Effects of phosphine substituents on CO and norbornene insertion rates into (P,N)-Pd-alkyl and -acyl bonds

Article abstract of DOI:10.1016/S0022-328X(98)00953-X

The synthesis is described of (P,N)-PdMe(X) (P,N=R₂P-o-C₆H₄CH₂NMe₂ (L1), Me₂N-o-C₆H₄CH₂-PR₂ (L2), 1-(dimethylamino)-8-(R₂phosphino)naphthalene (L3) with R₂=Ph₂ (a), Cy₂ (b), Me,Ph (c, for L1 only); X=Cl, OTf) and cationic (P,N)-PdMe(L)⁺ (L, MeCN, CO; anion, ^-B[3,5-Ph-(CF₃)₂]₄). The complexes react with CO to give (P,N)-Pd{C(O)Me}(X/L⁺) derivatives. Carbonylation is faster in complexes with the more basic phosphines of type 1, but slower in ligand type 2 and 3. Norbornene inserts slowly into the (P,N)-Pd-acetyl bond of the cationic borate complexes with L1-3, and for L2 with X=Cl, OTf. Cationic palladium complexes with phenyl substituted phosphine ligands are more reactive toward CO/norbornene mixtures than those with cyclohexyl and complexes of L2 and L1 react faster than that of L3. The observed reactivity is described in terms of differences in CO insertion rates and a rate determining isomerization or trapping by norbornene of the intermediates, trans-P (P,N)-Pd{C(O)Me}⁺ complexes.

Full text of DOI:10.1016/S0022-328X(98)00953-X

Journal of Organometallic Chemistry 572 (1999) 193–205
Effects of phosphine substituents on CO and norbornene insertion
rates into (P,N)–Pd–alkyl and –acyl bonds
Peter H.P. Brinkmann, Gerrit A. Luinstra *
Department of Chemistry, Uni6ersity of Konstanz, Postfach 5560 M738, D-78457 Konstanz, Germany
Received 11 May 1998
Abstract
The synthesis is described of (P,N)–PdMe(X) (P,N=R2P-o-C6H4CH2NMe2 (L1), Me2N-o-C6H4CH2–PR2 (L2), 1-(dimethyl-
amino)-8-(R2phosphino)naphthalene (L3) with R2=Ph2 (a), Cy2 (b), Me,Ph (c, for L1 only); X=Cl, OTf) and cationic
(P,N)–PdMe(L)+ (L, MeCN, CO; anion, −B[3,5-Ph–(CF3)2]4). The complexes react with CO to give (P,N)–Pd{C(O)Me}(X/L+
)
derivatives. Carbonylation is faster in complexes with the more basic phosphines of type 1, but slower in ligand type 2 and 3.
Norbornene inserts slowly into the (P,N)–Pd–acetyl bond of the cationic borate complexes with L1–3, and for L2 with X=Cl,
OTf. Cationic palladium complexes with phenyl substituted phosphine ligands are more reactive toward CO/norbornene mixtures
than those with cyclohexyl and complexes of L2 and L1 react faster than that of L3. The observed reactivity is described in terms
of differences in CO insertion rates and a rate determining isomerization or trapping by norbornene of the intermediates, trans-P
(P,N)–Pd{C(O)Me}+ complexes. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Palladium; Carbonylation; CO; Kinetics; Olefin insertion
1. Introduction
generally less active. Consecutive insertion of CO and
olefins in the palladium carbon bond in square planar
complexes with non-C2-symmetrical bidentate ligands is
complicated by possible pre- or post-insertion cis–
trans-isomerization (Scheme 1) [18–21]. We recently
reported that in (P,N)–PdR+ complexes, CO insertion
takes place in the trans-N Pd-alkyl isomers, and nor-
bornene insertion in the trans-P Pd-acyl derivatives
[22]. The chemoselectivity of the coordination sites for
both monomers results from two factors: CO insertion
into the Pd–Me bond is faster than trans-N to trans-P
pre-insertion isomerization and, norbornene insertion
occurs only in the reactive trans-P Pd{C(O)Me}+ iso-
mer [22–24]. Methyl migration in the trans-N (P,N)–
PdMe(CO)+ complex to yield the trans-P (P,N)–
Pd{C(O)Me}+ insertion product is fast and reversible,
and is followed by a rate-determining isomerization to
the trans-N (P,N)–Pd{C(O)Me}(CO)+ complex
(Scheme 1) [22]. The trans-P (P,N)–Pd{C(O)Me}+
complex is very reactive and may be trapped by norbor-
Olefin polymerization and copolymerization with
Pd(II) or Ni(II) catalysts [1–14] has recently become a
major field of catalysis research. The facile co- and
terpolymerization of CO and 1-olefins to perfectly alter-
nating polyketones is part of this research [2–11]. The
majority of metal complexes that are productive cata-
lysts for this polymerization contain cis-chelating lig-
ands with either two phosphorus- or two
nitrogen-donor atoms, for example, the series of
diphenyldiphosphines with C1 through C4 bridges first
exploited by Drent and coworkers [12], and bipy and
phen compounds utilized by a number of groups [13–
16].
Mixed (P,N)–Pd(II)-systems, received very minor at-
tention as polymerization catalysts [17,18] and are
* Corresponding author. Fax:
gl1@chclu.chemie.uni-konstanz.de.
+49-7531-883137; e-mail:
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(98)00953-X

194
P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
Scheme 1. Reaction pathways for the reaction of (P,N)–PdMe(CO)+ with CO and norbornene.
nene to give the trans-N (P,N)–Pd{-2-C7H10-3-
{C(O)Me}}+ complex in a fast reaction. Reaction of
the trans-N (P,N)–Pd{C(O)Me}+ complex with nor-
bornene is slow. Norbornene insertion into the (P,N)–
Pd–{C(O)Me}+ bond proceeds most likely through a
pre-insertion isomerization to a configuration with a
trans-P Pd–{C(O)Me} group.
Palladium complexes with P,N ligands form a very
interesting class of potential copolymerization catalysts:
The different coordination sites trans to the ligand
donor atoms have different affinities and reactivity
toward (dissimilar) substrates [25,26]. Both features are
desirable properties for the synthesis of alternating
copolymers through coordination polymerization
[27,28]. Based on the insight that CO and norbornene
insertion take place (independently) at different coordi-
nation sites, we started an investigation on the ligand
effects on both type of insertion rates with the objective
to prepare active CO–olefin copolymerization catalysts
through individually tailoring the phosphine and the
nitrogen donors [22]. It was therefore decided to deter-
mine the effects of substituent variation of the phos-
phine and the bridge of rigid C3-bridged ligands on the
2. Results
2.1. Synthesis and characterization of (P,N)–PdMeX
and -PdMe+(L) compounds
P,N ligands with a C3-bridge were chosen in this
study (Fig. 1). Ligands of type L1 have an N,N-
dimethyl-o-phosphino-benzylamine frame [29,30]. In
type L2, the frame has the ‘opposite’ constitution with
N,N-dimethyl-2-(phosphinomethylene) aniline as
a
backbone [31]. In type L3, a 1,8-substituted naph-
thalene forms the bridge between the phosphine and the
amine [19]. In ligands of type 1–3, the phosphorus
atom is substituted with either two phenyl (a type) or
two cyclohexyl substituents (b type), of type 1, the
mixed methyl phenyl substituted phosphine derivative
(L1c) was also included. Palladium methyl chloride
complexes, (P,N)–PdMe(Cl) (1a–3b, with L1a–L3b re-
spectively) were obtained from the reaction of the P,N
ligands with (COD)PdMe(Cl) [19,32]. CO insertion into
the palladium carbon bond in complexes with 1-
diphenylphosphino-8-dimethylamino-naphthalene (L3a)
was studied in depth before by Vrieze and coworkers
and forms one of the starting points of this study [19].
((1))
The corresponding triflate derivatives (P,N)–
PdMe(OTf) (T1a–3b; OTf: OSO2CF3)1 were prepared
through reaction of 1a–3b with AgOTf (Eq. (1)). One
individual steps of the CO/olefin insertion pathways of
Scheme 1. The reaction of the (P,N)–PdMe(X/L+)
complexes with CO, (P,N)–Pd{C(O)Me}(X/L+) with
norbornene as well as reaction of (P,N)–PdMe(L)+
with mixtures of CO and norbornene were studied
(X=Cl, OTf; L=CO, MeCN). Relative reaction rates
were determined from competition experiments and the
results were interpreted in terms of the steps in Scheme
1.
1 The following codes are used to allow an easy identification of the
complexes: prefix T is used for triflate, C for the CO complex of the
cations, A for CO insertion, N for norbornene insertion, and order of
A or N is such that the first character relates to the first insertion
process. Thus AN indicates subsequent insertion of CO and norbor-
nene.

