Mechanisms of cyclopalladation reactions in acetic acid: Not so simple one-pot processes

Article abstract of DOI:10.1002/(SICI)1099-0682(200001)2000:1<217::AID-EJIC217>3.0.CO;2-Q

The processes operating during the synthetic cyclopalladation reactions of imines in acetic acid have been studied from a kinetico-mechanistic point of view. These reactions include a fast initial coordination to the palladium through the N-donor atom of the imine, followed by the proper C-H bond activation to produce the acetato bridged dimeric species. At this point, the lability of the bridging acetato groups, the hydrolysis of the C-Pd bonds, and/or the hydrolysis of C=N exo bonds contribute to the generation of dark red polynuclear compounds. The processes occurring after the C-H activation have been followed kinetically, both from palladium acetate plus imine, and the synthetically pure isolated acetato dimers as starting materials. The kinetic and activation parameters have been found identical within experimental error whatever the starting material was (k³²³ = 1.5 x 10^-4 s^-1; ΔH(≠) = 51 kJ mol^-1; ΔS(≠) = -163 JK^-1 mol^-1 ΔV(≠) = +19 cm³ mol^-1 for the 4-ClC₆H₄-CH=N-CH₂-C₆H₅ imine derivative 1a). Acidolysis of C-Pd bonds has been found to occur in these polynuclear species. When alternative monomeric C(benzylic)-Pd bond-containing complexes are possible follow ups of the reactions produce them as final dead-end complexes (k³²³ = 2.2 x 10^-5 s^-1; ΔH(≠) = 61 kJ mol^-1; ΔS(≠) = JK^-1 mol^-1 ΔV(≠) ~ 0 cm³ cm^-1 for the [2,4,6-[CH₃]₃]C₆H₂-CH=N- CH₂-[2-[CH₃]C₆H₄] imine derivative 3d). The same study has been carried out with primary amines in order to check the validity of the data if C=N bond hydrolysis is taking place in the imine derivatives with exo C=N bonds. For complexes with similar type of metallacycles, the results agree reasonably well with the proposed mechanism [k³²³ = 1.2·10^-4 s^-1, ΔH(≠) = 46 kJ · mol^-1, ΔS(≠) = -180 J · K^-1mol^-1, ΔV(≠) = -16 cm³ · mol^-1 for the polynuclear formation of the C₆H₅-CH₂-NH₂ derivatives 4e; k³²³ = 3.0·10^-4 s^-1, ΔH(≠) = 55 kJ · mol^-1, ΔS((+)) = -147 J · K^-1 mol^-1, ΔV(≠) = -24 cm³ · mol^-1 for the polynuclear formation of the C₆H₅-CH₂-N=CH-2,3,6-(CH₃)₃C₆H₂ derivative 2c].

Full text of DOI:10.1002/(SICI)1099-0682(200001)2000:1<217::AID-EJIC217>3.0.CO;2-Q

FULL PAPER
Mechanisms of Cyclopalladation Reactions in Acetic Acid:
Not So Simple One-Pot Processes
[a]
[a]
[a]
´
Montserrat Gomez, Jaume Granell, and Manuel Martinez*
Keywords: Reaction mechanisms / Cyclopalladation / Acetic acid / Pd–C bond stability / Polynuclear species / Palladium
The processes operating during the synthetic cyclopalladation has been found to occur in these polynuclear species. When
reactions of imines in acetic acid have been studied from a
alternative
monomeric
CbenzylicϪPd
bond-containing
kinetico-mechanistic point of view. These reactions include a complexes are possible follow ups of the reactions produce
fast initial coordination to the palladium through the N-donor
atom of the imine, followed by the proper C–H bond activation
to produce the acetato bridged dimeric species. At this point,
the lability of the bridging acetato groups, the hydrolysis of the
them as final dead-end complexes (k323 = 2.2 ϫ 10Ϫ5 sϪ1
;
∆H° = 61 kJ molϪ1; ∆S° = JKϪ1 molϪ1 ∆V° ഠ 0 cm3 cmϪ1 for
the [2,4,6-(CH3)3]C6H2ϪCH=NϪCH2-[2-(CH3]C6H4] imine
derivative 3d). The same study has been carried out with
C–Pd bonds, and/or the hydrolysis of C=N exo bonds contribute primary amines in order to check the validity of the data if
to the generation of dark red polynuclear compounds. The C=N bond hydrolysis is taking place in the imine derivatives
processes occurring after the C–H activation have been
with exo C=N bonds. For complexes with similar type of
followed kinetically, both from palladium acetate plus imine, metallacycles, the results agree reasonably well with the
and the synthetically pure isolated acetato dimers as starting
materials. The kinetic and activation parameters have been
proposed mechanism [k323 = 1.2·10–4 s–1, ∆H° = 46 kJ·mol–1
∆S° –180 J·K–1mol–1 ∆V° –16 cm3·mol–1 for the
,
=
,
=
found identical within experimental error whatever the starting polynuclear formation of the C6H5–CH2–NH2 derivative 4e;
material was (k323 = 1.5 ϫ 10Ϫ4
s
Ϫ1; ∆H° = 51 kJ molϪ1; ∆S°
=
k323 = 3.0·10–4 s–1, ∆H° = 55 kJ·mol–1, ∆S‡ = –147 J·K–1mol–1
,
∆V° = –24 cm3·mol–1 for the polynuclear formation of the
C6H5–CH2–N=CH–2,3,6-(CH3)3C6H2 derivative 2c].
