Metal ion complexation in organometallic crown ethers : Structural and rate effects on the migratory insertion of carbon monoxide in indenyl and cyclopentadienyl iron (II) complexes

Article abstract of DOI:10.1016/S0022-328X(97)00521-4

The iron complexes [Fe(η⁵-ligand)(CO)₂R] [ligand=C₅H₅ (Cp), R=Me (1); ligand=C₅H₅, R=CHMe₂ (2); ligand=C₉H₇ (Ind), R=Me (3)] react with the crown ether functionalized phosphine [Ph₂P-CH₂-(aza-15-crown-5)] (L) to give the products of alkyl migratory insertion [Fe(η⁵-ligand)(COR)(CO)(L)] 4, 5 and 6, respectively. The cyclopentadienyl methyl complex [Fe(η⁵-C₅H₅)(CO)₂Me] reacts with L only in refluxing hexane. The reactions of 2 and 3, in acetonitrile or chloroform, have been followed by FT-IR spectroscopy, or by UV-Vis. The presence of sodium or calcium salts, NaI, NaPF₆, or CaI₂, does not affect the reactivity of the cyclopentadienyl complex 2, whereas it increases two to three times the rate of reaction of the indenyl complex 3. The acetyl complexes [Fe(η⁵-ligand)(COMe)(CO)(L)], (ligand=Cp, 4; ligand=Ind, 6) exhibit a shift of the stretching band ν(COMe) to lower frequencies, in the presence of metal salts, up to 40 cm^-1 in the case of 6·NaPF₆ in chloroform, and up to 46 cm^-1 in the case 6·KPF₆ in dichloromethane, as a result of coordination of the carbonyl oxygen to the crown ether held cation. This shift is absent or small in the more polar solvents tetrahydrofuran, acetonitrile, or dioxane, and it is not observed for the isopropyl complex 5, suggesting that steric hindrance of the alkyl group prevents acyl coordination to the metal ion. The terminal carbonyl group of complexes 4, 5 and 6 displays a shift of the ν(CO) band to higher frequencies in most solvents, up to 27 cm^-1 for 6·NaPF₆ in acetonitrile and to 26 cm^-1 for 6·LiClO₄ in dichloromethane, arising from electronic polarization by the metal ion and indicating electronic communication between the two metal centres.

Full text of DOI:10.1016/S0022-328X(97)00521-4

Ž
.
Journal of Organometallic Chemistry 551 1998 331–338
Metal ion complexation in organometallic crown ethers
Structural and rate effects on the migratory insertion of carbon monoxide
1
ž /
in indenyl and cyclopentadienyl iron II complexes
)
Valerio Berdini, Mauro Bassetti , Giovanna Mancini, Luigi Mandolini, Donato Monti
Centro C.N.R. di Studio sui Meccanismi di Reazione, cro Dipartimento di Chimica, UniÕersita ‘La Sapienza’1, Roma, Italy
´
Received 6 June 1997
Abstract
5
w
Ž
.Ž
.
x w
Ž
.
Ž .
Ž .
Ž
.
The iron complexes Fe h -ligand CO 2R ligandsC5H5 Cp , RsMe 1 ; ligandsC5H5, RsCHMe2 2 ; ligandsC9H7 Ind ,
Ž .x
3
w
Ž
.x Ž .
RsMe
x
migratory insertion Fe h –ligand COR CO L 4, 5 and 6, respectively. The cyclopentadienyl methyl complex Fe h -C5H5 CO 2Me
react with the crown ether functionalized phosphine Ph2P–CH2- aza-15-crown-5 L to give the products of alkyl
5
5
w
Ž
.Ž
.Ž .Ž .x
w
Ž
.Ž
.
reacts with L only in refluxing hexane. The reactions of 2 and 3, in acetonitrile or chloroform, have been followed by FT-IR
spectroscopy, or by UV–Vis. The presence of sodium or calcium salts, NaI, NaPF6, or CaI2, does not affect the reactivity of the
cyclopentadienyl complex 2, whereas it increases two to three times the rate of reaction of the indenyl complex 3. The acetyl complexes
5
w
Ž
.Ž
.Ž .Ž .x Ž
.
Ž
.
Fe h -ligand COMe CO L , ligandsCp, 4; ligandsInd, 6 exhibit a shift of the stretching band n COMe to lower frequencies, in
the presence of metal salts, up to 40 cmy in the case of 6PNaPF6 in chloroform, and up to 46 cmy in the case 6PKPF6 in
dichloromethane, as a result of coordination of the carbonyl oxygen to the crown ether held cation. This shift is absent or small in the
more polar solvents tetrahydrofuran, acetonitrile, or dioxane, and it is not observed for the isopropyl complex 5, suggesting that steric
hindrance of the alkyl group prevents acyl coordination to the metal ion. The terminal carbonyl group of complexes 4, 5 and 6 displays a
1
1
shift of the n CO band to higher frequencies in most solvents, up to 27 cm for 6PNaPF6 in acetonitrile and to 26 cmy1 for 6PLiClO4
in dichloromethane, arising from electronic polarization by the metal ion and indicating electronic communication between the two metal
centres. q 1998 Elsevier Science S.A.
y
1
Ž
.
Keywords: Crown ether; Carbon monoxide; Indenyl
1. Introduction
different donor molecules have also been reported for
pentamethylcyclopentadienyl complexes of rhodium and
5
w
Ž
.Ž .Ž .Ž .x Ž
The migratory insertion of carbon monoxide in tran-
sition metal complexes can be promoted both by donor
iridium M h -C5Me5 R CO X MsRh, Ir; Rs
.
w
x
Me, aryl; XsCl, Br, I, MeCO2 , CF3CO2 6–8 , as
w x
molecules and by Lewis acids 1 . We have reported in
well as comparative studies of the effect of the two
w x
recent years the reaction of indenyl and cyclopentadi-
different metals on rates of migratory insertion 9 .