P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
195
Table 1
31P-NMR (ppm) in CDCl3 at 298 K and IR (cm−1) data of complexes containing L1a–3b
Compound/Ligand
Ligand (L)
1a
−14.6
1b
−17.3
1c
−37.6
2a
−8.8
2b
6.9
3a
0.5
3b
6.6
PdMeCl
PdCOMe(Cl) (A)
IR (w(CꢀO))
37.7
19.3
1684
30.9
23.6
1702
15.1
4.6
1701
45.7
25.7
1686
57.2
42.8
1704
40.2
20.4
1691
44.5
36.8
1695
PdMe(OTf) (T)
PdCOMe(OTf) (TA)
IR (w(CꢀO))
PdMe(MeCN)+ (B)
IR (w(CꢁN))
37.7
19.6
1696
39.8
31.3
1701
21.0
9.9
1695
50.6
22.8
64.9
50.6
42.7
20.6
51.2
38.3
1704
49.4
2316
2289
29.5
1695
60.6
2315
2287
45.0
1691
41.7
2318
2291
22.2
1695
49.0
2317
2289
40.7
36.1
2317
2290
18.2
35.2
2317
2288
27.2
17.9
2319
2293
6.9
(w(CꢁN))
PdCOMe(MeCN)+ (BA)
PdMe(CO)+ (C)
34.2
18.2
34.6
27.3
16.7
5.2
48.4
27.8
59.8
42.7
39.0
21.9
43.3
39.3
PdCOMe(CO)+ (CA)
Dl(31P)a mean
Standard deviation
18.1
1.6
7.8
0.6
11.0
0.4
22.1
3.8
15.3
1.3
19.4
1.7
8.2
3.7
a l(31P): Difference in 31P-NMR chemical shift between acyl and methyl derivatives of (P,N)–Pd(R)(X/L+).
1
set of signals is observed in the H-, 13C- and 31P-NMR
spectrum for both the chloride and the triflate deriva-
tives, demonstrating that only one isomer is present in
solution [17,33,34]. The methyl group is bonded trans
to the nitrogen donor as indicated by a small scalar
coupling of the methyl hydrogens to P (J(H–P):3 Hz)
[35], as expected by the different trans influences of the
ligands in the coordination sphere of Pd [36].
acidity of the (P,N)–Pd compounds is higher than that
of the bisphosphine analogs.
The carbonyl complexes, (P,N)–PdMe(CO)+ (C1a–
3b) were obtained through reaction of 1a–3b with the
borate anion sodium salt, NaBAr%, in a CO saturated
4
chloroform solution. The carbonyl stretching frequency
is found at 2135 cm−1 for all compounds C1a–3b,
close to that of free CO in solution [19]. In a cationic
bisphosphine palladium alkyl CO complex, the stretch-
ing frequency was found at 2077 cm−1, again indicat-
ing the relative increase in Lewis acidity in the
(P,N)Pd-systems [39].
Cationic
complexes
of
the
type
(P,N)–
PdMe(NꢁCMe)+ −B[3,5-Ph-(CF3)2]4 (B1–3b) were pre-
pared in situ by reaction of the respective chlorides with
one equivalent of NaB[3,5-Ph–(CF3)2]4(NaBAr%) in the
4
((2))
2.2. CO insertion
presence of one equivalent of acetonitrile (Eq. (2))
[13,37]. Compared to free acetonitrile, it was found that
the CꢁN stretching frequency is shifted about 35 cm−1
to higher wavenumbers in the cationic complexes. The
shift is not dependent on the ligand L1a–3b. It is about
twice as large as was observed for trans
(PPh3)2PdMe+(NꢁCMe) [38], indicating that the Lewis
Treatment of chloride complexes 1a–3b with CO at 1
bar results in the formation of the acyl complexes
(A1a–3b), reaction times range from 1 to 50 h for
completion. The triflates T1a–3b react likewise with
CO to the acyl compounds (TA1a–3b). Carbonylation
of the triflate complexes is about ten times faster than
of the chloro compounds [19,40–42]. The IR absorp-
tion for the CꢀO group in palladium acyl chloride
compounds appears between 1685 and 1705 cm−1
,
which compares nicely to the 1693 cm−1 reported for
A3a [19] (Table 1). The acyl group coordinates trans to
the nitrogen donor as was inferred from the 31P-NMR
shift and the absence of a coupling between P and the
acyl carbon atom [19].
It is revealing to compare the difference of the 31P-
NMR chemical shift (Dl(31P)) between the acyl and
Fig. 1. (P,N)-Ligand systems.

196
P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
Fig. 2. 2-D Representation of the carbonylation rate of 1a–3b (A) and T1a–3b (B), axis according to the number of aryl substituents on the
phosphine and alkyl on the amine donor.
methyl complexes. In the carbonylated products, an
upfield shift is found with a magnitude that is charac-
teristic for the P,N ligand (Table 1), but hardly depen-
dent of the anion X or the neutral donor L. Large
upfield shift changes of 15–20 ppm are found for the
phenyl phosphino derivatives (with L1a, L2a and L3a).2
An upfield shift of 18 ppm was also reported for the
(1,3-bis(diphenylphoshino)propane)PdMe(X/L) system
[19], indicating that such a number is characteristic for
PPh2 donors [40,42]. For the electron richer cyclohexyl
phosphine groups (L1b, L2b and L3b), a change of
about half that magnitude is observed. The differences
Dl(31P) between phenyl and cyclohexyl derivatives can
be attributed to the basicity of the phosphine ligation.
The smaller change in chemical shift in the cyclohexyl
phosphine derivatives demonstrates that the change in
charge density on the electron richer alkyl phosphine is
smaller than in the phenyl derivatives. The 31P-NMR
shift change between acyl and alkyl derivatives is thus
characteristic for a ligand frame and may be used to
detect CO insertion (Table 1, vide infra).3
solved in CDCl3. The solution was saturated with CO
and the decrease in the concentration of the two methyl
complexes was determined by H-NMR spectroscopy.
The following series were established for the chloride
and the triflate derivatives:
1
1.5
10
1.5
2
1.5
1.5
2a\3a\1b\3b\2b\1c\1
5
5
2
2
2
1.5
T1b\T4a\T2\T1c\T4b\T1a\T2b.
The numbers above the inequality sign indicate the
relative enhancement in rate between two adjacent
compounds. The same competition experiments were
performed for in situ prepared palladium methyl CO
complexes C1a–3b (Eq. (2)):
4
2
2
1.5
C1b\C4a2\−3C2a\C1c\C4b\C1a\C2b.
The relative rates of carbonylation of C1a–3b coincide
with that of the triflate series. It may thus be assumed
that the dissociation of the triflate anion does not
significantly affect the relative carbonylation rates and
that the observed trends are characteristic for the P,N-
ligand frame. Treatment of the acetonitrile adducts
B1a–3b with CO yields the acyl derivatives (BA1a–3b).
The reactivity order of CO insertion was not measured
by competition experiments for these complexes, but
qualitatively follows the one established for the triflate
(T1a–3b) and cationic CO complexes (C1a–3b). This
was concluded from reaction rate of in situ prepared
B1a–3b with CO.
A more informative overview of the relative carbony-
lation rates is obtained by a simple two dimensional
representation, where the axes are categorized accord-
ing to the number of aryl substituents on phosphine
and alkyl substituents on nitrogen, respectively (Fig. 2).
From Fig. 2, the influence of the phosphine substituents
on the carbonylation rate can be extracted. In com-
pounds with ligand of type 1, we observe an increase in
the rate of Pd-acyl formation with the electron richer
2.3. Relati6e CO insertion rates
Relative carbonylation rates of chloro 1a–3b and the
triflate derivatives T1a–3b were determined from com-
petition experiments in order to obtain insight into the
influence of the bridge and phosphine substituents on
the CO insertion. Equimolar amounts of two methyl
complexes (P,N)–PdMe(X) (X=Cl, OTf) were dis-
2 The l(31P) can, like in 13C-NMR, be described through sub-
stituent increments [43], with the restriction that no major changes in
the geometry around P occurs. These are probably small here: the
difference between two structurally alike compounds is considered.
3 The change of Dl(31P) between the free ligand and coordinated to
(P,N)–PdMeX(/L) (X=Cl, OTf; L=MeC+N) shows a lesser de-
pendence on the phosphine substituents and a much larger shift
difference (ca. 50 ppm for type L1–3 and 40 ppm for L4).