Ϫ163 JKϪ1 molϪ1 ∆V° = +19 cm3 molϪ1 for the 4-ClC6H4ϪCH=
NϪCH2ϪC6H5 imine derivative 1a). Acidolysis of CϪPd bonds
organometallic species has been pointed out, as well as the
possible role of arenonium ions in both processes.[10]
Very little information is available about the nature of the
species existing in these reaction solutions. It is generally
assumed that cyclopalladated compounds maintain their di-
meric structure in solution, but recently it has been shown
by 1H-NMR spectroscopy that compounds like [Pd(µ-
Introduction
Cyclometallation reactions were one of the first known
examples of CϪH bond activation and cyclometallated
compounds of a wide variety of ligands (containing N, P,
As, O, or S as heteroatoms) have been described.[1] The
cyclopalladation of N-donor ligands has been extensively
studied by a number of research groups, and this reaction
has acquired great interest because of the applications of
metallacycles in many areas including organic synthesis,[2]
homogeneous catalysis,[3] the design of new metallomesog-
enes,[4] and antitumor drugs.[5] The factors that influence
the ease and mode of cyclopalladation of N-donor ligands
are not entirely understood, but in general the following
trends are accepted: intramolecular electrophilic attack of
“Pd2ϩ” at the carbon atom, strong tendency to form five-
membered metallacycles, and preferential activation of aro-
matic versus benzylic CϪH bonds.[1]
There are few studies on these reactions dealing with ki-
netico-mechanistic information for the cyclopalladation of
N-donor ligands like tertiary amines,[6] imines,[7] 2-phenyl-
pyridines,[8] or triamines.[9] For these reactions, a highly or-
dered transition state has been proposed, in which the leav-
ing proton is accepted by a base. Recently, the formal re-
lationship that may be established between the cyclometall-
ation reaction and the protonation of σ MϪC bonds of d8
Br){C6H4CH(CH3)NH2}]2
and
[Pd{1-CH2-2-(CHϭ
NCH2C6H4)-3,5-(CH3)2C6H2}(CH3COO)]2 adopt monon-
uclear structures in acetone[11] and acetic acid respec-
tively.[7a]
Following our research on the cyclopalladation of N-do-
nor ligands[12] and the mechanisms operating on organome-
tallic reactions involving the activation of CϪH and CϪX
bonds,[13] we present, in this paper, the study of the process
that occurs between imines and palladium acetate in an
acetic acid solution, both from a kinetic and a synthetic
point of view. The imines have been selected to allow com-
parison of the behaviour of endo- and exo-metallacycles
containing Carom.ϪPd or Cbenz.ϪPd bonds. As an extension
of this work, the action of palladium acetate on primary
amines has also been investigated.
Results and Discussion
The processes taking place when imines a, b, c, and d
(Scheme 1) are treated stoichiometrically with palladium
acetate in acetic acid have been studied both from a syn-
thetic and from a kinetic point of view. Although the for-
[a]
´
`
Departament de Quımica Inorganica, Universitat de Barcelona,
´
`
Martı i Franques 1Ϫ11, E-08028 Barcelona, Spain
Eur. J. Inorg. Chem. 2000, 217Ϫ224
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1434Ϫ1948/00/0101Ϫ0217 $ 17.50ϩ.50/0
217

´
M. Gomez, J. Grandell, M. Martinez
FULL PAPER
mation of cyclopalladated compounds is a well-established
Examination of Scheme 1 gives a very good viewpoint of
reaction, and even kinetico-mechanistic studies have been the cyclopalladation process. Reaction I takes place by the
carried out by us[7] and other groups,[6] when the com- initial fast formation of a 1:1 coordination compound in
pounds are prepared in an acetic acid solution, important stoichiometric conditions, followed by the proper CϪH
quantities of undefined material appear, and reduced yields bond activation.[19] Kinetic measurements both in toluene
are evident, compared with the reactions carried out in and acetic acid are available at variable temperature and
toluene (20% yield in acetic acid versus 60% in toluene for pressure, and the results have already been published.[7a]
compounds 2c and 2d).[7a,12a] The available kinetic studies
The crude yellow acetato-bridged complexes isolated
carried out in acetic acid also show the presence of further from the reaction I, 1a and 1b, are stable for days in the
processes, once the main CϪH electrophilic bond activation solid state, but when dissolved in acetic acid a deep red
has taken place.[7a] Previous studies have indicated that the solution is formed; the process (II) is complete after 3 h at
isolated acetato-bridged cyclopalladated complexes have 70°C. Kinetic studies at variable temperature and pressure
significant differences when considered in acetic acid or in have been carried out. Table 1 collects the available data
toluene. Furthermore, even though the nature of the com- together with those for all the other processes depicted in
plexes in solution is also dependent on the nature of the Scheme 1. Despite the similar reactivity, the values deter-
activated CϪH bond (benzylic in 3c and 3d, versus aro- mined for the volume of activation indicate a dramatic dif-
matic in 2c and 2d), differences in behaviour for endo, 1a ference between the mechanism for the endo, 1a and 1b,
and 1b, versus exo, 2c and 2d, compounds are not so cer- and exo, 2c and 2d, complexes. Although in both cases an
tain. In any case, it is clear that the lability of protonated induction period exists, for the exo complexes this period is
bridging or terminal acetato groups plays a key role in the much better defined.
overall process.
