
pressure; after TLC separation, only one complex could be
obtained (1c).
2
2
2
2 2
2 2 1/2
having Fo > 2σ(Fo ), Rw = [Σ(wFo – Fc ) /Σw(Fo ) ]
=
2
2 2
0.127, goodness-of-fit = [Σw(Fo – Fc ) /(No. of unique re-
flections – No. of parameters)]1/2 = 0.959. Programs used
were SHELXTL (9) for structure solution, refinement, and
molecular graphics, Bruker AXS SMART (diffractometer
control), SAINT (integration), and SADABS (absorption
correction) (10). The high R values are due to the poor qual-
ity of crystals; the intensity measurements were repeated on
other crystals with the same results. Therefore, the calcu-
lated hydrogen atom positions must be considered with cau-
tion.
Complex 1c
IR (cm–1): 2061 (m-s), 2018 (m-s), 1987 (vs), 1977 (s sh),
1
1946 (m). H NMR δ: 7.70–7.28 (mm, Ph), 3.58 (s, 1H),
3.51 (s, 1H), 2.31 (s, 1H), 2.19 (s, 3H). CI-MS: M+
=
584 m/z, loss of 6CO. Anal. calcd. (%): C 55.5, H 3.6, Fe
19.2; found: C 55.4, H 3.7, Fe 19.3.
Reactions of 3 with triphenylphosphine
Approximately 0.5 g of complex 3 (ca. 1.3 mmol) were
dissolved in heptane under N2 and a 2:1 molar excess of
PPh3 was added. The suspension was brought to reflux
(10 min) and allowed to reflux for 3 min. The colour of the
clear solution turned from orange to red. TLC purification
gave the following bands: dark yellow (3, ca. 10%), red
(complex 3a, ca. 40%) and impure PPh3.
Results and discussion
Spectroscopic characterization of complexes 5–7
Complex 5 is obtained only from the isomeric ligands
trimethylsilyl propargyl alcohol and propargyloxytrimethyl-
silane. The analytical and spectroscopic data (except for the
1H NMR) are closely comparable with those of complex 4.
However, in view of the high complexity of the NMR spec-
trum, we cannot provide a structure. Complex 5 has an iso-
meric formula with respect to 4 (elemental analysis and
mass spectrum) and the NMR spectrum could indicate that
“5” is a nonseparated mixture of the two isomers. Attempts
Complex 3a
IR (cm–1): 2042 (vs), 1991 (vs), 1977 (s), 1963 (m), 1936
1
(m b). H NMR δ: 7.55–7.25 (mm, Ph), 3.97 (s, 1H), 3.40
(s, 1H), 3.21 (s, 3H), 2.66 (s, 1H). 31P NMR δ: 30.8 (s).
Anal. calcd. (%): C 54.0, H 3.5, Fe 18.7; found: C 54.1, H
3.6, Fe 18.8.
© 2006 NRC Canada

Gervasio et al.
341
Scheme 3.
H NCH C CH
2
2
HC
CCH OH
2
HC CCH Cl
2
CH
2
C
Fe
(OC)
3
CH
2
Fe
(CO)
3
1
+
OCH
CH
3
C
O
C
Fe
(OC)
3
CH
2
Fe
(CO)
3
3
CH
(CH ) SiOCH C
3 3
2
(CH ) SiC CCH OH
3 3
2
C
2
H CCOOCH
3
CH
at obtaining crystals suitable for X-ray analysis were unsuc-
cessful.
was observed only in trace amounts in other reactions
leading to medium yields of 1. This could represent indirect
evidence for 2 being a labile intermediate in the formation
of 1 (as a by-product) and of 3 in a direct reaction. (ii) An
alternative possibility could be that 1 is an intermediate in
the formation of 3.
Complexes 6 and 7 are the homologues of 1 and 3, re-
spectively, and contain a phenyl at the site of one of the hy-
drogens of the C3 part of the organic ligand. The complexes
are obtained from 3-phenyl-2-propyn-1-ol, which apparently
behaves like the other alkynes, but does not lose the phenyl
group. Complex 7 shows the same analytical data, but a dif-
ferent NMR with respect to the complex reported in ref. 7.
Again, it could be an isomer obtained by a different position
of the phenyl group in the structure.
