Trans-influences in mononuclear cyclopalladated compounds containing a σ(Pd-Csp2, ferrocene) bond. X-ray crystal structures of [Pd{[(η5-C5H3)-CH=N-CH2-C 6H5]Fe(η5-C5H 5)}(X)(PPh3)] with X-=Br- and I-

Article abstract of DOI:10.1016/S0022-328X(00)00900-1

The syntheses, characterization and study of the properties of cyclopalladated compounds of general formula: [Pd{[(η⁵-C₅H₃)-CH=N-CH₂-C ₆H₅]Fe(η⁵-C₅H ₅)}(X)(PPh₃)] with X^-=Cl^-, Br^-, I^-, CN^-, SCN^- or AcO^- are reported. Compounds with X^-=Br^- and I^- were characterized by X-ray diffraction, and their crystal structures revealed the existence of a weak C-HX interaction involving one of the -CH₂- type hydrogen atoms bound to the imine nitrogen. Comparison of the spectroscopic properties of these compounds has allowed us to elucidate the influence of the monoanionic ligand, X^-, on a trans-arrangement to the metallated carbon atom upon the environment of the iron(II). Theoretical studies based on the semiempirical PM3(tm) were also been carried out to clarify the effects produced by the X^- ligand on the ring current of the metallated C₅H₃ ring of the ferrocenyl fragment.

Full text of DOI:10.1016/S0022-328X(00)00900-1

Journal of Organometallic Chemistry 625 (2001) 67–76
www.elsevier.nl/locate/jorganchem
Trans-influences in mononuclear cyclopalladated compounds
containing a s(Pd–C , ferrocene) bond. X-ray crystal structures
2
sp
of [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(X)(PPh3)]
with X−=Br− and I−
Sonia Pe´rez a, Ramon Bosque a, Concepcio´n Lo´pez a,*, Xavier Solans b,
Merce´ Font-Bardia b
a Departament de Quimica Inorga´nica, Facultat de Quimica, Uni6ersitat de Barcelona, Mart´ı i Franque`s 1-11, E-08028 Barcelona, Spain
b Departament de Cristal lografia, Mineralogia i Dipo`sits Minerals, Facultat de Geologia, Uni6ersitat de Barcelona, Mart´ı i Franque`s s/n,
E-08028 Barcelona, Spain
Received 2 October 2000; received in revised form 4 November 2000; accepted 4 November 2000
Abstract
The syntheses, characterization and study of the properties of cyclopalladated compounds of general formula: [Pd{[(h5-C5H3)–
CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(X)(PPh3)] with X−=Cl−, Br−, I−, CN−, SCN− or AcO− are reported. Compounds with
X
−=Br− and I− were characterized by X-ray diffraction, and their crystal structures revealed the existence of a weak C–H···X
interaction involving one of the –CH2– type hydrogen atoms bound to the imine nitrogen. Comparison of the spectroscopic
properties of these compounds has allowed us to elucidate the influence of the monoanionic ligand, X−, on a trans-arrangement
to the metallated carbon atom upon the environment of the iron(II). Theoretical studies based on the semiempirical PM3(tm) were
also been carried out to clarify the effects produced by the X− ligand on the ring current of the metallated C5H3 ring of the
ferrocenyl fragment. © 2001 Elsevier Science B.V. All rights reserved.
Keywords: Trans-influence; Cyclopalladates; Semiempirical theory; Ferrocene
1. Introduction
transition in the visible-ultraviolet spectrum and a de-
crease of the half-wave potential [E1/2(Fc)] of the ferro-
cenyl fragment, i.e. the iron becomes more prone to
oxidation [6].
Previous studies on cyclopalladated complexes con-
taining N-donor ferrocenyl moieties of general formula:
[Pd{[(h5-C5H3)–C(R)ꢀN–R%]Fe(h5–C5H5)}Cl(L)] (with
R=H, Me or Ph and R%=phenyl or benzyl groups,
and L=phosphine ligand) (Fig. 1) [1,2] have shown the
effect of the electronic properties of the neutral L group
upon the proclivity of the ferrrocenyl fragment to un-
dergo oxidation [3–6]. In particular, the replacement of
the PPh3 in: [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-
C5H5)}Cl(PPh3)] by a more basic phosphine ligand such
as PEt3 [7] causes a decrease of the energy of the d–d
* Corresponding author. Tel.: +34-93-4021274; fax: +34-93-
4907725.
E-mail address: conchi.lopez@qi.ub.es (C. Lo´pez).
Fig. 1. Schematic view of compounds under study.
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(00)00900-1

68
S. Pe´rez et al. / Journal of Organometallic Chemistry 625 (2001) 67–76
I−, CN− or SCN−} to the filtrate, and the subsequent
purification of the compound by SiO2-column chro-
matography (Scheme 1).
Table 1
Electronic and steric parameters of the monoanionic ligands, X a
X
sI
sR
ES–CH
This procedure allowed us to isolate compounds
[Pd{[(h5 -C5H3)–CHꢀN–CH2–C6H5]Fe(h5 -C5H5)}(X)-
(PPh3)], with X−=Br−, CN− and SCN− in fairly
reasonable yields, but [Pd{[(h5-C5H3)–CHꢀN–CH2–
C6H5]Fe(h5-C5H5)}(I)(PPh3)] was obtained in low yield
(ca. 15%). Attempts to prepare [Pd{[(h5-C5H3)–
Cl
Br
I
CN
SCN
AcO
0.51
0.50
0.44
0.61
0.55
0.44
−0.23
−0.23
−0.22
0.09
−0.05
−0.24
0.55
0.65
0.78
0.40
0.50
CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(AcO)(PPh3)]
using
a sI and sR are inductive (para) and mesomeric (para) values for
the X groups and the ES–CH values are Charton’s steric parameters
calculated according to structural data. All the parameters presented
in this table were obtained from Ref. [8].
this procedure failed.