5
Ž
.
w
Ž
enyl carbonyl iron II
complexes Fe h –
The activation of CO by strong Lewis acids has been
.Ž
.
x
Ž
.
w
x
ligand CO R ligandsC9 H7, C5H5; RsMe, iPr
clearly documented in the literature 10–13 . Interaction
of the acidic centre with the oxygen atom and the
resulting stabilizing effect in the acyl product is pre-
sumed to be the driving force, as supported by the
2
with phosphines to give the corresponding acyl com-
w
x w
x
plexes 2,3 4,5 . The occurrence of an associative
mechanism has been described, involving fast and re-
versible formation of equilibrium followed by a rate
determining carbon–carbon coupling. The reactions with
w
x
isolation of bimetallic m-COR complexes 10 . Rate
effects of different cations have been reported for the
migratory insertion process on the anionic carbonyl
qw xy
Ž
species M Fe CO R
Ž
.
MsLi, Na, Na–crown ether,
PPh3 2 N , and ion pairing equilibrium constants were
found to be higher in the corresponding acyl complexes
qw xy
. Ž
4
Ž
.
.
)
Corresponding author. Fax: q39-6-490421; e-mail:
bassetti@netmgr.ced.cnr.rm.it.
1 Dedicated to Professor Peter M. Maitlis, with gratitude, on the
occasion of his 65th birthday.
Ž .
Fe CO COR , the acyl group acting as the
4
M
w
x
cation binding site 14 .
0022-328Xr98r$19.00 q 1998 Elsevier Science S.A. All rights reserved.

(
)
332
V. Berdini et al.rJournal of Organometallic Chemistry 551 1998 331–338
w
with the crown ether functionalized phosphine Ph2 P–
Ž
.x Ž .
CH2- aza-15-crown-5 L in the presence of alkali and
alkaline-earth metal salts, as depicted in Scheme 1 for
complex 3.
Binding of metal ions to the cavity of L forms a
heterotopic macrocyclic species, which contains a donor
site at the phosphorus atom and a Lewis acid centre.
5
w
Ž
Reactions of alkyl carbonyl complexes Fe h -
.Ž
.
x
C5 H5 CO R with amphoteric ligands, namely
2
X
Ž
.
aluminoamine phosphines of the type Ph2 P-NR-AlR2 ,
w
x
have been reported as very rapid 18 . Extensive work
has also been described on the effect of cation binding
in molybdenum and tungsten complexes of phosphorus
donor crown ether ligands toward the activation of
nucleophilic addition to coordinated carbon monoxide
Fig. 1. Intramolecular acetyl oxygen–cation interaction in the com-
5
w
Ž
.Ž
.Ž .Ž .x
plexation of metal salts to Fe h -C5H5 COMe CO L .
The existence of acyl–cation interactions in solution
has been clearly demonstrated in the case of intramolec-
ular acyl bonding to a crown ether held cation. It has
been shown that alkali or alkaline-earth metal salts
w
x
19 . It has been reported that cobalt and rhodium
complexes of the aza–crown ether L accelerate the
hydroformylation of alkenes when co-complexed by
dissolve in dichloromethane solutions of the complex
Naq or Liq 20 .
The field of metallomacrocycles is an area of current
interest, because of different properties and applications
potentially deriving from combination of metal centres
w
x
5
w
Ž
.Ž
.Ž
.Ž
Fe h -C5 H5 COMe CO Ph2 P–CH2-aza-15-crown-
x
1
.
5 , producing shifts in the infrared and H NMR spec-
tra which are consistent with coordination of the acetyl
oxygen atom to the metal cation in the cavity of the
w
x
in organized systems 21 . In particular, structures which
w
x
Ž
.
aza–crown ether 15 Fig. 1 .
5
Ž
.
combine a p-face h -ligand for redox active transition
metals and a crown ether s-face for coordination of
group I and II metal ions may display electronic interac-
tion and attractive properties in the combination of hard
and soft metal centres 22 . In this context, the effect of
alkali and alkaline-earth cations and that of molecular
In view of our interest in the migratory insertion
reaction of carbonyl iron II complexes 2–9 , as well as
in the effect of alkali and alkaline earth cations on
Ž .
w
x
w
x
addition reactions to carbonyl compounds 16 , we have
addressed a key point underlying the study of acyl
oxygen-metal ions binding, i.e., the potential effect
played by such interaction on the rate of the carbon–
carbon forming step. Therefore, the scope of this work
is to describe whether the addition of metal ions may
affect the process of migratory insertion, when the
electrophilic metal centre is held in proximity of the
functional group which is undergoing a change from
terminal carbonyl to acyl. This possibility had been
w
x
Ž .
structural changes on the infrared features of the iron II
acyl complexes 4–6 are also described in this work.
2. Experimental
NMR spectra were recorded on a Bruker AM-300
w
x
Ž1
.
pointed out in the original communication 15 and
spectrophotometer, in CDCl3, at 300 H , or 121.7
MHz
respect to Me4Si or H3PO4, respectively. Infrared spec-
tra were taken with a FT-IR Nicolet 510 spectrophotom-
eter, interfaced with a Nicolet 620 workstation. Low
31
w x
w
x
Ž
Ä1
4
H
.
P and chemical shifts are reported with
further addressed in the literature 1 , p. 372, 17 . We
5
w
Ž
report here a study of the reaction of Fe h -
ligand CO R ligandsC5H5, RsMe 1 ; ligands
C5H5, Rs2CHMe2 2 ; ligandsC9 H7, RsMe 3
.Ž
.
x w
Ž .
Ž .
Ž .x
Scheme 1.

