Coordination insertion reactions of acrylonitrile into Pd-H and Pd-methyl bonds in a diimine-palladium(II) system

Article abstract of DOI:10.1016/j.jorganchem.2005.07.031

Acrylonitrile (AN) displaces the ethyl ether ligand of the cationic complex [Pd(N-N)Me(Et₂O)]⁺ (N-N = (2,6-(i-Pr) ₂C₆H₃)-NCMeCMeN-(2,6-(i-Pr)₂C ₆H₃)) to form the N-bonded AN complex [Pd(N-N)Me(AN)] ⁺, which exists as two interconverting rotamers. On standing or heating, [Pd(N-N)Me(AN)]⁺ undergoes 2,1-insertion to give [Pd(N-N)(CH(CN)CH₂Me)(AN)]⁺, which undergoes β-hydrogen elimination to give the intermediate hydride, [Pd(N-N)H(AN)]⁺, which in turn inserts AN to give the cyanoethyl complex [Pd(N-N)(CH(CN)Me)]⁺. Dimerization of the [Pd(N-N)(CH(CN)CH₂CH₃)]⁺ moiety via bridging nitrile groups also occurs, giving the dicationic species [Pd(N-N)(CH(CN)CH2Me)]22+. Although [Pd(N-N)Me(AN)]⁺ does behave as a typical Brookhart ethylene polymerization catalyst, it does not catalyze AN polymerization and in fact added AN suppresses ethylene polymerization.

Full text of DOI:10.1016/j.jorganchem.2005.07.031

Journal of Organometallic Chemistry 690 (2005) 4349–4355
www.elsevier.com/locate/jorganchem
Coordination insertion reactions of acrylonitrile into Pd–H
and Pd–methyl bonds in a diimine-palladium(II) system
*
Goran Stojcevic, Ernest M. Prokopchuk, Michael C. Baird
Department of Chemistry, QueenÕs University, Kingston, Ont., Canada K7L 3N6
Received 5 June 2005; received in revised form 4 July 2005; accepted 8 July 2005
Available online 11 August 2005
Abstract
Acrylonitrile (AN) displaces the ethyl ether ligand of the cationic complex [Pd(N-N)Me(Et2O)]+ (N-N = (2,6-(i-Pr)2C6H3)-
N@CMeCMe@N-(2,6-(i-Pr)2C6H3)) to form the N-bonded AN complex [Pd(N-N)Me(AN)]+, which exists as two interconverting
rotamers. On standing or heating, [Pd(N-N)Me(AN)]+ undergoes 2,1-insertion to give [Pd(N-N)(CH(CN)CH2Me)(AN)]+, which
undergoes b-hydrogen elimination to give the intermediate hydride, [Pd(N-N)H(AN)]+, which in turn inserts AN to give the cyano-
ethyl complex [Pd(N-N)(CH(CN)Me)]+. Dimerization of the [Pd(N-N)(CH(CN)CH2CH3)]+ moiety via bridging nitrile groups also
2þ
occurs, giving the dicationic species ½PdðN-NÞðCHðCNÞCH2MeÞꢀ . Although [Pd(N-N)Me(AN)]+ does behave as a typical Brook-
2
hart ethylene polymerization catalyst, it does not catalyze AN polymerization and in fact added AN suppresses ethylene
polymerization.
Ó 2005 Elsevier B.V. All rights reserved.
1. Introduction
has been found to be induced by cationic metallocene
species, the mechanism involves not coordination poly-
merization but rather a group transfer process [3].
While many polar monomers are readily polymerized
via radical and ionic processes [4], the resulting poly-
meric products are generally atactic in nature and no
highly stereoselective procedures giving largely iso- or
syndiotactic materials are as yet available. In addition,
since ethylene and propylene are not readily polymer-
ized via radical or ionic processes, there are as yet no
generally satisfactory routes to the formation of linear
ethylene- or propylene-polar monomer copolymers. Gi-
ven the properties which materials of these types are ex-
pected to exhibit [1c,2a,5], it is not surprising that
research in this area has been intense.
Although there have in recent years been numerous
investigations into the use of group 4 metallocene
complexes, of the type ½Cp02MRꢀþ (M = Ti, Zr, Hf;
R = alkyl; Cp0 = substituted cyclopentadienyl), as
homogenous catalysts for the coordination (Ziegler)
polymerization of olefins [1], this class of compounds
has not normally been found useful for either the hom-
opolymerization of polar olefins such as acrylates, meth-
acrylates, vinyl halides and vinyl amines or for the
copolymerization of these monomers with e.g. ethylene
or propylene [2]. In general, it seems, the Lewis base sites
of all such candidate monomers coordinate preferen-
tially to the highly Lewis acidic metal ions, thus preclud-
ing g2-coordination and the migratory insertion
processes necessary for coordination polymerization.
