New monocationic methylpalladium(II) compounds with several bidentate nitrogen-donor ligands: Synthesis, characterisation and reactivity with CO

Article abstract of DOI:10.1016/S0020-1693(01)00700-9

Two series of methylpalladium(II) compounds with mono and bidentate nitrogen-donor ligands, namely [Pd(N-N)₂(CH₃)][X] (N-N=phen (1a), dm-phen (1b) (dm-phen=4,7-dimethyl-1,10-phenanthroline), tm-phen 1c (tm-phen=3,4,7,8-tetramethyl-1,10-phenanthroline); X=OTf, PF₆^-) and [Pd(N-N)(L)(CH₃)][OTf] (N-N=phen and L=py (1ad) (py=pyridine), N-N=phen and L=2-Ph-py (1ae) (2-Ph-py=2-phenyl-pyridine), N-N=phen and L=BzQ (1af) (BzQ=7,8-benzoquinoline), N-N=tm-phen and L=BzQ (1cf)), have been synthesised and fully characterised both in solid state and in solution. The crystal structures of [Pd(phen)₂(CH₃)][PF₆] and [Pd(phen)(2-Ph-py)(CH₃)][OTf] show a square planar coordination geometry for palladium with the monodentate ligand (one phen molecule plays this role in 1a) bound to the metal with its plane almost perpendicular to the coordination plane. In both structures the Pd-N bond length trans to the methyl is remarkably affected by its trans influence. The behaviour in solution is characterised for the first series of compounds by a dynamic process which makes the two N-N ligands equivalent, as corroborated by the ¹⁵N NMR analysis: only one averaged signal is shown for all of the four nitrogen atoms. No fluxional process is present for the compounds of the second series, and three main crosspeaks are shown in the ¹⁵N-¹H HMQC spectra. In particular, the signal of the ¹⁵N trans to the methyl group has a typical chemical shift, which differs from those of two ¹⁵N trans to each other. Both series of complexes are reacted with carbon monoxide and the reaction products are studied by ¹H NMR spectroscopy and, when possible, by isolating the acyl derivatives. The products of this reaction are affected by the nature of the second molecule of N-ligand.

Full text of DOI:10.1016/S0020-1693(01)00700-9

www.elsevier.com/locate/ica
Inorganica Chimica Acta 327 (2002) 188–201
New monocationic methylpalladium(II) compounds with several
bidentate nitrogen-donor ligands: synthesis, characterisation and
reactivity with CO
Barbara Milani a,*, Angelica Marson a, Ennio Zangrando a, Giovanni Mestroni a,
Jan Meine Ernsting b, Cornelis J. Elsevier b
a Dipartimento di Scienze Chimiche, Uni6ersita` di Trieste, Via Licio Giorgieri 1, 34127 Trieste, Italy
b Institute of Molecular Chemistry, Organometallic and Coordination Chemistry, Uni6ersity of Amsterdam, Nieuwe Achtergracht 166,
1018 WV Amsterdam, The Netherlands
Received 4 July 2001; accepted 12 September 2001
Dedicated to Professor Kees Vrieze for his outstanding contribution to the development of coordination chemistry
Abstract
Two series of methylpalladium(II) compounds with mono and bidentate nitrogen-donor ligands, namely [Pd(N-N)2(CH3)][X]
(N-N=phen (1a), dm-phen (1b) (dm-phen=4,7-dimethyl-1,10-phenanthroline), tm-phen 1c (tm-phen=3,4,7,8-tetramethyl-1,10-
phenanthroline); X=OTf, PF6−) and [Pd(N-N)(L)(CH3)][OTf] (N-N=phen and L=py (1ad) (py=pyridine), N-N=phen and
L=2-Ph-py (1ae) (2-Ph-py=2-phenyl-pyridine), N-N=phen and L=BzQ (1af) (BzQ=7,8-benzoquinoline), N-N=tm-phen
and L=BzQ (1cf)), have been synthesised and fully characterised both in solid state and in solution. The crystal structures of
[Pd(phen)2(CH3)][PF6] and [Pd(phen)(2-Ph-py)(CH3)][OTf] show a square planar coordination geometry for palladium with the
monodentate ligand (one phen molecule plays this role in 1a) bound to the metal with its plane almost perpendicular to the
coordination plane. In both structures the PdꢀN bond length trans to the methyl is remarkably affected by its trans influence. The
behaviour in solution is characterised for the first series of compounds by a dynamic process which makes the two N-N ligands
equivalent, as corroborated by the 15N NMR analysis: only one averaged signal is shown for all of the four nitrogen atoms. No
fluxional process is present for the compounds of the second series, and three main crosspeaks are shown in the 15N–1H HMQC
spectra. In particular, the signal of the 15N trans to the methyl group has a typical chemical shift, which differs from those of two
1
15N trans to each other. Both series of complexes are reacted with carbon monoxide and the reaction products are studied by H
NMR spectroscopy and, when possible, by isolating the acyl derivatives. The products of this reaction are affected by the nature
of the second molecule of N-ligand. © 2002 Elsevier Science B.V. All rights reserved.
Keywords: Palladium; Nitrogen-donor ligands; Carbon monoxide; Insertion reactions; 15N NMR spectroscopy
1. Introduction
chemistry [2]. They embraces many different types of
reactions such as the reductive carbonylation of ni-
troaromatic compounds to carbamates [3], the oxida-
tive carbonylation of unsaturated substrates [4], the
amidocarbonylation [5], the synthesis of acetic acid [6],
and the CO/olefin copolymerisation [7].
The study of intimate mechanism of these reactions
has received a considerable interest for many years. In
this sense, the contribution of Professor Kees Vrieze to
this branch of coordination chemistry is widely recog-
Since the discovery of the oxo-synthesis by Otto
Roelen in 1938 [1] the carbon monoxide has found wide
application in homogeneous catalysis and the carbony-
lation reactions, from a more general point of view,
represent one of the main topics of organometallic
* Corresponding author.
E-mail address: milani@dsch.univ.trieste.it (B. Milani).
0020-1693/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S0020-1693(01)00700-9

