On factors influencing insertion of allylic substrates in Pd-Caryl bonds

Article abstract of DOI:10.1016/j.ica.2008.09.027

The main products of reactions between palladium(II) aryl complexes [PdArI(phen)] (1a, Ar = C₆H₅-; 2a, Ar = 4-MeO-C₆H₄-; 3a, Ar = 4-CF₃-C₆H₄-; phen = 1,10-phenanthroline), [Pd(C<

Full text of DOI:10.1016/j.ica.2008.09.027

Inorganica Chimica Acta 362 (2009) 2015–2019
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
On factors influencing insertion of allylic substrates in Pd–Caryl bonds
b
b
c
Vincenzo De Felice a, , Augusto de Renzi , Natascia Fraldi , Barbara Panunzi
*
a Dipartimento S.T.A.T., Università degli Studi del Molise, Contrada Fonte Lappone, Pesche (IS) 86090, Italy
b Dipartimento di Chimica ‘‘P. Corradini”, Università degli Studi di Napoli ‘‘Federico II”, Via Cintia, Napoli 80126, Italy
c Dipartimento di Scienze Alimentari, Università degli Studi di Napoli ‘‘Federico II”, Via Università, 100 – Portici (NA), Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 19 June 2008
Received in revised form 5 September 2008
Accepted 14 September 2008
Available online 21 September 2008
The main products of reactions between palladium(II) aryl complexes [PdArI(phen)] (1a, Ar = C6H5–; 2a,
Ar = 4-MeO–C6H4–; 3a, Ar = 4-CF3–C6H4–; phen = 1,10-phenanthroline), [Pd(C6H5)Cl(phen)] (1b) or
[PdAr(phen)(MeCN)]BF4 (1, Ar = C6H5–; 3, Ar = 4-CF3–C6H4–) with allylic substrates CH2@CHCH2A, where
þ
A = OH, OR, OCOR, CN or NMe3 (R = Et or C6H5), or 2,5-dihydrofuran were identified. The nature of the
products affords information on mechanistic features of arylation of allylic substrates prompted by Pd(II),
meanwhile allowing comparison with analogous Pt(II) reactivity. A most noteworthy difference concerns
the fate of the early insertion product derived from cationic complexes, containing the moiety ArCH2-
CH(MII)CH2A. The product can be obtained, in case metal is platinum, only in the absence of a coordinat-
ing solvent, and evolves via CAr,H oxidative addition followed by H,A elimination and formation of a
Keywords:
Palladium(II) aryl complexes
Allylic substrates
Insertion reactions
stable
r
,
p
-chelated arylallyl moiety, thus avoiding b-elimination. The corresponding palladium species,
-derivative prone to b-elimination. The choice
again in the absence of coordinating species, affords a
r
between the two possible elimination pathways is dictated by the relative stabilities of the end products
(both retaining the A function) of the two chains of sequential equilibrium steps.
Ó 2008 Elsevier B.V. All rights reserved.
1. Introduction
2. Experimental
The attainment of C–C bonds by metal-induced reaction of an
hydrocarbyl group with an unsaturated hydrocarbon is the main
achievement of organometal chemistry. Within this area interest
assessed by far [1] for arylation of allylic substrates is increased
even presently by the effective use of ionic solvents [2,3] or water
[4] and by the chance of obtaining new cyclic products [5]. Despite
the fact that palladium is the more useful metal in insertion pro-
motion, many mechanistic investigations on allylic substrates, also
from our laboratory [6–8], dealt in various grades with platinum
chemistry. In fact, the aim of rationalizing the factors of mechanis-
tic relevance gains substantial advantage by the expected minor
lability and easier accessibility as isolated compounds of platinum
intermediates in comparison with analogous palladium species.
However, even recognizing many similarities in the behaviour of
the two d8 ions, extension of results from Pt(II) to Pd(II) has to take
into account the possible substantial differences, as at least sug-
gested by the substantial lack of activity of Pt(II) [9] in Heck catal-
ysis. Thus, notwithstanding the expected failure in attempts to
isolate intermediates, a direct study of palladium activity in stoi-
chiometric reactions appeared of interest. This work aimed to gain
more insight in the insertion of allylic substrates in Pd-aryl moie-
ties and to compare results with platinum chemistry.