Where polymerization of, e.g., methyl methacrylate
One potential solution to the problem of coordina-
tion of a polar monomer via the Lewis base site is to uti-
lize relatively weakly electrophilic metal complexes, and
considerable research has been carried out with late
metal catalysts, especially of palladium [6]. Cationic a-
diimine complexes of the type [PdMe(L)(a-diimine)]+
*
Corresponding author. Fax: +1 613 533 6669.
E-mail address: bairdmc@chem.queensu.ca (M.C. Baird).
0022-328X/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.07.031

4350
G. Stojcevic et al. / Journal of Organometallic Chemistry 690 (2005) 4349–4355
(A: L = solvent, neutral ligand), which are effective cat-
alysts for the polymerization of ethylene, also catalyze
the copolymerization of ethylene with acrylates and
methacrylates [6].
contrast to the situation with vinyl chloride where facile
b-chloride migration occurs to give a catalytically inac-
tive species [7], b-cyano migration in a species of the
type Pd–CH2CH(CN)Me would probably be much less
facile because of the greater bond strength of the C–
CN bond (ꢁ132 kcal/mol) than of the C–Cl bond
(ꢁ94 kcal/mol) [11].
+
i-Pr
i-Pr
N
N
We have been investigating the chemistry of AN with
[Pd{(2,6-(i-Pr)2C6H3)-N@CMeCMe@N-(2,6-(i-Pr)2C6-
H3)}Me(OEt2)][B{(3,5-(CF3)2C6H3)4}4] (1), a complex
of type A (L = Et2O) and a very active catalyst for eth-
ylene, propylene and 1-hexene polymerization [6b]. We
find that the ethyl ether of 1 is readily substituted by
AN to give the isolable AN complex [Pd(N-
N)Me(AN)][B{(3,5-(CF3)2C6H3)4}4] (2), and that the
latter reacts further to give new complexes which con-
tain inserted AN molecules. As this work was being
completed, Wu et al. reported [12a] a somewhat similar
but more extensive study of the AN coordination and
insertion chemistry of 2 and related cationic complexes
while Groux et al. reported [12b] a complementary study
of neutral and anionic complexes. However, while the
former report is especially relevant to this work, the fo-
cus is quite different. Whereas Wu et al. [12a] broadly
investigate the room temperature insertion and coordi-
nation chemistry of a series of six cationic complexes,
of which compound 2 receives virtually no attention,
our work is focused specifically on 2. We report two no-
vel and unprecedented observations for this type of
compound, that the N-bonded AN ligand exists in solu-
tion as two interconverting rotamers and that ring cur-
rent effects of the aryl rings of the diimine ligand
generate unusual and heretofore unnoted chemical shifts
of the other coordinated ligands. We also impose rela-
tively forcing reaction conditions on 2, thereby inducing
and observing previously unreported AN insertion
chemistry, and we investigate the influence of AN
on ethylene polymerization by this catalyst system.
Henceforth, the diimine ligand (2,6-(i-Pr)2C6H3)-
N@CMeCMe@N-(2,6-(i-Pr)2C6H3) will be designated
Pd
i-Pr
Me
i-Pr
L
A
However, the efficacies of such catalyst systems are
affected adversely by the inclination of incorporated es-
ter-containing units to coordinate to the metal, thus
decreasing catalytic activity significantly. Attempts to
use similar catalysts to polymerize vinyl chloride also re-
sulted in an initial 1,2-vinyl chloride insertion into the
Pd–methyl bond, but the thus formed Pd–CH2CHClMe
species undergoes facile b-chloride migration to the me-
tal to form propylene and catalytically inactive chloro-
palladium compounds [7].
Given the at least partial functional group tolerance
of compounds of type A, we were interested in investi-
gating the chemistry of such a diiminepalladium(II)
complex with the polar monomer acrylonitrile (AN).
The commercially important material hydrogenated
acrylonitrile-butadiene rubber (HNBR) is formed by
the radical co-polymerization of AN with 1,3-butadiene
followed by hydrogenation of the backbone C@C bonds
[8]. HNBR is in effect a copolymer of AN with ethylene,
and a direct coordination polymerization route to AN–
ethylene co-polymers is therefore very desirable.