P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
197
Table 2
31P-NMR shifts and signal ratios of the reaction mixtures of five equivalents norbornene with C1a–3b under a CO atmosphere
Cx
x
l
31P (ppm)
Ratios
B/A/X
1/2/1
Estimated half-life (h)
Observed M+ (M+ calculated)
NAN(A)
33.95
NAN(B)
34.04
AN(X)
NAN
AN
1a
34.14
32.64b
18.53
44.7
0.3
2
ꢀ5
0.3
1ba
1ca,c
2ac
2b
668 (669.2)
500 (500.9)
656 (657.1)
668 (669.2)
692 (693.2)
704 (705.3)
574 (575.1)
594 (595.1)
562 (563.0)
574 (575.1)
(599.0)d
18.10
44.26
52.68
39.33
45.11
18.33
44.45
53.39
39.59b
45.44
3/2/4
1/6/2
3/1/5
5/1/–
4/3/2
53.39
3
3aa
3ba
\5
\10
d
45.69
610 (611.1)
a (P,N)–Pd{C(O)R}+ complex (R=Me (L3a,b), C7H10COC7H10Me (L1b,c)) is also formed.
b Less than three signals observed, but the proton NMR spectrum and mass spectrum indicate that all three products are formed.
c More products are formed.
d Trace amount.
nene insertion is generally slower than for the phenyl
derivatives a, but formation of (P,N)–Pd[3-{C(O)Me}-
2-C7H10]+ (AN1a–3b) is observed for the complexes
with the P,N ligands of Fig. 1.
phosphines b and c, whereas in both types 2 and 3 the
opposite effect is found.4
2.4. Norbornene insertion into the Pd–acyl bond
2.5. Reaction of Pd methyl deri6ati6es with mixtures of
CO and norbornene
The (L1a–3b)–Pd{C(O)Me}(X) (X=Cl, OTf)
derivatives have a low reactivity toward norbornene;
insertion was only found for ligand L2a and L2b. Sen
Reaction of (P,N)–PdMe+ compounds with mix-
tures of CO and norbornene results in the formation of
three major products, AN1a–3b and the exo,endo and
has
shown
in
compounds
of
the
type
{(PPh3)2Pd{C(O)R}(MeCN)+}−BF4 (R=Me, Ph)
that the insertion rate of norbornene is significantly
faster than in the corresponding chlorides [38]. Similar
observations were made by others [44]. Exchange of the
exo,exo
isomers
of
(P,N)–Pd{C7H10–CO–
C7H10Me}+, NAN1a–3b, the product of consecutive
norbornene–CO–norbornene insertion [22]. Formation
of the latter two becomes dominant at high norbornene
concentrations. The ratio of exo,exo- to exo,endo-in-
serted isomers of NAN is dependent on ligand L1a–3b
(Table 2). For compound C1b, a 31P-NMR signal was
observed at a characteristic position for an acyl deriva-
tive (ca. 10% of ꢀ (AN1b, NAN1b)) and is assigned to
NANA1b. Its formation is corroborated by the FAB-
MS spectrum of the reaction mixture. For complexes
C3a and C3b, the acyl complexes CA3a(b) are formed
in appreciable amounts (ca. 30% of the total of Pd-
complexes). This is explained by the generally lower
activity of complexes with ligands L3 toward mixtures
of CO and norbornene (Table 2) [22].
chloride for the non-coordinating BAr% anion in the
4
acyl complexes A1 and A3 resulted in insertion of
norbornene at an observable rate. Apparently, if a free
coordination site is available through the application of
non-coordinating anions, olefin insertion occurs readily
[22]. This is evident from the appearance of signals of
diastereotopic methyl groups for the NMe2 group in the
1H-NMR spectrum, and a characteristic 31P-NMR sig-
nal. An IR absorption in the range of 1620–1635 cm−1
indicates that the carbonyl of the norbornyl acyl entity
is coordinating to palladium. Competition experiments
as for the CO insertion were not performed because of
the low rates. Qualitatively, the order is CA2:CA1\
CA3. For cyclohexyl phosphine derivatives b norbor-
Several competition experiments were carried out to
establish the differences in reactivity of cationic com-
plexes with L1a–3b toward mixtures of CO and nor-
bornene. Equimolar amounts of compounds 1a and 1b
4 Note that geometry is of particular importance. A swap of
aryl-phosphine-benzyl-amine
1 to benzyl-phosphine-aryl-amine 2
takes us from the slowest to the fastest CO insertion rates in the
chloro derivatives, and in the cyclohexyl triflate derivatives T1b and
T2b from the fastest to the slowest carbonylation rate. The individual
type ligand frames respond with a different sensitivity to anion
substitution. The ratio of carbonylation rates between 1a/1b over
T1a/T1b increases by a factor of 30, whereas for ligand type 2 the
ratio is 0.3 and about 1 for L3. As a consequence, the absolute
reactivity order changes, but the order within one type of ligand
remains the same. Probably, (a preequilibrium) substitution of the
anion plays a role in the overall rate of carbonylation [19,40].
were activated with NaBAr% and reacted with a mixture
4
of CO and about one equivalent of norbornene. It was
observed that formation of AN1a is faster than that of
AN1b, contrary to formation of the respective acetyl
derivatives CA1a and CA1b. A similar competition
experiment with excess of norbornene analogously
shows that formation of NAN1a is faster than that of
NAN1b. Phenyl phosphine derivatives of L2 and L3

198
P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
also react faster with CO/norbornene mixtures than
their cyclohexyl derivatives. Relative reaction rates of
the three compounds with diphenyl substituted phos-
phines (a), (L1a–3a)–PdMe+ with mixtures of CO and
norbornene were measured for formation of AN in
reactions with about one equivalent per Pd as well as
for NAN in reactions with excess (five equivalents per
Pd) norbornene. For methyl complexes with ligands L
the order was found to be L2a:L1aꢀL3a in both
cases.
that the order of rates for the formation of the trans-N
(P,N)–Pd{C(O)Me}(X/L+) is the same within one type
of ligand, irrespective whether chloride, triflate, MeCN
or CO are coordinating trans to the phosphine suggests
that the rate determining step is the trans-P to trans-N
isomerization and that other possible reactions steps—
e.g. anion/donor displacements by CO, methyl migra-
tion, (partial) decoordination of the chelate—are of
kinetically less importance for the carbonylation rate.
Norbornene
insertion
into
the
(P,N)–Pd–
{C(O)Me}+ bond is faster for the phenyl substituted
derivatives of all ligand types. Insertion proceeds
through pre-insertion trans-N to trans-P isomerization
of the acetyl group, which is expected to be more
difficult in the cyclohexyl phosphines. It is assumed that
the trans-N to trans-P isomerization in Pd–acetyl
derivatives is the rate determining factor and not the
subsequent trapping by norbornene (vide infra, Fig. 3).
The order of reactivity thus corresponds to the concen-
tration of the trans-P Pd{C(O)Me}+ intermediates.
The order of reaction rate of palladium methyl
cations toward mixtures of CO and norbornene is the
same as for the rate of the palladium acetyl compounds
with norbornene with respect to phosphine substituents
as well as the ligand frames: complexes of L2 and L1
react faster than those with L3. This again relates to the
formation of the trans-P Pd{C(O)Me} intermediates.
The order of rates suggests that a partly flexible back-
bone in the ligand is favorable for isomerization and
migratory insertion of methyl to coordinated CO [19].
Finally, the experimental data and the analysis of the
insertion chemistry of cationic P,N-ligated palladium
complexes allow the following conclusions for consecu-
tive CO–olefin insertions. Migration of the methyl
group to coordinated CO is retarded by large electronic
differences between donor atoms [19,23], and has be-
come the rate determining factor in the reaction of the
(L1a–3b)–Pd complexes with mixtures of norbornene
and CO. The high rate of reaction of the highly reactive
trans-P Pd{C(O)Me}+ complexes with olefins is
thereby outweighed. It is an intriguing question whether
it would be possible to change this to a situation with
comparable rates of CO and olefin insertion. We are
currently investigating this matter.
3. Discussion and conclusions
An elaborate study was performed on the CO and
norbornene insertion chemistry of cationic palladium
complexes; the details are discussed below [45]. Com-
plexes with more basic phosphines (b,c) undergo a faster
carbonylation with ligands of type 1, and a slower one
for those of types 2 and 3. These differences in carbony-
lation rate between L1 and L2,3 can be rationalized
based on the model given in Scheme 1. CO insertion
into the Pd–Me bond of (P,N)–PdMe+ occurs through
reversible formation of the trans-P Pd{C(O)Me}+ fol-
lowed by a slow isomerization to the stable trans-N
Pd{C(O)Me}+ compound. The overall rate constant is
hence a function of the forward and reverse rate con-
stants for formation of the trans-P and its isomerization
rate to the trans-N Pd{C(O)Me}+ complex. It may be
inferred from the known trans influences and effects that
with increasing basicity of the phosphine [46], the Gibbs
energy and activation energy for formation of the trans-
P Pd{C(O)Me}+ species is increased. Thus, its concen-
tration is lower in complexes with more basic
phosphines. Trans-P Pd{C(O)Me}+ derivatives that
have a higher Gibbs free energy on the other hand, may
isomerize faster by virtue of the Hammond Postulate
[47]. Assuming Curtin–Hammett conditions, it follows
that the higher rate of carbonylation of C1b over C1a
results from a negative DDG‡iso for ligand type 1, and a
positive one for ligands of types 2 and 3 (DDG‡iso=DGiso
(cyclohexyl)−DG‡iso (phenyl)).
In the above explanation, it is assumed that the
(steady state) concentration of the reactive trans-P
Pd{C(O)Me}+ complex is effectively lower in all com-
plexes with cyclohexyl phosphines. This was substanti-
ated for complexes with L1b and L1a through the
competition experiment with mixtures of norbornene
and CO. Indeed, (L1a)–PdMe+ reacts faster than
(L1b)–PdMe+, indicati6e of a higher equilibrium con-
centration of the trans-P Pd{C(O)Me}+ species or a
faster migration of the methyl group to CO.5 The fact
4. Experimental section
4.1. General considerations
All operations were performed in an inert atmosphere
with rigorous exclusion of oxygen and moisture using
Schlenk, vacuum-line or glovebox techniques. Solvents
were thoroughly dried (ether and THF over Na/ben-
zophenone, pentane over Na/K alloy, toluene over Na,
CCl2H2 over CaH2) and distilled prior to use. Norbor-
5 This does not constitute a conclusive proof, since no data are
available on the relative rates of trapping of the reactive trans-P
Pd{C(O)Me}+ intermediates by norbornene.