2.1. Materials and methods
All air-sensitive compounds were prepared and handled under
a nitrogen atmosphere using standard Schlenk techniques. Anhy-
drous solvents were used in the reactions. Solvents were distilled
from drying agents or passed through columns under a nitrogen
atmosphere. Deuterochloroform (99.8%) and trideuteronitrome-
thane were purchased from Aldrich. 1H NMR spectra were obtained
on 200 MHz Varian or 300 MHz Gemini spectrometer. CDCl3
(CHCl3, d = 7.26 ppm as internal standard), CD3NO2 (CHD2NO2,
d = 4.33 ppm) and CD3CN (CHD2CN, d = 1.93 ppm) were used as
solvents. [Pd(4-R-C6H4)I(tmeda)] [10] and [Pd(C6H5)Cl(phen)]
(1b) [11] have been prepared as described.
2.2. Synthesis of 1a–3a
To a stirred solution of [Pd(4-R-C6H4)I(tmeda)] (1.36 mmol) in
CH2Cl2 (30 ml), 1,10-phenantroline (0.425 g, 2.72 mmol) was
added. After 36 h stirring at room temperature, the precipitated
yellow [Pd(4-R-C6H4)I(phen)] was recovered by filtration, washed
with CH2Cl2 (3 mL) and dried in vacuo. Yield 90%.
Compound 1a: Anal. Calc. for C18H13IN2Pd: C, 44.06; H, 2.67; I,
25.87; N, 5.71; Pd, 21.69. Found: C, 43.97; H, 2.64; I, 25.80; N,
5.68; Pd, 21.73%. 1H NMR in CDCl3: 9.95 (d, 1H, H2, phen,
3J = 4.2 Hz); 8.45 (d, 2H, H4, H7 phen, 3J = 4.5 Hz); 8.00–7.43 (m,
* Corresponding author.
E-mail address: defelice@unimol.it (V. De Felice).
0020-1693/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2008.09.027

2016
V. De Felice et al. / Inorganica Chimica Acta 362 (2009) 2015–2019
3H, H5, H6, H9, phen); 7.75–7.63 (m, 2H, H3,H8, phen); 7.58–6.80
(m, 5H, C6H5).
2.5. General procedure for the reactions of cationic complexes (1), (3)
Compound 2a: Anal. Calc. for C19H15IN2OPd: C, 43.83; H, 2.90; I,
24.37; N, 5.38; O, 3.07; Pd, 20.44. Found: C, 43.75; H, 2.88; I, 24.42;
N, 5.35; O, 3.09; Pd, 20.40%. 1H NMR in CDCl3: 9.85 (d, 1H, H2, phen,
3J = 4.2 Hz); 8.47 (d, 2H, H4, H7, phen, 3J = 4.5 Hz); 8.04–7.80 (m,
5H, H3, H5, H6, H8, H9, phen); 7.71–6.77 (m, 4H, C6H4); 3.80 (s,
3H, OCH3).
Compound 3a: Anal. Calc. for C19H12F3IN2Pd: C, 40.85; H, 2.17; F,
10.20; I, 22.72; N, 5.01; Pd, 19.05. Found: C, 40.76; H, 2.16; F,
10.18; I, 22.75; N, 5.98; Pd, 19.07%. 1H NMR in CDCl3: 9.90 (d,
1H, H2, phen, 3J = 4.2 Hz); 8.51 (d, 2H, H4, H7, phen, 3J = 4.5 Hz);
8.00–7.60 (m, 5H, H3, H5, H6, H8, H9, phen); 7.75–7.20 (m, 4H,
C6H4).