Where possible, spectroscopic and kinetic measurements
Scheme 1 shows a simplified view of the reactivity in- for the initiation have been made (Table 1). If the reaction
volved in these cyclopalladation reactions. Discussion of is carried out in the presence of palladium acetate (IIЈ) the
Scheme 1 has to be carried out very cautiously. Any of the intensity of the red colour increases with its concentration,
processes depicted (i.e. I, II, and III) does not necessarily and the induction period decreases. The kinetic results agree
have to be a single reaction; in fact, most of them have to with those of the experiments carried out in the absence of
be considered as rather complex reactions. Consequently, Pd(CH3COO)2, that is, the palladium acetate concentration
any rate constants and activation parameters discussed are does not affect the rate constant, only its extent. The deep
apparent values that can only be considered as such, despite red reaction mixture formed when the reaction was carried
the fact that can be very informative about the equivalence out with exo complexes, 2c and 2d, contains free aldehyde,
1
of processes between different systems.
as detected by H-NMR spectroscopy. For the endo com-
Scheme 1. Reactivity involved in the cyclopalladation reactions of imines in acetic acid
218
Eur. J. Inorg. Chem. 2000, 217Ϫ224

Mechanisms of Cyclopalladation Reactions in Acetic Acid
FULL PAPER
Table 1. Available kinetic data for the processes depicted in Scheme is always observed to some extent. Purification of the ob-
1. Apparent thermal activation parameters are obtained from 4Ϫ5
measures in a range of at least 35 degrees and apparent pressure
tained product by column chromatography produces the fi-
nal acetato-bridged dimer plus some quantities of an un-
identified red material. The existence, in the acetic acid re-
action medium, of palladium acetate and the final cyclome-
tallated complex should enable the formation of
polynuclear species according to reaction sequence I and
IIЈ. Kinetic measurements at variable temperature and
pressure are available for this reaction sequence starting
from imine and palladium acetate. Table 1 collects the data
corresponding to the process that follows the CϪH bond
activation already described in previous studies.[7] The re-
sults agree with those obtained from the reactions carried
out with the crude isolated yellow acetato dimers, 1a, 1b,
2c, and 2d (processes II and IIЈ).
activation parameters from 5Ϫ6 measures in a range of 1500 atm.
Parameters indicated for process labeled as I ϩ II correspond to the
process following the CϪH bond activation, i.e. process II, under
synthetic conditions starting from imine plus palladium acetate
mixtures
Imine[a] Process
k323
∆H°
∆S°
∆V°
sϪ1
kJ·molϪ1 J·KϪ1molϪ1 cm3·molϪ1
a
I ϩ II
II
1.5·10Ϫ4
1.4·10Ϫ4
8.5·10Ϫ5
1.7·10Ϫ4
3.0·10Ϫ5
1.5·10Ϫ5
1.0·10Ϫ4
9.7·10Ϫ5
51
Ϫ163
ϩ19
id I ϩ II id I ϩ II
id I ϩ II id I ϩ II
78
55
70
id II
70
66
IIЈ[b]
I ϩ II
II
b
c
Ϫ74
ϩ11
Ϫ24
ca. 0
Ϫ147
Ϫ123
id II
Ϫ108
Ϫ89
III
d
I ϩ II
II
Ϫ25
Figure 1 collects all the spectroscopic changes observed
in the UV/Vis spectrum for the sequence I ϩ II. For the
reactions carried out with exo compounds, 2c and 2d, keep-
ing these intense red solutions for 24 h at 80°C produces
the light yellow endo monomeric complexes, 3c and 3d, as
II initiation[c] 3.7·10Ϫ3
[d]
I ؉ II
2.2·10Ϫ5
61
Ϫ149
ca. 0
[a]
Only compounds with the best spectroscopic handles are inc-
luded. Ϫ [b] [Pd(CH3COO2)2] in the 5Ϫ15% range from that of the
cyclometallated complex. Ϫ [c] Initial fast process corresponding to
the appearance of aldehyde (see text). Ϫ [d] Reaction with too small indicated by NMR experiments in deuterated acetic acid
(process III).[7a] Kinetic measurements at variable tempera-
spectroscopic changes to be measured.
ture and pressure are available for this process starting from
plexes, 1a and 1b, the amount of aldehyde is less important, the isolated exo dimers, and these are also collected in Table
and appears after longer periods. Furthermore, practically 1. These endo, Cbenz.ϪPd bond containing, complexes seem
no free or metallated amine is found by NMR spectroscopy, to be the dead-end of the whole process given their stability
indicating that most of it remains coordinated to the pal- over very long periods in acetic acid.
ladium through the N atom. The presence of polynuclear
The results, although difficult to interpret unequivocally,
species in the red reaction mixture has been detected by can be rationalized by taking into account the different sta-
FAB and ion-spray (IS) mass-spectrometry. While mass bilities in acid medium of the Cbenz.ϪPd and Carom.ϪPd
spectra for the isolated yellow material (1a, 1b, 2c, and 2d) bonds, as well as the reactivity of the labile acetato bridging
indicate molecular peaks corresponding to the formulated groups. From the results presented, it is evident that, while
dimer complexes, the isolated red material following process the Carom.ϪPd bond-containing complexes can be hy-
II has peaks at masses that correspond to species containing drolyzed in acetic acid, the stability of the metallacycles
further acetato and palladium groups.