(
)
V. Berdini et al.rJournal of Organometallic Chemistry 551 1998 331–338
333
5
Ž
[
(
6
)(
)( )(
resolution mass spectra FAB, m-nitrobenzyl alcohol as
2.1.3. Fe h -C9 H7 COMe CO Ph2 P–CH2-aza-15-
.
)] ( )
crown-5
matrix have been obtained with a VG-Quattro instru-
ment, at the Universita di Tor Vergata, Roma.
Ž
.
Ž
Complex 3 0.55 g, 2.3 mmoles and L 1.53 g, 3.6
mmoles in a Schlenk tube, were stirred in hexane 5
ml at 508C, for 24 h. The dark brown solid, precipi-
tated out of the reaction mixture, was dissolved in
dichloromethane, then chromatographed with
dichloromethane first, then with increasing amounts of
tetrahydrofuran, up to a 1:5 ratio tetrahydrofuran:dichlo-
romethane. Evaporation of the solvent gave complex 6
`
.
Ž
All solvents were distilled under argon, tetrahydrofu-
ran and toluene from potassiumrbenzophenone, hexane
and pentane from sodium, and dichloromethane from
calcium hydride. Acetonitrile was distilled first from
calcium hydride, then from phosphorus pentoxide. Alu-
mina 3% water was used for the chromatographic
separations. All reactions were performed under argon,
using standard Schlenk line techniques.
.
Ž
.
Ž
.
as a dark orange solid 0.85 g, yield 48% . IR: 1919.5,
1611.9 cmy1 tetrahydrofuran . 1H NMR: d 7.6–7.1
Ž
.
Ž
.
The phosphine Ph2 P–CH2- aza-15-crown-5 , L, was
w
x
Ž
.
Ž
prepared as described in the literature 23 . It is easily
oxidized by air, either during preparation or in the
subsequent manipulations. We have never been able to
obtain samples with a purity higher than 95%, checked
m, 14 H, benzo and phenyl , 5.28, 5.01, 4.78, s, 3 H,
indenyl C2 H and C1,3H , 3.8–3.2 m, 16 H, OCH2 ,
.
Ž
.
Ž
.
Ž
3.41 d, 2H, JPyH s3.8 Hz, N-CH2–P , 2.58 s, 3 H,
COMe , 2.35 m, 4 H, N- CH2–
ppm. 31 P NMR: d
.
Ž
Ž
. .
2
by 31 P NMR dsy28.1 ppm , with respect to the
q
Ž
.
Ž
.
.
70.8 ppm. Mass spectrum: mrz 658.9, 661.0 M
Ž
corresponding phosphine oxide derivative ds26.3
.
ppm . Different attempts to purify it by further manipu-
2.2. Spectroscopic and kinetic experiments
lations failed.
Samples for spectroscopic and kinetic measurements
were always prepared under argon. The metal salt was
added in excess to a concentrated solution of acyl
2.1. Synthesis
5
[
(
4
)(
)( )(
2.1.1. Fe h -C5 H5 COMe CO Ph2 P–CH2-aza-15-
Ž
.
complex 4–6 and the mixture was sonicated for a few
minutes. Solutions were then filtered into a 0.1 mm
CaF2 cell and analyzed by FT-IR. If water was present
in the salt or solvent, a band appeared in the range
)] ( )
crown-5
Ž
.
Ž
Complex 1 0.78 g, 4.1 mmoles and L 1.35 g, 3.2
.
Ž
mmoles in a Schlenk tube, were stirred in hexane 10
.
ml at 708C, for 3 days. A dark brown solid precipitated
1615–1640 cmy1
.
out of the reaction mixture. Hexane was distilled off,
the residue was recovered with dichloromethane and
chromatographed. Excess starting material was eluted
first with dichloromethane, then a dark band was eluted
with increasing amount of tetrahydrofuran, up to a 1:6
ratio tetrahydrofuran:dichloromethane, to obtain 0.46 g
In a typical experiment for UV–Vis kinetic measure-
Ž
.
ments, solutions of L 2 ml were mixed with solutions
Ž
.
of complex 0.1–0.5 ml in 1 cm quartz cells, to obtain
initial concentrations of 2 or 3 of ca. 10y4. For the
experiments in the presence of metal salts, a weighted
amount was previously added to the solution of L or
Ž
.
23% yield of a thick brown oil. IR: 1923.8, 1610.7
Ž
directly into the cell e.g., 93 mg of NaI in 2.4 ml of
1
cmy1 cyclohexane . H NMR: d 7.4–7.3 m, 10 H,
Ž
.
Ž
.
acetonitrile to give a clear solution 0.26 M . For more
accurate comparisons of rate effects, measurements 3–
.
.
Ž
.
Ž
phenyl , 5.28 s, 5 H, C5H5 , 3.7 y3.3 m, 16 H,
Ž
Ž
.
OCH2 , 3.45 d, 2H, JPyH s4.4 Hz, N-CH2-P , 2.52
.
4
were performed simultaneously in the instrument
Ž
Ž
. .
Ž
.
tr, 4 H, Js5.9 Hz, N- CH2– , 2.47 s, 3 H, COMe
2
from the same batch of L, and grouped together in the
tables. The increase in absorbance associated with prod-
uct formation was followed with time, in the range
380–400 nm.
ppm. 31 P NMR: d 73.4 ppm. Mass spectrum: mrz
q
Ž
.
.
608.9, 611.2 M
5
[
(
)(
)( )(
2.1.2. Fe h -C5 H5 COCHMe2 CO Ph2 P–CH2-aza-
Samples for kinetic measurements by FT-IR were
prepared by addition of L, metal salt and complex in a 2
ml flask, to obtain an initial concentration of 2 or 3 of
)] ( )
15-crown-5
5
Ž
.