To a stirred solution of the complex (0.2 mmol) in CD3NO2
(1.5 ml), an equimolar amount of the allylic substrate dissolved
in CDCl3 (0.20 mg/ll) was added. The mixture was stirred at room
temperature with the periodical monitoring of the reaction
advancement by 1H NMR spectra. After reaction completion part
of the mixture was filtered on a small bed of Florisil. 1H NMR spec-
troscopy of isolated fractions allowed the identification of prod-
ucts. Hydrogenolysis of another part with excess NaBH4 and
filtration as above allowed to identify the organic moieties still
bounded to metal. For a better assessment of eventual presence
of unsaturated bonds in residual moieties, protonolysis with HCl
was made in some case.
Selected 1H NMR signals of types I and II species from the reac-
tion of 1 and 3 with CH2@CH–CH2O–C6H5:
2.3. Synthesis of [Pd(4-R-C6H4)(CH3CN)(phen)]+BF4ꢀ (1, R = ꢀH; 3,
R = ꢀCF3)
From 1:
I, (CD3NO2): d 7.50–7.00 (m, 5H, C6H5); 7.20–6.90 (m, 5H, O–
The new trifluoromethyl derivative 3 has been prepared as re-
ported for phenyl compound 1 [6]. Both complexes were stored
in nitrogen at ꢀ20 °C.
C6H5); 4.56 (t, 1H, C6H5–CH, 3J = 6.9 Hz); 4.25 (dt, 2H, CH2O,
3
3J = 4.4); 2.37 (app q, 2H, –CH2–, 3J = 6.9 Hz, J0 = 4.4 Hz).
II, (CD3CN): d 7.40–7.10 (m, 5H, C6H5); 7.20–6.90 (m, 5H, O–
C6H5); 5.20 (dd, 1H, C6H5–O–CH, 3J = 4.1); 3.10 (m, 2H, C6H5-
CH2); 2.40 (m, 2H, –CH2–).
Compound 3: Anal. Calc. for C21H15BF7N3Pd: C, 45.07; H, 2.70; B,
1.93; F, 23.77; N, 7.51; Pd, 19.02. Found: C, 45.00; H, 2.69; B, 1.93;
F, 23.75; N, 7.49; Pd, 19.04%. 1H NMR (CD3NO2): d 9.05 (d, 1H, H2,
phen, 3J = 3.5 Hz); 8.82 (dd, 2H, H4, H7, phen, 3J = 4.5 Hz, 3J0 = 3 Hz);
8.26–8.10 (m, 4H, H3, H5, H6, H8, phen); 7.82 (app q, 1H, H9, phen);
7.69–7.43 (m, 5H, C6H5); 2.55 (s, 3H, CH3CN).
From 3:
I, (CD3CN): d7.80–7.20 (m, 4H, CF3–C6H4); 7.00–6.80 (m, 5H, O–
C6H5); 4.90 (t, 1H, CF3–C6H4–CH, 3J = 6.9 Hz); 4.30 (dt, 2H, CH2O,
3
3J = 4.4); 2.37 (app q, 2H, –CH2–, 3J = 6.9 Hz, J0 = 4.4 Hz)
II, (CD3NO2): d 7.80–7.20 (m, 4H, CF3–C6H4); 7.00–6.80 (m, 5H,
O–C6H5); 5.22 (dd, 1H, C6H5–O–CH, 3J = 4.1); 3.20 (m, 2H, CF3–
C6H4–CH2); 2.49 (m, 2H, –CH2–).
2.4. General procedure for the reactions of neutral complexes (1a, 2a,
3a, 1b)
To a stirred suspension of the complex (0.2 mmol) in CDCl3
(1.5 ml), an equimolar amount of the allylic substrate dissolved
3. Results and discussion
in CDCl3 (0.20 mg/ll) was added. The mixture was stirred at room
3.1. Reagents and general reaction procedures
temperature with the periodical monitoring of the reaction
advancement by the recording of 1H NMR spectra. After reaction
completion Pd metal was removed by centrifugation. After filtra-
tion on a small bed of Florisil, 1H NMR spectroscopy allowed the
identification of 4-R-C6H4CH2CH2CHO obtained with yields ranging
65–90% (see Table 1).