To improve the yield of the preparation of [Pd{[(h5-
C5H3)–CHꢀN–CH2-C6H5]Fe(h5–C5H5)}(I)(PPh3)], an
alternative strategy was adopted. This consisted in the
conversion of [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe-
(h5-C5H5)}(m-Cl)]2 into the corresponding di-m-iodo-
bridged dinuclear complex: [Pd{[(h5-C5H3)–CHꢀN–
CH2–C6H5]Fe(h5-C5H5)}(m-I)]2, and its subsequent
treatment with PPh3 in a 1:2 molar ratio (Scheme 2).
[Pd{[(h5 - C5H3)–CHꢀN–CH2–C6H5]Fe(h5 - C5H5)}-
(AcO)(PPh3)] was synthesised successfully with a two-
step sequence of reactions (Scheme 3). The first step
was reaction of the ferrocenylimine [(h5-C5H5)Fe{(h5-
C5H3)–CHꢀN–CH2–C6H5}] [1] with the equimolar
amounts of Pd(AcO)2 and NaAcO·3H2O, which yielded
[Pd{[(h5 - C5H3)–CHꢀN–CH2–C6H5]Fe(h5 - C5H5)}-
(m-AcO)]2. This reaction involved the activation of the
s(Csp2, ferrocene–H) bond. In the second step, the
addition of PPh3 to a suspension of dimeric complex in
acetone, produced the cleavage of the ‘Pd(m-AcO)2Pd’
However, the influence of the nature of the monoan-
ionic ligand (X−), in a trans-arrangement to the metal-
lated carbon atom of the ferrocenyl fragment, on the
spectroscopic and structural properties of this sort of
compound, has not yet been clarified. On this basis we
decided to compare the properties of [Pd{[(h5-C5H3)–
CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(Cl)(PPh3)] reported
previously [1], with properties of the new derivatives:
[Pd{[(h5 -C5H3)–CHꢀN–CH2–C6H5]Fe(h5 -C5H5)}(X)-
(PPh3)], containing X−=Br−, I−, CN−, SCN− or
AcO−, which differ exclusively in the electronic and
steric properties (Table 1) [8] of the monoanionic ligand
X− bound to the palladium.
2. Results and discussion
units
to
give:
[Pd{[(h5-C5H3)–CHꢀN–CH2–
C6H5]Fe(h5-C5H5)}(AcO)(PPh3)].
2.1. Syntheses of the compounds
2.2. Characterization
The new monomeric complexes [Pd{[(h5-C5H3)–
CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(X)(PPh3)], with X−
=Br−, I−, CN− and SCN− were obtained by a ligand
exchange procedure which consisted in the reaction of
[Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(Cl)-
(PPh3)] [1,2] with the stoichiometric amount of AgNO3
in acetone to produce the precipitation of AgCl, fol-
lowed by the addition of the stoichiometric amount of
the corresponding potassium salt, KX {with X−=Br−,
All the compounds prepared in this study were or-
ange or red solids at room temperature. The
monomeric compounds are highly soluble in CH2Cl2,
CHCl3, benzene and acetone, but they have lower
solubility in alcohols and alkanes. The di-m-iodo- and
the di-m-acetato-bridged cyclopalladated complexes
have low solubility in the most common solvents. All
the compounds were characterized by elemental analy-
Scheme 1.

S. Pe´rez et al. / Journal of Organometallic Chemistry 625 (2001) 67–76
69
Scheme 2.
Scheme 3.
ses and in all cases the experimental data confirmed the
proposed formula (see Section 3). The most outstand-
ing features observed in the infrared spectra of these
compounds is the presence of a sharp and intense band
in the range 1570–1595 cm−1, which is assigned to the
asymmetric stretching of the ꢁCꢀN– group. This ab-
sorption appears at lower wavenumbers than in the free
ferrocenylaldimine [w(ꢁCꢀN–)=1625 cm−1] [1,2] due
to the coordination of the ligand. In all cases, the
typical bands of the coordinated triphenylphosphine [9]
were also observed and the IR spectra of compounds
[Pd{[(h5 -C5H3)–CHꢀN–CH2–C6H5]Fe(h5 -C5H5)}(X)-
(PPh3)] with X−=CN− or SCN− also showed the
bands due to these functional groups (at ca. 2091 and
2145 cm−1, respectively). The IR spectra of: [Pd{[(h5-
C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(AcO)(PPh3)]
showed wasym. (CO2) and wsym. (CO2) separated by 278
cm−1, which is consistent with unidentate acetate coor-
dination [10]. The mononuclear compounds were char-
Table 2
Selected 1H-NMR data (in ppm) for compounds for [Pd{[(h5-C5H3)–
CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(X)(PPh3)] (labelling of the atoms
refers to those shown in Schemes 1–3)
a
X
C5H5
3.67
3.65
3.67
3.80
3.72
3.69
H3
H4
H5
–CH2–
–CHꢀN– b
7.90
Cl
3.31
3.36
3.42
3.52
3.29
3.45
4.02
3.99
3.99
4.09
4.06
4.02
4.29
4.27
4.27
4.33
4.31
4.25
4.63
5.41
4.77
5.50
5.03
5.67
4.57
4.73
4.64
5.00
4.51
4.82
Br
7.93
I c
7.07
CN
SCN c
AcO
7.86
7.93
d
1
acterized by H and and 13C{1H}-NMR spectroscopy
a The two protons of this group are diastereotopic.
(Tables 2 and 3). The signals were assigned with the aid
of two-dimensional heteronuclear {1H-13C}-NMR ex-
periments. In all cases, the number of signals observed
b Doublet due to phosphorus coupling, 3J(P–H) in the range
7.5–8.0 Hz.
c Additional singlet at ca. 5.12 ppm due to the CH2Cl2 of crystal-
1
in the H-NMR spectra and their multiplicities (Table
lization.
d Overlapped by the multiplet due to the aromatic protons of the
PPh3 ligand.