Ž
Complex 2 0.56 g, 2.52 mmoles and L 1.03 g,
ca. 10y2 e.g. 137 mg of L, 60 mg of NaPF6 in 2ml of
.
Ž
2.47 mmoles were dissolved in a Schlenk tube in
.
toluene, and heated under stirring at 508C, for 40 h. The
crude reaction mixture was chromatographed, by elution
with hexane first and then with dichloromethane, to
acetonitrile . In chloroform, the metal salt was added to
a solution of L, the mixture was sonicated and filtered
into the reaction flask. A 0.5 mm CaF2 cell fitted with a
thermostatted jacket was filled with freshly prepared
reaction mixture, under argon, sealed and placed in the
sample compartment of the spectrophotometer. The re-
action progress was followed by monitoring the decay
Ž
.
yield 0.71 g of product yield 45% , as a brown oil. IR:
1
1919.2, 1604.6 cmy1 cyclohexane . H NMR: d 7.5–
Ž
.
.
Ž
Ž
.
Ž
7.3 m, 10 H, phenyl , 5.28 s, 5 H, C5H5 , 3.6 y3.4
Ž
.
m, 18 H, OCH2 and N-CH2–P , 2.91 septet, 1H,
.
Ž
Ž
Ž
.
Js3.6 Hz, COC H Me2 , 2.56 t, 4 H, Js5.8 Hz,
of the intense n CO bands of the of 2 and 3, and the
growth of the carbonyl band of 5 and 6, respectively.
Ž
. .
.
N- CH2– , 0.97–0.64 d, 6 H, COCH Me2 ppm.
2

(
)
334
V. Berdini et al.rJournal of Organometallic Chemistry 551 1998 331–338
Table 1
First-order rate constants, kobs, were obtained by fitting
the first order rate equation to the experimental data
using a non linear least squares regression program.
Runs were carried out under pseudo-first order condi-
tions, using a large excess of L with respect to complex
2 or 3. The runs were duplicated and were reproducible
to within 10%. Slightly larger deviations were observed
in measurements obtained from different batches of L.
The actual concentration of L was corrected from the
phosphine oxide by-product, determined by 31 P NMR.
We previously checked that the presence of phosphine
oxide does not affect the reactivity of the iron com-
y1
Ž
.
Terminal carbonyl and acyl stretching frequencies n, "0.2 cm
of the cyclopentadienyl complexes Fe h -C5H5 COR CO L 4,
RsMe; 5, RsCHMe2 in the presence of metal salts
5
w
Ž
.Ž
.Ž .Ž .x Ž
.
Ž
.
Ž
.
Salt
n CO
n COR
5
[
(
)(
)( )( )]
Complex Fe h -C5 H5 COMe CO L
Dichloromethane
4
none
NaPF6
CaI2
none
NaPF6
CaI2
none
NaPF6
CaI2
none
NaPF6
1917.5
1931.9
1925.9
1916.8
1930.6
1926.3
1916.9
1928.2
1927.5
1917.7
1941.3
1596
1565
1515
1591
1561
1515
1598
1567
1514
1605
y
Chloroform
1,2-dichloroethane
w
x
plexes toward phosphines 2–5 .
Tetrahydrofuran
3. Results and discussion
5
[
(
)(
)( )( )]
Complex Fe h -C5 H5 COCHMe2 CO L
5
Dichloromethane
none
1910.2
1926.9
1910.8, 1959sh
1917.5, 1950sh
1909.0
1594
1589
1595
1603
1595
3.1. Reactions
NaPF6
CaBr2
CaI2
Complexes 1–3 react with the aza–crown ether phos-
Acetonitrile
none
w
Ž
.x Ž .
phine Ph2 P–CH2- aza-15-crown-5 L to give the
Ž
.
products of alkyl migratory insertion Scheme 2 .
The reaction of the cyclopentadienyl complex
5
w
Ž
.Ž
.
x Ž .
Fe h -C5H5 CO Me 1 with L was reported previ-
2
3.2. Metal ion complexation
oulsy, although experimental details regarding the
preparation of the acetyl product were not provided
The complexation of alkali or alkaline-earth metal
salts to the crown ether functionalized acyl complexes
has been studied by FT-IR spectroscopy. Infrared data
for the cyclopentadienyl complexes 4, 5 and for the
indenyl complex 6 are reported in Tables 1 and 2,
respectively. When dichloromethane solutions of acyl
w
x
15 . The complex is extremely reluctant toward reac-
tion with L, under the conditions described for the
w
x
reaction of 1 with phosphines and phosphites 24 . In
fact, decomposition of starting materials was observed
after prolonged reaction times in refluxing polar sol-
vents. After several attempts, we found that the reaction
proceeded in boiling hexane, from which the acyl prod-
uct 4 separated out as a dark brown oil. Since these
conditions were not suitable for a kinetic study, we have
searched for structural changes which would raise the
reactivity of the iron complex. This rate effect can be
obtained by increasing the steric hyndrance of the mi-
grating alkyl group or by changing the nature of the
pentahapto ligand. In fact, complexes 2 and 3 reacted
smoothly with L in different solvents and moderate
temperatures, in agreement with the higher migratory
aptitude of the isopropyl group, with respect to methyl
Table 2
"
Terminal carbonyl and acyl stretching frequencies n, 0.2 cmy1 of
Ž
.
5
w
Ž
.Ž
.Ž .Ž .x Ž .
the indenyl complex Fe h -C9 H7 COMe CO L 6 in the pres-
ence of metal salts
Ž
.
Ž
.