The unsaturated substrates were of the type CH2@CHCH2A,
where A = OH, OR, OCOR, CN or NMeþ3 (R = ethyl or phenyl). Also
a cyclic allylic system, i.e., 2,5-dihydrofuran, was used.
The choice of the complexes was dictated by the aim of homo-
geneous comparison between the wanted results and the previous
findings on platinum complexes [6,7] with the same coordinative
environment.
Table 1
Three neutral complexes of general formula PdArI(phen) (1a,
Ar = C6H5–; 1b, Ar = 4-MeO–C6H4–; 1c, Ar = 4-CF3–C6H4–; phen =
1,10-phenanthroline) were used. The simple phenyl derivative
has been quoted before [6], but no synthetic detail or product char-
acterization was given. The three complexes were obtained as sug-
gested by previous results [9] according to Eq. (1)
Reaction of allyl substrate with palladium(II) aryl complexes
Allyl substrate
Complex
Organic product
Yielda
CH2@CHCH2OH
1a
2a
3a
1b
1
C6H5(CH2)2CHO
90
85
75
80
60
30
4-MeO–C6H4(CH2)2CHO
4-CF3–C6H4(CH2)2CHO
C6H5(CH2)2CHO
C6H5(CH2)2CHO
½PdArIðMe2NCH2Þ2ꢁ þ phen ¼ ½PdArIðphenÞꢁ þ ðMe2NCH2Þ2
ð1Þ
C6H5CHMeCHO
The tetramethylethylenediamine precursors were obtained as
previously reported [9] also in the case of the new trifluoromethyl
complex. The complex [Pd(C6H5)Cl(phen)] and the corresponding
cationic species 1 were previously obtained in our laboratory [8].
The synthesis of cationic complexes of general formula [PdAr(-
phen)(MeCN)]BF4 was attempted by the standard reaction of the
suited iodo neutral complex with Ag+ in the presence of acetoni-
trile. While compounds 1 and 3 could be isolated, respectively,
from 1a and 3a, halide abstraction from 2a afforded decomposition
with the release of C6H5OMe and (4-MeO–C6H4–)2 [12].
3
4-CF3–C6H4(CH2)2CHO
4-CF3–C6H4CHMeCHO
C6H5(CH2)2CH2O C6H5
4-CF3–C6H4–(CH2)2CH2O C6H5
C6H5(CH2)2CH2OEt
60
30
CH2@CHCH2OC6H5
1
3
1
1
1
1
1
90b
85
CH2@CHCH2OEt
90b
90b
90b
65c
65c
CH2@CHCH2OCOMe
CH2@CHCH2OCO C6H5
CH2@CHCH2CN
C6H5(CH2)2CH2OCOMe
C6H5(CH2)2CH2OCO C6H5
C6H5(CH2)2CH2CN
CH2 ¼ CHCH2Meþ3
C6H5ðCH2Þ2CH2NMe3þ
Ph
80c
The 1H NMR spectra of all complexes show that the two halves
of the chelated phen are not equivalent, showing also that in the
cationic species the exchange between the nitrogen atoms of the
chelate is slow on the NMR time scale.
1
O
O
1H NMR yield based on the amount of allyl substrate used.
Based on Pd-derivatives.
Based on isolated product.
a
b
c
The general procedure here adopted for the reactions involved a
1.35 M solution of the complex in the suited deuterated solvent

V. De Felice et al. / Inorganica Chimica Acta 362 (2009) 2015–2019
2017
and an equimolar amount of the allylic substrate (dissolved in a
minimum volume of the same deuterated solvent) at room tem-
perature. Solubility of complexes required the use of a different
solvent for neutral (chloroform) and cationic (nitromethane) spe-
cies. Although the 1H NMR spectra of the reaction mixture ap-
peared too complicated for sound identification of the products
in some case, at least monitoring of a suited reagent signal allowed
to assess the reaction completion. Part of the reaction mixture was
analyzed by chromatography in order to isolate and identify the di-
rectly released organic compounds. Another part was treated with
NaBH4 to recover organic fragments not eliminated from the metal.