The formation of complexes 1 and 3 has been therefore
taken into account. The results show that complexes 1 and 3
are obtained in medium yields starting from propargylic al-
cohol, propargyl chloride, trimethylsilyl propargyl alcohol,
propargyloxytrimethylsilane, monopropargylamine, and
propargyl acetate (see Table 1). These alkynes can release
-CϵCCH2OH or HCϵCCH2-, three-carbon atom fragments
that could rearrange (and uptake hydrogen atoms) to form
the H2C=C=CH2 ligand observed in 1 and 1a. In contrast,
complexes 1 and 3 are not obtained from trimethylsilyl acet-
ylene and 1,1-diphenyl-2-propyn-1-ol, which can release
only C2 fragments. The reaction of Fe3(CO)12 with 1,1-
diphenyl-2-propyn-1-ol in the same conditions reported for
the other alkyne ligands yields only two products in trace
amounts (not collected). The reaction of Fe3(CO)12 with
trimethylsilyl acetylene gives mostly Fe3(CO)12 and very
small yields of a orange, yellow, and red bands showing IR
Reactions leading to complexes 1 and 3
As previously discussed, the reactions of Fe3(CO)12 with
propargylic alcohol and propargyl chloride give, as the main
products, complexes 1, 3, and 4 and much smaller amounts
of the allenylidene complex 27. No unequivocal evidence,
both for the structure of 1 and for its role in the reaction
pathways leading to 3, was obtained. As shown in Scheme 2,
two hypotheses can be advanced: (i) 1 would be a fragmen-
tation product derived from 2; however, complex 2 was ob-
tained in yields not allowing a reactivity study, only from the
reactions of propargylic alcohol and propargyl chloride. It
© 2006 NRC Canada

342
Can. J. Chem. Vol. 84, 2006
Scheme 4.
CH
CH
2
2
PPh
3
C
(PPh )(CO)
Fe
3 2
C
Fe
(OC)
3
+ 1b
CH
2
CH
2
Fe
(CO)
Fe
(CO)
3
3
1
a
1
CH OH
3
CH OH
3
OCH
3
HC
CH
CH
C
O
CH
(PPh )(CO)
3
Fe
2
Fe(CO)
3
C
(OC)
Fe
3
1c
O
CH
2
Fe
(CO)
CH
3
3
3
PPh
3
OCH
CH
3
C
O
C
(PPh )(CO)
2Fe
3
3a
CH
2
Fe
(CO)
3
and NMR spectra different from those of 1 and of 3. The
“parent” alkynes forming complexes 1 and 3 are shown in
Scheme 3.
the reaction conditions adopted was reported in the literature
(1); (ii) its presence during the synthesis of Fe3(CO)12 start-
ing from Fe(CO)5 is well-established (12). It is also possible
that, in the above reaction conditions, the radical anion
[Fe3(CO)11]·– (13) is present.
Interestingly, 3-phenyl-2-propyn-1-ol, which would not
easily release the C6H5 substituent, gives complexes 6 and 7,
which are the phenyl-containing homologues of 1 and 3.
It is also worth noting that 1 and 3 are obtained from
monopropargylamine: this also indicates that the NH2 func-
tionality can be lost. Finally, complex 3 is also obtained
from propargyl acetate. Mathieu and co-workers (11) have
described the synthesis of the acetate open-cluster derivative
Fe3(CO)10(PPh2)[CHCCHC(OCH3)O] containing a hetero-
cyclic ring comparable with that of 3. The complex was ob-
tained from [HFe3(CO)11]– and (CH3COO)CH2CϵCPPh2.
It is also interesting to observe that 1 (and 3) are obtained
when reacting propargylic alcohol and propargyl chloride
both with Fe3(CO)12 and with Fe(CO)5 in basic methanolic
conditions. The intermediacy of the [HFe3(CO)11]– anion in
these reactions can be hypothesized on the grounds of the
following observations: (i) its formation from Fe3(CO)12 in
Reactions leading to 1a and 1b — Proposed structure
of complex 1 on the basis of the X-ray structure of
complex 1a
The reaction of complex 1 with triphenylphosphine leads
to the monosubstituted 1a, whose structure is discussed in
the following, and to the disubstituted complex 1b presum-
ably bearing the two phosphine ligands on different iron at-
oms, as indicated by the 31P NMR spectrum.
Complex 1, when dissolved in heptane, heptane–toluene,
or other solvents and kept at –30 °C, gives large and well-
formed orange “crystals”; however, when the solvent is re-
moved, even at low temperature, the crystals collapse quickly
forming a thick oil. For 1 a structure is therefore proposed
on the basis of analytical and spectroscopic data and on the
© 2006 NRC Canada

Gervasio et al.
343
3
Fig. 1. ORTEP plot of Fe2(CO)6(PPh3)[µ-η -(H2CCCH2)] (com-
plex 1a) with thermal ellipsoids (30% probability).