2) were consistent with the patterns expected for 1,2-
disubstituted ferrocenyl ligands having a s(Pd–Csp
)
2

70
S. Pe´rez et al. / Journal of Organometallic Chemistry 625 (2001) 67–76
Table 3
Selected 13C{1H}-NMR data (in ppm) for compounds [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(X)(PPh3)] (labelling of the atoms refers
those shown in Schemes 1–3)
X
13C-NMR
C5H5
31P-NMR
C1
C2
C3
C4
C5
–CH2–
–CHꢀN–
l(31P)
Cl
Br
I
CN
SCN
AcO
69.80
69.82
69.93
70.20
69.83
69.79
87.57
87.68
87.72
86.91
87.01
87.41
104.18
107.03
108.72
96.90
97.20
98.00
77.38
77.40
76.54
77.20
77.30
77.40
69.35
69.14
68.94
77.34
77.14
77.55
66.48
66.34
66.23
66.77
66.93
66.36
59.10
62.90
64.25
58.70
61.32
58.98
172.38
172.52
172.88
171.82
172.05
172.12
37.54
37.75
37.93
34.93
35.78
36.47
a
a An additional signal at 23.42 ppm due to the CH3 group of the acetato ligand was observed in the 13C-NMR spectrum of this compound.
bond [1–4]. The resonance of the N–CH2– protons
appeared as a doublet of doublets due to diastereopic-
ity. The chemical shifts and the separation of these
signals were strongly affected by the nature of the
ligand X.
2.3. Description of the crystal structures of
[Pd{[(p5-C5H3)–CHꢀN–CH2–C6H5]Fe(p5-C5H5)}(X)-
(PPh3)] with X−=Br− or I−
Perspective drawings of the molecular structures of
[Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(Br)-
(PPh3)] and [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-
C5H5)}(I)(PPh3)], together with their atom labelling
schemes are depicted in Figs. 2 and 3. A selection of
bond lengths and angles is shown in Table 4.
These structures consist of discrete molecules of
[Pd{[(h5 -C5H3)–CHꢀN–CH2–C6H5]Fe(h5 -C5H5)}(X)-
(PPh3)] with X−=Br− or I−, separated by van der
Waals contacts. In each molecule the palladium is in a
slightly distorted square-planar environment coordi-
nated with the bromo or iodo ligands, the phosphorus
of the PPh3 ligand, the nitrogen and the C10 atom (for
X−=Br−) or the C6 atom (for X− =I−) of the
ferrocenyl group. The PPh3 ligand and the imine nitro-
gen are in a trans-arrangement in close accord with the
results obtained by 31P-NMR spectroscopy.
Comparison of 13C{1H}-NMR data revealed that the
position of the signal due to the metallated carbon
atom, C2, is strongly sensitive (ca. 12 ppm) to the
nature of the monoanionic group (X−) on a trans-ar-
rangement (Table 3). The chemical shifts of this signal
increased according to the sequence:
CN− BSCN− BAcO− BCl− BBr− BI−
(1)
Similar trends were also found for the resonances of
the imine carbon and of C1, but in these cases the
variations for all the compounds were smaller.
31P-NMR spectra of compounds [Pd{[(h5-C5H3)–
CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(X)(PPh3)] showed a
singlet in the range 30–40 ppm (Table 3). This is
consistent, according to the literature [1,2,6,11] with a
trans-arrangement of the phosphine ligand and the
nitrogen of the ferrocenyl Schiff base. As can be easily
seen in Table 3, changes in the monoanionic group,
X−, in a cis-arrangement to the PPh3 ligand, produced
a shift of the signal to higher fields, which follows the
trend:
In
compounds:
[Pd{[(h5-C5H3)–CHꢀN–CH2–
C6H5]Fe(h5-C5H5)}(X)(PPh3)], the five-membered pal-
ladacycles, are practically planar, contain the imine
group (endocyclic) and form angles of ca. 7.1° (for
X−=Br−) and 5.4° (for X−=I−) with the C5H3 ring
of the ferrocenyl.
In the two structures the imine adopts the anti-con-
CN− BSCN− BAcO− BCl− BBr− BI−
(2)
,
formation and the ꢁCꢀN– bond lengths [1.281(5) A
−
−
−
−
,
(for X =Br ) and 1.285(5) A (for X =I )] does not
differ significantly from the length reported for the free
Although in this case the maximum variation is only
of ca. 3 ppm, this sequence is formally identical to that
found for the chemical shift of the metallated carbon
atom (Eq. (1)). These equations provide a rough first
approach to order the X− ligands according to: (a)
their trans- influence upon the metallated carbon,
which is directly involved in the ring current of the
ferrocenyl fragment, and (b) the effect produced by
these groups on the electronic environment of the phos-
phorus atom of the PPh3 ligand in a cis-orientation
(cis-influences).
,
ligand [1.262(7) A] [12].
The phenyl ring bound to the –N–CH2– portion is
planar and the normal to its main plane forms angles of
ca. 99.5° (for X−=Br−) or 101.96° (for X−=I−) with
the imine moiety plane.
An interesting feature of the structures of [Pd{[(h5-
C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)}X(PPh3)]
is
the short distance between the bromo or iodo ligands
and the H12A atom of the –N–CH2– group [2.86 and
−
−
2.95 A for X =Br or I−, respectively], which is
,

S. Pe´rez et al. / Journal of Organometallic Chemistry 625 (2001) 67–76
71
bond lengths and angles of complexes [Pd{[(h5-C5H3)–
CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(X)(PPh3)] with X−
=Cl− [14], Br− and I− was undertaken. Data shown
in Table 4 revealed that, except for the Pd–X bond
length and the Pd–N bond length, which increases
according to the sequence Cl−BBr−BI−, the remain-
ing Pd-ligands’ bond distances [Pd–P and Pd–C] are
very similar (the differences do not clearly exceed 3|).