B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
189
2. Experimental
nised, as witnessed quite recently at the occasion of the
25th anniversary of his appointment to the Chair of
Inorganic Chemistry at the University of Amsterdam
[8].
2.1. Materials
The nitrogen-donor ligands (Aldrich) together with
the analytical grade solvents (Carlo Erba) were used
without further purification for synthetic and spectro-
scopic purposes. The dichloromethane used for the
synthesis of complexes was purified through distillation
over CaCl2 and stored under inert atmosphere.
Pd(CH3COO)2 was a lone from Johnson Matthey. Car-
bon monoxide (CP grade, 99.9%) was supplied by
SIAD.
Among the carbonylation reactions cited above we
focused our attention on the CO/olefin co- and terpoly-
merisation reactions promoted by palladium(II) com-
pounds. In particular, we found that the dicationic
bischelated Pd(II) derivatives of general formula [Pd(N-
N)(L-L)][PF6]2 (N-N=2,2%-bipyridine (bipy), 1,10-
phenanthroline (phen) and their substituted derivatives;
L-L=N-N or L-L"N-N, therefore L-L=1,3-bis-
(diphenylphosphino)propane (dppp)), are very efficient
catalyst precursors for these reactions and the corre-
sponding polyketones are obtained in high yield, with
high molecular weight and with negligible decomposi-
tion of the active species to palladium metal [9,10].
Unlike the catalytic systems reported in literature
[7,11–15], these complexes are characterised by the
presence of two molecules of (potentially) chelating
ligand in the coordination sphere of the metal centre.
Moreover, up to now all the mechanistic investiga-
tions reported for the CO/olefins copolymerisation are
based on model systems characterised again by a palla-
dium to ancillary ligand ratio of 1 [16–20]; only one
paper has appeared related to the reactivity with CO
of organometallic palladium(II) species with two
molecules of bidentate ancillary ligand in the coordina-
tion sphere of the metal, [Pd(p-An-BIAN)(L-L)-
2.2. Instrumentation
IR spectra were recorded in a Perkin–Elmer 983G
1
spectrometer in KBr pellets. H and 13C NMR spectra
were recorded at 400 and 100.5 MHz, respectively, in a
JEOL EX 400 spectrometer; the resonances were refer-
enced to the solvent peak versus TMS (CDCl3 at 7.26 l
1
1
for H and 77.0 l for 13C; CD2Cl2 at 5.33 l for H and
53.8 l for 13C). Two-dimensional correlation spectra
(COSY) were obtained with the automatic program of
the instrument. The NOE experiments were run with a
1H pulse of 90° of 12.3 ms.
The 15N NMR spectra were recorded using the PFG
HMQC sequence [22] in a Bruker DRX 300 spectrome-
ter equipped with a 5 mm triple resonance inverse
probe with z-gradient, operating at 30.42 MHz 15N
frequency, a second 300 W X decoupler giving a 90°
15N pulse of 10 ms. Spectra were recorded without
decoupling 15N in f2, using a spectral width in f2 (1H) of
10 ppm, an acquisition time of 0.4 s giving a digital
resolution of 1.2 Hz per point. In the first experiments
up to three values for J{1H, 15N} (2.5, 5 and 8 Hz) were
used, these experiments were recorded with a spectral
width in f1 of 500 ppm and 32 increments only. The
final experiment was done with a spectral width of 20
ppm (15N) with 256 increments and 16–1024 scans per
increment depending on the concentration of the com-
pound. This provided, after linear prediction to 1024, a
digital resolution of 0.6 Hz per point. The relaxation
delay was 1 s. For experiments using 256 increments
and 16 scans, data collection required 2 h. Chemical
shifts were referenced to external nitromethane=0
ppm, negative chemical shifts were reported for lower
frequencies. Spectra were recorded at 296 K.
(CH3)][OTf]
phthene; L-L=p-An-BIAN, phen, 2,9-dimethyl-1,10-
phenanthroline, 1,2-bis(diphenylphosphino)ethane,
(p-An-BIAN=bis(anisylimino)acena-
dppp; OTf=triflate) [21].
One of the questions Professor Vrieze often asked us
is ‘‘What is the role of the second molecule of nitrogen-
donor ligand in the bischelated-Pd(II) system?’’ With
the aim to clarify this aspect we synthesised and charac-
terised two series of monocationic organometallic palla-
dium(II) compounds of general formula [Pd(N-N)2-
(CH3)][X] (N-N=phen (1a), dm-phen (1b) (dm-phen=
4,7-dimethyl-1,10-phenanthroline), tm-phen (1c) (tm-
phen=3,4,7,8-tetramethyl-1,10-phenanthroline); X=
OTf, PF6−) and [Pd(N-N)(L)(CH3)][OTf] (N-N=phen
and L=py (1ad) (py=pyridine), N-N=phen and L=
2-Ph-py (1ae) (2-Ph-py=2-phenyl-pyridine), N-N=
phen and L=BzQ (1af) (BzQ=7,8-benzoquinoline),
N-N=tm-phen and L=BzQ (1cf)). In these two series
of complexes two molecules of nitrogen-donor ligand
are bound to the same palladium centre: one is always
a bidentate ligand, while the other can be either a
bidentate or a monodentate one. In this paper we
report the synthesis and characterisation of these com-
plexes, together with their reactivity with carbon
monoxide.
2.3. Synthesis of complexes
All manipulations were carried out in argon atmo-
sphere by using Schlenk techniques and at room tem-
perature, unless otherwise stated.
Elemental analysis (C, H, N), performed by Dr E.
Cebulec (Dipartimento di Scienze Chimiche, Trieste,

190
B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
Italy), were in perfect agreement with the proposed
stoichiometry (Tables 1 and 2).
[Pd(CH3COO)2], used as a starting material, was
synthesised mainly following the method reported in
the literature [16b], but with different amount of sol-
vents. In particular, to a suspension of 0.40 mmol of
[Pd(N-N)(CH3)Cl] in 10 ml of dichloromethane a solu-
tion of AgOTf (0.46 mmol in 8 ml of anhydrous
acetonitrile) was added. After 30 min the AgCl formed
is filtrated and the solution obtained is concentrated
under vacuum yielding a pale yellow solid upon addi-
tion of diethyl ether. The solid is filtered, washed with
diethyl ether and dried under vacuum. Average yield:
75%.
transformed in trans-[Pd(PhCN)2Cl2] following
a
modified procedure with respect to that reported in the
literature [23]. [Pd(COD)Cl2] was obtained from trans-
[Pd(PhCN)2Cl2] [24]. The synthesis of [Pd(N-N)-
(CH3)Cl] (N-N=phen and its substituted derivatives)
from [Pd(COD)Cl2] was done in analogy to that of
[Pd(p-An-BIAN)(CH3)Cl] [25].
2.3.1. Synthesis of trans-[Pd(PhCN)2Cl2]
[Pd(CH3COO)2] (2 g) was dissolved in 60 ml of
benzonitrile together with 2.7 ml of hydrochloric acid
(37%). The solution was heated to 90 °C for 30 min.
To the cooled orange–red solution 35 ml of petrol
ether was added and it was filtered over Celite. The
product was recovered after 2 days at 277 K. It was
filtered under vacuum, washed with petrol ether and
dried under vacuum. From the mother liquor another
fraction precipitated upon addition of another portion
of petrol ether and after 1 week at 277 K. Elemental
analysis: C, 44.0; H, 2.54; N, 7.31. Calc. values for
C14H10N2Cl2Pd: C, 43.8; H, 2.62; N, 7.30%.
2.3.3. Synthesis of [Pd(N-N)2(CH3)][OTf ]
(N-N=phen (1a), dm-phen (1b), tm-phen (1c))
To a suspension of 0.20 mmol of [Pd(N-N)(CH3)-
(CH3CN)][OTf] in 20 ml of chloroform, 0.26 mmol of
the proper N-N ligand was added (Pd–N-N=1:1.3),
yielding an orange/yellow solution. After 10 min it is
filtered over fine paper and concentrated under vacuum
to induce precipitation of the product as an orange
solid. The solid was removed by filtration, washed with
diethyl ether and vacuum dried. Average yield: 80%.
2.3.4. Synthesis of [Pd(phen)(py)(CH3)][OTf ] (1ad)
To a suspension of 0.20 mmol of Pd(N-N)(CH3)-
(CH3CN)][OTf] in 20 ml of dichloromethane, 2.0 mmol
of pyridine was added (Pd–py=1:10), yielding immedi-
ately a solution. After 10 min it is filtered over fine
paper and concentrated under vacuum to induce pre-
cipitation of the product as a white solid. The solid was
removed by filtration, washed with diethyl ether and
2.3.2. Synthesis of [Pd(N-N)(CH3)(CH3CN)][OTf ]
(N-N=phen, dm-phen, tm-phen)
The compound [Pd(N-N)(CH3)(CH3CN)][OTf] was
Table 1
Elemental analysis for compounds [Pd(N-N)2(CH3)][OTf] (1a–1c) and
[Pd(N-N)(L)(CH3)][OTf] (1ad–1cf) a
1
vacuum dried. Average yield: 90%. H NMR (CD2Cl2,
Compound
C (%)
H (%)
N (%)
293 K, l ppm): 8.97 (dd, 1H, H2), 8.88 (dt, 2H, Ha),
8.70 (dd, 1H, H4), 8.64 (dd, 1H, H7), 8.09 (m, 3H,
[Pd(phen)2(CH3)][OTf]
49.6 (49.5) 2.99 (3.03) 8.87 (8.88)
52.6 (52.4) 3.94 (3.96) 8.17 (8.15)
53.2 (54.9) 4.61 (4.75) 7.34 (7.54)
41.5 (43.1) 2.87 (3.04) 7.68 (7.93)
[Pd(dm-phen)2(CH3)][OTf]
[Pd(tm-phen)2(CH3)][OTf]
[Pd(phen)(py)(CH3)][OTf]
H
5,6+Hg), 8.04 (dd, 1H, H9), 7.97 (dd, 1H, H3), 7.87
(dd, 1H, H8), 7.68 (td, 2H, Hb), 1.10 (s, 3H, CH3).
[Pd(phen)(2-Ph-py)(CH3)][OTf] 49.4 (49.6) 3.20 (3.33) 6.92 (6.93)
[Pd(phen)(BzQ)(CH3)][OTf]·
BzQ
[Pd(tm-phen)(BzQ)(CH3)]-
[OTf]·BzQ
49.3 (49.3) 3.60 (3.61) 6.92 (6.92)
2.3.5. Synthesis of [Pd(N-N)(L)(CH3)][OTf ]
(N-N=phen, L=2-Ph-py (1ae); N-N=phen,
L=BzQ (1af); N-N=tm-phen, L=BzQ (1cf))
56.5 (61.1) 4.12 (4.31) 6.52 (6.47)
The procedure is the same reported above for com-
pound 1ad, but the solvent used is chloroform instead
of dichloromethane. Average yield: 90%. 1H NMR
(CD2Cl2, 293 K, l ppm): for 1ae 9.09 (d, 1H, H6%), 8.86
(dd, 1H, H2), 8.67 (dd, 1H, H4), 8.60 (dd, 1H, H7), 8.18
(td, 1H, H4%), 8.05 (m, 4H, H5,6+Hb), 8.01 (dd, 1H,
H9), 7.92 (m, 2H, H3+H3%), 7.82 (dd, 1H, H8), 7.65 (td,
1H, H5%), 7.33 (m, 3H, Ha+Hg), 0.98 (s, 3H, CH3); for
1af 12.11 (d, 1H, H10%), 9.86 (d, 1H, H2%), 9.09 (d, 1H,
H2), 8.76 (d, 1H, H4), 8.63 (d, 1H, H4%), 8.54 (d, 1H,
H7), 8.10–7.55 (signals overlapped: H5,6, H3, H8, H9,
H9%, H8%, H7%, H6%, H5%, H3%), 1.21 (s, 3H, CH3); for 1cf
12.05 (d, 1H, H10%), 9.91 (d, 1H, H2%), 8.71 (s, 1H, H2),
8.64 (d, 1H, H4%), 8.26 (s, 2H, H5,6), 8.04–7.55 (over-
lapped signals: H9%, H8%, H7%, H6%, H5%), 7.35 (s, 1H, H9),
a Calculated values are reported in parenthesis.
Table 2
Elemental analysis for compounds [Pd(N-N)2(C(O)CH3)][OTf] (2a–
2c) and [Pd(N-N)(L)(C(O)CH3)][OTf] (2ad–2ae) a
Compound
C (%)
H (%)
N (%)
[Pd(phen)2(C(O)CH3)][OTf]
49.4 (49.6) 3.20 (3.33) 6.92 (6.93)
[Pd(dm-phen)2(C(O)CH3)][OTf] 50.3 (52.1) 3.68 (3.81) 7.52 (7.83)
[Pd(tm-phen)2(C(O)CH3)][OTf] 51.1 (54.5) 4.32 (4.57) 6.83 (7.26)
[Pd(phen)(py)(C(O)CH3)][OTf] 41.8 (43.1) 2.71 (2.89) 7.42 (7.53)
[Pd(phen)(2-Ph-py)(C(O)CH3)]- 47.7 (49.3) 3.02 (3.18) 6.59 (6.63)
[OTf]
a Calculated values are reported in parenthesis.