Solvent
Dichloromethane none
LiClO4
Salt
n CO
n COMe
1919.3
1945.8, 1932.6 1585
1945.2, 1938.3 1589
1604
LiI
NaI
1924.1
1924.5
1941.8
1940.6
1941.4
1927.3
1942.4
1919.6, 2066a
1924.3
1942.7
1919.2
1919.5
1940.8
1917.5
1944.6
1572
1574, 1606 m
1558
1604
1570, 1528
1570
1520
1605
1600
1559
NaPF6
KPF6
NH4 PF6
w
x
Ž .
4,25 , and the higher reactivity of iron II indenyl
complexes with respect to their cyclopentadienyl analogs
Ž
.
Mg ClO4
2
CaI2
CaI2
w
x
5,12 . Before studying the kinetic behaviour of the iron
complexes 2 and 3, we have checked the coordination
properties of the acetyl complex 4 toward cations in
different solvents, and studied those of the new com-
plexes 5 and 6.
Ž
.
Ba SCN
2
Chloroform
none
NaPF6
LaCl3
none
NaPF6
none
NaPF6
KPF6
CaI2
1599
1612
y
1606
Tetrahydrofuran
Acetonitrile
1589
1915.1, 1938sh 1592sh
1915.2, 1927
y
a
Ž
.
Scheme 2.
n SCN .

(
)
V. Berdini et al.rJournal of Organometallic Chemistry 551 1998 331–338
335
complex are stirred in the presence of solid metal salts,
the iron complex may extract the salt by complexation
of the metal ion in the cavity of the aza–crown ether.
This information is obtained by an infrared analysis of
the resulting filtered solution, by comparing the values
of the stretching frequencies of the terminal carbonyl
n CO and of the acyl carbonyl n COR groups with
those obtained in the absence of metal salt, and accord-
ing to the following criteria. When the aza–crown ether
coordinates the metal ion, the terminal carbonyl group
exhibits a shift to higher frequencies, suggesting intra-
molecular electronic interaction between crown-com-
salts dissolve easily but band shifts are not observed as
well, except for shoulders, indicating competition be-
tween solvation and complexation.
5
[
(
)(
)( )( )] (
)
3.2.2. Fe h -C5 H5 COCHMe2 CO L 5, Table 1
In dichloromethane solutions, a consistent high fre-
Ž
.
Ž
.
Ž
.
quency shift is observed for the n CO band in the
y1
Ž
.
Ž
presence of NaPF6 q16 cm
and of CaI2 q7
cmy1, q40 cmy1 as a strong shoulder . On the other
hand, significant changes in the frequencies of the acyl
carbonyl group are not observed. It appears that com-
plex 5 coordinates metal ions in the aza–crown ether
cavity, but that intramolecular oxygen–metal ion inter-
action of the acetyl group does not occur. This is likely
to depend on the bulkiness of the isopropyl group which
hinders a suitable molecular conformation for the ex-
pected interaction.
.
w
x
plexed metal ions and the iron centre 22 , as the result
of reduced electron density on iron and hence reduced
p-back bonding toward CO. The stretching of the acyl
Ž
.
group n COR is either not affected by complexation or
strongly shifted to lower frequencies when the acyl
oxygen binds the coordinated metal ion, as depicted in
Fig. 1. We have checked this interpretation by blank
5
[
(
)( )(
)( )] (
)
3.2.3. Fe h -C9 H7 CO COMe L 6, Table 2
experiments. For instance, a dichlomethane solution of
The infrared spectra of complex 6 have been ob-
tained in the presence of different alkali and alkaline-
earth metal salts. Changes in the infrared frequencies of
both terminal and acetyl carbonyl groups suggest struc-
tural arrangements similar to those of the cyclopentadi-
5
w
Ž
.Ž .Ž
.Ž
.x
the complex Fe h -C9 H7 CO COMe PMePh2 was
stirred in the presence of NaI, and significant changes in
the carbonyl stretching bands were not observed; the
5
w
Ž
corresponding aza – crown complex Fe h -
.Ž .Ž
.Ž .
x
Ž
.
Ž
.
C9 H7 CO COMe L showed a n CO shift of q4.3
enyl -complex 4 Fig. 2 .
cmy1 and a lower frequency n COMe shift of y32
Ž
.
Ž
.
The higher energy shift of the n CO band is large
cmy1, after addition of NaI.
upon complexation of Liq, Kq, Mg2q and Ca2q up to
Ž
q26 cmy1 for Liq in dichloromethane. The ammo-
.
5
nium ion NHq4 yields a shift as well q21 cm , in
y1
[
(
)(
)( )( )] (
)
Ž
.
3.2.1. Fe h -C5 H5 COMe CO L 4, Table 1
Our study was limited to sodium and calcium salts.
Infrared spectra in the presence of litium and magne-
sium salts as well were reported by McLain, in
agreement with the ability of the cation to fit into the
q
w
x
aza-15-crown-5 ring 26 . For Na , we have observed
larger shifts in chloroform or tetrahydrofuran than in
dichloromethane. Complexation of Ba2q is observed
w
x
dichloromethane 15 . Our data are in good agreement
with the reported values, except that we did not observe
such a large high frequency shift of the terminal car-
y1
Ž
.
only as a high frequency shoulder ;1940 cm
the band due to uncomplexed 6 1919.6 cm , present
of
y1
Ž
.