Computational investigations suggest that palladium
complexes of type A, where L = AN prefer r-coordina-
tion via nitrogen over p-bonding via the C@C bond
(henceforth g2-coordination) of AN by about 13 kcal/
mol, although interestingly the calculated barrier for
insertion of g2-coordinated AN is similar to that of eth-
ylene [9]. However, these computed relative energies are
very sensitive to the nature of the coordinated ligands
[9c]. and it seems quite possible that the effects of chang-
ing ligands, solvents and/or counteranions could reduce
the preference for N-coordination by a sufficient amount
that coordination polymerization processes involving
AN could become feasible. Furthermore, complexes of
type A, where L = AN and aryl nitriles have been used
as catalysts for the copolymerization of ethylene and
1-alkenes with methyl methacrylate, implying that sub-
stitution of N-bonded nitriles by olefins in this system
is facile [6d], and it has long been known that even such
weakly coordinating alkenes such as isobutene, pinene,
camphene and styrene readily displace the benzonitrile
ligands from PdCl2(PhCN)2 to give alkene complexes
of the type [PdCl2(alkene)]2 [10]. We note also that, in
as N–N, [B{(3,5-(CF3)2C6H3)4}4] as [BAr0 ].
4
2. Experimental
All reactions were carried out under purified argon
using standard Schlenk line techniques. Dichlorometh-
ane was dried by passing through a dried alumina col-
umn, CD2Cl2 over calcium hydride and DMSO-d6
1
over 4A molecular sieves. All H NMR spectra were
run on a Bruker AV500 spectrometer, chemical shifts
being referenced to TMS via the residual proton signals
of the deuterated solvents. Electrospray mass spectro-
metry (ESMS) experiments were run in positive and
negative ion modes on Quattro VG and Applied Biosys-
tems/MDS Sciex QSTAR XL instruments at a cone

G. Stojcevic et al. / Journal of Organometallic Chemistry 690 (2005) 4349–4355
4351
voltage typically of 20 V; nitrogen was used as the neb-
+
+
Hc
Ar
Ar
ulizing gas, CH2Cl2 as the carrier solvent. The ions were
in all cases complex multiplets, and the m/e values
quoted below are for the strongest line in each multiplet.
Elemental analyses were carried out by Canadian
Microanalytical Services of Delta, BC, and molecular
mechanics calculations were carried out using PCModel
version 8.0 (Serena Software, Bloomington, IN). Acry-
lonitrile was purchased from Aldrich and purified of
moisture and inhibitor by stirring over CaH2 for a
day; it was then vacuum distilled into a Schlenk tube
and stored in a refrigerator (3 °C). The compound
N
Ha
Hc
N
Hb
Pd
N
Pd
N C
C
N
N
Hb
Ar
Ha
Ar
CH3
CH3
2a
2b
Ar = 2,6-(i-Pr)2C6H3
The molecular ion of 2 (578 Da) was readily observed
in an ES mass spectrum, the isotope pattern of the
molecular ion agreeing closely with a calculated spec-
trum and clearly confirming the formulation of the com-
plex as a coordination complex in which AN had
replaced the ether ligand of 1. Carrying out the reaction
of 1 with AN in dichloromethane or ethyl ether resulted
in the same product being formed.
Wu et al. [12a] have recently prepared (via the
reaction of Pd(N-N)MeCl with [Li(Et2O)2.8][B(C6F5)4])
but not isolated the analogous compound [Pd(N-
N)Me(AN)][B(C6F5)4], containing the same cation but
a different anion. Aside from the additional [BAr04]ꢂ res-
½PdðN-NÞMeðOEt2Þꢀ½BAr0 ꢀ was prepared as in the liter-
4
ature [6b].
Synthesis of 2.
A
solution of 200 mg of
½PdðN-NÞMeðOEt2Þꢀ½BAr0 ꢀ (0.137 mmol) in 2 mL AN
4
was stirred under argon for 1 h. The AN was then re-
moved to give a dark orange/brown oil, which was
washed with hexanes and dried under reduced pressure
to give a yellow solid. This was washed by repeated
extractions with hexanes, and was dried overnight under
reduced pressure. Yield: 177 mg, 90%. 1H NMR
(CD2Cl2, 298 K) d 7.74 (s, 8H, BAr0 Ho), 7.58 (s, 4H,
4
1
BAr0 Hp), 7.43 (m, 2H, Haryl), 7.37 (m, 4H, Haryl), 6.22
onances in the H NMR spectrum of 2, the room tem-
4
(d, 1H, 3J(HH) = 12 Hz, CHHCHCN), 5.84 (d, 1H,
perature NMR spectra of the two compounds
compare well. Wu et al. [12a] and Groux et al. [12b] have
also identified in solution a number of similar Pd–AN
complexes, but in all cases, 2,1-insertion of the AN into
the Pd–Me bond occurred relatively rapidly and no cat-
ionic AN complexes were actually isolated.