P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
199
Fig. 3. Qualitative energy diagram for formation of (P,N)–Pd(COMe)+
.
4.2. L1c
nene was distilled from CaH2. CDCl3 was vacuum
transferred from CaH2. AgOTf was recrystallized
through diffusion of pentane into a concentrated
toluene solution. CO (5 N) was obtained from Messer-
Griesheim and was passed over blue silica prior to use.
PMePhCl [48], PCy2Cl [49], NaPPh2 [50], o-lithio-N,N-
dimethylbenzylamine [51], 1-dimethylamino-8-lithio-
A solution of ClPMePh (310 mg, 1.96 mmol) in 20
ml of THF was added to solid o-lithio-N,N-dimethyl-
benzylamine (0276 mg, 1.96 mmol) at room tempera-
ture. The mixture turned yellow within minutes. After
20 min, the solvent was exchanged for ether (30 ml) and
the resulting suspension was filtered. The ether was
removed in vacuum to leave a colorless oil that was
pure enough for further reaction. Yield 290 mg (1.08
naphthalene [52] and NaBAr% [53] and (COD)PdMeCl
4
[54], L1a [26], L1b [27], L2a [28] and L3a [26] were
prepared as published.
1
IR spectra were recorded on a Mattson Galaxy spec-
trometer as Nujol mulls between KBr disks. NMR
spectra were recorded on Bruker WM250, AC250,
DXR 600 Avance or JEOL FX-90Q, JNM GX400
spectrometers. Chemical shifts are reported in ppm and
referenced to residual protons in deuterated solvents
mmol, 55%). H-NMR (250 MHz, CDCl3): l=1.53 (d,
J(H–P)=4.3 Hz, 3H, P–Me), 2.14 (s, 6H, NMe2), 3.42
(dd, J(H–P)=2.9 Hz, J(H–H)=13.1 Hz, 1H, N–
CH), 3.74 (dd, J(H–P)=2.9 Hz, J(H–H)=13.1 Hz,
1H, N–CH), 7.2–7.45 (m, 9H, aryl) ppm.
1
(CDCl3: l=7.24 ppm) for H-NMR and to character-
4.3. L2b
istic multiplets for 13C-NMR (CDCl3: l=77.0 ppm).
31P shifts are reported against external H3PO4 at 0 ppm.
Mass spectra were obtained with Finnigan MAT 312 or
Finnigan MAT 312/AMD5000 instruments. Elemental
analyses were carried out at the Micro Analytical De-
partment of the University of Konstanz.
2-Methyl-N,N-dimethylaniline (4.8 ml, 0.03 mol) was
added to a mixture of 6 ml TMEDA (0.04 mol) and 24
ml of a 1.6 N solution of BuLi (0.038 mol) in hexane.
After refluxing for 3 h, the yellow reaction mixture was
cooled to −60°C and ClPCy2 (7.66 g, 0.033 mol) was

200
P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
C, 53.93; H, 7.68; N, 3.16. Calc. for C22H37ClNPPd: C,
54.10; H, 7.63; N, 2.87%. H-NMR (90 MHz, CDCl3):
l=0.73 (d, J(H–P)=2.9 Hz, 3H, Pd–Me), 1.0-2.5 (m,
22H, Cy), 2.47 (s, br, 6H, NMe2), 3.85 (s, br, 2H,
NCH2), 7.17 (m, 1H, aryl), 7.4 (m, 2H, aryl), 7.64 (m,
1H, aryl) ppm.
added. The mixture was slowly warmed to room tem-
perature. Water (50 ml) was added (exothermic reac-
tion). The organic layer was separated and the water
phase was extracted twice with ether. The ether and the
organic phase were combined and dried over Na2SO4.
The solvents were removed in vacuum to yield a yellow-
ish crystalline material that was not further purified.
Yield: 6.0 g (0.018 mol, 55%). Anal. Found: C, 75.81;
H, 10.61; N. 4.42. Calc. for C21H34NP: C, 76.09; H,
1
4.7. Compound 1c
1
Scale 0.76 mmol, yield 71%. The color of the solution
becomes brown, the solid is brown. Anal. Found: C,
48.08; H, 5.37; N, 2.90. Calc. for C17H23ClNPPd: C,
10.34; N, 4.22%. H-NMR (90 MHz, CDCl3): l=0.8–
2.0 (m, 22H, Cy), 2.87 (d, 2H, J(H–P)=2.4 Hz, CH2),
2.65 (s, 6H, NMe2), 6.91–7.25 (m, 3H, aryl), 7.40 (d,
J(H–H)=7 Hz, 1H, aryl) ppm.
1
49.29; H, 5.59; N, 3.38%. H-NMR (250 MHz, CDCl3):
l=0.80 (d, J(H–P)=3.6 Hz, 3H, Me–Pd), 1.83 (d,
J(H–P)=10 Hz, 3H, Me–P), 2.30 and 2.74 (s, 3H,
NMe2), 2.94 (d, br, J(H–H)=12 Hz, 1H, NCH), 3.46
(d, J(H–H)=12 Hz, 1H, NCH), 7.2–7.75 (m, 9H,
aryl) ppm.
4.4. L3b
A
THF (50 ml) solution of 8-lithium-(1-N,N-
dimethyl)naphthylamine etherate (1.3 g, 5.04 mmol)
was treated with dicyclohexylphosphine chloride (1.17
g, 5.04 mmol). The solution first became yellow and
subsequently brown. After stirring for 48 h, the
volatiles were removed in vacuum and the residue
extracted with 50 ml of ether. The ether was removed
and the residue was washed several times with pentane.
Yield: 0.82 g (2.2 mmol, 44%). Anal. Found: C, 78.68;
H, 9.15; N, 3.52. Calc. for C24H34NP: C, 78.43; H, 9.32;
4.8. Compound 2a
Scale 8.15 mmol, yield 82%. The color of the solution
becomes dark-red, the solid is ochre. Anal. Found: C,
54.48; H, 5.14; N, 3.27. Calc. for C22H25ClNPPd: C,
1
55.48; H, 5.29; N, 2.94%. H-NMR (250 MHz, CDCl3):
l=0.65 (d, J(H–P)=3.3 Hz, 3H, Pd–Me), 3.22 (s,
6H, NMe2), 3.75 (d, J(H–P)=11.1 Hz, 2H, P–CH2),
6.81 (m, 1H, aryl), 6.85 (m, 1H, aryl), 7.17 (m, 1H,
aryl), 7.27 (m, 1H, aryl), 7.33–7.41 (m, 6H, aryl), 7.54
(m, 4H, aryl) ppm.
1
N, 3.81%. H-NMR (600 MHz, CDCl3): l=1.1–1.3
(m, 10H, Cy), 1.63 (m, 2H, Cy), 1.69 (m, 6H, Cy), 1.79
(m, 2H, Cy), 1.88 (m, 2H, Cy), 2.68 (s, 6H, NMe2), 7.28
(d, J(H–H/P)=0.9 Hz, 1H, aryl-H7), 7.37 (m, 2H,
aryl-H3/6), 7.56 (m, 2H, aryl-H4/5), 7.73 (d, J(H–H)=
7 Hz, 1H, aryl-H2) ppm.
4.9. Compound 2b
Scale 3.55 mmol, yield 95%. The solution becomes
brown, reaction time is 2 h, the solid is grayish. Anal.
Found: C, 54.49; H, 7.64; N, 2.90. Calc. for
4.5. Compound 1a
Pd(COD)MeCl (150 mg, 0.56 mmol) was added to a
solution of L1a (180 mg, 0.56 mmol) in 15 ml of ether.
The color changed to faintly yellow, and after stirring
for 1 h, the volatiles were removed in vacuum. The
off-white residue was washed three times with 20 ml of
ether and dried. Yield: 170 mg (0.36 mmol, 64%). Anal.
Found: C, 55.65; H, 5.42; N, 2.96. Calc. for
1
C22H37ClNPPd: C, 54.15; H, 7.67; N, 2.87%. H-NMR
(250 MHz, CDCl3): l=0.72 (d, J(H–P)=2.5 Hz, 3H,
Pd–Me), 1.03–2.02 (m, 22H, Cy), 3.08 (s, 6H, NMe2),
3.14 (d, J(H–P)=10.8 Hz, 2H, P–CH2), 7.09–7.51 (m,
4H, aryl) ppm.
1
4.10. Compound 3b
C22H25ClNPPd: C, 55.48; H, 5.29; N, 2.94%. H-NMR
(600 MHz, CDCl3): l=0.58 (d, 3H, J(H–P)=3.4 Hz,
Pd–Me), 2.59 (s, br, 6H, NMe2), 3.12 (s, br, 2H, CH2),
6.86 (ps. t, J(H–H)=9 Hz, 1H, aryl-H6), 7.19 (dd,
J(H–H/P)=7 and 4.1 Hz, 1H, aryl-H3), 7.30 (ps. t,
J(H–H)=8 Hz, 1H, aryl-H5), 7.40 (m, 1H, aryl-H4),
7.4–7.5 (m, 10H, aryl) ppm. Compounds 1b–3b were
prepared analogously.
Scale 0.84 mmol, 74%. The color of the solution
becomes yellowish, reaction time is 48h and the solid is
white. Anal. Found: C, 57.82; H, 7.52; N, 3.27. Calc.
1
for C25H37ClNPPd: C, 57.26; H, 7.11; N, 2.67%. H-
NMR (600 MHz, CDCl3): l=0.81 (d, J(H–P)=2.3
Hz, 3H, Pd–Me), 0.44–2.34 (m, 22H, Cy), 3.0 (s, 6H,
NMe2), 7.32 (d, J(H–H)=7.5 Hz, 1H, aryl), 7.44 (t,
J(H–H)=7.5 Hz, 1H), 7.51 (t, J(H–H)=7.5 Hz, 1H,
aryl), 7.60 (d, J(H–H)=7.5 Hz, 1H, aryl), 7.75 (t,
J(H–H)=7.5 Hz, 1H, aryl), 7.94 (d, J(H–H)=7.5 Hz,
1H) ppm.
4.6. Compound 1b
Scale 2.63 mmol, yield 78%. The color of the solution
becomes faintly yellow, the solid is white. Anal. Found:

