Carbonylation reactions: 8. Chlorocarbonylation reactions: Catalytic formation of chloroacyl chlorides without phosgene and application to the synthesis of 2(H)-5-furanone

Article abstract of DOI:10.1016/S1381-1169(98)00386-0

Chloroacyl halides are obtained under mild conditions by catalytic carbonylation in halogenated solvents of allylic halides or ethylene and allyl chloride mixtures under pressure of CO and anhydrous HCl in the presence of PdCl₂; excellent selec

Full text of DOI:10.1016/S1381-1169(98)00386-0

Journal of Molecular Catalysis A: Chemical 143 Ž1999. 181–195
www.elsevier.comrlocatermolcata
Carbonylation reactions
8
. Chlorocarbonylation reactions: catalytic formation of
chloroacyl chlorides without phosgene and application to the
ž /
synthesis of 2 H -5-furanone
1
)
Michel C. Bonnet , Nathalie Carmona, Igor Tkatchenko
Laboratoire de Chimie de Coordination, CNRS, 205 route de Narbonne, 31077 Toulouse Cedex, France
Received 27 July 1998; accepted 28 September 1998
Abstract
Chloroacyl halides are obtained under mild conditions by catalytic carbonylation in halogenated solvents of allylic
halides or ethylene and allyl chloride mixtures under pressure of CO and anhydrous HCl in the presence of PdCl ; excellent
2
selectivities for dichloroacyl and chloroacyl derivatives are achieved. Allyl chloride is partly consumed for the regeneration
of palladium chloride. The product of oxidative addition of HCl to the oligomeric wPdŽCO.Clx formed in situ by reaction of
n
CO on PdCl2 is proposed as the first step towards the generation of the active species. q 1999 Elsevier Science B.V. All
rights reserved.
Keywords: Chlorocarbonylation; Chloroacyl chlorides; Allyl halides; Ethylene; Palladium; Reaction mechanisms
1
. Introduction
stepwise incorporation of carbon monoxide and
chlorine to an unsaturated substrate or through
the formal insertion of carbon monoxide into a
C–Cl bond w3x. Indeed, acyl halides are reported
Acyl halides and their halogenated deriva-
tives are worthy intermediates for the synthesis
of fine chemicals w1,2x. The most attractive way
to prepare these compounds could be by transi-
tion metal-catalysed carbonylation, either via
as active intermediates in several processes
leading to carboxylic acids w4x and esters w5x.
Carbonylation of allyl halides is well-docu-
3
mented
Ž
Scheme 1
.
and involves
Ž
h -allyl
.
com-
plexes which can generate acyl halides, espe-
cially in the case of palladium w6,7x or different
)
Corresponding author. Present address: Laboratoire de
families of acids and ketoacids and their esters
when further reactions with ethylene or acety-
lene are possible, as in the case of nickel w8x.
Synth e` se et Electrosynth e` se Organom e´ tallique, UMR 5632
CNRS-Universit e´ de Bourgogne, Facult e´ des Sciences Gabriel, 6
Boulevard Gabriel, F-21004 Dijon Cedex, France. Tel.: q33-380-
3
96-072; Fax: q33-380-396-098; E-mail:
The reaction of terminal alkenes Eq. 1
Ž Ž .. is
tkatchen@u-bourgogne.fr
1
much less facile. Of all the elements of groups
Present address: Institut de Recherches sur la Catalyse, CNRS,
2
Avenue Albert Einstein, 69626 Villeurbanne Cedex, France.
8–10 examined, palladium appears to be the
1
381-1169r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
PII: S1381-1169Ž98.00386-0