containing Cbenz.ϪPd bond is much higher. This stability
The IS mass spectrum of the red solid obtained from the can be related both to the nature of the CϪPd bond, and
reaction between palladium acetate and imine a shows to the final form of the complex in acetic acid (dimeric or
peaks up to 1121, while the molecular weight of the di- polynuclear in the case of Carom.ϪPd bond-containing com-
nuclear compound 1a is 786. Furthermore, proton NMR plexes and monomeric for the Cbenz.ϪPd bond-containing
spectroscopy of this red material shows an extremely com- complexes). The presence of monomeric cyclometallated
plicated pattern in the 2.0 ppm zone, indicating that many complexes containing Cbenz.ϪPd bonds in acetic acid that
different types of acetato ligands are present in the material, do not evolve during long periods indicate that these are
as should correspond to an undefined polynuclear com- the most thermodynamically favourable compounds under
pound. Further confirmation of our assumptions comes these conditions. These results can be related with the li-
from the fact that in spite of pure dinuclear acetato bridged gand exchange reaction of N-donor ligands on cyclopallad-
compounds usually being obtained in solid state after a ated compounds. It has been shown that PdII effectively mi-
purification process (recrystallization or column chroma- grates to the ligand, forming the palladacycle less suscep-
tography), the action of palladium acetate on 2-dimeth- tible to acidolysis.[16]
ylaminotoluene in acetic acid affords a trinuclear species, of
As for the details of the reactivity, from the kinetic and
general formula [Pd3(CϪN)2(µ2-O2CCH3)4] (CϪNϵmetal- activation parameters data, endo and exo cyclometallated
lated amine), as a mixture of cis and trans isomers [in re-
Carom.ϪPd bond-containing complexes present dramatic
lation to the Pd3(CH3COO)4 moiety].[14] In addition, the differences in their reaction mechanism. This can be attri-
formation of trinuclear species containing acetato bridges, buted to the greater ease of PdϪC acidolysis of exo-metal-
isostructural to palladium acetate, has recently also been lacycles[16c] and the hydrolysis of the CϭN imine bond, as
described for ligands containing C,S donor atoms.[15]
seen by NMR spectroscopy. Given the fact that the reaction
Process I ϩ II corresponds, de facto, to the standard syn- mixture contains “all the necessary ingredients”, and that
thetic procedure in acetic acid. Under these conditions, be- the cyclopalladated amine derivatives react with benzal-
sides cyclometallated complexes, the red coloured solution dehydes to form imine endo-metallacycles (see below), long
Eur. J. Inorg. Chem. 2000, 217Ϫ224
219

´
M. Gomez, J. Grandell, M. Martinez
FULL PAPER
Figure 1. UV/Vis spectroscopic changes for the synthetic process, I ϩ II, at 50°C between palladium acetate (1·10Ϫ3 ) and imine a
(1·10Ϫ3 ) depicted in Scheme 1; above, metallation process (I); below, polynuclear formation process (II)
standing periods at high temperature of these red solutions good yields. These results have been explained by the gener-
produce the final dead-end monomeric endo Cbenz.ϪPd ation of coordinatively unsaturated species that experience
bond-containing complex. On the other hand, the induction the metallation.[11][17] The cyclopalladated derivative of
period for the endo metallacycle for the same process has benzylamine was prepared by refluxing a 1:1 mixture of pal-
to be the formation of palladium acetate via the initial ladium acetate and the amine in acetonitrile, as has been
C
arom.ϪPd endo bond hydrolysis, given the fact that the reported,[17e] and the cyclopalladation of the 2-methylben-
presence of palladium acetate seems to somehow reduce zylamine was accomplished under the same reaction con-
this induction period. From then on, the complex reaction ditions. The dimeric compound obtained in this latter case
to produce polynuclear red material (and a certain degree [Pd(CH3COO)(3-CH3C6H3CH2NH2)]2 (4f) was charac-
of decomposition) takes place with kinetic and activation terized by elemental analyses and NMR spectra. The
parameters independent of the presence of palladium acet- monomeric
triphenylphosphane
complex
[PdBr(3-
ate in the reaction medium. The pressure activation param- CH3C6H3CH2NH2)(PPh3)] (5f) was prepared by reaction of
eters derived for the processes agree with this differen- the acetate bridged derivative with LiBr and PPh3 in ace-
tiation. While the activation volume for the polynuclear for- tone.
mation process for the a and b imines endo Carom.ϪPd
bond-containing complexes is positive, in spite of the ex-
The action of palladium acetate on these primary amines
tremely negative activation entropy, the value measured for in acetic acid was also studied; in this case the addition of
the process that takes place for the exo Carom.ϪPd bond- deuterated pyridine to a deuterated chloroform solution of
containing complexes is very negative. Kinetic factors seem the compounds obtained, shows the formation of the
to be dominant in the differentiation of these two processes [Pd(CH3COO)(CϪN)([D5]py)] (CϪN ϵ e or f metallated
as seen for the inverse cyclometallation reaction already amine, see Scheme 2) derivatives, indicating that the metall-
studied.
ation of these ligands can take place even in acetic acid.
In view of the interest of the processes labeled as II and The proton NMR spectra of the compounds obtained in
III for 2c and 2d, the action of palladium acetate on pri- absence of deuterated pyridine show, in addition to the sig-
mary amines has also been studied. It is generally accepted nals of the dimeric derivatives, new signals at δ ϭ 2.2Ϫ1.7
that primary amines are inert towards cyclometallation re- suggesting that, under these reaction conditions, polynu-
actions, but they can undergo cyclopalladation if exper- clear acetato bridged complexes are also formed. This fact
imental conditions are appropriate. The action of AgClO4 has been confirmed by mass spectra. The overall process
on coordination compounds [PdCl2L2] (Lϭprimary amine), depicted in Scheme 2 and its study from a kinetic point of
or the action of palladium acetate on the amines in a 1:1 view has also been carried out. Again, reaction IV takes
ratio leads to the cyclopalladation of primary amines with place by the initial fast formation of the 1:1 coordination
220
Eur. J. Inorg. Chem. 2000, 217Ϫ224

Mechanisms of Cyclopalladation Reactions in Acetic Acid
FULL PAPER
Scheme 2. Reactivity involved in the cyclopalladation reactions of amines in acetic acid
compound in stoichiometric conditions, followed by the
Table 2. Kinetic and activation parameters for the cyclometallation
reaction of palladium acetate with the amines e and f indicated in
Scheme 2 in toluene and acetic acid solutions (process IV); litera-
ture data from the related imines c and d from Scheme 1 are also
included.[7]
C
arom.ϪH bond activation.