Ž
bonyl band in the presence of CaI2 1952 vs. 1917
as the main species. The fact that the lanthanum salt
cmy1 15 but a rather smaller change 1926 vs. 1917.5
.
w
x
Ž
w
Ž
. x
LaCl3 P H2O did not induce any band shift may be
x
cmy1 . However, we do observe a consistent shoulder
.
due to the effect of the strong lattice energy of lan-
thanum chloride, preventing complexation by the aza–
at higher frequency of the main band, both in
dichloromethane and in chloroform, which can be at-
tributed to the species characterized by the 1952 cmy1
absorbance. We have also observed the strong shift to
Ž
.
crown ether system. Shifts of the acetyl n COMe band
are large for Naq, Kq, Ca2q and Mg2q. The lack of
acetyl shift in the presence of the ammonium salt
Ž
.
lower frequencies of the n COMe band, of ca. y30
cmy1 for Naq and ca. y80 cmy1 for Caq, indicative
of intramolecular coordination of the acetyl oxygen to
the crown ether held cation. These data regard chlori-
nated solvents. An intermediate situation is observed in
Ž
.
tetrahydrofuran. The n CO band shows a large high
frequency shift of q24 cmy1 upon complexation of the
sodium cation in the crown ether, whereas the absorb-
Ž
.
tion of the n COMe band appears disturbed by mois-
ture, which probably hinders coordination of the acetyl
Ž
oxygen. In the apolar solvent cyclohexane 1923.8,
1611 cmy1 , the salt cannot be extracted by the acetyl
.
complex and band shifts are not observed. In the very
polar solvent 1,4 dioxane 1915.2, 1603 cm
Fig. 2. Intramolecular acetyl oxygen–cation interaction in the com-
y1
5
Ž
.
metal
w
Ž
.Ž
.Ž .Ž .x Ž .
plexation of metal salts to Fe h -C9 H7 COMe CO L 6 .

(
)
336
V. Berdini et al.rJournal of Organometallic Chemistry 551 1998 331–338
represents a clean example of discrimination between
aza–crown ether and acetyl oxygen complexation to the
added cation. In the polar solvent acetonitrile, coordina-
tion of the metal cation in the aza–crown ring is
Ž
.
indicated by the shift of the n CO band, as in the
presence of NaPF6, although shifts of the acetyl band
Ž
.
are small or absent. Fig. 3 shows the shift of the n CO
band of complex 6 in the presence of NaPF6 and KPF6.
Ž
.
The existence of shoulders of the main n CO band or
the values observed in different sample preparations
Ž
.
entries 9, 10 in Table 2 indicate the presence of both
coordinated and uncoordinated species or different com-
plex conformations in solution.
In summary, the changes of acetyl stretching fre-
quencies vary up to y80 cmy1 for the cyclopentadienyl
complex 4 and up to y84 cmy1 for the indenyl com-
plex 6 in the presence of CaI2. The changes of the
terminal carbonyl stretching frequencies vary up to
y1
Ž
. Ž
versus concentration of Ph2 P–CH2- aza-
Fig. 4. Plot of kobs
s
5
. Ž .
w
Ž
.Ž
.
x
15-crown-5 L for the reaction of complex Fe h -C9 H7 CO 2 Me
3 , in acetonitrile at 408C.
Ž .
q40 cmy1 for complex 4 upon interaction with MgI2
y1
w
x
in dichloromethane 15 , and up to 26 cm
for com-
3.3. Kinetic measurements
plex 6 in the presence of LiClO4. The shifts of the CO
group are indicative of effective electronic interactions
between the two metal centres, and are expecially sig-
nificant when compared to those in other literature
Since the acetyl complex 4 formed only in refluxing
hexane under heterogeneous conditions, rate measure-
ments of the reaction of complex 1 with L were not
attempted. The kinetic features of the reactions of the
cyclopentadienyl isopropyl complex 2 and of the in-
denyl methyl complex 3 with the aza–crown ether
phosphine L were first studied in the absence of metal
ions Scheme 2 . The progress of the reaction was
followed by UV–Vis spectroscopy, monitoring the
growth of absorbance, due to product formation, around
Ž
.
reports. For instance, Cr CO -substituted benzocrown
3
ethers used as FT-IR-readable molecular sensors for
y1
Ž
.
metal ions display shifts of n CO up to 7 cm after
w
x
complexation of group I metal ions 27,28 . The cy-
mantrene complex Mn CO h -2-indenyl-aza-15-
crown-5 shows a 2 cm shift upon complexation with
5
Ž
. Ž
3
Ž
.
y1
.
LiClO4, and only protonation in strong acidic media of
the crown ether nitrogen atom produces a desired change
w
x
400 nm 2–5 , or by FT-IR. The phosphine, L, was used
in large excess with respect to the iron complexes, in
order to ensure pseudo-first order conditions. In order to
y1
Ž
.
w x
of n CO by 36 cm to higher frequency 22 .
determine the rate dependence on L, rate measurements
5
w
Ž
for the reaction of the indenyl methyl complex Fe h -
.Ž
.
x
C9 H7 CO Me were obtained at different concentra-
tions of L. A plot of first order observed rate constants,
2
y1
Ž
. w x
, versus L is shown in Fig. 4 and values of
kobs
s
kobs are reported in Table 3.
Table 3
Values of observed rate constants, kobs, for the reaction of complex
5
w
Ž
.Ž
.
x Ž .
Fe h -C9 H7 CO 2 Me
3 with the aza-crown phosphine Ph2 P-
Ž
. Ž .
CH2- aza-15-crown-5 L , and with PPh3 in acetonitrile, at 408C, by
UV–Visa
Ž
.
w x
L , M
Salt M
kobs, secy1
None
None
None
None
0.41
0.25
0.15
0.09
3.9=10y5
2.4=10y5
1.4=10y5
0.9=10y5
w
x
PPh3 , M
None
0.08
0.08
0.08
2.8=10y5
3.5=10y5
2.1=10y5
Ž
.
NaI 0.14
5
w
Ž
.Ž
.Ž .Ž .x
Fig. 3. Infrared spectra of complex Fe h -C9 H7 COMe CO L
6
Ž
.
CaI2 0.11
Ž .
Ž .
Ž .
a , complex 6PKPF6 b , and complex 6PNaPF6 c in acetonitrile, in
the region of the terminal carbonyl stretching band n CO .
aAbsorbance growth at 400 nm.