3J(HH) = 18 Hz, CHHCHCN), 5.44 (dd, 1H, J(HH) =
3
12, 18 Hz, CH2CHCN), 2.92 (m, 4H, CHMe2), 2.26 (s,
3H, N@CCH3), 2.24 (s, 3H, N@CCH3), 1.38 and 1.26
(m, 12H each, CH(CH3)2), 0.58 (s, 3H, PdCH3). Anal.
Calc. for C64H58BF24N3Pd: C, 53.29; H, 4.05; N, 2.91.
Found C, 53.21; H, 4.07; N, 2.80%.
The olefinic chemical shifts in a room temperature 1H
NMR spectrum of 2 are informative with respect to
structure. On coordination, the resonance of Ha shifts
upfield from d 5.71 in free AN to d 5.45 (0.26 ppm), that
of Hb downfield from d 6.09 to d 6.21 (0.12 ppm) and
that of Hc upfield from d 6.22 to d 5.85 (0.37 ppm).
These coordination induced changes are very different
from those of known g2–AN complexes of metals in
the +2 oxidation state [13], as well as those of the anal-
ogous g2–propene complex (A, L = MeCH@CH2); in
these, all three vinyl hydrogen resonances shift upfield
by 0.75 ppm or more [6b]. In view of these observations,
in addition to the above mentioned computational re-
sults [9], we conclude that the AN of 2 is r-coordinated
via the nitrogen of AN. This conclusion is further sup-
ported by the nitrile stretching frequency shifting by
49 cmꢂ1 to a higher frequency in the IR spectrum [14].
Interestingly, variable temperature NMR studies of 2
suggest that this complex exists in solution as a mixture
of at least two interconverting, isomeric forms. Reduc-
ing the temperature from 298 to 193 K results in revers-
ible changes in the NMR spectrum, the Pd–methyl
resonance and the resonances of Hb and Hc shifting up-
field by 0.2, 0.07 and 0.3 ppm, respectively, although the
resonances of Ha and the diimine ligand change rela-
tively little.
Reactions of 2 with acrylonitrile. In a typical reaction,
35–45 mg of 2 were dissolved in 1 mL of AN and the
solution was stirred under argon (a) at room tempera-
ture for 12 h, (b) at 50 °C for 2 h, or (c) at reflux
(77 °C) for 12 h. As polyacrylonitrile was formed in
(c), a blank run, in the absence of 2, was also run. In
all cases, the AN was then removed under reduced pres-
sure and the resulting oily materials were washed with
hexanes, dried overnight under vacuum and character-
1
ized by H NMR spectroscopy and ESMS.
3. Results and discussion
Synthesis and structure of 2. Dissolution of the
orange complex 1 in AN resulted in an immediate color
change to yellow, suggesting displacement of the
ether by AN to produce the cationic nitrile complex
[Pd(N-N)Me(AN)]+ (2). After stirring for 1 h, the
solvent was removed under reduced pressure and the
1
resulting yellow powder was shown by H NMR spec-
troscopy, ESMS and elemental analyses to be
½PdðN-NÞMeðANÞꢀ½BAr40 ꢀ which may exist in two con-
formations, 2a and 2b (to be discussed below).

4352
G. Stojcevic et al. / Journal of Organometallic Chemistry 690 (2005) 4349–4355
We rationalize these observations in terms of rapid
50 °C for 2 h gave yellowish powders whose NMR spec-
tra and ES mass spectra were very similar. Thus, the 1H
NMR spectra of both products exhibited, in addition to
the resonances of unreacted 2, a new set of olefinic res-
onances at d 6.32 (d, J = 12.1 Hz), apparently corre-
sponding to Hb of a second AN complex, and a new
Ha resonance at d ꢁ5.45 (m, confirmed by a COSY
experiment). The new Hc resonance overlaps exactly
with the analogous resonance of 2 at d 5.85 (confirmed
by relative integrations and by a COSY experiment).
There were also new ligand resonances in the region d
1.0–3.3 and several new resonances in the region d
0–1.0 (an apparent singlet Pd–Me resonance at d 0.42,
a 1:2:1 triplet at d 0.23, a partially obscured multiplet
at ꢁd 0.41 and multiplets at ꢁd 0.8–0.9). Thus, the reac-
tions clearly produced complex mixtures.
ES mass spectra were run in an attempt to identify
the species in solution, and the mass spectra of the
two product mixtures were also found to be similar. In
addition to the molecular ion of 2 at 578 Da, both mass
spectra exhibited singly charged ions at 592, 631 and
647 Da and several doubly charged ions (ꢁ553, ꢁ564,
ꢁ572 and ꢁ586 Da), characterized by their 0.5 Da spac-
ings and very different isotope distribution patterns.