B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
191
2.87 (s, 3H, (CH3)4), 2.69 (s, 3H, (CH3)7), 2.67 (s, 3H,
difference Fourier synthesis revealed the presence of a
solvent molecule of chloroform. In 1ae the triflate anion
(disordered over two sites with occupancy of 0.5 each)
was refined using a fixed model, leading to an overall
structural determination of low accuracy.
(CH3)3), 2.20 (s, 3H, (CH3)8), 1.01 (s, 3H, PdꢀCH3).
2.3.6. Synthesis of [Pd(N-N)2(C(O)CH3)][OTf ]
(N-N=phen (2a), dm-phen (2b), tm-phen (2c)) and of
[Pd(phen)(L)(C(O)CH3)][OTf ] (L=py (2ad),
All the calculations were performed using the WINGX
System, Ver 1.63 [28].
L=2-Ph-py (2ae))
Carbon monoxide was bubbled through a solution of
0.05 mmol of [Pd(N-N)2(CH3)][OTf] or of [Pd(phen)-
(L)(CH3)][OTf] in 10 ml of dichloromethane for 20 min,
at 273 K. Concentration of the solution and addition of
diethyl ether induced the precipitation of the product as
a brown solid, that was removed by filtration, washed
with diethyl ether and dried under vacuum. The solid
2.5.1. Crystal data for 1a
C26H20Cl3F6N4PPd, M=746.18, triclinic, space
(
,
group P1, a=10.887(3), b=11.781(3), c=12.421(5) A,
h=117.37(3), i=95.03(3), k=92.51(2)°, V=1402.9(8)
3
A , Z=2, zcalc=1.766 g cm−3, v (Mo Ka)=1.070
,
mm−1, F(000)=740. Final R=0.0638, wR2=0.1176,
S=1.021 for 371 parameters and 5676 reflections, 5423
unique [Rint=0.0424], of which 2170 with I\2|(I),
1
must be stored at 277 K. Average yield: 75%. H NMR
(CD2Cl2, 273 K, l ppm): for 2a 8.77 (d, 4H, H2,9), 8.56
(d, 4H, H4,7), 8.08 (s, 4H, H5,6), 7.77 (dd, 4H, H3,8), 2.34
(s, 3H, CH3); for 2b 8.58 (d, 4H, H2,9), 8.24 (s, 4H,
H5,6), 7.55 (d, 4H, H3,8), 2.86 (s, 12H, (CH3)4,7), 2.27 (s,
3H, CH3); for 2c 8.44 (s, 4H, H2,9), 8.22 (s, 4H, H5,6),
2.73 (s, 12H, (CH3)4,7), 2.40 (s, 12H, (CH3)3,8), 2.22 (s,
3H, CH3); at 163 K: for 2ad 8.95 (d, 2H, Ha), 8.70 (d,
1H, H4), 8.60 (d, 1H, H7), 8.54 (d, 1H, H2), 8.12 (t, 1H,
Hg), 8.07 (s, 2H, H5,6), 7.93 (dd, 1H, H3), 7.89 (d, 1H,
H9), 7.81 (dd, 1H, H8), 7.72 (t, 2H, Hb), 2.52 (s, 3H,
CH3); for 2ae 9.33 (d, 1H, H6%), 8.67 (m, 2H, H4,7), 8.45
(d, 1H, H2), 8.22 (t, 1H, H5%), 8.08 (b, 4H, H5,6+Hb),
7.98–7.80 (b, 4H, H3%+H9+H8+H3), 7.73 (t, 1H,
H4%), 7.45 (b, 3H, Ha+Hg), 2.03 (s, 3H, CH3).
maximum positive and negative peaks in DF map 0.398
−3
,
and −0.652 e A
.
2.5.2. Crystal data for 1ae
C25H20F3N3O3PdS, M=605.90, monoclinic, space
group C2/c, a=23.842(5), b=13.855(4), c=14.971(4)
3
,
,
A, i=98.82(1)°, V=4887(2) A , Z=8, zcalc=1.647
g cm−3, v (Mo Ka)=0.901 mm−1, F(000)=2432. Fi-
nal R=0.0785, wR2=0.1838, S=1.176 for 246
parameters and 5005 reflections, 4808 unique [Rint
0.0671], of which 1513 with I\2|(I), maximum posi-
=
tive and negative peaks in DF map 0.740 and −0.836
−3
,
e A
.
2.4. NMR in situ experiments
3. Results and discussion
The reactivity of both series of complexes, 1a–1c and
1ad–1cf, with CO was investigated in situ by running
NMR experiments. CD2Cl2 (0.7 ml) was added to a 5
mm NMR tube charged with the complex (7×10−3
mmol). After cooling the solution to 273 K, CO was
bubbled for 5 min via a needle inserted through a
rubber cap into the NMR tube. The sample was placed
3.1. Synthesis of complexes [Pd(N-N)2(CH3)][OTf ]
and [Pd(N-N)(L)(CH3)][OTf ]
The synthesis of the two series of complexes [Pd(N-
N)2(CH3)][OTf] (1a–1c) and [Pd(N-N)(L)(CH3)][OTf]
(1ad–1cf), starting from Pd(CH3COO)2, consists of
six steps, five of which are reported in the literature
for similar derivatives, like [Pd(p-An-BIAN)(CH3)-
(CH3CN)][OTf] [16b,25] and [Pd(N-N)(CH3)(CH3CN)]-
1
in a precooled NMR probe and the H NMR spectrum
was obtained after 15 min.
2.5. X-ray structure determination
[BAr%] (N-N=bipy, phen; Ar%=3,5-(CF3)2C6H3) [12a].
4
The main difference consists in the low solubility of our
system, which is instead guaranteed, in the literature
systems, by the p-An-BIAN ligand in one case and by
the particular anion (BAr%) in the other. In the com-
plexes studied this problem was overcome by increasing
the amount of acetonitrile in the reaction mixture (see
Section 2). The desired compounds [Pd(N-N)2(CH3)]-
[OTf] and [Pd(N-N)(L)(CH3)][OTf] are easily synthe-
sised from [Pd(N-N)(CH3)(CH3CN)][OTf] through an
exchange reaction of the coordinated acetonitrile
with a second molecule of the nitrogen-donor ligand
(Scheme 1), with the only care to add an excess of
the monodentate ligand with respect to palladium
Diffraction experiments were carried out at room
temperature using the ꢀ–2q scan technique on a CAD4
Enraf–Nonius single-crystal diffractometer equipped
with graphite monochromator and Mo Ka radiation
,
(u=0.71073 A). Throughout the data collection, no
noticeable variation in intensity was observed. The
reflections were corrected for Lorentz–polarisation ef-
fects and for absorption, based on an empirical „-scan
method. The structures were solved by conventional
Patterson [26] and Fourier analyses and refined on F2
by full-matrix anisotropic least-squares method using
the SHELXL-97 package [27]. In the crystal of 1a, a