Ž
.

(
)
V. Berdini et al.rJournal of Organometallic Chemistry 551 1998 331–338
337
Table 4
The linear dependence of rate versus phosphine con-
centration indicates a clean second order behaviour,
which implies that the phosphine L is part of the rate
limiting step. We checked whether metal salts may
change the reactivity of the iron complexes by effects
not related to complexation in the aza–crown ether
system, and did not observe significant rate effects for
the reaction of complex 3 with triphenylphosphine in
acetonitrile upon addition of NaI or CaI2 Table 3 .
Rate measurements of the reaction with L were
obtained in the presence of metal salts, and carried out
by FT-IR spectroscopy, in acetonitrile and in chloro-
form. Due to the high solubility of metal salts in
acetonitrile, this solvent ensures the presence of
equimolar amounts of L and of the salt, and the exis-
tence of 1:1 aggregates, due to the high affinity of
Values of observed rate constants, kobs, for the reaction of complexes
5
5
w
Ž
.Ž
.
x Ž .
w
Ž
.Ž x Ž .
.
Fe h -C5H5 CO 2CHMe2
2
and Fe h -C9 H7 CO 2 Me 3
Ž
. Ž .
L
with the aza-crown phosphine Ph2 P–CH2- aza-15-crown-5
Method
T, 8C Solvent
Salt M
L , M kobs, secy1
Ž
.
w x
Complex 2
FT-IRa
30
30
40
40
acetonitrile none
acetonitrile CaI2 0.15
acetonitrile none
0.15
0.11
0.16
5.6=10y5
5.2=10y5
1.5=10y4
1.7=10y4
FT-IRa
Ž
.
FT-IRa
FT-IRa
acetonitrile NaPF6 0.17 0.16
Ž
.
Ž
.
Complex 3
UV–Visb 40
UV–Visb 40
acetonitrile none
acetonitrile NaI 0.26
acetonitrile CaI2 0.19
0.10
0.10
0.13
0.18
0.21
0.24
0.6=10y5
1.6=10y5
1.5=10y5
0.9=10y5
1.5=10y5
1.6=10y5
Ž
.
FT-IRc
FT-IRd
FT-IRd
FT-IRd
40
40
40
40
Ž
.
chloroform none
chloroform NaPF6e
chloroform CaIe2
a
b
c
w
x
Growth of the n CO band of 5 at 1912 cmy1 ; 400 nm; decay of
monoaza-15-crown-5 for the group I and II cations 26 .
On the other hand, the infrared spectra in chloroform
showed shifts of both carbonyl and acyl stretching
bands as the result of aggregation, and acylic coordina-
tion, between complexes 5, 6 and metal cations. The
presence of NaPF6 in deuterated chloroform solution of
phosphine L was also observed in the 31 P NMR spec-
trum, showing a septet signal due to the hexafluo-
Ž
.
d
the n CO band at 2002 cmy1
;
decay of the n CO band at 2005
Ž
.
.
e
cmy1 ; saturated solutions.
The progress of the reaction is characterized by the
Ž
.
disapperance of the n CO bands of the starting com-
y1
Ž
.
plex e.g.: 1996.5, 1938 cm , 2 in acetonitrile and by
y1
Ž
.
Ž
.
rophosphate counterion d y145 ppm, JPF s713 Hz .
The experiments were carried out in the presence of
sodium or calcium salts, since these metal ions exhib-
ited the strongest band shifts and were therefore ex-
pected to display rate effects, if any.
the growth of the terminal carbonyl band 1912 cm
of the product. A sequence of spectra for the reaction of
2 with L in the presence of NaPF6 is shown in Fig. 5.
The disappearance of the iron complexes 2 or 3 dis-
played a clean first order time dependence.
5
w
Ž
.Ž
.
x Ž .
Ž
.
Fig. 5. A series of infrared spectra recorded during the reaction of complex Fe h -C5H5 CO 2CHMe2 2 with L 0.16 M in acetonitrile, at
Ž
.
408C. The solution contained NaPF6 0.18 M . Bands are assigned to complexes 2 and 5 as indicated.

(
)
338
V. Berdini et al.rJournal of Organometallic Chemistry 551 1998 331–338
Values of kobs are reported in Table 4. The indenyl
Quımicas, Isla de la Cartuja, Sevilla, during the prepara-
´
complex 3 exhibits an acceleration up to three times in
the presence of sodium or calcium salts, both in aceto-
nitrile and in chloroform. In the latter solvent, where
aza–crown ether complexation and acetyl oxygen coor-
dination to the cation were observed, somewhat higher
rate enhancements were expected in the case of elec-
trophilic activation by the metal ion. These effects are
rather small, although indicative of overall favorable
induction by the metal ions. It is possible that when the
expected oxygen–metal ion interaction is present in the
transition state, it is then counterbalanced by other
effects coming into play. For instance, coordination of a
metal ion in the aza–crown ether cavity may reduce the
donor abilty of the phoshine by polarization. Further-
more, the reacting ion pair LMqPXy may introduce
more stringent steric requirements as well as more
extensive solvent reorganization during reaction, pro-
ducing counteracting effects on reactivity. The reactiv-
ity of the isopropyl complex 2 does not appear to be
affected by the presence of metal salts, in acetonitrile,
which is in line with the lack of interaction between
acyl oxygen and cations, as observed by FT-IR spec-
troscopy.
tion of this manuscript, and Dr. Anthony Haynes, Uni-
versity of Sheffield, for the infrared kinetic programs
used throughout these years.
References
w x
1
J.P. Collman, L.S. Hegedus, J.R. Norton, R.G. Finke, Principles
and Applications of Organotransition Metal Chemistry, Univer-
sity Science Books, Mill Valley, CA, 1987.
w x
Ž
.