This also applies for the bond angles involving the
metallated carbon bond. The major differences were
found in the bond angles involving the X− ligand. For
instance, the P–Pd–X bond angle in the chloro deriva-
tive [95.4(1)°] is greater than in the bromo or iodo
analogues [94.99(6)° (for X−=Br−) and 94.81(5)° (for
X−=I−)]. This variation is compensated by a narrow-
ing of the N–Pd–Cl bond angle [92.8(1)°] from the
N–Pd–Br [93.75(9)°] or N–Pd–I [94.68(10)°] bond
angles.
clearly less than the sum of the van der Waals radii of
,
the two atoms involved {H, 1.20 A and Br, 1.95 or
,
I=2.15 A [13]}. This suggests a weak intramolecular
C12–H12A···X {X=Br or I} interaction.
The two rings of the ferrocenyl unit are planar,
nearly parallel [tilt angles of 4.2° (for X−=Br−) and
2.5° (for X−=I−)] and they deviate from the ideal
eclipsed conformation by 9.2 and 4.6° for X−=Br−
and I−, respectively. The distances between the palla-
−
−
,
dium and the iron(II) are: 3.6028 A (for X =Br ) and
−
−
,
3.6750 A (for X =I ), which suggests that there is no
direct interaction between them.
In a first attempt to elucidate whether the replace-
ment of a Cl− group in [Pd{[(h5-C5H3)–CHꢀN–CH2–
C6H5]Fe(h5-C5H5)}(Cl)(PPh3)] by bulkier and less
electronegative (Table 1) X− ligand (X−=Br− or I−)
could cause significant modifications in the structures of
this sort of compounds, a comparative study of selected
Fig. 2. Molecular structure and atom labelling scheme for [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(Br)(PPh3)].
Fig. 3. Molecular structure and atom labelling scheme for for [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(I)(PPh3)].

72
S. Pe´rez et al. / Journal of Organometallic Chemistry 625 (2001) 67–76
Table 4
intramolecular C–H···Cl interaction. This may be re-
lated to the differences detected in the values of the
bond angles: N–Pd–X and P–Pd–X described above.
Since the effective bulks of the Br− and I− groups are
[ES–CH=0.65 (for Br−) and 0.78 (for I−) [8]] greater
than the bulk of the Cl− [ES–CH=0.55 [8]] group, one
would expect that the Br− or the I− groups would tend
to move further away from the PPh3 to reduce steric
hindrance. However, in the bromo or iodo derivatives
the P–Pd–X bond angles are [94.99(6)° (for X−=Br−)
and 94.81(5)° (for X−=I−)] smaller than for the
chloro analog [95.4(1)°]. The narrowing of the N–Pd–
Cl bond angle causes, the Cl− ligand to approach the
H12A atom, which is involved in the intramolecular
C–H···Cl interaction.
,
Selected bond lengths (A) and bond angles (°) for compounds
[Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(X)(PPh3)]
{with
−=Br−, I− and Cl− a} (standard deviations are given in parenthe-
X
sis)
X
−=Br−
X
−=I−
X
−=Cl−
Bond lengths
Pd–Cmet
Pd–N
Pd–P
b
2.012(3)
2.159(3)
2.2469(14)
2.5040(17)
1.281(5)
2.012(4)
2.172(3)
2.2532(16)
2.6640(13)
1.285(5)
2.004(5)
2.146(6)
2.247(2)
2.368(2)
1.279(7)
2.051(7)
1.417(10)
Pd–X
N–C(11)
Fe–C c
2.045(8)
1.416(13)
2.047(9)
1.421(13)
C–C(ring) c,b
Bond angles
N–Pd–Cmet
P–Pd–Cmet.
N–Pd(1)–X
P(4)–Pd–X
b
80.50(13)
90.74(11)
93.75(9)
94.99(6)
80.66(14)
89.83(11)
94.68(10)
94.81(5)
80.8(2)
91.0(2)
92.8(1)
95.4(1)
b
2.4. Theoretical approaches to clarify the influence of
the ligand in a trans-arrangement to the metallated
carbon upon the electronic en6ironment of the Fe(II)
a Data from Ref. [19].
b Cmet. represents the metallated carbon atom: C(10) for X−=Br−
and C(6) for X−=I−
.
In a first attempt to understand the trends detected in
the NMR spectra of the compounds holding different
X− groups upon the ferrocenyl unit, we decided to
perform theoretical calculations to find out whether
these monoanionic groups could cause significant varia-
tions in the electronic density of the ferrocenyl frag-
ment. Previous studies have shown that for the
c Average values for the ferrocenyl moiety.
ferrocenylamine
[(h5-C5H5)Fe{(h5–C5H4)–CH2–
N(CH3)2}], the Schiff bases [(h5-C5H5)}Fe{(h5-C5H4)–
C(CH3)ꢀN–R}] {R=phenyl or benzyl groups} and
their cyclopalladated complexes, the PM3(tm) model of
the SPARTAN 5.0 program [15] provided optimized ge-
ometries which were in excellent agreement with the
structural data obtained by X-ray diffraction [6]. On
this basis, we assumed that this procedure could also be
used for compounds [Pd{[(h5-C5H3)–CHꢀN–CH2-
C6H5]Fe(h5-C5H5)}(Br)(PPh3)].
For derivatives holding X−=Cl−, Br− and I−,
whose structural data are available, the crystallographic
coordinates were used as input for the optimization of
the geometry, but no significant variations in either
bond lengths, angles or other structural parameters
were detected, thus indicating that the PM3(tm) model
could also be extended to these systems.