192
B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
Scheme 1. Synthesis of complexes 1a–1cf.
([L]/[Pd]=10) for the derivatives 1ad–1cf. In particu-
lar, for the [Pd(N-N)(L)(CH3)][OTf] derivatives this
procedure represents a good alternative to the method
reported in literature for the compound [Pd(bipy)(g-pi-
coline)(CH3)][BF4] [29].
All the final products are obtained in high yield and
high purity. Following the synthetic procedure dis-
cussed above and in agreement with the elemental
analysis, the complexes with benzoquinoline, 1af and
1cf, are always characterised by a BzQ to Pd ratio of 2.
When a modified synthetic method was applied ([BzQ]/
[Pd]B10) a non-well defined solid was obtained.
nitrogen in a pseudo-apical position with respect to the
,
square planar geometry of the metal, at 2.737(8) A.
Following the rationalisation reported by Vrieze and
co-workers [21], which indicate that an axial/equatorial
distances ratio higher than 1.3 is distinctive of a square
planar coordination rather than a square pyramidal
one, we are inclined to indicate the present complex as
square planar, although a metal displacement of
,
0.075(5) A from the coordination plane towards N(4)
(see Fig. 1) should be indicative of a weak interaction
between palladium and this nitrogen.
While the PdꢀN distances in trans positions present
,
values comparable to those of about 2.04 A found in
3.2. Characterisation of complexes
[Pd(N-N)2(CH3)][OTf ] (1a–1c)
the bischelated [Pd(phen)2]2+ cations [10b,30], the
s-donor power of the methyl group provides an elec-
tron-rich metal centre affecting the value observed for
the PdꢀN(4) distance, which is longer in comparison
The complexes [Pd(N-N)2(CH3)][OTf] (1a–1c) were
fully characterised both in solid state and in solution.
In order to obtain single crystals suitable for X-ray
analysis the replacement of triflate with hexafluoro-
phosphate was necessary and crystals of [Pd(phen)2-
(CH3)][PF6] were directly obtained from the synthesis in
,
with that of 2.674(3) A, found in the corresponding
derivative containing a nitromethyl instead of CH3 [31].
chloroform
CN)][PF6].
starting
from
[Pd(phen)(CH3)(CH3-
The molecular structure of the [Pd(phen)2(CH3)]+
cation with the atom labelling scheme is depicted in
Fig. 1. The metal displays a slightly distorted square
planar geometry with one phen molecule acting as a
monodentate ligand, the other one presenting the usual
chelating feature and they are practically perpendicular
to each other, the best fit planes making a dihedral
angle of 89.4(1)°.
The coordination distances for the difunctional base
are considerably different (PdꢀN(1) 2.036(8), PdꢀN(2)
Fig. 1. ORTEP drawing (thermal ellipsoids at 40% probability level) of
,
the cation 1a. Selected bond lengths (A) and angles (°): PdꢀC(13)
2.018(9); PdꢀN(1) 2.036(8); PdꢀN(2) 2.115(8); PdꢀN(3) 2.040(8);
PdꢀN(4) 2.737(8); N(1)ꢀPdꢀN(2) 80.5(3); N(3)ꢀPdꢀN(4) 69.5(3);
C(13)ꢀPdꢀN(1) 94.4(4); C(13)ꢀPdꢀN(2) 173.9(4); C(13)ꢀPdꢀN(3)
,
2.115(8) A), the latter being affected by the trans influ-
ence exerted by the methyl group. On the other hand,
the monodentate phen ligand, unsymmetrically bound
87.9(4);
C(13)ꢀPdꢀN(4)
97.6(4);
C(14)ꢀN(3)ꢀPd
116.2(6);
,
to Pd through N(3) (2.040(8) A), presents the second
C(18)ꢀN(3)ꢀPd 126.8(6).

B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
193
Table 3
Relevant 1H, 13C and 15N NMR chemical shifts of complexes 1a–1c and free N-N ligands a
Ligand/compound
N-N (1H)
H2,9
CH3
(
13C)
15N b N1, N10
H4,7
H5,6
H3,8
phen
1a
dm-phen
1b
tm-phen
1c
9.20 (dd)
8.83 (dd)
9.05 (d)
8.67 (d)
8.93 (s)
8.55 (s)
8.27 (dd)
8.60 (dd)
7.82 (s)
8.08 (s)
8.04 (s)
8.24 (s)
8.04 (s)
8.25 (s)
7.65 (q)
7.83 (dd)
7.46 (d)
7.59 (d)
−74.4 (8.8)
−123.6 c
−77.3 (12.2)
n.d.
0.93 (s) (0.99)
0.82 (s) (0.65)
0.75 (s) (0.35)
−78.3 (12.5)
−130.8 c
a Measured at 400 MHz at 20 °C in CDCl3; l, ppm; s=singlet, d=doublet, dd=double doublet, q=quartet.
b Measured at 30.42 MHz at 23 °C in CDCl3; l, ppm; 2J(15N,1H) in parentheses.
c Averaged signal (see text).
On the other hand, the PdꢀN(4) apical distance agrees
by recording the spectra at low temperature in CD2Cl2,
which result in a progressive broadening of the aro-
matic frequencies. However, even at the lowest temper-
ature reached (T=163 K), the signals are too broad to
allow any assignment. After the addition of 1 equiv. of
free phen to a solution of [Pd(phen)2(CH3)][OTf] both
the signals of coordinated and free phen are shown in
the spectrum at room temperature, thus indicating that,
if an exchange process with the free ligand is present, its
rate is very low on the NMR timescale. Thus, the
fluxional process, which renders both of the two
molecules of N-N ligand as well as both halves of each
of them equivalent, may involve two different equi-
libria, which alternate to each other: (a) the exchange
between the unbound nitrogen atom of the monocoor-
dinated N-N ligand (i.e. N2) and that of the chelating
N-N molecule trans to the methyl group (i.e. N3; open-
ing–closure) (Scheme 2); (b) the exchange of the two
nitrogen atoms (i.e. N1 and N2) at the same coordina-
tion site ( flipping) (Scheme 2). At room temperature,
the whole dynamic process is fast on the 1H NMR
timescale.
,
with that of 2.714(6) A, found in the [Pd(p-An-
BIAN)2(CH3)]+ [21], for the p-An-BIAN ligand coordi-
nated in a unidentate fashion.
,
The PdꢀC methyl distance, 2.018(9) A, is close to
,
those of 2.024(3) and 2.036(6) A, detected in [Pd(phen)-
(CH3)((Et)2O)]+ [32] and in [Pd(bipy)(CH3)(g-picol-
ine)]+ [29] cations, respectively, making allowance for
the different e.s.d.
Finally, the crystal packing shows the cations ar-
ranged head-to-tail around a symmetry centre with
,
chelating phen stacking at about 3.6 A.
The behaviour in solution for the [Pd(N-N)2(CH3)]-
[OTf] compounds (1a–1c) was studied by 1H NMR
spectroscopy in CDCl3 (Table 3). At room temperature
the spectra show, for the nitrogen ligands, four signals
integrating for a quarter of the number of chemically
equivalent groups, indicating the equivalence of both
the two ligands bound to palladium and of the two
halves of each ligand. The signal attribution to proton
was done in agreement with the literature [30,31]. All
the signals are shifted with respect to the free ligand
and no signal due to the latter is present. In particular,
the resonances of the ‘probe-protons’ (H2,9) are upfield
shifted with respect to the same signal in the free
ligand. The singlet due to methyl falls in the range
3.3. Characterisation of complexes
[Pd(N-N)(L)(CH3)][OTf ] (1ad–1cf)
1
between 1.00 and 0.50 ppm in the H NMR spectra,
To the best of our knowledge a systematic study of
the organometallic Pd(II)-derivatives [Pd(N-N)(L)-
(CH3)][OTf] (1ad–1cf), characterised by one bidentate
and one monodentate ligand bound to the same palla-
dium centre, has never been reported. As monodentate
ligands we chose pyridine (py), as it is the simplest,
2-phenyl-pyridine (2-Ph-py) and 7,8-benzoquinoline
(BzQ) since they resemble to 2,2%-bipyridine and to
1,10-phenanthroline, respectively, coordinated in a
monodentate fashion to palladium.
and between 1.00 and 0.1 ppm in the 13C NMR spectra.
Its chemical shift is upfield-shifted on going from phen
to tm-phen. This trend may be related to the electron-
donor properties of the ligand, which increase in the
same order. The same trend was observed for the signal
of the CH2NO2 methylenic protons in the complexes
[Pd(N-N)2(CH2NO2)][PF6] [31]. Moreover, as already
reported for the [Pd(p-An-BIAN)(L-L)(CH3)][OTf]
compounds [21] and for the nitromethyl derivatives
[31], the most important feature in the spectra is the
equivalence, in solution, of the two ligands, which
appears in contrast with the structure in solid state. The
presence of a dynamic process has been demonstrated
The molecular structure of the [Pd(phen)(2-Ph-
py)(CH3)]+ cation is shown in Fig. 2. The metal dis-
plays a square planar coordination geometry with the
phen molecule acting as a chelating ligand. The coordi-