Although the assignments of the doubly charged species
are as yet unclear, the weak ion at 631 Da corresponds
to the mass of a cation of stoichiometry [Pd(N-
N)(CH2CHCN)2Me]+, containing two AN units in
addition to the basic [Pd(N-N)Me]+ moiety. The com-
plex presumably results from insertion of an AN unit
followed by coordination of a second N-bonded AN
to give a complex such as [Pd(N-N)(CHCNCH2-
Me)(CH2CHCN)]+ (3) (Eq. (1)).
interconversion between two or more conformations
with different chemical shifts. Molecular mechanics
(MM) calculations suggest that, because of the steric
requirements of the bulky isopropyl groups, the plane
of the coordinated AN ligand prefers to be positioned
approximately in the molecular plane. Two such confor-
mations are possible, one with Hc approximately eclips-
ing the Pd–methyl group (2a), the other with Ha
approximately eclipsing the Pd–methyl group (2b;
shown as a space filling model in Fig. 1). In the latter,
Hc would interact with the p face of the aryl ring and
hence experience significant shielding because of ring
current effects [15]. Indeed, our MM calculations suggest
that conformation 2b represents the energy minimum,
with close contacts between Hc and several of the ring
carbon atoms. Such is not the case for Ha and Hb in
either conformation, and thus changing the relative pop-
ulations of the possible conformations would result in
the greatest ring current effects being on the chemical
shift of the resonance of Hc, which is indeed observed.
Increasing the population of 2b apparently results in
an enhanced rate of rotation about the Pd–N bond
and increases in the population of rotamers other than
that shown in Fig. 1, resulting an upfield shift of the res-
onance of Hc. Thus, the observed temperature effects on
the chemical shift of Hc suggests that 2b is the more sta-
ble rotamer, consistent with the MM calculations.
Reactions of 2 with AN. Although solutions of the
[B(C6F5)4]ꢂ salt of 2 are reported to be stable with re-
spect to insertion at 23 °C for several days [12a], we find
that compound 2 does in fact engage in insertion reac-
tions within hours in the presence of free AN and that
secondary reactions occur under more forcing condi-
tions. Interestingly, reactions in which 2 was stirred in
neat AN either at room temperature for 12 h or at
+
+
Me
AN
N
N
N
N
CH(CN)Et
AN
AN
Pd
Pd
ð1Þ
2
3
A number of similar cationic species have been re-
ported recently, and all involve 2,1-insertions to give
1-cyanopropyl complexes for which d(CH) = 1.7–2.8
(m), d(CH2) = 1.0–1.7 (m) and d(Me) = 0.65–1.00 (t)
[12a]. While the spectra obtained here were obscured
in the range d 1.0–3.3 by very intense diimine reso-
nances, both did exhibit a 1:2:1 triplet at d 0.23. We ten-
tatively assign this triplet and the new set of AN olefinic
resonances to 3, and thus the NMR and ESMS data are
consistent with 3 being the product of a 2,1-insertion of
AN. A chemical shift of d 0.23 is highly unusual for the
methyl group of a 1-cyanopropyl ligand [12a], but may
be rationalized on the basis of ring current effects as dis-
cussed above for compound 2. Similar upfield shifts
have been reported for propylene insertion products of
the type [Pd(N-N)(CH2@CHMe)(CHMeCH2)nMe] [6b]
and, although not commented on at the time, may pos-
sibly be rationalized in the same way. In contrast, the
series of analogous complexes by Wu et al. [12a] and
Fig. 1. Space filling model of 2 showing how Hc lies close to an aryl
ring in the preferred conformation 2b. The Pd and nitrile N atoms and
the Pd–CH3 group are labeled to assist in orientation of the viewer.

G. Stojcevic et al. / Journal of Organometallic Chemistry 690 (2005) 4349–4355
4353
Groux et al. [12b], in which there are no ligands which
can induce ring current effects, exhibit normal chemical
shifts for the methyl group of the 1-cyanopropyl ligands.
The medium intensity ion at 592 Da exhibits an
isotope pattern appropriate for a complex of the stoichi-
ometry [Pd(N-N)(C4H5N)Me]+ (4; C4H5N = crotono-
nitrile, MeCH@CHCN). Crotononitrile is the product
of b-hydrogen elimination from complex 3 (Eq. (2)),
and the formation of 4 must involve release of the coor-
dinated crotononitrile followed by substitution by it of
the AN of a molecule of 2 (Eqs. (3) and (4)).
this species although a close examination of the 1H
NMR spectra in the region d 0 to ꢂ20 revealed no hy-
dride resonance. Thus, the species in solution may well
be the insertion product, [Pd(N-N)(CHCNMe)]+ (5,
Eq. (5)), again with the vacant site on the metal occupied
by an agostic interaction as with 4.