The metallation process has been studied kinetically in
toluene solution at variable temperature and pressure. Table
2 collects the kinetic and thermal and pressure activation
parameters for this process together with relevant compara-
tive data; the results in solution agree with those obtained
for the already published[7] metallation reactions. For the
reactions studied in acetic acid solution, the activation par-
ameters show important differences with respect to the sys-
tems already published. These observations can be ex-
plained by the fact that the metallation product obtained in
acetic acid, although containing the palladacycle depicted
in Scheme 2, does not correspond to that isolated according
to literature procedures. As in the processes depicted in
Scheme 1, the isolated solid dimer prepared according to
the literature procedure is stable indefinitely, but when dis-
solved in acetic acid evolves to produce intense red coloured
solutions. This reaction has been studied, both in neat
acetic acid and in an acetic acid solution of palladium acet-
ate, on the crude isolated acetato dimer (processes V and
VЈ respectively). After an induction period, the formation
Ligand Solvent
104k323 ∆H°
∆S°
∆V°
sϪ1
kJ·molϪ1 J·KϪ1molϪ1
cm3·molϪ1
e
Toluene
2.6
73
100
55
67
49
62
69
Ϫ91
ϩ3
Ϫ145
Ϫ100
Ϫ138
Ϫ120
Ϫ91
Ϫ16
Ϫ11
Ϫ20
Ϫ12
Ϫ13
Ϫ17
Ϫ13
Acetic acid[a] 76
f
c
Toluene
Toluene
Acetic acid
Toluene
Acetic acid
2.5
70
13[b]
9.2
d
2.2[b]
[a] In acetic acid solution process IV cannot be completely separated
from process V, consequently the kinetic and activation parameters
[b]
indicated have a limited reliability (see text). Ϫ
298 K.
Measured at
Table 3. Available kinetic data for the processes depicted in Scheme
2. Apparent thermal activation parameters are obtained from 4Ϫ5
measures in a range of at least 35 degrees and apparent pressure
activation parameters from 5Ϫ6 measures in a range of 1500 atm.
Parameters indicated for process labeled as IV ؉ V correspond to
of an intense red coloured solution takes place; the kinetic the process following the CϪH bond activation, i.e. process V, un-
der synthetic conditions starting from amine plus palladium ace-
and activation parameters determined for this process are
the same, within experimental error, in both cases and are
tate mixtures
characterized by extremely negative activation entropies
and a negative volume of activation. Table 3 collects these
data together with those for all the processes depicted in
Scheme 2.
The induction period can be detected spectrophotometr-
ically, although it implies very small absorbance changes;
the kinetic and thermal activation parameters associated
indicate dramatic changes with respect to the other pro-
cesses involved in the overall reaction scheme. As seen in
Amine[a] Process
k323
∆H°
∆S°
∆V°
sϪ1
kJ·molϪ1 J·KϪ1molϪ1 cm3·molϪ1
e
IV ϩ V
V initiation[b]
V
2.5·10Ϫ4
3.4·10Ϫ3
1.2·10Ϫ4
82
121
46
Ϫ63
80
Ϫ180
ϩ14
>>0
Ϫ16
VЈ initiation[b,c] id V initiation[b]
VЈ[c]
id V
[a]
Only compounds with the best spectroscopic handles are inc-
[b]
[c]
luded. Ϫ Initial fast process (see text). Ϫ [Pd(CH3COO2)2] in
Table 3, the activation enthalpy is extremely large and the the 5Ϫ15% range from that of the cyclometallated complex.
Eur. J. Inorg. Chem. 2000, 217Ϫ224
221

´
M. Gomez, J. Grandell, M. Martinez
FULL PAPER
value found for the activation entropy is very positive. As proven to be partly due to the hydrolysis reaction of the
for the activation volume, an extremely large positive value CϭN bond, for compounds 4e and 4f this reaction has to
is estimated from the experiments. Its determination has not be related solely to the hydrolysis of the main Carom.ϪPd
been possible due to the fact that there is an important bond. It is also interesting to note that, although in the
pressure retardation of the initiation process with small presence of palladium acetate in the acetic acid solution
spectroscopic changes implied, and that the above men- medium this initiation also exists, it involves smaller spec-
tioned process has a large pressure acceleration with im- tral changes. It seems evident that the initiation process in
portant differences in the spectra. The presence of polynu- this reaction needs the presence of Pd(CH3COO)2 in order
clear species in the red reaction mixture is clearly indicated to start the formation of the red polynuclear material. In
in solid state by mass spectrometry, as well as in solution this respect, the values measured for the activation param-
(1H-NMR spectroscopy). Peaks up to 757 are observed in eters (specially the activation volume) for the synthetic pro-
the FAB mass spectrum of the red solid obtained from the cess IV ϩ V, have to be looked into more carefully. Given
reaction of palladium acetate and amine e. The molecular the fact that the Carom.ϪH bond activation of the amines
weight of the dinuclear compound 4e is 543. From gas chro- (process IV) is much slower (see Table 2) than the fast in-
matography analysis of the crude red solution small quan- itiation process of reaction V, detected when the isolated
tities of free amine are also detected. Under synthetic con- acetato dimer 4e is dissolved in acetic acid (see Table 3),
ditions in acetic acid solution (process IV ϩ V) the final the values determined for the polynuclear formation process
red coloured solution is always observed. Kinetic measure- starting from amine and palladium acetate mixtures, prob-
ments have been carried out at variable temperature and ably also contain a contribution from the CϪH bond acti-
pressure for the reaction following the CϪH bond acti- vation reaction, IV.
vation process; the results obtained for both thermal and
pressure activation parameters and collected in Table 3.