Fig. 4. HOMO of compounds [Pd{[(h5-C5H3)–CHꢀN–CH2–
C6H5]Fe(h5-C5H5)}(X)(PPh3)] {with X=Br− and CN−} (A) and
(B), respectively. The drawings labelled as (C) and (D) correspond to
As shown in Fig. 4A and B, the metallated carbon
atom is directly involved in the HOMO, in addition to
the contribution of the orbital(s) of the X− ligand: a p
orbital of the Cl−, Br−, I−, or the sulfur of the
SCN−groups or the y-orbital of the CN− ligand (Fig.
4A,B). These findings suggested that the electronic den-
sity of the metallated ring could be directly affected by
the characteristics of the X− ligand.
the LUMO of compounds with X−=Br− or CN−
.
A careful study of the structural data reported for
compound [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-
C5H5)}(Cl)(PPh3)] [19] suggests that, in this case too,
the distance between the Cl− and one of the protons of
In a first approach to clarifying the effect of the X−
group on the ring current of the C5H3 ring, the average
values of Mulliken’s bond orders {here in after MBO}
[16] for the C–C bonds of this ring were calculated for
,
the –N–CH2– fragment is [2.83 A] smaller than the
sum of the van der Waals radii of these two atoms {Cl,
,
1.75 and H, 1.20 A [13]}. This also suggests a weak

S. Pe´rez et al. / Journal of Organometallic Chemistry 625 (2001) 67–76
73
all the compounds as well as for ferrocene and the free
imine. In these systems the geometry chosen was the
same as the geometry of these groups in the cyclometal-
lated compound. The highest value of MBO corre-
sponded to ferrocene, the intermediate was for the free
imine and the lowest for the cyclometallated complexes.
This is consistent with the electron-pulling nature of the
imine and of the ‘[PdX(PPh3)]’ moieties [14]. For the
cyclopalladated derivatives, the averaged value of Mul-
liken’s bond order [16] increased upon changing the
anionic group according to the sequence: CN−B
SCN−BCl−BBr−BI−. This trend, which is formally
identical to that shown in Eq. (1), reflects the increase
of the electronic density of the C5H3, which is consis-
tent with the lowfield shift of the signal due to the
metallated carbon atom, C2, in the 13C{1H}-NMR
spectra.
Pd–N–C12–H12A caused a significant increase of the
formation enthalpy, DHform., of these complexes. Thus,
these C–H···X intramolecular interactions appear to
contribute to the stabilisation of these molecules.
3. Experimental
3.1. Materials and synthesis
The ligand [(h5-C5H5)Fe{(h5-C5H3)–CHꢀN–CH2–
C6H5}], the di-m-chloro-bridged cyclopalladated com-
plex
[Pd{[(h5-C5H3)–CHꢀN–CH2-C6H5]Fe(h5–C5-
H5)}(m-Cl)]2 and compound [Pd{[(h5-C5H3)–CHꢀN–
CH2–C6H5]Fe(h5-C5H5)}(Cl)(PPh3)] were prepared as
described previously [1].
Elemental analyses {C, H, N and S (for SCN–
derivative)} were carried out at the Serveis Cientifico-
Te`cnics de la Universitat de Barcelona. Infrared spectra
were obtained with a Nicolet Impact 400 instrument
using KBr disks. Proton and two-dimensional {1H-13C}
heteronuclear NMR experiments were run at 500 MHz
with either a Varian 500 or a Bruker advance 500DMX
instrument. The solvent used for the NMR experiments
was CDCl3 (99.9%) and SiMe4 was the internal refer-
ence. 31P{1H}-NMR spectra were recorded with a
Bruker 250DXR instrument using CDCl3 (99.9%) as
In all cases, the LUMO is mainly formed by the
combination of two y* orbitals, one of the ferrocenyl
fragment and the other of the imine group (Fig. 4 C,D).
These results are consistent with those reported for:
[Pd{[(h5 - C5H3)–C(CH3)ꢀN–C6H5]Fe(h5 - C5H5)}(Cl)-
(PH3)] [6].
2.5. Conclusions
The results presented here have allowed us to estab-
lish trans-influence of the monoanionic ligand X− {Cl−,
Br−, I−, CN−, SCN− or AcO−} on the spectroscopic
properties of cyclopalladated derivatives of general for-
solvent
and
P(OMe)3
as
internal
reference
[l{P(OMe)3=140.17 ppm]. In all cases, the chemical
shifts (l) are given in ppm and the coupling constants
(J) in Hz.
mula:
[Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-
C5H5)}(X)(PPh3)]. The variations observed in the NMR
spectra have been rationalized with the aid of semiem-
pirical calculations performed with the SPARTAN 5.0
suite of programs [15]. The results revealed that
changes of the X− ligand in a trans-arrangement to the
metallated carbon made the electronic density of the
C5H3 ring of the ferrocenyl moiety increase according
to the sequence: CN−BSCN−BCl−BBr−BI−. This
is consistent with the shift of the signal due metallated
carbon atom, C2, of the ferrocenyl fragment.
3.1.1. Preparation of [Pd{[(p5-C5H3)–CHꢀN–CH2–
C6H5]Fe(p5-C5H5)}(v-I)]2
The di-m-chloro-bridged cyclopalladated complex
[Pd{[(h5 - C5H3)–CHꢀN–CH2–C6H5]Fe(h5 - C5H5)}-
(m-Cl)]2 [1] (157 mg, 1.67×10−4 mol) was suspended in
30 cm3 of methanol. Then a solution containing 56 mg
(3.37×10−4 mol) of KI in methanol (25 cm3) was
added. The resulting mixture was stirred at room tem-
perature (r.t.) ca. 20°C, for 1 h. The solid formed was
collected by filtration and air-dried (yield: 79.7%).
The comparison of the structural data for com-
pounds:
[Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-
Characterization
data:
Anal.