In some case, also protonolysis with HCl was made with the same
purpose. Identification of organic products was obtained by the
comparison of 1H NMR and mass spectra with those of authentic
samples or the literature data.
sive tautomerization of ArCH2–CH@CHOH is the step that
definitively removes the organic product. The alternative pathway
affording ArCH@CHCH2OH is not observed, owing to the known
reversibility of the b-elimination process in palladium chemistry,
and eventually feeds the irreversible selective formation of the lin-
ear arylpropanal.
3.3. Reaction of cationic complexes with allyl alcohol
Monitoring by 1H NMR, the reaction between equimolar
amounts of the complex (1 or 3) and allyl alcohol in CD3NO2 indi-
cated that the reaction was complete in ca. 30 min. No appreciable
precipitation of metal was observed. Analysis of the reaction mix-
ture disclosed the presence of the linear aldheyde Ar(CH2)2CHO
and of the isomer ArCHMeCHO (Ar = C6H5 or 4-CF3–C6H4–) in ca.
3:1 ratio [13]. Minor presence of propanal was also detected. The
presence of the latter product seems to find a rationale in the ab-
sence of the halide, allowing a longer life to the reactive hydride.
The higher durability of it is in keeping with the production of
propanal by isomerization of the alcohol, and with the addition
of other steps (ii–iv, Scheme 1), affording the branched product,
to those (previous section) affording the linear aldheyde.
As a general observation on the results, it is to be noted that
neutral complexes were efficient promoters of the reaction only
in case the unsaturated substrate was allyl alcohol. Otherwise only
the cationic species afforded the products in sufficiently short time
to prevent unwanted secondary transformation of the early inser-
tion products.
3.2. Reactions of allyl alcohol with neutral complexes
The additional sequential equilibria are similar to the ones in-
voked for the attainment of C6H5CHMeCH2CHO in palladium-cata-
lyzed phenylation of 3-butenol [14].
During the reaction of equimolar amounts of a neutral complex
(1a–3a) and the alcohol in CDCl3, slow separation of palladium me-
tal was observed. The time required for the completion of the reac-
tion was significantly influenced by the nature of the complex. In
fact ca.12 h was required for 1a, ca. 24 h for 1b, while for the tri-
fluoromethyl derivative about 10 days were needed, with conver-
sion less than 10% after 24 h.
However, decrease in chemoselectivity with respect to the per-
formance of neutral complexes needs explanation, since the early
reactive species should be in both cases [PdAr(CH2@CHCH2OH)-
(phen)]+. The finding seems to be correlated to the difference in
dielectric constant of the two solvents and to the presence of the
halide. For a neutral complex, the low dielectric constant of chloro-
In all cases, the analysis of the mixture as such indicated that
the process has the pattern depicted by Eq. (2)
form, together with the presence of the good r
-donor Xꢀ, is ex-
pected to favor the presence of ion pairs of type [{PdAr
(CH2@CHCH2OH)(phen)}+Xꢀ] apt to influence the insertion regio-
½PdXArðphenÞꢁ þ CH2@CHCH2OH
¼ ArCH2CH2CHO þ phen þ HX þ Pd
chemistry [15]. In nitromethane and in the absence of a good
r-
ð2Þ
donor the alternative path for insertion reasonably involves the
intramolecular assistance of OH via a 5-terms ring [16] and
bonding of the allylic terminal carbon to the metal.
r-
Very little trace of propanal was also observed.
The mechanism claimed to be operated in this type of processes
is well known and involves steps: (i) substitution of halide by the
unsaturated species CH2@CHCH2A; (ii) insertion to afford the frag-
ment CH(CH2Ar)CH2A; (iii) hydrogen shift to yield a metal hydride
The substantial absence of palladium metal in the final mixture
points to a fair ‘‘stability” of the metal hydride. In fact, the use of
excess allyl alcohol in the reaction or addition of more alcohol after
reaction completion affords substantial amounts of propanal, thus
confirming the presence of hydride with isomerising ability. Final
loss of this ability is attained by the formation of the stable
with
p-coordinate ArCH2CH@CHA; (iv) release of the arylated
product; (v) tautomerization to give the final product.