When para-chlorobenzaldehyde is added to the crude deep
Conclusions
red solution in acetic acid, condensation with the amine
In summary, our mechanistic studies have shown that the
takes place, producing the endo cyclometallated compound, reaction between imines and palladium acetate, in acetic
1a, derived from the corresponding imine as detected by acid solution, is a rather complex process in which several
NMR spectroscopy.
different reactions, such as acidolysis of PdϪC and CϭN
The results from the data in Scheme 2 have to be inter- bonds, and substitution on the acetato bridging positions
preted together with those for the imine derivatives. It is take place. It has also been found that cyclopalladated am-
noteworthy to indicate that the behaviour of the cyclometal- ine compounds react with benzaldehydes to afford the cor-
lated amine has to be directly related to the reactivity of responding metallated imines in the endo form. Even the
the exo cyclometallated complexes, 2c and 2d, containing structures generally assumed for all these complexes are not
Carom.ϪPd bonds, once the initial CϭN bond hydrolysis so simple and they strongly depend on the reaction con-
has taken place (see above). The fact that formation of an ditions. Under smooth conditions (25°C, 1 h), dimeric spec-
intense red coloured solution is observed, both in synthetic ies containing five-membered metallacycles in endo (1a and
conditions in acetic acid and by further treatment of the 1b) and exo (2c and 2d) conformations, are formed and can
isolated acetato dimers in neat acetic acid or acetic acid be isolated as yellow solids in good yields. Longer reaction
solutions of palladium acetate, is a clear indication that the periods or higher temperatures induce the formation of po-
reactivity is related to the acidolysis of the Carom.ϪPd bond lynuclear red materials due to the reactivity of the acetato
in the complexes. Again mass spectra indicate that import- bridging groups, as well as the reactivity of the CϪPd and
ant amounts of polynuclear species are present in the iso- CϭN bonds. Obviously these more drastic reaction con-
lated red material. Nevertheless, even in these intense red ditions favour the formation of the more thermo-
polynuclear containing solutions, the presence of cyclomet- dynamically stable metallacyclic species (endo six-mem-
allated complexes is always detected, no full decomposition bered monomeric cycles versus exo five-membered dimeric
of the cyclopalladated compound is taking place.
cycles, for ligands c and d). In any case, the reaction mixture
Reaction of these red solutions with the corresponding contains polynuclear species that can be detected, both in
aldehyde produces solutions where the endo cyclometallated solution by NMR spectroscopy and in solid state by mass
complex 1a, already described, is detected. We assume that spectrometry. The cyclometallation of amines e and f in
the presence of the free amine in the polynuclear material acetic acid has also been accomplished for the first time; in
solutions allows for the amine-aldehyde condensation;[20] an analogous way, dimeric cyclopalladated compounds are
thereafter the cyclometallation process, already described initially formed, but the existence of dynamic processes af-
for the imine derivatives, takes place. As for the kinetic and fords some final polynuclear species.
activation parameters measured, an interesting fact is that
Deeper interpretation of the kinetic and activation data
those related to reaction V are in perfect agreement with available would be speculative at this point, given the multi-
those measured for the appearance of the polynuclear-con- tude of steps taking place in any of the processes depicted
taining red solutions for the reactions of the exo imine de- in the reaction schemes. Further detailed studies for some
rivative complexes, II. In both cases, an initiation reaction of these systems are currently under way in order to estab-
is detected, while for the exo imine complexes it has been lish a better kinetic understanding.
222
Eur. J. Inorg. Chem. 2000, 217Ϫ224

Mechanisms of Cyclopalladation Reactions in Acetic Acid
FULL PAPER
instrument equipped with a multicell transport, thermostated
(±0.1°C) with a circulation bath. Observed rate constants were de-
rived from the absorbance versus time traces at the wavelengths
where a maximum increase or decrease of absorbance was ob-
served. The general kinetic technique is that previously de-
scribed.[13] The concentration range for the different compounds
used in this work is the same as that in previous papers;[7] no de-
pendence on the different concentrations has been found for any
of the rate constants in the 0.7Ϫ1.3·10Ϫ3  range. For runs at high
pressure, a previously described pressurizing system and high press-
ure cell were used.[7][13] In these cases, the absorbance versus time
traces were recorded on a J&M TIDAS instrument. Rate constants
were derived from exponential least-square fitting by the standard
routines. Least-square errors for the rate constants were always in
the range of 10Ϫ15% of the calculated value. All post-run fittings
to rate laws were done by standard, commercially available fitting
programmes.
Experimental Section
General: 1H- and 31P{1H}-NMR spectra were recorded on a Varian
XL-200 (200 MHz) or Bruker 250 DRX (101.25 MHz) spec-
trometer. Chemical shifts (in ppm) are relative to SiMe4 for 1H and
to 85% H3PO4 for 31P spectra. The solvents used were CDCl3 for
1H spectra and CHCl3 for 31P spectra. Ϫ IR spectra were recorded
as KBr disks on a Nicolet 520 FT-IR spectrometer. Ϫ GC analyses
were performed on a Hewlett-Packard 5890 Series II gas chromato-
graph (50-m Ultra 2 capillary column 5% phenylmethylsilicone and
95% dimethylsilicone) with a FID detector. Ϫ Microanalyses were
´
´
`
performed by the Institut de Quımica Bio-Organica de Barcelona
`
(CSIC) and by the Serveis Cientıfico-Tecnics de la Universitat de
Barcelona. Ϫ Mass spectra (FAB and IS) were recorded on a Mic-
romass VG-Quattro spectrometer; the samples were introduced in
a matrix of 2-nitrobenzyl alcohol for FAB spectra and a mixture
of water/acetonitrile was used as eluent for ion-spray spectra. Ϫ
Imines a؊d and cyclopalladated compounds 1a, 1b, 2c, 2d, and 4e
were prepared according to literature procedures.[12a,17e,18]
[Pd(CH3COO)2(4-ClC6H4CH؍NCH2C6H5)2]: 0.100 g (0.45·10Ϫ3
mol) of palladium acetate and 0.21 g of a (0.90·10Ϫ3 mol) were
dissolved in 10 cm3 of toluene and stirred at room temperature for
2 hours, affording a pale yellow precipitate. The solid was filtered
off and washed successively with toluene and diethyl ether, and
Acknowledgments
We acknowledge financial support from the DGICYT (Projects
PB97Ϫ0407-C05Ϫ04, PB96-0164, PB97-0914), CIRIT and the Di-
dried under reduced pressure. Yield: 0.24 g (80%).