194
B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
nation distances for 1ae fall in the expected range for
these types of complexes, although the low accuracy of
crystallographic data does not allow a detailed analysis.
However, the geometrical features of the 2-Ph-py coor-
dination deserve a comment. The 2-Ph-py ligand ex-
hibits a tilted conformation with a N(3)ꢀC(18)ꢀC(19)ꢀ
C(20) torsion angle of 41(2)°, which excludes any conju-
gation between the rings. The mean plane through the
Fig. 3. Perspective view of the crystal packing of 1ae (triflate anions
are not shown for sake of clarity).
py ring forms an angle of 64.2(4)° with that of phenan-
throline, an orientation remarkably distorted from the
expected ideal value of 90°, but close to 62.1° found in
the g-picoline derivative [29]. This conformation, likely
dictated by packing forces, does not prevent the py ring
to affect the magnetic behaviour of H9 (Fig. 2, see
below).
In the crystal the cations, piled along the crystallo-
graphic c axis, are related alternatively by a symmetry
centre and a two-fold axis, in such a way to allow the
,
chelating phen to stack at about 3.5 A (Fig. 3).
The characterisation in solution of this series of
1
complexes was done by mono- and bidimensional H
and 15N NMR spectroscopy (Table 4; numbering
scheme is reported in Scheme 3).
The spectra of the four derivatives 1ad–1cf share
some common features: (i) the ratio between phen and
the coordinated L ligand is 1; (ii) the most significant
resonances of the two ligands are well separated allow-
ing a clear assignment to the corresponding protons by
means of simple COSY experiments. In particular, the
number of signals of phen and their integration confirm
that phenanthroline is bound to palladium in a non-
symmetrical chemical environment and it is not in-
volved in any dynamic process. No signals due to the
free ligands, either N-N or L, are present. The protons
of the methyl group generate a singlet in a narrow
range of frequencies. Thanks to their NOE effect, it is
possible to assign the signal of H2 (Table 4) and start-
ing from it all of the other NꢀN ligand resonances.
Again through an NOE experiment it is possible to
distinguish between the frequency of H7 (NOE with
H5,6) and H9. For all the four complexes, the signal of
H9 is remarkably upfield shifted (more than 1 ppm)
with respect to the same signal in the free ligand, in
agreement with the fact that H9 falls in the shielding
cone of the aromatic nitrogen ligand in cis position to
it, as clearly evidenced by the crystal structure of 1ae
Scheme 2. Proposed mechanism for the exchange process of com-
plexes 1a–1c.
Fig. 2. ORTEP drawing (thermal ellipsoids at 40% probability level) of
,
the cation 1ae. Selected bond lengths (A) and angles (°): PdꢀC(13)
2.048(12); PdꢀN(1) 2.096(13); PdꢀN(2) 2.110(14); PdꢀN(3) 2.078(11);
C(13)ꢀPdꢀN(1) 93.0(6); C(13)ꢀPdꢀN(2) 174.7(6); N(1)ꢀPdꢀN(2)
82.2(6);
C(13)ꢀPdꢀN(3)
88.2(5);
N(3)ꢀPdꢀN(1)
178.7(5);
N(3)ꢀPdꢀN(2) 96.6(6); C(14)ꢀN(3)ꢀPd 119.1(12); C(18)ꢀN(3)ꢀPd
122.3(11).

B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
195
(Fig. 2). This shift, already observed for the Pd–(g-pi-
coline) compound [29], confirms that in 1ad–1cf the
monodentate ligand is bound to palladium with its
plane non-coplanar with the coordination plane. More-
over, in the case of pyridine the square-planar coordi-
nation plane represents a plane of symmetry, too.
Some differences should be noted between the spec-
tra of 1ad, 1ae and those of complexes with benzo-
quinoline, 1af and 1cf. Since the aromatic region of the
spectrum of [Pd(phen)(BzQ)(CH3)][OTf]·BzQ is very
crowded, the derivative with tm-phen was synthesised
and its signals were easily attributed through NOE
experiments performed by irradiating the frequencies of
the different methyl groups. The number of signals of
benzoquinoline and their integration evidence the pres-
ence of two molecules of BzQ for each molecule of N-N
ligand, a result in agreement with the stoichiometry of
the complexes derived by the elemental analyses (Table
1). On comparison with the spectrum of free benzo-
quinoline, one set of signals was attributed to the latter,
while the other resonances are due to the ligand bound
to palladium. Among its several signals, those related to
the protons of the heterocyclic aromatic ring (H2%, H3%
and H4%) are shifted to high frequency with respect to
the free ligand and they are apparently unaffected by
the nature of the N-N ligand. In both complexes 1af
and 1cf the frequency at the lowest field is the doublet
around 12.00 ppm (12.11 ppm for 1af and 12.05 ppm
for 1cf) with a coordination-induced shifts (CIS) value
of almost 2.75 ppm. This signal is attributed to H10% (see
Scheme 3). Studies on [Pd(dmp)(L)(OH2)][ClO4]
(dmp=N,N-dimethyl-benzylamine;
L=8-methyl-
quinoline, 7,8-benzoquinoline) and on trans-[PtCl2-
(BzQ)(PX3)] (X=Et, Ph, Tol) evidence that a down-
field shift of the H10% signal of more than 1 ppm is
indicative of a weak interaction M’HꢀC [33]. There-
fore, in agreement with the literature, the coordinated
benzoquinoline has the hydrogen atom (H10%) approach-
ing the palladium centre and its plane is perpendicular
to the coordination plane, which is no more a plane of
symmetry. The above results indicate that no metalla-
tion process occurs on the benzoquinoline, which be-
haves like a real monodentate ligand.
Moreover, when a saturation transfer experiment is
run by irradiating the signal due to H10% of the bound
ligand a clear decrease in intensity was observed for the
Table 4
Relevant 1H and 15N NMR data for the compounds [Pd(N-N)(L)(CH3)][OTf] (1ad–1cf) a
b
Compound
N-N
H2
CH3
15N
N1
H9
N10
NL
1ad
1ae
1af
1cf
8.97 (dd)
8.86 (dd)
9.09 (dd)
8.71 (s)
8.04 (dd)
8.01 (dd)
7.45 c
1.10 (s)
0.98 (s)
1.21 (s)
1.01 (s)
−155.7 (−81.3)
−155.9 (−81.5)
−157.6 (−83.2)
n.d.
−120.3 (−45.9)
−121.2 (−46.8)
−121.5 (−47.1)
n.d.
−142.7 (−77.0)
−141.9 (−66.9)
−155.9 (−79.1) d
n.d.
7.35 (s)
a Measured at 400 MHz at 293 K in CD2Cl2; l, ppm; s=singlet, d=doublet, dd=double doublet, q=quartet.
b Measured at 30.42 MHz at 296 K in CD2Cl2; l, ppm; CIS values in parenthesis.
c Overlapped to other signals (see Section 2).
d Associated BzQ −75.5 ppm.
Scheme 3. Numbering scheme for complexes 1ad–1cf.