2
w x
3
D. Monti, M. Bassetti, J. Am. Chem. Soc. 115 1993 4658.
L. Ambrosi, M. Bassetti, P. Buttiglieri, L. Mannina, D. Monti,
Ž
.
J. Organomet. Chem. 455 1993 167.
w x
4
Ž
.
M. Bassetti, L. Mannina, D. Monti, Organometallics 13 1994
3293.
w x
5
M. Allevi, M. Bassetti, C. Lo Sterzo, D. Monti, J. Chem. Soc.,
Ž
.
Dalton Trans., 1996 3527.
w x
6
M. Bassetti, G.J. Sunley, F.P. Fanizzi, P.M. Maitlis, J. Chem.
Ž
.
Soc., Dalton Trans., 1990 1799.
w x
7
D. Monti, M. Bassetti, G.J. Sunley, P. Ellis, P.M. Maitlis,
Ž
.
Organometallics 10 1991 4015.
w x
8
D. Monti, G. Frachey, M. Bassetti, A. Haynes, G.J. Sunley,
P.M. Maitlis, A. Cantoni, G. Bocelli, Inorg. Chim. Acta 240
Ž
.
1995 485.
w x
9
M. Bassetti, D. Monti, A. Haynes, J.M. Pearson, I.A. Stan-
bridge, P.M. Maitlis, Gazz. Chim. Ital. 122 1992 391.
Ž
.
w
w
x
10 S.B. Butts, S.H. Strauss, E.M. Holt, R.E. Stimson, N.W. Al-
cock, D.F. Shriver, J. Am. Chem. Soc. 102 1980 5093.
11 T.G. Richmond, F. Basolo, D.F. Shriver, Inorg. Chem. 21
Ž
.
x
4. Conclusions
Ž
.
1982 1272.
w
w
w
x
Ž
.
12 T.C. Forschner, A.R. Cutler, Organometallics 4 1985 4015.
13 S.J. La Croce, A.L. Cutler, J. Am. Chem. Soc. 104 1982 2312.
14 J.P. Collman, R.G. Finke, J.N. Cawse, J.I. Brauman, J. Am.
Chem. Soc. 100 1978 4766.
15 S.J. McLain, J. Am. Chem. Soc. 105 1983 6355.
16 R. Cacciapaglia, L. Mandolini, Chem. Soc. Rev. 1993 271.
17 G. Braca, A.M. Raspolli Galletti, G. Sbrana, Catalysis by Metal
Complexation of group I and II metal ions within the
x
Ž
.
cavity of the aza-15-crown-5 unity of indenyl and cy-
x
5
w
Ž
clopentadienyl acetyl com plexes
ligand COMe CO Ph2 P–CH2-aza-15-crown-5
Fe h -
Ž
.
.Ž .Ž .Ž
.
x Ž
4,
w
w
w
x
Ž
.
x
Ž
.
.
6 raises the stretching energy of the terminal carbonyl
group CO, as a result of electronic polarization of the
iron centre and consequent reduction of p-back bond-
ing. Shifts to lower energy of the COMe stretching band
indicate intramolecular coordination of the acetyl oxy-
gen to the complexed cation. Such intramolecular con-
tact is hindered by changing methyl for isopropyl in the
x
Complexes Vol. 16, Oxygenates by Homologation or CO Hy-
Ž
.
drogenation by Metal Complexes, G. Braca Ed. , Kluwer Aca-
demic Publisher, Dordrecht 1994, page 176.
18 J.A. Labinger, J.N. Bonfiglio, D.L. Grimmet, S.T. Masuo, E.
w
w
w
w
x
Ž
.
Shearin, J.S. Miller, Organometallics 2 1983 733.
x
Ž
.
19 J. Powell, A. Lough, F. Wang, Organometallics 11 1992 2289,
and references therein.
20 S.J. McLain, F.J. Waller, U.S. patent 4, 1984 , 432, 904;
Ž .
acyl group 5 , or in polar aprotic solvents. The reaction
x
Ž
.
5
w
Ž
.Ž
.
x
of the indenyl complex Fe h -C9 H7 CO Me with
Ž
.
Chem. Abstr. 100, 1984 210158a.
2
.
Ž
x
the phosphine Ph2 P–CH2- aza-15-crown-5 is only
moderately affected by the presence of metal salts in
21 F.C.J.M. van Veggel, W. Verboom, D.N. Reinhoudt, Chem.
Ž
.
Rev. 94 1994 279.
w
w
w
w
x
Ž
.
22 H. Plenio, D. Burth, Organometallics 15 1996 1151.
23 S.J. McLain, Inorg. Chem. 25 1986 3124.
24 J.P. Bibler, A. Wojcicki, Inorg. Chem. 5 1966 889.
solution, whereas that of the cyclopentadienyl complex
x
Ž
.
5
w
Ž
.Ž
.
x
Fe h -C5H5 CO CHMe2 does not show any salt
2
x
Ž
.
dependence.
x
Ž
.
25 J.D. Cotton, G.T. Crisp, L. Latif, Inorg. Chim. Acta 47 1981
171.
w
w
w
x
26 R.M. Izatt, K. Pawlak, J.S. Bradshaw, R.L. Bruening, Chem.
Ž
.
Rev. 91 1991 1721.
Acknowledgements
x
27 C.E. Anson, C.S. Creaser, G.R. Stephenson, J. Chem. Soc.,
Ž
.
Chem. Comm., 1994 2175.
M.B. wishes to thank Dr. Margarita Paneque for her
kind hospitality at the Instituto de Investigaciones
x
28 M. Salmain, A. Vessieres, G. Jaouen, I.S. Butler, Anal. Chem.
Ž
.
63 1991 2323.