Ϫ
´
reccio General dЈUniversitats (Generalitat de Catalunya). Helpful
C32H30Cl2N2O4Pd (683.91): calcd. C 56.22, H 4.42, N 4.10; found
comments from Drs. Joan Albert and Margarita Crespo are grate-
fully acknowledged.
1
C 56.26, H 4.48, N 4.38. Ϫ H NMR: δ ϭ 9.06 (br. s, 1 H, CHϭ
N), 7.70Ϫ7.10 (m, 9 H, aromatic), 5.39 (s, 2 H, CH2N), 1.60 (s,
3 H,CH3).
Synthesis of 1a: 0.080 g of 4e (0.15·10Ϫ3 mol) and 0.082 g of 4-
chlorobenzaldehyde (0.60·10Ϫ3 mol) were treated with 5 cm3 of
acetic acid at 75°C overnight. The solution was filtered over celite,
and the solution was concentrated under reduced pressure. After
addition of diethyl ether, a yellow solid was precipitated. The solid
1a was filtered off and washed successively with hexane and diethyl
ether, and dried in vacuo.
[1] [1a]
M. I. Bruce, Angew. Chem. Int. Ed. Engl. 1977, 16, 73. Ϫ
[1b]
G. R. Newkome, W. E. Puckett, W. K. Gupta, G. E. Kiefer,
[1c]
Chem. Rev. 1986, 86, 451. Ϫ
I. Omae, Coord. Chem. Rev.
[1d]
1988, 83, 137. Ϫ
V. Dunina, O. A. Zalevskaya, V. M. Po-
tatov, Russ. Chem. Rev. 1988, 57, 250.
[2] [2a]
[2b]
A. D. Ryabov, Synthesis 1985, 233. Ϫ
Trav. Chim. Pays Bas. 1990, 109, 567.
M. Pfeffer, Recl.
[3]
[3a] M. Camargo, P. Dani, J. Dupont, R. F. de Souza, M. Pfeffer,
I. Tkatchenko, J. Mol. Catal. 1996, 109, 127. Ϫ [3b] R. Navarro,
E. P: Urriolabeitia, C. Cativiela, M. D. Diaz de Villegas, M. P.
[Pd(CH3COO)(3-CH3C6H3CH2NH2)]2 (4f): 0.3 g (0.25·10Ϫ3 mol)
of 2-methylbenzylamine, f, were treated with palladium acetate
(0.25·10Ϫ3 mol, 0.55 g) in acetonitrile at reflux for 4 h and then
filtered. The solution was concentrated in vacuo, and the solid,
obtained after addition of diethyl ether was recrystallised from
chloroform/diethyl ether to obtain 4f in 60% yield (0.4 g). Anal.
[found (calcd. for C20H26N2O4Pd2 (571.24): C 42.3 (42.05), H 4.7
´
Lopez, E. Alonso, J. Mol. Catal. 1996, 105, 111.
[4]
[5]
[6]
P. Espinet, M. A. Esteruelas, L. A. Oro, J. L. Serrano, E. Sola,
Coord. Chem. Rev. 1992, 117, 215.
´
´
C. Navarro-Ranninger, I. Lopez-Solera, J. M. Perez, J. R. Masa-
guer, C. Alonso, Appl. Organomet. Chem. 1993, 7, 57
[6a] A. D. Ryabov, Chem. Rev. 1990, 90, 403. Ϫ [6b] A. D. Ryabov,
I. K. Sakodinskaya, A. K. Yatsimirsky, J. Chem Soc., Dalton
Trans. 1985, 2629.
1
(4.59), N 5.0 (4.90)%. Ϫ H NMR: δ ϭ 7.25 (m, 1 H, aromatic),
[7] [7a]
´
M. Gomez, J. Granell, M. Martinez, J. Chem. Soc., Dalton
[7b]
6.80 (br m, 2 H, aromatic), 4.00 (br m, 2 H), 3.05 (br. s, 2 H), 2.05
(br. s, 6 H, CH3ϪC6H3, CH3COO). 1H NMR (4fϩ[D5]py): δ ϭ
7.15 (m, 1 H, aromatic), 7.05 (br m, 1 H, aromatic), 6.0 (br d, 1 H,
aromatic), 4.62 (br. s, 2 H, NH2), 4.42 (s, 2 H, CH2N), 2.26 (s, 3
H, CH3), 1.98 (s, 3 H, CH3COO).
[PdBr(3-MeC6H3CH2NH2)(PPh3)] (5f): 0.15 g (0.25·10Ϫ3 mol) of
4f were treated with PPh3 (0.50·10Ϫ3 mol, 0.14 g) and LiBr
(0.5·10Ϫ3 mol, 0.046 g) in acetone at room temperature for 45 min
and then filtered. The solution was concentrated in vacuo, and the
solid, obtained after addition of diethyl ether was recrystallized
´
M. Gomez, J. Granell, M. Martinez,
Trans. 1998, 37. Ϫ
Organometallics 1997, 16, 2539.