P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
201
4.16. Compound A3a
4.11. Compound A1a
Scale 0.9 mmol, yield 48%, reaction time 12 h, ochre.
Anal. Found: C, 56.47; H, 4.31; N, 3.04. Calc. for
C26H25NOPPd: C, 57.79; H, 4.66; N, 2.59%. H-NMR
(600 MHz, CDCl3): l=2.21 (s, 3H, COMe), 3.21 (s,
6H, NMe2), 7.36–7.74 (m, 14H, aryl), 7.90 (d, J(H–
H)=7 Hz, 1H, aryl), 8.13 (d, J(H–H)=7 Hz, 1H,
aryl) ppm.
CO was admitted to a dichloromethane solution (15
ml)
of
(o-diphenylphosphino–dimethylbenzy-
1
lamine)PdMeCl (1a) (100 mg, 0.21 mmol). The mixture
was stirred for 2 days. The solution was evaporated to
dryness to give an ochre solid (95 mg, 0.19 mmol, 89%).
Anal. Found: C, 52.57; H, 4.99; N, 3.51. Calc. for
C23H25ClNOPPd: C, 54.77; H, 4.99; N, 2.37%. 1H-
NMR (250 MHz, CDCl3): 2.08 (s, 3H, Pd–COMe),
2.46 (s, br, 6H, NMe2), 3.18 (s, br, 2H, NCH2), 7.06–
7.62 (m, 14H, aryl) ppm. Other derivatives were pre-
pared analogously.
4.17. Compound A3b
Scale 0.44 mmol, yield 58%, reaction time 30 h,
white. Anal. Found: C, 55.22; H, 6.84; N, 2.75. Calc.
for C26H37ClNOPPd: C, 56.55; H, 6.75; N, 2.53%.
1H-NMR (600 MHz, CDCl3): 0.2–2.5 (m, 20H, Cy),
2.27 (m, 2H, Cy), 2.70 (s, 3H, COMe), 2.92 (s, 6H,
NMe2), 7.25 (d, J(H–H)=8 Hz, 1H, aryl), 7.43 (t,
J(H–H)=8 Hz, 1H, aryl), 7.60 (d, J(H–H)=8 Hz,
1H, aryl), 7.92 (d, J(H–H)=8 Hz, 1H, aryl), 7.49 (t,
J(H–H)=8 Hz, 1H, aryl), 7.76 (t, J(H–H)=8 Hz,
1H, aryl) ppm.
4.12. Compound A1b
Scale 0.39 mmol, yield 91%, reaction time 2 days,
ochre. Anal. Found: C, 53.25; H, 7.32; N, 3.06. Calc.
for C23H37ClNOPPd: C, 53.49; H, 7.22; N, 2.71%.
1H-NMR (90 MHz, CDCl3): l=0.8–2.4 (m, 22H, Cy),
2.37 (s, 3H, COMe), 2.91 (s, 6H, NMe2), 3.72 (s, br,
2H, NCH2), 7.1–7.7 (m, 14H, aryl) ppm.
4.18. Compound T1a
4.13. Compound A1c
Compound 1a (0.89 g, 1.87 mmol) was dissolved in
15 ml of CH2Cl2 and AgOTf (0.41 g, 0.61 mmol) was
added. After stirring for 48 h, a yellow solution had
formed that was filtered and evaporated to dryness.
Yield: 0.85 g (1.38 mmol, 74%). Anal. Found: C, 47.05;
H, 4.41; N, 2.69. Calc. for C22H25F3NO3PPdS: C,
Scale 0.15 mmol, yield 89%, reaction time 4 days,
ochre. Anal. Found: C, 47.10; H, 5.16; N, 3.29. Calc.
for C18H29ClNPPd: C, 48.88; H, 5.24; N, 3.17%. H-
NMR (250 MHz, CDCl3): 1.75 (d, J(H–P)=9.3 Hz,
3H, PMe), 2.18 (s, 3H, COMe), 2.18 and 2.58 (s, 3H,
NMe2), 2.83 (d, br, J(H–H)=12 Hz, 1H, NCH), 3.38
(d, J(H–H)=12 Hz, 1H, NCH), 7.2–7.8 (m, 9H, aryl)
ppm.
1
1
46.62; H, 4.59; N, 2.36%. H-NMR (250 MHz, CDCl3):
l=0.65 (d, J(H–P)=1.8 Hz, 3H, Pd–Me), 2.53 (s, br,
6H, NMe2), 3.32 (s, br, 2H, NCH2), 7.15–7.56 (m, 13H,
aryl), 6.84 (ps. dt, J(H–H)=1.2 and 6.5 Hz, 1H, aryl)
ppm. Other derivatives were prepared analogously.
4.14. Compound A2a
4.19. Compound T1b
Scale 0.63 mmol, yield 98%, reaction time 1 day,
yellow-orange. Anal. Found: C, 54.10; H, 5.11; N, 3.13.
Calc. for C23H25ClNOPPd: C, 54.77; H, 4.99; N, 2.77%.
1H-NMR (250 MHz, CDCl3): l=1.94 (s, 3H, COMe),
3.08 (s, 6H, NMe2), 3.82 (d, J(H–P)=10.8 Hz), 6.71
(d, J(H–H)=8.0 Hz, 1H, aryl), 6.78 (t, J(H–H)=8.0
Hz, 1H, aryl), 7.18-7.55 (m, 12H, aryl) ppm.
Scale 0.76 mmol, yield 84%, reaction time 2 h, faintly
yellow. Anal. Found: C, 44.15; H, 5.93; N, 2.04. Calc.
for C23H37F3NO3PPdS: C, 45.89; H, 6.19 N, 2.32%.
1H-NMR (90 MHz, CDCl3): l=0.78 (d, J(H-P)=0.9
Hz, 3H, Pd–Me), 1.0–2.1 (m, 22H, Cy), 2.41 (s, br, 6H,
NMe2), 3.85 (s, br, 2H, NCH2), 7.1–7.9 (m, 4H, aryl)
ppm.
4.15. Compound A2b
4.20. Compound T1c
Scale 0.8 mmol, yield 96%, reaction time 1 day, gold.
Anal. Found: C, 52.12; H, 7.25; N, 2.72. Calc. for
C23H37ClNOPPd: C, 53.49; H, 7.22; N, 2.71%. 1H-
NMR (600 MHz, CDCl3): l=0.85–2.15 (m, 22H, Cy),
2.54 (s, br, 3H, COMe), 2.94 (s, 6H, NMe2), 3.09 (d,
J(H–P)=11 Hz, 2H, P–CH2), 7.05 (t, J(H–H)=7.5
Hz, 1H, aryl), 7.10 (d, J(H–H)=7.5 Hz, 1H, aryl),
7.25 (m, 2H, aryl) ppm.
Scale 0.24 mmol, yield 76%, reaction time 24 h, the
brown product was washed three times with 20 ml of
ether. Anal. Found: C, 40.22; H, 4.26; N, 2.10; S, 5.68.
Calc. for C18H23F3NO3PPdS (527.83): C, 40.96; H,
4.39; N, 2.65; S, 6.07%. H-NMR (250 MHz, CDCl3):
0.84 (d, J(H–P)=1 Hz, 3H, Pd–Me), 1.89 (d, J(H–
P)=10.9 Hz, 3H, PMe), 2.25 and 2.62 (s, 3H, NMe2),
1