+
+
N
N
N
N
H
Pd
Pd
ð5Þ
AN
CHMeCN
5
A reaction was also carried out in which 2 was re-
fluxed in neat AN (77 °C) for 12 h and the resulting mix-
ture worked up as above. The 1H NMR spectrum of the
products formed under these much more forcing condi-
tions exhibited many resonances in the region d 1.0–3.3,
suggesting the presence of a complex mixture. Interest-
ingly, there were no resonances in the olefinic region
and the above mentioned triplet at d 0.23 and multiplet
(now seen to be a triplet) at d 0.41 and were much more
apparent while the singlet at d 0.42 had disappeared.
The ES mass spectrum exhibited multiplets at 509 (m),
564 (m), 578 (s), 598 (w), 605 (w), 631 (w), 649 (w)
and 663 (w) Da, in addition to several others at higher
masses. The multiplet at 509 Da exhibited an isotope
pattern consistent with its formulation as 6, i.e., a com-
pound in which metallation of one of the isopropyl
groups has occurred. This multiplet was not present in
the mass spectra of the reaction mixtures obtained under
less forcing conditions, and thus is probably a product
of decomposition. As with 4 and 5, it is possible that
the vacant site on the metal is occupied by an agostic
interaction.
+
+
N
N
N
N
H
CH(CN)Et
AN
-AN
Pd
Pd
NCH=CHMe
3
ð2Þ
+
N
N
H
Pd
NCH=CHMe
+
N
N
H
+AN
ð3Þ
Pd
+ MeCH = CHCN
AN
+
N
N
Me
Pd
+ MeCH=CHCN
AN
2
N
N
+
CH(CN)CHMe2
-AN
ð4Þ
Pd
4
It is difficult to convincingly rationalize the conver-
sion of 2–4 in the presence of excess AN, but crotononit-
rile does have a much higher boiling point (120 °C) than
does acrylonitrile (77 °C) and its relative concentration
would increase as the volatiles were removed during
workup. In any case a high resolution ES mass spectrum
was run to obtain an exact mass for the peak at 592 Da
and the experimental value, 592.2903 Da, differed by
only 3.3785 ppm from the calculated value for
C33H48N3Pd, 592.2882 Da. No other reasonable stoichi-
ometry was within 17 ppm of the experimental value,
and thus attribution of the multiplet at 592 Da to 4
seems firm. Since we observed no unexplained reso-
nances in the olefinic region of the NMR spectrum, it
seems likely that 4 contains an inserted C4H5N moiety,
as in [Pd(N-N)(CHCNCHMe2)]+ (Eq. (4)). It is possible
that the vacant site on the metal is occupied by an ago-
stic interaction with one of the i-Pr groups or with a
methyl of the CHCNCHMe2 group; ring strain would
seem to rule out chelation via the nitrile group.
+
i-Pr
i-Pr
i-Pr
N
N
Pd
C
H
C
H
H
Me
6
The peak at 564 Da, attributed above to 5, is much
stronger than it is in the mass spectra discussed above.
1
Again the H NMR spectrum of the reaction mixture
exhibited neither hydride nor olefinic resonances, and
thus this species seems best formulated as the product
of insertion. The multiplet at 631 Da has been dis-
cussed above, but those at 649 and 663 Da remain
unidentified.
Interestingly, while the main set of peaks of the mul-
tiplets at 578, 598 and 631 Da exhibit the isotope pat-
terns expected, all also contain sets of weaker peaks at
masses differing from those of the main peaks by
0.5 Da and exhibiting the isotope patterns expected for
the corresponding, dicationic dimers. Since the 1H
NMR spectrum of the reaction mixture showed clearly
The initially formed palladium-containing product of
b-hydrogen elimination from 3 (Eq. (2)) would be the
hydride complex [Pd(N-N)(C4H5N)H]+, isomeric with
2 and hence indistinguishable on the basis of mass spec-
trometry. A secondary product, [Pd(N-N)(AN)H]+,
would arise from substitution of the crotonitrile by
AN (Eq. (3)), and a weak peak in the ES mass spectrum
at 564 Da may be taken as evidence for the presence of

4354
G. Stojcevic et al. / Journal of Organometallic Chemistry 690 (2005) 4349–4355
that the molecules in solution did not contain coordi-
nated AN, the new species must be dimeric, as in 7.
lonitrile and they are shifted upfield slightly from those
in the NMR spectrum of 2. Thus, the resting state in this
catalytic system seems to be a complex of the type
[Pd(N-N)(AN)(polymeryl)]+.