196
B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
Fig. 4. 1H,15N-HMQC NMR spectrum of compound 1af, recorded in CD2Cl2 at 296 K.
same signal of the free BzQ. An analogous result was
obtained upon irradiation of the frequency of H2% for
bound BzQ, indicating that the coordinated and the
unbound benzoquinoline are in slow exchange on the
1H NMR timescale at room temperature.
HMQC spectra for one of the compounds was first run
at several input values of this parameter, which were
subsequently used for the other experiments (see Sec-
tion 2).
The 15N NMR data and CIS have been compiled in
Tables 3 and 4. The data were first obtained using a
delay corresponding to J(15N,1H)=2.5 Hz. For most
compounds, all relevant 15N,1H crosspeaks were ob-
served. In some cases, additional recording using a
delay corresponding to J(15N,1H)=5 or 8 Hz, empha-
sising crosspeaks due to some of the larger couplings,
gave complementary results.
1
3.4. PFG H,15N-HMQC
In order to substantiate whether or not the various
N-ligands in compounds 1a–1c and 1ad–1af are coor-
dinated to palladium, we have acquired their PFG
{1H,15N}-HMQC NMR spectra, at natural abundance
of the 15N isotope. The nuclear properties of 15N, i.e. its
very low abundance and its rather low and negative
gyromagnetic ratio, render acquisition of 15N data em-
ploying the PFG {1H,15N}-HMQC scheme the method
of choice, provided a suitable nJ(15N,1H) is present. For
the compounds studied, the required scalar coupling is
The 15N chemical shifts of N1 and N10 of compounds
1ad–1af were attributed by their correlations with H2
and H9, respectively, in the HMQC spectra. Also, clear
3
correlations due to J(15N,1H) of H3 and H8 and even
in some cases with H4 and H7 were observed. More-
over, an additional crosspeak with PdꢀCH3 was ob-
served for each compound, further substantiating the
attribution of the 15N signal due to the trans-disposed
2
indeed available; J(15N,1H) ranging from about 8 to
about 14 Hz for the free N-ligands employed in this
study and having unknown but probably reasonable
values for their respective palladium compounds. Addi-
1
N10. As an example, the H,15N-HMQC spectrum of
1af is shown (Fig. 4).
3
tionally, a small but distinct J(15N,1H) was expected
Concerning compounds 1ad–1af, it is seen from
Table 4 that the nuclei N1 and NL have 15N chemical
shifts between −142 and −157 ppm, but the reso-
nance of N10 is found at higher frequencies, around
−120 ppm. All 15N resonances fall within the expected
chemical shift ranges, which were estimated prior to
between the protons of the methyl group at Pd and the
15N nucleus in trans position. The expected crosspeak
due to the latter coupling should be an aid in assigning
the observed 15N resonances. Since the exact values of
2J(15N,1H) in the complexes were unknown, the PFG-

B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
197
Table 5
the respective bidentate N-ligands are in fast exchange
1H NMR chemical shifts for methyl group of compounds [Pd(N-
N)2(R)][OTf] (1a–2c) and [Pd(N-N)(L)(R)][OTf] (1ad–2ae) (R=CH3,
C(O)CH3) and values of w(CO) in cm−1 a
1
on the 15N NMR as well as on the H NMR timescale
(see also above), and the equilibria reported in Scheme
2 account for the complete equivalence of the four
nitrogen atoms of the two N-N ligands. The chemical
shifts observed are indeed very close to the calculated
values, which can be obtained from estimates of the
chemical shifts of the individual N-atoms, which are in
turn based on the shifts of the free ligands (Table 3)
and the CIS values observed for 1ad–1af (Table 4).
Estimated values, being the average of the expected 15N
shifts of the four N-atoms belonging to one bidentate
and one monodentate coordinated phenanthroline
(−156, −121, −156 and −75 ppm) and analogously
for tetra-methyl-phenanthroline (−161, −125, −161
and −78 ppm), amount to −127 ppm for 1a and
−131 ppm for 1c. These values are very close to the
data observed for 1a and 1c and corroborate that
indeed exchange of all four nitrogen atoms of the two
bidentate ligands in these complexes occurs.
Compound
CH3 C(O)CH3 Dl
w(CO)
[Pd(phen)2(R)][OTf]
0.93 2.34
0.82 2.27
0.75 2.22
1.11 2.51
1.00 2.20
1.41 1686
1.45 1687
1.47 1681
1.40 1691
1.20 1686
[Pd(dm-phen)2(R)][OTf]
[Pd(tm-phen)2(R)][OTf]
[Pd(phen)(py)(R)][OTf]
[Pd(phen)(2-Ph-py)(R)][OTf]
[Pd(phen)(BzQ)(R)][OTf]·BzQ 1.18 2.97
[Pd(tm-phen)(BzQ)(R)][OTf]
1.79
2.02
n.d.
n.d.
1.01 3.03
·BzQ
a NMR spectra recorded at 400 MHz at 293 K in CD2Cl2; IR
spectra recorded in KBr pellets.
measurement from the known 15N chemical shifts of the
ligands and their approximate CIS values (−75 to
−90 ppm for N trans to N, −40 to −50 ppm for N
trans to C) based on previously obtained data, for e.g.
[Pd(CH3)(N-N-N)]+ (N-N-N=2-(2-((2%-pyridylmethyl-
ene)amino)ethyl)pyridine) [25], Pd(CH3)2(Ar-BIAN)
[34], and [Pd(p-An-BIAN)(L-L)(CH3)]+ [21]. Indeed,
except for NL of 1ae, the CIS for N1 and NL in these
compounds amounts to approximately −80 ppm and
for N10 the CIS is about −47 ppm.
3.5. Reacti6ity with carbon monoxide
All the complexes of both series were treated with
carbon monoxide and the reactions were studied both
by 1H NMR spectroscopy and by isolating the insertion
products, when possible.
The second BzQ ligand present in compound 1af is
not coordinated, since it gave a 15N resonance correlat-
ing to H2 and H3 of this ligand at −75.5 ppm (see Fig.
4), which differs only slightly from the value (−76.8
ppm) for the pure BzQ ligand in the same solvent. The
exchange process between the free and the unbound
BzQ, witnessed by saturation transfer experiments on
When carbon monoxide was bubbled through a solu-
tion of [Pd(N-N)2(CH3)][OTf] (1a–1c), in CD2Cl2, at
1
273 K, in the H NMR spectra, recorded after 15 min,
the singlet of the methyl group of the precursor is not
present anymore and it is substituted by a new signal at
higher frequency (Table 5). In agreement with the liter-
ature [21], a downfield shift of 1.4 ppm with respect to
the methyl signal of the precursor is indicative of the
CO insertion in the palladium–methyl bond, yielding
the palladium–acyl derivative [Pd(N-N)2(C(O)CH3)]-
[OTf] (2a–2c) (Scheme 4). This is also confirmed by the
1
the H NMR timescale (see Section 3.3), was not ob-
servable on the 15N NMR timescale.
Compounds 1a and 1c both exhibit averaged 15N
signals, which are observed at −123.6 and −130.5
ppm, respectively (Table 3). Apparently, all N-atoms of
Scheme 4. Reaction scheme of complexes 1a–1c with CO.