202
P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
2.99 (d, br, J(H–H)=12 Hz, 1H, NCH), 3.47 (d,
J(H–H)=12 Hz, 1H, NCH), 7.2–7.75 (m, 9H, aryl)
ppm.
Calc. for C24H25F3NO4PPdS: C, 46.65; H, 4.07; N,
2.66%. 1H-NMR (250 MHz, CDCl3): l=2.12 (d, J(H–
P)=0.8 Hz, 3H, COMe), 2.44 (s, br, 6H, NMe2), 3.18
(s, br, 2H, CH2), 7.06–7.62 (m, 14H, aryl) ppm. Other
derivatives were prepared analogously.
4.21. Compound T2a
Scale 1.12 mmol, yield 89%, reaction time 24 h, the
red-brown product was washed three times with 20 ml
of ether. Anal. Found: C, 47.39; H, 4.48; N, 2.63. Calc.
for C23H25F3NO3PPdS: C, 46.83; H, 4.27; N, 2.37%.
1H-NMR (600 MHz, CDCl3): l=0.65 (s, br, 3H, Pd–
Me), 3.12 (s, 6H, NMe2), 3.73 (d, J(H–P)=11.8 Hz,
2H, PCH2), 6.81 (d, J(H–H)=7 Hz, 1H, aryl), 6.88 (t,
J(H–H)=7 Hz, 1H, aryl), 7.2–7.54 (m, 12H, aryl)
ppm.
4.26. Compound TA1b
Scale 0.24 mmol, yield 67%, reaction time 24 h,
brown-yellow. Anal. Found: C, 44.59; H, 6.21; N, 2.16;
S, 5.03. Calc. for C24H37F3NO4PPdS: C, 45.75; H, 5.92;
1
N, 2.22; S, 5.09%. H-NMR (90 MHz, CDCl3): l=
1.0–2.5 (m, 22H, Cy), 2.34 (s, br, 6H, NMe2), 2.65 (s,
3H, COMe), 3.75 (s, br, 2H, NCH2), 7.1 (m, 1H, aryl),
7.4 (m, 2H, aryl), 7.6 (m, 1H, aryl) ppm.
4.22. Compound T2b
4.27. Compound TA1c
Scale 1.54 mmol, yield 88%, reaction time 24 h,
greyish. Anal. Found: C, 46.33; H, 6.29; N, 2.47. Calc.
for C23H37F3NO3PPdS: C, 45.89; H, 6.79; N, 2.32%.
1H-NMR (250 MHz, CDCl3): l=0.69 (s, br, 3H, Pd–
Me), 1.02–1.99 (m, 22H, Cy), 2.98 (s, 6H, NMe2), 3.15
(d, J(H–P)=11.6 Hz, 2H, PCH2), 7.11–7.73 (m, 4H,
aryl) ppm.
Scale 0.07 mmol, yield 94%, reaction time 24 h,
brown). Anal. Found: C, 40.81; H, 4.79; N, 2.07; S,
5.01%. Calc. for C19H23F3NO4PPdS: C, 41.06; H, 4.17;
1
N, 2.52; S, 5.76%. H-NMR (250 MHz, CDCl3): l=
1.79 (d, J(H–P)=9.9 Hz, 3H, PMe), 2.15 and 2.56 (s,
3H, NMe2), 2.58 (d, J(H–P)=0.9 Hz, 3H, COMe),
2.87 and 3.42 (d, J(H–H)=11.5 Hz, 1H, NCH2),
7.2–7.9 (m, 9H, aryl) ppm.
4.23. Compound T3a
4.28. Compound TA2a
Scale 1.8 mmol, yield 39%, reaction time 12 h,
brown-orange. Anal. Found: C, 50.52; H, 4.27; N, 2.56.
Calc. for C26H25F3NO3PPdS: C, 49.89; H, 4.03; N,
2.24%. 1H-NMR (250 MHz, CDCl3): l=0.63 (d, J(H–
P)=3.1 Hz, 3H, Pd–Me), 2.92 (s, 6H, NMe2), 7.21–
7.72 (m, 14H, aryl), 7.90 (m, 1H, aryl), 8.13 (m, 1H,
aryl) ppm.
Scale 0.93 mmol, yield 94%, reaction time 48 h,
dark-ochre. Anal. Found: C, 46.17; H, 4.00; N, 2.41.
Calc. for C24H25F3NO4PPdS: C, 46.65, H, 4.07; N,
1
2.26%. H-NMR (250 MHz, CDCl3): l=1.97 (s, 3H,
COMe), 3.04 (s, 6H, NMe2), 3.82 (d, J(H–P)=11.6
Hz, 2H, P–CH2), 6.69 (d, J(H–H)=7.5 Hz, 1H, aryl),
6.81 (t, J(H–H)=7.5 Hz, 1H), 7.10–7.60 (m, 12H,
aryl) ppm.
4.24. Compound T3
Scale 1.1 mmol, yield 52%, reaction time 48 h, the
yellow solid was washed three times with 20 ml of
ether). Anal. Found: C, 48.58, H, 6.04; N, 2.63. Calc.
for C26H37F3NO3PPdS: C, 48.94; H, 5.84; N, 2.19%.
1H-NMR (250 MHz, CDCl3): l=0.45 (m, 2H, Cy),
0.82 (s, 3H, Pd–Me), 0.75–2.05 (m, 18H, Cy), 2.3 (m,
2H, Cy), 2.90 (s, 6H, NMe2), 7.35 (d, J(H–H)=7 Hz,
1H, aryl), 7.46 (m, 2H, aryl), 7.65 (d, J(H–H)=7 Hz,
1H, aryl), 7.74 (t, J(H–H)=7 Hz, 1H, aryl), 7.96 (d,
J(H–H)=7 Hz, 1H, aryl) ppm.
4.29. Compound TA2b
Scale 0.8 mmol, yield 96%, reaction time 24 h, gold.
Anal. Found: C, 46.28; H, 6.19; N, 2.26. Calc. for
C24H37F3NO4PPdS: C, 45.75 H, 5.91; N, 2.22%. H-
NMR (250 MHz, CDCl3): l=0.8–2.1 (m, 22H, Cy),
2.54 (s, 3H, COMe), 2.91 (s, 6H, NMe2), 3.12 (d,
J(H–P)=11.6 Hz, 2H, PCH2), 7.05–7.4 (m, 4H, aryl)
ppm.
1
4.30. Compound TA3a
4.25. Compound TA1a
Scale 0.9 mmol, yield 83%, reaction time 12 h, yel-
low-brown. Anal. Found: C, 50.72; H, 3.88; N, 1.99.
Calc. for C27H25F3NO4PPdS: C, 49.59; H, 3.85; N,
2.14% H-NMR (250 MHz, CDCl3): l=1.88 (s, 2H,
COMe), 2.76 (s, 6H, NMe2), 7.36–7.76 (m, 15H, aryl),
Compound T1a (610 mg, 1.1 mmol) was stirred for 2
days in 15 ml of CH2Cl2 in a CO atmosphere. The
solution was filtered from Pd and evaporated to dry-
ness. A brown-yellow solid resulted. Yield: 0.61 g (1.0
mmol, 96%). Anal. Found: C, 46.52; H, 4.60; N, 2.29.
1
8.04 (d, J(H–H)=8.2 Hz, 1H, aryl) ppm.