Calc.
for:
C5H5)}(X)(PPh3)] with X−=Cl−, Br−, I−, shows that
although the Br− and I− groups are bulkier than the
Cl− ligand (Table 1) the replacement of the Cl− ligand
by a Br− or a I−, does not produce significant varia-
tions in the bond lengths and angles of the metallacycle.
In the three cases, the short distances between the X−
ligand and one of the hydrogens, H12A, of the –N–
CH2– unit suggest a weak C12–H12A···X {X−=Cl−,
Br− or I−} intramolecular interaction. The use of the
SPARTAN 5.0 computer program [15], also indicated
that for these compounds, this sort of interaction is
expected to be attractive, since small changes (910°) in
the N–Pd–X bond angles or in the torsion angles
C36H32Fe2I2N2Pd2 (Found): C, 40.37 (40.4); H, 3.01
(3.1) and N, 2.62 (2.55)%. IR: w(ꢁCꢀN–)=1580
cm−1
.
3.1.2. Preparation of [Pd{[(p5-C5H3)–CHꢀN–
CH2–C6H5]Fe(p5-C5H5)}(v–AcO)]2
The ferrocenylketimine: [(h5-C5H5)Fe{(h5-C5H3)–
CHꢀN–CH2–C6H5}] (1.01 g, 3.3×10−3 mol),
Pd(AcO)2 (0.747 g, 3.3×10−3 mol) and NaAcO·3H2O
(0.450 g, 3.3×10−3 mol) were suspended in 40 cm3 of
methanol. The reaction mixture was protected from
light with aluminium foil and stirred at r.t. (ca. 20°C)
for 24 h. After this period the red solid formed was

74
S. Pe´rez et al. / Journal of Organometallic Chemistry 625 (2001) 67–76
collected by filtration and washed with three portions
of 5 cm3 of methanol. The solid was air-dried and then
dried in vacuum (yield: 67%). Characterization data:
Anal. Calc. for: C40H38Fe2N2O4Pd2 (Found): C, 51.37
(51.45); 4.10 (4.15) and N, 3.00 (2.9)%. IR: w(ꢁCꢀN–)
=1590 cm−1 (br).
CH2–C6H5]Fe(h5-C5H5)}(Br)(PPh3)], but using the sto-
ichiometric amount of KCN or KSCN (yields: 68 and
62%, respectively). Characterization data: for [Pd{[(h5-
C5H3)–CHꢀN–CH2–C6H5]Fe(h5 - C5H5)}(CN)(PPh3)]:
Anal. Calc. for C37H31N2FePPd (Found): C, 63.77
(63.9); H, 4.48 (4.5) and N, 4.01 (3.9)%. IR: w(ꢁCꢀN–)
=1597 cm−1 and w(SCN)=2145 cm−1 and. For
[Pd{[(h5 - C5H3)–CHꢀN–CH2–C6H5]Fe(h5 - C5H5)}-
(SCN)(PPh3)]: Anal. Calc. for C37H31N2FePdPS·1/
2CH2Cl2 (Found): C, 58.39 (58.55); H, 4.18 (4.25); N,
3.63 (3.7) and S, 4.16 (4.2)%. IR: w(ꢁCꢀN−)=1590
3.1.3. Preparation of [Pd{[(p5-C5H3)–CHꢀN–
CH2–C6H5]Fe(p5-C5H5}(Br)(PPh3)]
A 105 mg (1.50×10−4 mol) amount of [Pd{[(h5-
C5H3)–CHꢀN–CH2–C6H5]Fe(h5-C5H5)(Cl)(PPh3)] was
dissolved in 20 cm3 of acetone, then AgNO3 (25.5 mg,
1.50×10−4 mol) was added. The reaction mixture was
protected from the light with aluminium foil and stirred
at rt. ca. 20°C for 30 min. After this, the solution was
filtered out to remove the AgCl formed and the filtrate
was treated with 150×10−4 mol of KBr. The reaction
mixture was stirred at r.t. for 2 h. The resulting red
solution was filtered out and the filtrate was then
concentrated to dryness on a rotary evaporator. The
residue was dissolved in the minimum amount of
CH2Cl2 and passed through an SiO2 column using
CH2Cl2 as eluant. The red band released was collected
and concentrated in a rotary evaporator to ca. 2 cm3.
The red cubes formed upon cooling were collected by
filtration and air-dried (yield: 72%). Characterization
data: Anal. Calc. for: C36H31BrFeNPPd (Found): C,
57.59 (57.6); H, 4.16 (4.2) and N, 1.87 (1.9)%. IR:
cm−1 and w(CN)=2091 cm−1
.
3.1.6. Preparation of [Pd{[(p5-C5H3)–CHꢀN–
CH2–C6H5]Fe(p5-C5H5)}(AcO)(PPh3)]
Compound [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe-
(h5-C5H5)}(m-AcO)]2 (130 mg, 1.39×10−4 mol) was
suspended in 40 cm3 of acetone. Then, 73 mg (2.79×
10−4 mol) of PPh3 were added. The reaction mixture
was refluxed for 30 min. After cooling to room temper-
ature (ca. 20°C), it was filtered out and the filtrate was
concentrated to dryness on a rotary evaporator giving a
deep red residue, which was later dissolved in the
minimum amount of CH2Cl2 and passed through an
SiO2 column chromatography. Elution with CH2Cl2
released a red band which was collected and concen-
trated to dryness on a rotary evaporator. The solid
formed was collected and air-dried (yield: 58%). Char-
acterization data: Anal. Calc. for: C38H34NFeO2PPd
(Found): C, 62.53 (62.6), H, 4.70 (4.75) and N, 1.92
(2.0)%. IR: w(ꢁCꢀN–)= 1594 cm−1, w(CO2)=1619
and 1341 cm−1
w(ꢁCꢀN–)=1585 cm−1
.