H
Et
CN
2+
On bubbling ethylene for 3 min through solutions of
2 in several solvents in the presence of a few equivalents
of AN, we observe not copolymer formation but rather
serious inhibition of ethylene polymerization. The addi-
tion of a large excess of AN completely shuts down eth-
ylene polymerization. Complementary experiments in
which we attempted to homopolymerize AN in the pres-
ence of BPh3 also failed; the Lewis acidic triarylborane
apparently does not bind the nitrogen of the AN suffi-
ciently strongly to allow the olefinic moiety to coordi-
nate preferentially.
N
N
C
N
N
Pd
Pd
NC
Et
C
H
7
Thus, dicationic, dinuclear species are readily formed
in this system and probably involve bridging nitrile li-
gands as shown; similar results have been observed else-
where and, in fact, dinuclear species have been shown to
be a common secondary product following 2,1-insertion
of AN into a Pd–Me bond [12]. We therefore tentatively
assign the second methyl triplet resonance in the NMR
spectrum to the dimer 7.
4. Conclusions
The question arises then as to whether the monomeric
complexes exist in solution or whether they are formed
via dimer dissociation within the mass spectrometer. This
question cannot be answered, of course, but ES mass
spectra of similar platinum complexes have been shown
to reflect the species in solution [16a], and the monomeric
species may well be stabilized by agostic interactions as
above. Somewhat similar platinum containing species
have been noted elsewhere [16b]. Unfortunately, the iden-
tities of many of the species giving rise to the ions detected
in the mass spectrum are as yet unknown.
Polymerization reactions involving 2. Interestingly, it
was found that trace amounts of polyacrylonitrile were
formed during the 12 h reflux reaction; a 1H NMR spec-
trum of the resulting white precipitate in DMSO-d6
exhibited broad resonances at d 2.1 and 3.2, characteris-
tic of polyacrylonitrile [17]. However, in none of the
mass spectra obtained to date have there been any series
of peaks separated by 53 Da, as would be expected if
multiple insertions of AN at palladium were occurring
in this diimine system. Thus, while a single insertion into
a Pd–Me or Pd–H bond seems to occur slowly, subse-
quent insertions appear not to be achievable, a conclu-
sion also reached elsewhere [12]. The small amounts of
polyacrylonitrile obtained during the long reflux reac-
tion are presumably a result of radical polymerization
processes, possibly initiated by homolysis of Pd–CHC-
NEt bonds. Polyacrylonitrile does not form during con-
trol experiments, i.e., in the absence of 2.
We have demonstrated that coordination of acryloni-
trile in the diimine palladium complex 2 results in
formation of complexes 3–7, in addition to other reac-
tions whose products have not been identified. Trace
amounts of polyacrylonitrile are formed under forcing
conditions, probably via conventional free radical poly-
merization initiated by homolysis of the Pd–alkyl bonds
of 3 or 4 rather than via coordination polymerization.
While 2 does catalyze the coordination polymerization
of ethylene, excess acrylonitrile inhibits ethylene poly-
merization and 2 does not catalyze the copolymerization
of ethylene and acrylonitrile.
Acknowledgments
We gratefully acknowledge Lanxess Inc. of Sarnia,
Ontario, the Natural Sciences and Engineering Research
Council and QueenÕs University for financial support,
and Dr. Bernd Keller for assistance with the mass spec-
trometry. We also thank Johnson Matthey Inc. for a
very generous loan of palladium dichloride.
References
[1] (a) M. Bochmann, J. Chem. Soc., Dalton Trans. (1996) 255;
(b) W. Kaminsky, M. Arndt, Adv. Polym. Sci. 127 (1997) 143;
(c) J.C.W. Chien, in: T.J. Marks, J.C. Stevens (Eds.), Topics in
Catalysis, vol. 7, Baltzer AG, Science Publishers, 1999, p. 125;
(d) H.G. Alt, A. Ko¨ppl, Chem. Rev. 100 (2000) 1205;
(e) E.Y.-X. Chen, T.J. Marks, Chem. Rev. 100 (2000) 1391.
[2] (a) L.S. Boffa, B.M. Novak, Chem. Rev. 100 (2000) 1479;
(b) V.C. Gibson, S.K. Spitzmesser, Chem. Rev. 103 (2003) 283.