P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
203
4.31. Compound TA3b
7.50; N, 2.27. Calc. for C30H45ClNOPPd: C, 59.21; H,
7.45; N, 2.30%. H-NMR (600 MHz, CDCl3): l=0.2
1
Scale 0.5 mmol, yield 96%, reaction time 24 h, white).
Anal. Found: C, 53.89; H, 4.15; N, 3.15. Calc. for
C27H37F3NO4PPdS: C, 53.24; H, 6.12; N, 2.29%. H-
NMR (250 MHz, CDCl3): l=0.35 (m, 2H, Cy), 0.8–
2.1 (m, 18H, Cy), 2.33 (m, 2H, Cy), 2.70 (s, 3H,
COMe), 2.88 (s, 6H, NMe2), 7.29 (d, J(H–H)=7 Hz,
1H, aryl), 7.4–7.55 (m, 2H, aryl), 7.64 (d, J(H–H)=8
Hz, 1H, aryl), 7.78 (t, J(H–H):7 Hz, 1H, aryl), 7.97
(d, J(H–H)=8 Hz, 1H, aryl) ppm.
(m, 1H, NB), 0.8–2.2 (m, 29H, NB, Cy), 2.48 (s, 3H,
COMe), 2.84 and 3.25 (s, br, 3H, NMe2), 3.13 and 3.24
(dd, J(H–P)=20 Hz, J(H–H)=12.7 Hz, 2H, PCH2),
7.18–7.65 (m, 4H, aryl) ppm. 31P{1H}-NMR (242.9
MHz, CDCl3): l=53.21 ppm.
1
4.35. Compound TAN2a
Compound T2a (200 mg, 0.34 mmol) and norbornene
(40 mg, 0.43 mmol) were stirred for 2.5 days in 30 ml of
CH2Cl2. The solvent was removed in vacuum and the
resulting solid was washed several times with 20 ml
portions of ether to give a brown product. Yield: 220 mg
(0.31 mmol, 91%). Anal. Found: C, 55.98; H, 5.37; N,
2.06; S, 4.07. Calc. for C31H35F3NO4PPdS: C, 52.29; H,
4.32. Compounds B1a–3b, BA1a–3b, C1a–3b,
CA1a–3b and B1a
Compound 1a (P,N)PdMeCl (6.35 mg, 0.0133 mmol),
11.8 mg NaBAr% (0.0133 mmol) and 0.72 ml MeCN
4
1
(0.0133 mmol) were mixed in 0.5 ml CDCl3 to yield a
yellow-brown solution and a dark precipitate. 1H-NMR
(400 MHz, CDCl3): l=0.36 (d, J(H–P)=1.5 Hz, Pd–
Me), 1.98 (s, 3H, NCMe), 2.30 (s, 6H, NMe2), 3.20 (s,
br, 2H, N–CH2), 6.8–7.6 (m, 16H, aryl) ppm. Other
borate complexes were prepared in the same way. The
respective acyl compounds (BA1a–3b) were obtained by
saturating thus prepared solutions with CO. Reaction
times are 2–8 h. The same procedure was used to obtain
C1a–3b and CA1a–3b, respectively, but no MeCN was
present and reactions were performed in a CO atmo-
sphere with CO saturated solvent.
4.95; N, 1.97; S, 4.50%. H-NMR (250 MHz, CDCl3):
l=0.3 (m, 1H, NB), 0.9–1.75 (m, 8H, NB), 2.43 (s, 3H,
COMe), 2.86 (d, J(H–H)=6.5 Hz, 1H, NB), 3.09 (s, br,
6H, NMe2), 3.64 (ps. t, J(H–H) /J(H–P)=14 and 13
Hz, 1H, PCH), 4.31 (dd, J(H–H)/J(H–P)=14 and 8.1
Hz, 1H, PCH), 6.8–7.6 (m, 14H, aryl) ppm. 31P{1H}-
NMR (101 MHz, CDCl3): l=45.35 ppm.
4.36. Compound TAN2b
As TAN2a starting from TA2b (scale 0.15 mmol,
reaction time 3 days, yield 55%, color brown). Anal.
Found: C, 51.30; H, 6.71; N, 2.14. Calc. for
1
4.33. Compound AN2a
C31H43F3NO4PPdS: C, 51.52; H, 6.27; N, 1.94%. H-
NMR (250 MHz, CDCl3): l=0.2 (m, 1H, NB), 0.75–
2.05 (m, 29H, NB, Cy), 2.49 (s, 3H, COMe), 2.86 and
3.25 (s, 3H, NMe2), 3.25 (m, 2H, PCH2), 7.20–7.36 (m,
4H, aryl) ppm. 31P{1H}-NMR (242.9 MHz, CDCl3):
l=56.21 ppm.
Compound A2a (150 mg, 0.3 mmol) was dissolved in
30 ml of CCl2H2 and norbornene (50 mg, 0.6 mmol) was
added at room temperature. After stirring for 5 days,
the volatiles were removed in vacuum. A brown solid
resulted which was isolated. Yield: 160 mg (0.28 mmol,
93%). Anal. Found: C, 57.54; H, 5.55; N, 2.13. Calc. for
C30H35ClNOPPd: C, 60.21; H, 5.89; N, 2.34%. 1H-
NMR (250 MHz, CDCl3): l= −0.21 (m, 1H, NB),
0.67 (m, 1H, NB), 0.92 (d, J(H–H)=9.4 Hz, 2H, NB),
1.1–1.45 (m, 3H, NB), 1.71 (d, J(H–H)=9.7 Hz, 1H,
NB), 2.31 (s, 6H, NMe2/COMe), 2.34 (s, 3H, NMe),
2.57 (d, J(H-H)=6.5 Hz, 1H, NB), 4.22 and 4.33 (dd,
J(H–H)/J(H–P)=14 Hz, 1H, PCH2), 6.8–7.6 ppm (m,
11H, aryl), 7.85 (ps. t, J(H–H)=8 Hz, 2H, aryl), 8.17
(d, J(H–H)=8 Hz, 1H, aryl) ppm. 31P{1H}-NMR (101
MHz, CDCl3): l=40.21 ppm.
4.37. Reactions of CA1a–3b with norbornene
A typical experiment is described. Compound AN1b.
Complex A1b (7.5 mg, 0.012 mmol), NaBAr% (10.6 mg,
4
0.012 mmol) and 2-norbornene (2 mg, 0.021 mmol) were
dissolved in CDCl3. After 4 h, the following spectrum
1
was recorded. H-NMR (400 MHz, CDCl3): l=0.68
(m, 1H, norbornyl), 0.8–2.2 (m, 29H, Cy/NB), 2.05 (s,
3H, COMe), 2.32 (s, br, 3H, NMe), 2.52 (m, 1H, NB),
2.75 (s, 3H, N–Me), 2.93 (m, 1H, NB), 3.17 (d, J(H–
H)=13.2 Hz, 1H, NCH), 4.46 (d, J(H–H)=13.2 Hz,
1H, NCH), 7.1–7.7 (m, 16H, aryl) ppm. 31P{1H}-NMR
(161 MHz, CDCl3): l=32.62 ppm.
4.34. Compound AN2b
Compound A2b (200 mg, 0.32 mmol) and 60 mg
norbornene (0.64 mmol) were dissolved in 30 ml of
CH2Cl2 and the mixture was stirred for 2 days. The
solvent was removed in vacuum to yield a gray solid
(172 mg, 0.28 mmol, 89%). Anal. Found: C, 55.29; H,
4.38. Competition experiments for determination of
relati6e carbonylation rates
A typical experiment is described for T2a and T3a.
Compound T2a (5.53 mg, 0.01 mmol) and T3a (6.3 mg,