198
B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
strong CO stretching band at 1693 cm−1 present in the
IR spectrum recorded in dichloromethane solution. The
protons of N-N ligands generate four signals integrat-
ing for a quarter of the number of chemically equiva-
lent groups, indicating that the equivalence of the
nitrogen ligand, observed in the precursor, is retained in
the acyl derivative. The only difference with respect to
the precursor is the upfield shift (Dl=0.1 ppm) of the
signal due to H2,9, which is also quite broad at 273 K.
Lowering the temperature, up to 163 K, results only in
a progressive broadening of the signals and no attribu-
tion signal to proton was possible. Under our reaction
conditions, no difference in the rate of CO insertion
could be observed on going from phen to tm-phen, as
well as no effect of the amount of N-N ligand was
evidenced by addition of increasing amounts of free
phen (1, 2, 4 and 8 equiv.) either before or after gas
bubbling. Moreover, the spectrum of the solution ob-
tained in the presence of 1 equiv. of free phen and CO
shows the broad signals of the free ligand, indicating
the existence of an exchange process between the coor-
dinated and the free ligand, as confirmed also by a
saturation transfer experiment. The rate of this equi-
librium is low on the NMR timescale.
Therefore, on the basis of these results it is reason-
able to assume that the acyl derivatives [Pd(N-
N)2(C(O)CH3)][OTf] are involved in the same fluxional
processes already discussed for the precursors [Pd(N-
N)2(CH3)][OTf]. However, the presence in solution of
an exchange equilibrium between the monodentate N-N
ligand and CO for the fourth coordination site, giving
the acyl-carbonyl species [Pd(N-N)(C(O)CH3)(CO)]-
[OTf], cannot be ruled out (Scheme 4). The [Pd(N-
N)(C(O)CH3)(CO)]+ species has been already reported
and characterised, even by X-ray analysis, by
Brookhart and co-workers [17]; whereas its formation
was not mentioned in the study of the reactivity of
[Pd(p-An-BIAN)(L-L)(CH3)][OTf] complexes with CO
[21].
The palladium–acyl derivatives 2a–2c were isolated
by bubbling CO in a dichloromethane solution of
[Pd(N-N)2(CH3)][OTf] at 273 K (see Section 2). Their
1H NMR spectra show the same signals obtained in the
in situ NMR experiments.
The carbonylation reaction on the compounds
[Pd(N-N)(L)(CH3)][OTf] (1ad–1cf) needs to be dis-
cussed separately depending on the L ligand.
When L is the pyridine, the spectrum recorded after
15 min from the treatment with CO shows the methyl
signal shifted at higher frequencies of 1.4 ppm with
respect to the PdꢀCH3 precursor, indicating the forma-
tion of the Pd–acyl species [Pd(phen)(py)(C(O)-
CH3)][OTf] (Table 5). The signals in the aromatic re-
gion of the spectrum, at 273 K, are broad, but, unlike
2a–2c complexes, they become better resolved on low-
ering the temperature, reaching a complete decoales-
cence at 163 K, which allows a clear attribution signal
to the aromatic protons through a COSY experiment.
It should be noted that both the signals of H2 and H9
are shifted to low frequency with respect to the precur-
sor, and the signal related to H9 is at lower frequency
due to the shielding cone of the pyridine cis to it. The
signal of the methyl is apparently unaffected by varia-
tion of temperature. Moreover, at 163 K no signal due
to free phen or py is present nor are any other minor
signals visible. These data suggest that the broadening
of the signals at 273 K may be due to a dynamic
process involving the N-N ligand, while not involving
the equilibrium between the acyl [Pd(phen)(py)-
(C(O)CH3)]+ and the acyl–carbonyl species [Pd(phen)-
(CO)(C(O)CH3)]+.
When the compound [Pd(phen)(2-Ph-py)(CH3)][OTf]
(1ae) is reacted with CO, in the spectrum at 273 K, the
methyl signal, at 2.20 ppm, is shifted to higher frequen-
cies by 1.20 ppm (Table 5) and it is broad as well as the
resonances of the aromatic protons (Fig. 5). The band
shape analysis performed by decreasing the temperature
up to 163 K shows changes for the signals of both the
aromatic protons and the methyl (Fig. 5). At 163 K the
decoalescence is reached, as for the pyridine derivative,
but at this temperature in the aromatic region there are
two sets of signals, one of which of very low intensity.
Also the signal of the methyl is split into two reso-
nances, at 2.03 and 2.95 ppm, with a 6.25:1 ratio. In
agreement with the literature [17], the singlet at 2.95
ppm is attributed to the acyl–carbonyl species
[Pd(phen)(CO)(C(O)CH3)]+, while the other one is as-
signed to the acyl derivative [Pd(phen)(2-Phpy)-
(C(O)CH3)]+. With regard to the aromatic protons
region, the main resonances, corresponding to the acyl
derivative, were attributed through a COSY experiment
and, as already observed for the pyridine derivative,
both the signals of H2 and H9 are shifted to lower
frequencies with respect to the PdꢀCH3 precursor.
Therefore, the reaction of the 2-Ph-py derivative with
CO gives rise to the equilibrium between the acyl
species and the acyl–carbonyl one, whose rate, at room
1
temperature, is intermediate on the H NMR timescale.
This equilibrium is considerably shifted towards the
acyl derivative [Pd(phen)(2-Ph-py)(C(O)CH3)]+.
Finally, when both the compounds with the 7,8-ben-
zoquinoline, [Pd(phen)(BzQ)(CH3)][OTf]·BzQ (1af) and
[Pd(tm-phen)(BzQ)(CH3)][OTf]·BzQ (1cf), are reacted
with CO, the shift of the methyl signal to higher
frequencies amounts to about 2.00 ppm (1.80 ppm for
1af and 2.02 ppm for 1cf; Table 5), which is much
higher than in the case of 1ad and 1ae. The resonances
of the aromatic protons are quite broad and, in partic-
ular for 1af those of phen correspond to that of
[Pd(phen)(CO)(C(O)CH3)]+ [17]. Moreover, in the case
of 1cf the signals of BzQ correspond exactly to those of
free ligand. Therefore, the reaction of the BzQ deriva-
tives with CO leads to the acyl–carbonyl species [Pd(N-

B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
199
1
Fig. 5. H NMR spectra in CD2Cl2 for the system [Pd(phen)(2-Ph-py)(CH3)][OTf]+CO: variation with temperature; region of signals of aromatic
protons and of CH3 group.

200
B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
Scheme 5. Reaction scheme of complexes 1ad–1cf with CO.
N)(CO)(C(O)CH3)]+ even though the ratio [BzQ]/[Pd]
is 2.
spectra confirm a non-symmetrical chemical environ-
ment for the N-N ligand.
In agreement with the in situ reactivity it was possi-
ble to isolate the acyl species [Pd(phen)(py)(C(O)CH3)]-
[OTf] (2ad) and [Pd(phen)(2-Ph-py)(C(O)CH3)][OTf]
(2ae), but not the BzQ derivatives (Scheme 5).
Therefore, the reactivity with CO of complexes 1ad–
1cf is different depending on the nature of the L ligand,
which can be related to their Lewis basicity, which
decreases on going from py to 2-Ph-py and to BzQ [35].
In agreement with the pKHB trend, the acyl/acyl–car-
bonyl equilibrium is evident for the 2-phenyl-pyridine
derivative, being, instead, totally shifted towards the
acyl species for the pyridine one and towards the
acyl–carbonyl for the benzoquinoline compound.
For all these complexes the insertion of CO in the
palladium–methyl bond is a very fast reaction leading
to different products depending on the nature of the
second molecule of N-ligand, which can compete with
carbon monoxide for the fourth coordination site. It is
straightforward to note that the decomposition of the
palladium(II) species to metal was observed in any case.
5. Supplementary material
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Centre, CCDC Nos. 166125 and 166126 for com-
plexes 1a and 1ae, respectively. Copies of this informa-
tion may be obtained free of charge from The director,
CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK
(fax: +44-1223-336-033; e-mail: deposit@ccdc.cam.
ac.uk or www: http://www.ccdc.cam.ac.uk).
4. Conclusions
The complexes described above represent a develop-
ment of the class of organometallic Pd(II) derivatives
with two nitrogen-donor ligands bound to the same
metallic centre. When both of the two N-ligands are
bidentate molecules, one of them is coordinated in a
unidentate fashion. Indeed, the crystal structure of one
exponent of each series confirms the coordination of
the monodentate N-ligand (either N-N behaving in this
way or L) with its plane almost perpendicular to the
square planar coordination plane.
Acknowledgements
This work was supported by Ministero dell’Univer-
sita` e della Ricerca Scientifica (MURST—Rome)
Grant No. 9903153427. The authors thank COST Ac-
tion D17 for a STSM of B.M. at the Amsterdam
University. Johnson Matthey is gratefully acknowl-
edged for a generous loan of Pd(CH3COO)2. The
Netherlands Organization for Scientific Research,
Chemical Sciences (NWO-CW), the National Research
school Combination Catalysis (NRSC-C) and the John
van Geuns Foundation is thanked for generous finan-
cial support concerning NMR equipment.
1
A detailed H and 15N NMR investigation evidences
the presence of a dynamic process in solution for the
[Pd(N-N)2(CH3)]+, which consists of alternate flipping
and opening–closure processes, which exchange all the
four nitrogen atoms. On the other hand the compounds
[Pd(N-N)(L)(CH3)]+ do not show this kind of be-
1
haviour in solution and both their H and 15N NMR