Our results prompt three comments. First, the reactivity trend
of the complexes correlates with the inductive effect of the 4-sub-
stituent on the ability of the metal to release (i) the anionic X.
It is also to be noted that the presence of the halide in the start-
ing complex is expected to prompt an easy removal of the reactive
hydride through the elimination of HX. This appears as the ratio-
nale for the high chemoselectivity, as well as for the very scarce
production of propanal (prompted by the hydride). Finally, it is
to be noted that the observed chemoselectivity could in principle
be ascribed to the stereoselectivity of both insertion (ii) and b-
elimination (iii) steps.
[Pd(phen)(g
3-CH2CHCH2)]+, identified by 1H NMR spectroscopy
and reasonably produced by water elimination from the hydride/
allyl alcohol complex.
3.4. Reaction of cationic complexes with CH2@CHCH2A
(A = OEt, OC6H5, OCOMe, OCOC6H5)
The reactions of cationic complexes 1 have been performed in
CD3NO2 following the general procedure. Within 30 min substan-
tial disappearing of the organic reagent was observed, in the ab-
sence of metal precipitation. Treatment of the reaction mixture
with NaBH4 afforded C6H5(CH2)3A as in the schematic Eq. (3) in
yields exceeding 80%.
In fact, the C@C bond polarization by the OH group prompts an
easier formation of the Pd–C
r-bond involving the central allyl car-
bon in (ii). On the other hand, the substantially irreversible conclu-
ð3Þ

2018
V. De Felice et al. / Inorganica Chimica Acta 362 (2009) 2015–2019
Scheme 1.
Attempts to isolate the original palladium species containing
On the other hand, in the presence of a solvent apt to fill the
coordinative unsaturation subsequent to the insertion step, it is ex-
pected (steps v–vii in the scheme) that the preferred b-elimination
involves the methylene near the OC6H5 group, affording double
bond shift from the allylic to the vinylic position [18]. Accordingly,
reinsertion of hydride affords a type II, instead of I, stable end
product.
the organic moiety failed. Anyway, fair grounds for structure
assignment to the major intermediate were given by 1H NMR spec-
tra of the reaction mixture. Particularly, in case A = OC6H5, spectra
indicated the presence of the moiety Pd–CH(C6H5)CH2CH2OC6H5.
In detail, signals at d 4.56 (1H, t, C6H5CH), 4.25 (2H, dt, CH2O part.
overlap with CHD2NO2), and 2.37 (2H, app.q, CH2) were observed.
Phenyl signals occurred as well-resolved groups at 6.9 and 7.2 d
(5H, OC6H5), while a broad group (5H) detected in the 7–7.5 range
could be ascribed to C–C6H5. Further information derived from
addition to the final mixture, again in case A = OC6H5, of an equal
volume of deuteroacetonitrile. Slow transformation of the early
product in a complex displaying 1H NMR signals was observed sup-
porting the sequence Pd–CH(OC6H5)(CH2)2C6H5: d 5.2 (1H, dd,
C6H5OCH), 3.1 (2H, m, CH2C6H5), 2.4 (2H, m, CH2). Moreover, the
addition of excess allylphenyl ether to the mixture prompted the
fast release of C6H5CH@CHCH2OC6H5, while the added allylphenyl
ether gradually isomerized to MeCH@CHOC6H5.
The use of 3 in reaction with allylphenyl ether in nitrometh-
ane, carried as for 1, affords directly a type II species, as indicated
by the 1H NMR spectra of the reaction mixture. Treatment with
NaBH4 afforded 4-CF3–C6H4–(CH2)3OC6H5 in 75% yield. This find-
ing is of further support to the attribution of the g3 bonding in
I. In fact, it is reasonable that the presence of the electron-with-
drawing CF3 and CHPd+ on the benzene ring is a more significant
drawback for
g
3-coordination than for oxopalladacyclus forma-
tion, enhancing importance of path to II even in the presence of
only one equimolar amount of MeCN. In this case, type I species
has been obtained in
a separate experiment by generating
On the ground of above results, the reaction mechanism de-
picted in Scheme 2 appears quite plausible.
in situ ‘‘[(1,10-phen)Pd–C6H4CF3]+” through the reaction of 3a
with AgBF4 followed by the addition of CH2@CHCH2OC6H5. As ex-
pected, subsequent addition of CD3CN afforded the type II
compound.