204
P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
Catalysts for Olefin Polymerization (Akron, OH, 1986), Transi-
tion Metals and Organometallics as Catalysts for Olefin Poly-
merization, Springer Verlag, Berlin, Heidelberg, 1988, p. 349.
[2] E. Drent, Eur. Pat. Appl. (1984) 121,965. ibid. Eur. Pat. Appl.
(1986) 408,229.
[3] A. Batistini, G. Consiglio, U.W. Suter, Angew. Chem. 104
(1992) 306.
[4] M. Barsacchi, A. Batistini, G. Consiglio, U.W. Suter, Macro-
molecules 25 (1992) 3604.
0.01 mmol) were dissolved in 0.5 ml of CDCl3. The
solution was saturated with CO, and the decrease in
concentration of the methyl complexes was monitored
by integrating the 1H-NMR signals of the methyl
groups against silicon grease as internal standard, and
by appearance of the respective acyl signals. The com-
petition experiments were performed for various combi-
nations of both the chlorides 1a–3b and triflates
T1a–3b. Two arbitrary compounds were mixed and the
one with the faster carbonylation rate was identified.
Then, the fastest was run against another. If it was still
the fastest the other was also run against the slower.
This procedure was repeated until the order of all
compounds was established. The absolute rates allow
for a convenient monitoring at room temperature.
[5] A. Sen, Acc. Chem. Res. 26 (1993) 303.
[6] N. Alperowicz, Chem. Week (1995) 22.
[7] A.S. Abu-Surrah, R. Wursche, B. Rieger, G. Eckert, W. Pech-
hold, Macromolecules 29 (1996) 4806.
[8] A. Vavasori, L. Toniolo, J. Mol. Cat. 110 (1996) 13.
[9] B. Milani, L. Vicentini, A. Sommazzi, F. Garbassi, E.
Chiarparin, E. Zangrando, G. Mestroni, J. Chem. Soc., Dalton
Trans. (1996) 3139.
[10] C.M. Killian, D.J. Tempel, L.K. Johnson, M. Brookhart, J. Am.
Chem. Soc. 118 (1996) 11664.
[11] E. Drent, P.H.M. Budzelaar, Chem. Rev. 96 (1996) 663.
[12] E. Drent, J.A.M. van Broekhoven, M.J. Doyle, J. Organomet.
Chem. 417 (1991) 235.
4.39. Competition experiments with mixtures of
norbornene and CO
[13] M. Brookhart, F.C. Rix, J.M. DeSimone, J. Am. Chem. Soc. 114
(1992) 5894.
[14] B.A. Markies, D. Kruis, M.H.P. Rietveld, K.A.N. Verkerk, J.
Boersma, H. Kooijman, M.T. Lakin, A.L. Spek, G. van Koten,
J. Am. Chem. Soc. 117 (1995) 5263.
[15] A. Sen, Z. Jiang, Macromolecules 26 (1993) 911.
[16] R.E. Ru¨lke, J.G.P. Delis, A.M. Groot, C.J. Elsevier, P.W.N.M.
van Leeuwen, K. Vrieze, K. Goubitz, H. Schenk, J. Organomet.
Chem. 508 (1996) 109.
[17] For a recent review, see K.J. Cavell, Coord. Chem. Rev. 155
(1996) 209.
[18] M. Sperrle, A. Aeby, G. Consiglio, A. Pfaltz, Helv. Chim. Acta
79 (1996) 1387 and references therein.
[19] G.P.C.M. Dekker, A. Buijs, C.J. Elsevier, K. Vrieze, P.W.N.M.
van Leeuwen, W.J.J. Smeets, A.L. Spek, Y.F. Wang, C.H. Stam,
Organometallics 11 (1992) 1937.
[20] H.A. Ankersmit, N. Veldman, A.L. Spek, K. Eriksen, K. Goub-
itz, K. Vrieze, G. van Koten, Inorg. Chim. Acta 252 (1996) 203.
[21] H.A. Ankersmit, B.H. Løken, H. Kooijman, A.L. Spek, K.
Vrieze, G. van Koten, Inorg. Chim. Acta 252 (1996) 141.
[22] P.H.P. Brinkmann, G.A. Luinstra, submitted to Organometallics
[23] P.W.N.M. van Leeuwen, K.F. Roobeek, Recl. Trav. Chim.
Pays-Bas 114 (1995) 73.
A typical procedure is described. Compound 1a (5.9
mg, 0.0124 mmol), compound 3a (6.4 mg, 0.0124
mmol), norbornene (2.2 mg, 0.023 mmol) and NaBAr%
4
(22 mg, 0.0242 mmol) were dissolved in 0.5 ml of
CDCl3. A brown suspension forms, which was analyzed
1
by 31P- and H-NMR spectroscopy. It was found that
the resonances assigned to C1a disappear and of AN1a
appear considerably faster than the corresponding ones
of C3a and AN3a. Similar experiments were performed
to establish the order of reactivity of 1a/2a and 2a/3a
toward mixtures of norbornene and CO. It was also
found that 1a and 2a react at about the same rate with
a mixture of excess norbornene (ca. ten equivalents/
ꢀPd); both are faster than 3.
5. Supplementary material available
Spectroscopic properties (13C-NMR shifts, IR) of
compounds L2b, L3b, 1a–3b, A1a–3b, T1a–3b, TA1a–
3b, AN2a,b, TAN2a,b, B1a–3b, BA1a–3b, C1a–3b,
CA1a–3b, AN1c–3b, and the data on the statistical
analysis of Dl(31P) (11 pages).
[24] P.W.N.M. van Leeuwen, C.F. Roobeek, H. van der Heijden, J.
Am. Chem. Soc. 116 (1994) 12117.
[25] P. Sennhenn, B. Gabler, G. Helmchen, Tetrahedron Lett. 35
(1994) 8595.
[26] A. Togni, U. Burckhardt, V. Gramlich, P.S. Pregosin, R. Salz-
mann, J. Am. Chem. Soc. 118 (1996) 1031.
[27] W. Kaminsky, M. Arndt, I. Beulich, Polym. Mater. Sci. Eng. 76
(1997) 18.
Acknowledgements
[28] M.K. Leclerc, R.M. Waymouth, Angew. Chem. 110 (1998) 964.
[29] W. de Graaf, S. Harder, J. Boersma, G. van Koten, J.A.
Kanters, J. Organomet. Chem. 358 (1988) 545.
[30] P. Barbaro, A. Currao, J. Herrmann, R. Nesper, P.S. Pregosin,
R. Salzmann, Organometallics 15 (1996) 1879.
Support from Degussa AG in the form of a generous
gift of PdCl2 is gratefully acknowledged. This investiga-
tion was supported by BMBF (c 03 N 1028 0).
[31] G.P. Schiemenz, E. Papageorgiou, Phosphorus Sulfur 13 (1982)
41.
[32] R.E. Ru¨lke, I.M. Han, C.J. Elsevier, K. Vrieze, P.W.N.M. van
Leeuwen, C.F. Roobeek, M.C. Zoutberg, Y.F. Wang, C.H.
Stam, Inorg. Chim. Acta 169 (1990) 5.
References
[1] K.A. Ostoja Starzewski, J. Witte, K.H. Reichertand, G. Vasil-
iou, in: W. Kaminsky, H. Sinn (Eds), Proceedings of the Interna-
tional Symposium on Transition Metals and Organometallics as
[33] K. Nozaki, N. Sato, Y. Tonomura, M. Yasutomi, H. Takaya, T.
Hiyama, T. Matsubara, N. Koga, J. Am. Chem. Soc. 119 (1997)
12779.

P.H.P. Brinkmann, G.A. Luinstra / Journal of Organometallic Chemistry 572 (1999) 193–205
205
[44] G.P.C.M. Dekker, C.J. Elsevier, K. Vrieze, P.W.N.M. van
Leeuwen, C.F. Roobeek, J. Organomet. Chem. 430 (1992) 357
and references therein.
[45] For monodentate systems: G.K. Anderson, R.J. Cross, Acc.
Chem. Res. 17 (1984) 67.
[46] F.A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, 3rd
ed., J. Wiley and Sons, New York, 1972, p. 368.
[47] G.S. Hammond, J. Am. Chem. Soc. 77 (1955) 334.
[48] H.J. Bestmann, J. Lienert, E. Heid, Chem. Ber. 115 (1982) 3875.
[49] K. Issleib, W. Seidel, Chem. Ber. 92 (1959) 2681.
[50] R.D. Feltham, H.G. Metzger, J. Organomet. Chem. 33 (1971)
347.
[51] L. Brandsma, H. Verkruijsse, Preparative Polar Organometallic
Chemistry, Springer Verlag, Berlin, 1987, p. 203.
[52] J.T.B.H. Jastrzebski, C.T. Knaap, G. van Koten, J. Organomet.
Chem. 255 (1983) 287.
[34] H. Jin, K.J. Cavell, B.W. Skelton, A.H. White, J. Chem. Soc.,
Dalton Trans. (1995) 2159.
[35] T.G. Appleton, M.A. Bennett, I.B. Tomkins, J. Chem. Soc.,
Dalton Trans. (1976) 439.
[36] F.R. Hartley, Chem. Soc. Rev. 2 (1973) 163.
[37] M. Brookhart, M.I. Wagner, J. Am. Chem. Soc. 118 (1996)
7219.
[38] J.S. Brumbaugh, R.R. Whittle, M. Parvez, A. Sen, Organometal-
lics 9 (1990) 1735.
[39] I. To´th, C.J. Elsevier, J. Am. Chem. Soc. 115 (1993) 10388.
[40] G.P.C.M. Dekker, C.J. Elsevier, K. Vrieze, P.W.N.M. van
Leeuwen, Organometallics 11 (1992) 1598.
[41] R. van Asselt, E.E.C.G. Gielens, R.E. Ru¨lke, K. Vrieze, C.J.
Elsevier, J. Am. Chem. Soc. 116 (1994) 977.
[42] R.E. Ru¨lke, V.E. Kaasjager, P. Wehman, C.J. Elsevier,
P.W.N.M. van Leeuwen, K. Vrieze, J. Fraanje, K. Goubitz, A.L.
Spek, Organometallics 15 (1996) 3022.
[43] S. Berger, S. Braun, H.-O. Kalinowski, NMR-Spektroskopie von
Nichtmetallen, Thieme Verlag, Stuttgart, New York, 1993.
[53] H. Nishida, N. Takada, M. Yoshimura, T. Sonoda, H.
Kobayashi, Bull. Chem. Soc. Jpn. 57 (1984) 2600.
[54] R.E. Ru¨lke, A.L. Spek, C.J. Elsevier, K. Vrieze, Inorg. Chem. 32
(1993) 5769.
.