3.1.4. Preparation of [Pd{[(p5-C5H3)–CHꢀN–
CH2–C6H5]Fe(p5-C5H5)}I(PPh3)]
Triphenylphosphine (45 mg, 1.72×10−4 mol) was
added to a suspension formed by 93 mg (8.9×10−5
mol) of [Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-
C5H5)}(m-I)]2 and 10 cm3 of acetone. The resulting
mixture was refluxed for 30 min. After this period the
undissolved materials were removed by filtration and
discarded, and the filtrate was concentrated to dryness
on a rotary evaporator. The residue was then dissolved
in the minimum amount of CH2Cl2 and passed through
a SiO2 column chromatography. Elution with CH2Cl2
produced the release of a red band which was collected
and concentrated in a rotary evaporator to ca. 3 cm3.
The solution was then treated with n-hexane (ca. 5
cm3). Slow evaporation of the solvent produced small
red cubes which were collected and air-dried (yield:
58%). Anal. Calc. for C36H31NFeIPdP·1/4CH2Cl2
(found): C, 53.11 (52.9); H, 3.90 (3.85) and N, 1.70
3.2. Crystallography
A prismatic crystal (sizes in Table 5) of [Pd{[(h5-
C5H3)–CHꢀN–CH2–C6H5]Fe(h5 - C5H5)}(X)(PPh3)]
with X−=Br− or I− was selected and mounted on a
Enraf-Nonius CAD4 four-circle diffractometer. Unit-
cell parameters were determined from automatic cen-
tring of 25 reflections in the range 12=[=21° and
refined by least-squares method. Intensities were col-
lected with graphite monochromatized Mo–Ka radia-
tion using ꢀ/2[ scan technique. For the bromo
derivative, 8985 reflections were measured in the range
2.325[529.96°, of which 8924 were non-equivalent
by symmetry [Rint(on I)=0.016]; while for: [Pd{[(h5-
C5H3)–CHꢀN–CH2–C6H5]Fe(h5
- C5H5)}(I)(PPh3)],
9012 reflections were collected in the range 2.385[5
29.96°, of which 8950 were non-equivalent by symmetry
[Rint(on I)=0.017]. The number of reflections assumed
as observed, applying the condition I\2|(I), was 6062
for X−=Br− and 6065 for X−=I−. In the two cases,
three reflections were measured every 2 h as orientation
and intensity control, and no significant intensity decay
was observed. Lorentz polarization and absorption cor-
rections were made.
(1.8)%. IR: w(ꢁCꢀN–)=1587 cm−1
.
3.1.5. Preparation of [Pd{[(p5-C5H3)–CHꢀN–
CH2–C6H5]Fe(p5-C5H5)}(X)(PPh3)] with X−=CN− or
SCN−
These compounds were prepared according to the
procedure described above for [Pd{[(h5-C5H3)–CHꢀN–

S. Pe´rez et al. / Journal of Organometallic Chemistry 625 (2001) 67–76
75
The structures were solved by direct methods using
the SHELXS computer program [17] and refined by
full-matrix least-squares method with the SHELX-97
computer program [18] using 8924 reflections (very
negative intensities were not assumed). The function
gen atoms were located from a difference synthesies
and refined with an overall isotropic temperature fac-
tor. The remaining four hydrogen atoms were com-
puted and refined with an overall isotropic temperature
factor equals to 1.2 times the equivalent isotropic tem-
perature factor of the atoms attached using a riding
model. The final R (on F) factors were 0.043 (for
2
2 2
minimized was ꢀwꢁꢁFoꢁ −ꢁFcꢁ ꢁ , where w=[[2(I)+
(0.0475P)2+0.0403P]−1 (for X−=Br−) and [s2(I)+
2
2
2
(0.0631P)2]−1 (for X−=I−) and P=(ꢁFoꢁ +2ꢁFcꢁ )/3.
ƒ, ƒ% and ƒ¦ were taken from the literature [19]. For
[Pd{[(h5−C5H3)–CHꢀN–CH2–C6H5]Fe(h5 - C5H5)}-
(Br)(PPh3)], all hydrogen atoms were located from a
difference synthesis and refined with an overall
isotropic temperature factor, while for [Pd{[(h5-C5H3)–
CHꢀN–CH2–C6H5]Fe(h5-C5H5)}(Br)(PPh3)] 27 hydro-
X−=Br−) and 0.042 (for X−=I−) and wR (on ꢁFꢁ )
values were 0.087 and 0.098 for compounds having
X−=Br− and I−, respectively. The goodness of fit was
1.025 (for X−=Br−) and 1.034 (for X−=I−) for all
the observed reflections. Further details concerning the
resolution and the refinement of the crystal structure
are shown in Table 5.
Table 5
3.3. Computational details
Crystal data and details of the refinement of the crystal structure of
compounds
[Pd{[(h5-C5H3)–CHꢀN–CH2–C6H5]Fe(h5-
C5H5)}(X)(PPh3)] with X−=Br− or I−
The calculations were performed using the SPARTAN
5.0 suite of programs [15] on a Silicon Graphics work-
station (INDIGO-2 power ZX). The PM3(tm) method
was used with the default parameters provided by the
program. The crystallographic coordinates of [Pd{[(h5-
C5 H3) – CHꢀN – CH2 – C6 H5]Fe(h5 -C5 H5)}(X)(PPh3)]
(X=Br− or I−) were used as input for the program,
and the Br− group was then replaced by the corre-
sponding monoanionic ligand {Cl−, I−, CN− or SCN−}.
Then, the geometry of all the compounds was opti-
mized before the calculations. No significant variations
were detected in the bond lengths and angles of the
optimized geometry and the structural data available
for the complexes [Pd{[(h5-C5H3)–CHꢀN–CH2–
C6H5]Fe(h5-C5H5)}(X)(PPh3)] (with X−=Cl− [19],
Br− or I−). Geometrical restrictions were not imposed
in any case.