According to the scheme, the fate of the early product of the
reversible hydride insertion/deinsertion would mainly depend on
the stabilization of the final complex. In pure nitromethane (steps
i–iv), the b-elimination from methylene adjacent to the unsubsti-
tuted phenyl is suited to afford a stable type I intermediate that
on treatment with NaBH4 produces C6H5(CH2)3OC6H5, while in
the presence of free CH2@CHCH2OC6H5 affords C6H5CH@CHCH2O-
C6H5. In this latter case, the excess of ether is isomerized to
CH3CH@CHOC6H5.
As for structure attribution to I, we recall that analogous assign-
ments in palladium [7] or nichel chemistry [16] had been sug-
gested previously. In fact, 1H NMR spectra do not allow to rule
out the alternative assignment of an oxopalladacycle structure
[17]. However, this hypothesis finds less convincing ground in
the behaviour of 3 (below).
3.5. Reactions of cationic complex 1 with CH2@CHCH2A
ðA ¼ CN; NMeþ3 Þ
In the same general conditions as in the previous section, in
nitromethane, the allylic substrate disappeared in about 30 min.
No isolated palladium compounds could be obtained, but spectro-
scopic 1H NMR spectra pointed to a type I structure. Addition of
NaBH4 to the reaction mixture allowed to isolate, respectively,
C6H5(CH2)3CN and [C6H5(CH2)3NMe3]Cl in ca. 60% yield. As far as
we know, no synthetic use of palladium-prompted allyl phenyla-
tion had been previously reported to obtain the aryl derivatives
Ar(CH2)3A reported here.
Scheme 2.

V. De Felice et al. / Inorganica Chimica Acta 362 (2009) 2015–2019
2019
Scheme 3.
3.6. Reaction of 1 with 2,5 dihydrofuran
It is to be noted that the linear organic products are obtained in
all cases, but for the reaction of allyl alcohol with cationic com-
plexes, with very high regioselectivity. It is also noteworthy that
the product deriving from the reaction of 2,5-dihydrofurane is 4-
phenyl-2,3-dihydrofuran, while Heck catalytic phenylation pro-
duces 3-phenyl-2,3-dihydrofuran.
The reaction of 1 with the cyclic allylic system 2,5-dihydrofuran
in deuteronitromethane was accompanied by the fast release of
palladium metal. The reaction mixture contained 4-phenyl-2,3-
dihydrofuran in ca. 70% yield. We recall that the Heck catalytic
phenylation of 2,5-dihydrofuran instead affords 3-phenyl-2,3-
dihydrofuran [19] through insertion of the double bond in Pd–Car
and b-elimination on the oxygen side. Our unexpected result can
find a rationale according to Scheme 3.
In fact, only in case b-elimination could occur on the phenyl
side, the pathway affording the observed product could be acti-
vated. However, this type of elimination has to involve also in
our case the early formation of 3-phenyl-2,3-dihydrofuran, other-
wise the anti position of the H in the putative 4-phenyl-2,3-dihy-
drofuran palladium hydride derivative would forbid elimination.
Thus, dissociation/re-association equilibria with configurational
inversion at the coordinated double bond appear to be a needful
feature of the mechanism.
Acknowledgments
This work was financially supported by the Ministero dell’Uni-
versita‘ e della Ricerca Scientifica e Tecnologica (PRIN 2004). The
authors thank the Centro Interdipartimentale di Metodologie
Chimico-Fisiche (C.I.M.C.F.) of the University of Napoli ‘‘Federico
II” for NMR facilities.