B. Milani et al. / Inorganica Chimica Acta 327 (2002) 188–201
201
[19] (a) J.H. Groen, C.J. Elsevier, K. Vrieze, W.J.J. Smeets, A.L.
Spek, Organometallics 15 (1996) 3445;
References
(b) J.H. Groen, J.G.P. Delis, P.W.N.M. van Leeuwen, K. Vrieze,
Organometallics 16 (1997) 68;
[1] O. Roelen, Deutsches Patent Schrift 849 (1938–1952) 548.
[2] B. Cornils, W.A. Herrmann (Eds.), Applied Homogeneous
Catalysis with Organometallic Compounds, VCH, Weinheim,
1996.
[3] A.M. Tafesh, J. Weiguny, Chem. Rev. 96 (1996) 2035.
[4] T. Yukawa, S. Tsutsumi, J. Org. Chem. 34 (1968) 738.
[5] H. Wakamatsu, J. Uda, N. Yamakami, J. Chem. Soc., Chem.
Commun. (1971) 1540.
(c) J.G.P. Delis, J.H. Groen, K. Vrieze, P.W.N.M. van Leeuwen,
N. Veldman, A.L. Spek, Organometallics 16 (1997) 551;
(d) R.A. Klein, P. Witte, R. van Belzen, J. Fraanje, K. Goubitz,
M. Numan, H. Schenk, J.M. Ernsting, C.J. Elsevier, Eur. J.
Inorg. Chem. (1998) 319.
[20] C. Carfagna, M. Formica, G. Gatti, A. Musco, A. Pierleoni,
Chem. Commun. (1998) 1113.
[21] J.H. Groen, B.J. de Jong, J.M. Ernsting, P.W.N.M. van
Leeuwen, K. Vrieze, W.J.J. Smeets, A.L. Spek, J. Organomet.
Chem. 573 (1999) 3.
[22] (a) R.E. Hurs, B.K. John, J. Magn. Reson. 91 (1991) 648;
(b) J. Ruiz-Cabello, G.W. Vuister, C.T.W. Moonen, P. van
Gelderen, J.S. Cohen, P.C.M. van Zijl, J. Magn. Reson. 100
(1992) 282;
[6] F.E. Paulik, J.F. Roth, J. Chem. Soc., Chem. Commun. (1968)
1578.
[7] E. Drent, P.H.M. Budzelaar, Chem. Rev. 96 (1996) 613.
[8] K.J. Cavell, Coord. Chem. Rev. 155 (1996) 209.
[9] B. Milani, L. Vicentini, A. Sommazzi, F. Garbassi, E.
Chiarparin, E. Zangrando, G. Mestroni, J. Chem. Soc., Dalton
Trans. (1996) 3139.
[10] (a) B. Milani, G. Corso, G. Mestroni, C. Carfagna, M. Formica,
R. Seraglia, Organometallics 19 (2000) 3435;
(b) B. Milani, A. Anzilutti, L. Vicentini, A. Sessanta o Santi, E.
Zangrando, S. Geremia, G. Mestroni, Organometallics 16 (1997)
5064.
[11] (a) A. Aeby, G. Consiglio, Inorg. Chim. Acta 296 (1999) 45;
(b) M. Barsacchi, G. Consiglio, L. Medici, G. Petrucci, U.W.
Suter, Angew. Chem., Int. Ed. Engl. 30 (1991) 989;
(c) C. Pisano, A. Mezzetti, G. Consiglio, Organometallics 11
(1992) 20.
[12] (a) M. Brookhart, F.C. Rix, J.M. De Simone, J.C. Baroborak, J.
Am. Chem. Soc. 114 (1992) 5894;
(b) M. Brookhart, M.I. Wagner, J. Am. Chem. Soc. 116 (1994)
3641.
[13] K. Nozaki, N. Sato, H. Takaya, J. Am. Chem. Soc. 117 (1995)
9911.
[14] C. Bianchini, H.M. Lee, A. Meli, W. Oberhauser, F. Vizza, P.
Bru¨ggeller, R. Haid, C. Langes, Chem. Commun. (2000) 777.
[15] S. Doherty, G.R. Eastham, R.P. Tooze, T.H. Scanlan, D.
Williams, M.R.J. Elsegood, W. Clegg, Organometallics 18 (1999)
3558.
[16] (a) R. van Asselt, E.E.C.G. Gielens, R.E. Ru¨lke, C.J. Elsevier, J.
Chem. Soc., Chem. Commun. (1993) 1203;
(b) R. van Asselt, E.E.C.G. Gielens, R.E. Ru¨lke, K. Vrieze, C.J.
Elsevier, J. Am. Chem. Soc. 116 (1994) 977.
[17] F.C. Rix, M. Brookhart, P.S. White, J. Am. Chem. Soc. 118
(1996) 4746.
(c) W. Wilker, D. Leibfritz, R. Kerssebaum, W. Bermel, Magn.
Reson. Chem. 31 (1993) 287.
[23] F.R. Hartley, Organomet. Chem. Rev., A 6 (1970) 119.
[24] L. Chatt, L.M. Vallarino, L.M. Venanzi, J. Chem. Soc. (1957)
3413.
[25] R.E. Ru¨lke, J.M. Ernsting, A.L. Spek, C.J. Elsevier, P.W.N.M.
van Leeuwen, K. Vrieze, Inorg. Chem. 32 (1993) 5769.
[26] G.M. Sheldrick, Acta Crystallogr., Sect.
A 46 (1990) 467
(
SHELXS-86, Program for Structure Solution).
[27] G.M. Sheldrick, SHELXL-97, Program for Crystal Structure
Refinement, University of Go¨ttingen, Go¨ttingen, Germany,
1997.
[28] L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837 (WINGX—A
Windows Program for Crystal Structure Analysis).
[29] P.K. Byers, A.J. Canty, B.W. Skelton, A.H. White, J.
Organomet. Chem. 393 (1990) 299.
[30] B. Milani, E. Alessio, G. Mestroni, A. Sommazzi, F. Garbassi,
E. Zangrando, N. Bresciani-Pahor, L. Randaccio, J. Chem. Soc.,
Dalton Trans. (1994) 1903.
[31] B. Milani, G. Corso, E. Zangrando, L. Randaccio, G. Mestroni,
Eur. J. Inorg. Chem. (1999) 2085.
[32] F.C. Rix, M. Brookhart, P.S. White, J. Am. Chem. Soc. 118
(1996) 2436.
[33] (a) A. Albinati, P.S. Pregosin, F. Wombacher, Inorg. Chem. 29
(1990) 1812;
(b) A.J. Deeming, J.P. Rothwell, M.B. Hursthouse, L. New, J.
Chem. Soc., Dalton Trans. (1978) 1490.
[34] J.M. Ernsting, C.J. Elsevier, unpublished results.
[35] M. Berthelot, C. Laurence, M. Safar, F. Besseau, J. Chem. Soc.,
Perkin Trans. 2 (1998) 283.
[18] K. Nozaki, H. Komaki, Y. Kawashima, T. Hiyama, T. Matsub-
ara, J. Am. Chem. Soc. 123 (2001) 534.