X
−=Br−
X
−=I−
Empirical fomula
Formula weight
Crystal size (mm)
Temperature (K)
C36H31BrFeNPPd C36H31IFeNPPd
750.75
0.1×0.1×0.2
293(2)
0.71069
797.74
0.1×0.1×0.2
293(2)
0.71069
Triclinic
,
Wavelength (A)
Crystal system
Triclinic
P1
(
(
Space group
P1
,
a (A)
10.908(9)
11.066(2)
13.258(4)
85.66(2)
88.93(4)
73.67(4)
1531.4(14)
2
11.037(2)
11.061(5)
13.412(11)
89.15(5)
81.15(4)
73.07(3)
1562.8(15)
2
,
b (A)
,
c (A)
a (°)
i (°)
k (°)
−3
,
V (A
)
Z
Dcalc (mg m−3
)
1.628
2.444
1695
2.104
Absorption coefficient
(mm−1
F(000)
)
752
788
q Range for data
collection (°)
2.32 to 29.96
2.38 to 29.96
4. Supplementary material
Index ranges
−155h515,
−155k515,
05l518
−155h515,
−155k515,
0 5l518
9012
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Centre, CCDC nos. 149875 and 149876 for com-
pounds with X−=Br− or I−, respectively. Copies of
this information may be obtained from The Director,
CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK
No. of reflections
collected
No. of unique reflections 8924
[Rint=0.0164]
Completeness to
2[=29.96°
Refinement method
8985
8950
[Rint=0.0171]
98.3%
100.0%
(fax:
+44-1233-336033;
e-mail:
deposit@ccdc.
cam.ac.uk or www: http://www.ccdc.cam.ac.uk).
Full-matrix
least-squares
8924
Full-matrix
least-squares
8950
No. of data
No. of parameters
Goodness-of-fit on F2
R indices [I=2s(I)]
494
1.025
478
1.034
R1=0.0428,
wR=0.0981
R1= 0.0739,
wR2=0.1094
Acknowledgements
R1=0.0431,
wR=0.0877
R1= 0.0949,
wR2=0.1010
0.883 and −0.686 0.556 and
−0.425
We are indebted to the Ministerio de Educacio´n y
Cultura of Spain and to the Generalitat de Catalunya
(grants BQU2000-0652 and 1999-SGR-00045) for finan-
cial support. SP is also grateful to the Ministerio de
Educacio´n y Cultura for a fellowship.
R indices (all data)
Largest difference peak
−3
,
and hole (e A
)

76
S. Pe´rez et al. / Journal of Organometallic Chemistry 625 (2001) 67–76
227. (b) G.B. Deacon, F. Huber, R.J. Phillips, Inorg. Chim. Acta
References
104 (1985) 41.
[11] (a) J. Albert, J. Granell, J. Sales, M. Font-Bard´ıa, X. Solans,
Organometallics 14 (1995) 1393. (b) J. Albert, J. Granell, R.
Moragas, M. Font-Bard´ıa, X. Solans, J. Organomet. Chem. 522
(1996) 59. (c) J. Albert, M. Go´mez, J. Granell, J. Sales,
Organometallics 9 (1990) 1405.
[1] C. Lo´pez, J. Sales, X. Solans, R. Zquiak, J. Chem. Soc. Dalton
Trans. (1992) 2321.
[2] R. Bosque, C. Lo´pez, J. Sales, X. Solans, M. Font-Bard´ıa, J.
Chem. Soc. Dalton Trans. (1994) 735.
[3] R. Bosque, M. Font-Bard´ıa, C. Lo´pez, J. Sales, X. Solans, J.
Silver, J. Chem. Soc. Dalton Trans. (1994) 746.
[12] C. Lo´pez, R. Bosque, X. Solans, M. Font-Bard´ıa, New. J. Chem.
20 (1996) 1285.
[13] (a) A. Bondi, J. Phys. Chem. 68 (1964) 441. (b) A.I. Kitaigorod-
skii, Molecular Crystals and Molecules, Academic Press, Lon-
don, 1973.
[4] C. Lo´pez, R. Bosque, X. Solans, M. Font-Bard´ıa, D. Tramuns,
G. Fern, J. Silver, J. Chem. Soc. Dalton Trans. (1994) 3039.
[5] R. Bosque, C. Lo´pez, J. Sales, Inorg. Chim. Acta 244 (1996) 141.
[6] C. Lo´pez, R. Bosque, X. Solans, M. Font-Bard´ıa, New J. Chem.
22 (1998) 977.
[7] W.A. Henderson, C.A. Screuli, J. Am. Chem. Soc. 82 (1960)
5791.
[8] C. Hansch, A. Leo, D. Hoekman, Exploring QSAR: Hydropho-
bic, Electronic and Steric Constants, American Chemical Soci-
ety, Washington, DC, 1995.
[14] C. Lo´pez, R. Bosque, X. Solans, M. Font-Bard´ıa, D. Tramuns,
G. Fern, J. Silver, J. Chem. Soc. Dalton Trans. (1994) 747.
[15] SPARTAN 5.0. Wavefunction Inc. Irvine, CA, 1995.
[16] I. Mayer, Chem. Phys. Lett. 97 (1983) 270.
[17] G.M. Sheldrick, SHELXS, A computer program for determination
of crystal structure, University of Go¨ttingen, Germany, 1997.
[18] G.M. Sheldrick, SHELX-97, A computer program for determina-
tion of crystal structure, University of Go¨ttingen, Germany,
1997.
[9] K. Nakamoto, Infrared Spectra of Inorganic and Coordination
Compounds, 3rd edn., Wiley, New York, 1978.
[10] (a) G.B. Deacon, R.J. Phillips, Coord. Chem. Rev. 33 (1989)
[19] International Tables of X-ray Crystallography, vol. IV, Kynoch
Press (Ed.), Birmingham, 1974, pp. 99, 100 and 149
.