References
[1] R.F. Heck, Adv. Catal. 26 (1977) 323.
[2] S. Bouquillon, B. Ganchegui, B. Estrine, F. Henin, J. Muzart, J. Organomet. Chem.
634 (2001) 153.
As far as we know, no synthetic use of palladium-prompted allyl
phenylation to produce 4-phenyl-2,3-dihydrofuran has been
reported.
[3] V. Calò, A. Nacci, A. Monopoli, V. Ferola, J. Org. Chem. 72 (2007) 2596.
[4] H. Zhao, M. Cai, R. Hu, C. Song, Synth. Commun. 31 (2001) 3665.
[5] J. Rathin, S. Shubhankar, R. Jayanta, Tetrahedron Lett. 49 (2008) 851.
[6] V. De Felice, A. de Renzi, D. Tesauro, A. Vitagliano, Organometallics 11 (1992)
3669.
[7] V. De Felice, M.E. Cucciolito, A. de Renzi, F. Ruffo, D. Tesauro, J. Organomet.
Chem. 493 (1994) 1.
4. Conclusion
[8] F.E. Hahn, M.C. Jahnke, T. Pape, Organometallics 25 (2006) 5927.
[9] Allowisng that heterogeneous platinum-prompted processes are under
investigation, e.g.: M.L. Kantam, M. Roy, S. Roy, M.S. Subhas, B. Sreedhar,
B.M. Choudary, Synlett 14 (2006) 2266.
[10] D. Kruis, B.A. Markies, A.J. Canty, J. Boersma, G. Van Koten, J. Organomet. Chem.
532 (1997) 235.
[11] M.E. Cucciolito, A. de Renzi, F. Giordano, F. Ruffo, Organometallics 14 (1995)
5410.
[12] The mediocre stability of cationic aryl Pd(II) complexes (T. Yagyu, M. Hamada,
K. Osakada, T. Yamamoto, Organometallics 20 (2001) 1087) is reasonably
expected to be even lower if electron-donor inductive effect increases
reactivity of the metal towards oxydative addition.
[13] We recall that in water a 4.5 to 1 ratio was founded: V. Didier, N. Belkacem,
Procedings of ESOC-1 and -2; First and second International Electronic
Conference on Synthetic Organic Chemistry (1999) 30.
[14] R.C. Larock, W.J. Leung, S. Stolz, Tetrahedron Lett. 30 (1989) 6629.
[15] A. Macchioni, Chem. Rev. 105 (2005) 2039.
[16] H. Lemkuhl, T. Keil, R. Benn, A. Rufinska, C. Kruger, J. Poplawska, M.
Bellenbaum, Chem. Ber. 121 (1988) 1931.
[17] C. Fernandez-Rivas, D.J. Cardenas, B. Martin-Matute, A. Monge, E. Gutierrez-
Puebla, A.M. Echavarren, Organometallics 20 (2001) 2998.
[18] T. Bieg, W. Szeja, J. Carbohydr. Chem. 4 (1985) 441.
[19] R.C. Larock, B.E. Baker, Tetrahedron Lett. 29 (1988) 905.
The main difference in reactivity of PdII and PtII emerging by
the above results, beside the expected higher reactivity of palla-
dium complexes (that in contrast with platinum can eventually
react also as neutral species), concerns the fate of the early
insertion product including the sequence ArCH2CH(MII)CH2A. As
for platinum, even the attainment of the coordination of the
allylic substrate requires the absence of a coordinating solvent.
In this case, the C–HAr bond is activated, thus prompting ortho-
metallation of the aryl ring, followed by HA elimination and for-
mation of a
r,p-chelated arylallyl moiety depleted of the A func-
tion. Also in the absence of coordinating species, the early
insertion of palladium product, through reversible deinsertion/
reinsertion steps, affords a stable derivative, possibly a
p-allyl
species, where the organic moiety retains the A function. If a
coordinating solvent is present, the more stable end product of
the sequential equilibrium steps is instead a
retaining the A function.
r-derivative, also