Table 2. Some relevant bond distances (Å)
and angles (°) for complex 1a,
Fe2(CO)6(PPh3)[µ-η -(H2CCCH2)].
3
Bond distances (Å)
Fe(1)—C(2)
Fe(1)—C(3)
Fe(1)—C(1)
Fe(1)—Fe(2)
Fe(2)—C(2)
Fe(2)—P(1)
P(1)—C(51)
P(1)—C(31)
P(1)—C(41)
C(1)—C(2)
C(2)—C(3)
1.952(5)
2.120(5)
2.123(6)
2.6479(10)
1.977(5)
2.2514(14)
1.824(4)
1.824(5)
1.835(5)
1.448(8)
1.384(8)
Bond angles (°)
C(3)-C(2)-C(1)
C(3)-C(2)-Fe(2)
C(1)-C(2)-Fe(2)
112.5(6)
119.4(5)
117.9(5)
following observations: (i) 1 is formed from several alkynes
that can loose a three-carbon atom fragment; (ii) in a recent
report propargylic alcohol has been shown to react with
triruthenium benzoquinoline complexes releasing a (coordi-
nated) C3H4 unit (14), albeit different from that proposed for
phine substitution for CO occurs first at the iron atom σ
bound to the organic ring (15).
The crystal structure of 1a
The structure of 1a is shown in Fig. 1 and significant dis-
tances and angles are given in Table 2. The complex is
formed by two nonequivalent iron atoms, with an Fe—Fe
distance of 2.648(1) Å.
1
1; (iii) complex 1 shows a H NMR spectrum indicating the
presence of two nonequivalent CH2 groups. Mass spectra of
samples coming from reactions with different alkynes all
showed a P+ = 348 m/z corresponding to Fe2(CO)7(C3H4). In
the light of the structure of 1a discussed in the following, we
now think that the more probable structure is that shown in
Scheme 3.
The C3H4 ligand is coordinated both to Fe(1), as a formal
3e– donor, and to Fe(2), as a formal 1e– donor. Six COs are
terminally bound to Fe(1) and to Fe(2), which is also bound
to the PPh3 ligand disposed cis with respect to C(2), in the
less sterically and electronic demanding situation. When
considering the C3H4 as a four-electron donor and the pres-
ence of an Fe—Fe bond, each iron atom would reach the 18-
electron configuration. The Fe(2)-C(1)-C(2)-C(3) moiety de-
viates from planarity with a mean deviation of 0.10 Å. The
Fe—C bond distances and their geometry agree with a for-
mal description of a π interaction of Fe(1) with C(1)—C(2)
and with C(2)—C(3) elongated double bonds and with a
Fe(2)—C(2) σ interaction. This situation corresponds to a π-
allyl moiety where a hydrogen atom is substituted by an iron
atom. Allyl groups coordinated in this way are commonly
found, especially in the nickel triad chemistry. Complex 1a
has been very presumably obtained upon simple substitution
of a phosphine for a CO, suggesting a similar formula for 1.3
Reactions of complexes 1 and 1a with CH3OH
Complex 1
These reactions have shown that under thermal conditions
complex 1 does not react very easily with methanol. Only
trace amounts of complex 3 were observed. Thus, complex 1
should be considered a side product, rather than an interme-
diate, in the formation of 3.
Complex 1a
From this reaction, one can see that complex 1a can add
methanol, giving, however, a complex (1c) different from
3a. For complex 1c, we propose the structure shown in
Scheme 4 on the basis of the analytical and spectroscopic re-
sults and by analogy with the previous reports of Mathieu
and co-workers (11). Thus, complex 1a is not an intermedi-
ate in the formation of the phosphine-substituted 3a.
Conclusions
The results obtained allow the unequivocal characteriza-
tion of complex 1. We have found that 1 is a by-product in
the synthesis of 3; it is presumably a fragmentation product
of the allenylidenic complex 2. Complex 1 does not react
with methanol to give 3. In contrast, 1a gives a product (1c)
Reaction of complex 3 with PPh3
The monosubstituted complex 3a was isolated and charac-
terized using spectroscopy; complex 3a probably follows the
behaviour of other ferrole-like complexes, where the phos-
3 Supplementary data for this article are available on the journal Web site (http://canjchem.nrc.ca) or may be purchased from the Depository
of Unpublished Data, Document Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0R6, Canada. DUD 5001. For more
information on obtaining material refer to http://cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml. CCDC 273281 contains the crystallographic
data for this manuscript. These data can be obtained, free of charge, via http://www.ccdc.cam.ac.uk/conts/retrieving.html (Or from the Cam-
bridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax +44 1223 336033; or deposit@ccdc.cam.ac.uk).