[3] (a) D.Y. Sogah, W.R. Hertler, O.W. Webster, G.M. Cohen,
Macromolecules 20 (1987) 1473;
However, we have investigated the possibility of eth-
ylene–AN co-polymerization by 2. Bubbling ethylene
through a solution of 2 in CD2Cl2 for 2 min results in
a color change from yellow to orange, disappearance
of the Pd–Me resonance and appearance of the peaks
of the type of branched polyethylene reported previ-
ously [6b]. The spectrum exhibits no resonance for free
ethylene and the diimine resonances are changed very
little, but there are vinyl resonances of coordinated acry-
(b) O.W. Webster, W.R. Hertler, D.Y. Sogah, W.B. Farnham,
T.V. Rajan Babu, J. Am. Chem. Soc. 105 (1983) 5706;
(c) Y. Li, D.G. Ward, S.S. Reddy, S. Collins, Macromolecules 30
(1997) 1875;
(d) S. Collins, D.G. Ward, J. Am. Chem. Soc. 114 (1992) 5460.

G. Stojcevic et al. / Journal of Organometallic Chemistry 690 (2005) 4349–4355
4355
[4] (a) G.G. Odian, Principles of Polymerization, fourth ed., Wiley–
Interscience, Hoboken, NJ, 2004;
(d) M.J. Szabo, R.F. Jordan, A. Michalak, W.E. Piers, T. Weiss,
S.-Y. Yang, T. Ziegler, Organometallics 23 (2004) 5565;
(e) S.-Y. Yang, M.J. Szabo, A. Michalak, T. Weiss, W.E. Piers,
R.F. Jordan, T. Ziegler, Organometallics 24 (2005) 1242.
[10] M.S. Kharasch, R.C. Seyler, F.R. Mayo, J. Am. Chem. Soc. 60
(1938) 882.
[11] These bond strengths were calculated using thermodynamic data
available at http:webbook.nist.gov/chemistry. We thank Prof.
J.A. Stone for informing us of this source.
(b) K. Matyjaszewski (Ed.), Advances in Controlled/living Rad-
ical Polymerization, ACS Symposium Series, vol. 854, American
Chemical Society, Washington, DC, 2003;
(c) K. Matyjaszewski, T.P. Davis (Eds.), Handbook of Radical
Polymerization, Wiley–Interscience, Hoboken, NJ, 2002;
(d) R.P. Quirk (Ed.), Applications of Anionic Polymerization
Research, ACS Symposium Series, vol. 696, American Chemical
Society, Washington, DC, 1998.
[12] (a) F. Wu, S.R. Foley, C.T. Burns, R.F. Jordan, J. Am. Chem.
Soc. 127 (2005) 1841;
[5] T.C.M. Chung, Functionalization of Polyolefins, Academic Press,
San Diego, 2002.
[6] (a) S.D. Ittel, L.K. Johnson, M. Brookhart, Chem. Rev. 100
(2000) 1169;
(b) L.F. Groux, T. Weiss, D.N. Reddy, P.A. Chase, W.E. Piers,
T. Ziegler, M. Parvez, J. Benet-Bucholz, J. Am. Chem. Soc. 127
(2005) 1854.
(b) L.K. Johnson, C.M. Killian, M. Brookhart, J. Am. Chem.
Soc. 117 (1995) 6414;
[13] (a) V.G. Albano, C. Castellari, M.E. Cucciolito, A. Panunzi, A.
Vitagliano, Organometallics 9 (1990) 1269;
´
(c) L.K. Johnson, S. Mecking, M. Brookhart, J. Am. Chem. Soc.
118 (1996) 267;
(d) S. Mecking, L.K. Johnson, L. Wang, M. Brookhart, J. Am.
Chem. Soc. 120 (1998) 888.
(b) I.D. del Rıo, R.A. Gossage, M.S. Hannu, M. Lutz, A.L.
Spek, G. van Koten, Organometallics 18 (1999) 1097.
[14] K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Coordination Compounds, fourth ed., Wiley and Sons, New
York, 1986, p. 280.
[15] F.A. Bovey, Nuclear Magnetic Resonance Spectroscopy, Aca-
demic Press, New York, 1969, pp. 64–71.
[7] S.R. Foley, R.A. Stockland, H. Shen, R.F. Jordan, J. Am. Chem.
Soc. 125 (2003) 4350.
[8] Th. Kemperman, S. Koch, J. Sumner (Eds.), Manual for
the Rubber Industry, second ed.
1993.
[9] (a) D.V. Deubel, T. Ziegler, Organometallics 21 (2002) 1603;
(b) D.V. Deubel, T. Ziegler, Organometallics 21 (2002) 4432;
(c) T. Ziegler, private communication;
,
Bayer AG, Leverkusen,
[16] (a) G. Gerdes, P. Chen, Organometallics 22 (2003) 2217;
(b) W. Baratta, S. Stoccoro, A. Doppiu, E. Herdtweck, A. Zucca,
P. Rigo, Angew. Chem., Int. Ed. 42 (2003) 105.
[17] P. Yang, B.C.K. Chan, M.C. Baird, Organometallics 23 (2004)
2752.