[8]
[9]
R. P. Thummel, Y. Jahng, J. Org. Chem. 1987, 52, 73.
T. Yagyu, S. Iwatsuki, S. Aizawa, S. Funahashi, Bull. Chem.
Soc. Jpn. 1998, 71, 1857.
[10]
[11]
A. J. Canty, G. van Koten, Acc. Chem. Res. 1995, 28, 406.
J. Vicente, I. Saura-Llamas, P. G. Jones, J. Chem. Soc., Dalton
Trans. 1993, 3619.
[12] [12a]
´
J. Albert, R. M. Ceder, M. Gomez, J. Granell, J. Sales,
[12b]
Organometallics 1992, 11, 1536. Ϫ
J. Albert, J. Granell, J.
´
Sales, M. Font-Bardıa, X. Solans, Organometallics 1995, 14,
1393. Ϫ [12c] J. Albert, A. Gonzalez, J. Granell, R. Moragas, X.
´
´
Solans, M. Font-Bardıa, J. Chem. Soc., Dalton Trans. 1998,
from chloroform to obtain 5f in 80% yield (0.23 g).
Ϫ
1781.
[13] [13a]
C26H25BrNPPd (568.77): calcd. C 54.91, H 4.43, N 2.46; found C
54.5, H 4.5, N 2.60. Ϫ 1H NMR: δ ϭ 7.90Ϫ7.70 (m, 6 H, aro-
matic),7.60 7.40 (m, 9 H, aromatic), 6.75 (d, 7.2 Hz, 1 H, aromatic),
6.46 (t, 7.2 Hz, 1 H, aromatic), 6.36 (br t, 1 H, aromatic), 4.46 (s,
2 H, CH2N), 4.05 (br. s, 2 H, NH2), 2.35 (s, 3 H, CH3). Ϫ 31P{1H}
NMR: δ ϭ 41.5 (s).
M. Crespo, M. Martinez, J. Sales, X. Solans, M. Font-Bar-
dia, Organometallics 1992, 11, 1288. Ϫ [13b] M. Crespo, M. Mar-
tinez, J. Sales, Organometallics 1993, 12, 4297. Ϫ [13c] G. Gonza-
´
lez, P. Lahuerta, M. Martinez, M. Sanau, E. Peris, J. Chem.
[13d]
Soc., Dalton Trans. 1994, 545. Ϫ
F. Estevan, P. Lahuerta,
´
´
´
E. Peris, M. A. Ubeda, S. Garcıa-Granda, F. Gomez-Beltran,
´
´
E. Perez Carren˜o, G. Gonzalez, M. Martinez, Inorg. Chim. Acta
1984, 218, 189. Ϫ [13e] F. Estevan, G. Gonzalez, P. Lahuerta, M.
´
Kinetics: The reactions at atmospheric pressure were followed by
UV/Vis spectroscopy in the full 750Ϫ300 nm range on a HP8542A
Martinez, E. Peris, R. van Eldik, J. Chem. Soc., Dalton Trans.
1996, 1045.
Eur. J. Inorg. Chem. 2000, 217Ϫ224
223

´
M. Gomez, J. Grandell, M. Martinez
FULL PAPER
[14]
´
M. Pfeffer, E. Wehman, G. van Koten, J. Organomet. Chem.
Vicente, I. Saura-Llamas, M. Palin, P. G. Jones, M. C. Ramırez
1985, 282, 127.
de Arellano, Organometallics 1997, 16, 826.
J. Albert, J. Granell, J. Sales, J. Organomet. Chem. 1984, 273,
393.
[15]
[18]
[19]
M. Basato, C. Bertani, M. Zena, A. Grassi, G. Valle, J. Or-
ganomet. Chem. 1999, 575, 163.
[16] [16a]
[16b]
A. D. Ryabov, Inorg. Chem. 1987, 26, 1252. Ϫ
A. D.
Formation of bis-coordination complexes, [Pd(CH3COO)2(im-
ine)2], has also been achieved by precipitation from toluene
solution when a 1:2 Pd/imine ratio is used. in addition, further
treatment of [Pd(CH3COO)2(a)2] in acetic acid for 3 h at 70 °C
produces mainly mono-cyclometallated acetato dimer plus free
imine.
The direct condensation of e plus p-clorobenzaldehyde in acetic
acid yields quantitatively imine a. In addition, after 3 hours in
refluxing acetic acid only a 3% of decomposition of imine a is
observed as monitored by GC chromatography.
Received May 27, 1999
[16c]
Ryabov, A. K. Yatsimirky, Inorg. Chem. 1984, 23, 789. Ϫ
´
R. M. Ceder, M. Gomez, J. Sales, J. Organomet. Chem. 1989,
[16d]
361, 391. Ϫ
J. Granell, D. Sainz, J. Sales, X. Solans, M.
Font-Altaba, J. Chem. Soc., Dalton Trans. 1986, 1785.
[17] [17a]
A. Avshu, R. D. OЈSullivan, A. W. Parkins, N. W. Alcock,
R. M. Countryman, J. Chem. Soc., Dalton Trans. 1983, 1619.
[20]
Ϫ
[17b] Y. Fuchita, H. Tsuchiya, A. Miyafuji, Inorg. Chim. Acta,
[17c]
1995, 233, 91. Ϫ
J. Vicente, I. Saura-Llamas, M. G. Palin,
[17d]
P. G. Jones, J. Chem. Soc., Dalton Trans. 1995, 2535. Ϫ
J.
´
Albert, J. Granell, A. Luque, J. Mınguez, R. Moragas, M. Font-
[17e]
´
Bardıa, X. Solans, J. Organomet. Chem. 1986, 522, 87. Ϫ
J.
[I99195]
224
Eur. J. Inorg. Chem. 2000, 217Ϫ224