© 2006 NRC Canada

344
Can. J. Chem. Vol. 84, 2006
Oro, and P.R. Raithby. Wiley-VCH, Weinheim. 1999. Chap. 8.
p. 796.
2. See, for example: J.P.H. Charmant, P. Crawford, P.J. King, R.
Quesada-Pato, and E. Sappa. J. Chem. Soc. Dalton Trans.
4390 (2000).
3. G. Gervasio and E. Sappa. J. Organomet. Chem. 498, 73 (1995).
4. See, for example: (a) J.P.H. Charmant, G. Davies, P.J. King,
J.R. Wigginton, and E. Sappa. Organometallics, 19 (2000) 2330;
(b) J.P.H. Charmant, P.J. King, R. Quesada-Pato, E. Sappa, and
C. Schaefer. J. Chem. Soc. Dalton Trans. 46 (2001).
5. G. Gervasio and E. Sappa. Organometallics, 12, 1458 (1993).
6. (a) G. Gervasio, D. Marabello, P.J. King, E. Sappa, and A.
Secco. J. Organomet. Chem. 671, 137 (2003);(b) G. Gervasio,
D. Marabello, E. Sappa, and A. Secco. J. Organomet. Chem.
689, 35 (2004).
containing the fragments of methanol, albeit coordinated dif-
ferently than in 3 or 3a.
These results confirm the mechanism already proposed for
the formation of 3 (see Scheme 2), that is, the intermediacy
of complex 2 only. Formation of complexes containing
methoxycarbonyl groups upon activation of CO under basic
methanolic conditions seems to be a general trend. For ex-
ample, recently we have observed that Ru3(CO)12 reacts with
1,4-dichloro-but-2-yne in basic methanolic conditions form-
ing the complex H2Ru3(CO)9[H2C=CCϵCC(=O)OCH3]
(16); the formation of this complex requires loss of chlorine
and activation of CO. In our opinion, the mechanisms lead-
ing to this type of complex deserve further efforts towards
being fully understood, especially in view of organometallic-
mediated organic syntheses.
7. G. Gervasio, D. Marabello, E. Sappa, and A. Secco. J.
Last but not least, from the reactions described in this pa-
per, complex 1a was obtained. The structure of this complex
is unprecedented and shows, once again, that alkynol-cluster
chemistry may lead to unprecedented bonding modes for
hydrocarbyls on metal clusters.
Organomet. Chem. 690, 3755 (2005).
8. G. Gervasio, D. Marabello, and E. Sappa. J. Chem. Soc. Dal-
ton Trans.1851 (1997).
9. G.M. Sheldrick. SHELXTL. Version 5.1 [computer program].
Bruker AXS Inc., Madison, Wisconsin. 1997.
10. SMART, SAINT, SADABS, and XPREP software for CCD
diffractometers. Bruker AXS Inc., Madison, Wisconsin.
11. A. Elarraoui, J. Ros, R. Mathieu, and R. Yanez. J. Organomet.
Chem. 678, 117 (2003).
12. (a) R.B. King, F.G.A. Stone, R.S. Summitt, and W.E. Edgell.
Inorg. Synth. VII, 193 (1965); (b) W. McFarlane, G.
Wilkinson, and W. Hubel. Inorg. Synth. VIII, 181 (1966).
13. F. Ragaini, J.-S. Song, D.L. Ramage, G.L. Geoffroy, G.A.P.
Yap, and A.L. Rheingold. Organometallics, 14, 378 (1995).
14. J.A. Cabeza, I. del Rio, S. Garcia-Granda, V. Riera, and M.
Suarez. Organometallics, 23, 3501 (2004).
Acknowledgements
Financial support of this work was obtained (in part) from
the Consorzio Interuniversitario Scienza e Tecnologia dei
Materiali (INSTM) and from the Universita’ del Piemonte
Orientale, and from the Ministero per l’Istruzione,
l’Università e la Ricerca (MIUR, Rome) as a part of the
COFIN 2003 programme.
15. L.J. Todd, J.P. Huckey, J.R. Wilkinson, J.C. Huffman, and K.
References
Folting. J. Organomet. Chem. 112, 167 (1976).
16. G. Gervasio, D. Marabello, E. Sappa, and A. Secco. J.
Organomet. Chem. 690, 1594 (2005).
1. See, for example: S. Deabate, P.J. King, and E. Sappa. In
Metal clusters in chemistry. Vol. 2. Edited by P. Braunstein, L.A.
© 2006 NRC Canada