1
82
M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
2
. Experimental
2
.1. Materials
Scheme 1.
Chlorine
ide Air Liquide N20 and N60
Liquide N25 , hydrogen chloride
electronic grade , and phosgene
Ž
Air Liquide N27
.
, carbon monox-
, ethylene Air
Air Liquide,
SNPE were
Ž
.
Ž
most efficient, although the reactions proceed at
relatively high pressures w9,10x except in the
presence of additional Lewis acids like GeCl2
or SnCl w11x. Yet, for all the cases reported,
the yields are low, due mostly to a poor selec-
tivity for the chlorocarbonylation process.
.
Ž
Ž
.
.
used as such. Buta-1,3-diene is dried over
molecular sieves; allylic halides are distilled
under argon over phosphorous anhydride. Ace-
tonitrile, dichloromethane, and dichloro-1,2-
ethane are distilled under argon over phospho-
rous anhydride; methanol is distilled under ar-
gon in the presence of magnesium chips; diethyl
ether and toluene are distilled under argon over
2
1
Ž .
The simultaneous chlorination–carbonylation
of olefins has been mostly exemplified by Tsuji
et al. b-Chloroacyl chlorides w12x and the corre-
sponding esters w13x are produced from terminal
sodiumrbenzophenone. Anhydrous copper
chloride Janssen, 98% and palladium chloride
Johnson Matthey are used as such. 10% PdrC
Aldrich has been analysed for palladium
7.98% and chlorine 1.06% and is used as
PhCN.2, w h -C H5.PdClx , PdCl -
PPh3.2 and Pd
PPh
II
Ž .
Ž
.
Ž
Ž
Ž
.
.
olefins in low yields, based on PdCl2 Eq. 2 .
Ž Ž ..
.
Ž
.
In addition, some 3,4-dichlorobutanoates are
formed when allyl chloride is the starting prod-
uct w14x. It is noteworthy that the reaction of
olefins with CO in the presence of carbon tetra-
chloride is catalysed by binuclear metal car-
bonyls and affords b-trichloromethylacyl chlo-
rides in satisfactory yields w15,16x. Therefore, if
a halogen source is available, a catalytic process
can be achieved. In fact, addition of excess
3
such. PdCl2
Ž
Ž
3
2
2
Ž
Ž
.
are prepared according
MeCN.4
3 4
to Refs. w20–23x; wPd
Ž
x
Ž
BF
.
is ob-
4 2
tained by protonation of Pd
Žacac.2 with HBF4 P
OEt in acetonitrile w24x.
2
2
.2. Catalytic chlorocarbonylations
Caution! All the following experiments are
copper
favours the catalytic conversion of ethylene into
-chloropropionyl chloride, albeit under drastic
conditions.
II
Ž . chloride w17x or of phosgene w18x
run in a dedicated facility equipped with effi-
cient fume cupboards, carbon monoxide, chlo-
rine, and phosgene detectors.
3
All catalytic runs are performed in a 100-ml
Hastelloy C276 autoclave equipped with a dou-
ble envelope for heating with an oil-circulating
bath. The cover of the autoclave is supplied
with a thermocouple probe; a stopcock for the
introduction of the catalyst solution and two
valves for introduction of gases and sampling,
2
Ž .
We report herein a new catalytic system for
such a double functionalisation reaction leading
to chloroacyl and dichloroacyl chlorides which
use hydrogen chloride as a source of chlorine,
CO, and Pd compounds and the application of
respectively; a manometer 0–200 bar ; and a
Ž .
safety valve adjusted to 100 bar. The agitation
these intermediates for the preparation of 2
-furanone, a useful building-block for bioac-
tive molecules. A preliminary account of part of
this work has been given w19x.
H -
Ž .
is ensured with a magnet bar.
5
In the standard test, PdCl2
and allyl chloride 7.65 g, 100 mmol
in a Schlenk tube under argon. The suspension
Ž
355 mg, 2 mmol
.
Ž
.
are placed

M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
183
is introduced in the autoclave and the Schlenk
tube rinsed with dichloromethane 2=3 ml,
=2 ml . The autoclave is pressurised with
HCl 9 bar and CO 24 bar successively. The
autoclave is heated to 1008C for 4 h. The reac-
tion is first quenched by cooling in a water bath
MS
Ž
EI, 70 eV
100
26.2
.
mrz: 143
Ž
10
2.6
4.2 , 103
.
, 141
Ž
67
15.7
.
,
,
q
Ž
139
111
Ž
.
wC H Cl CO x; 115
Ž
.
, 113
Ž
Ž
.
3
5
2
q
2
.
Ž
.
wC H Cl x; 105
Ž
95
.
13.8
.
3
5
2
q
q
Ž
.
Ž
.
wC H ClCO x; 77
Ž
32
.
, 75
Ž
.
wC H Cl x; 39
3
4
3
4
q
Ž
931
. wC H x.
3
3
H NMR
Ž
200 MHz, CDCl3 dd,
.
Ž
: d 3.3
Ž
2
2
Ž
ca. 208C
The remaining gases are vented through a dry
ice trap, and methanol 4 ml, 100 mmol is
.
, then in an acetone bath
Ž
y208C
.
.
Js17.8, 3Js8.5 Hz, 1H
.
, 3.65 dd, Js17.8
3
2
3
Js4.2 Hz, 1H
, 3.86 dd, Js11.6 Js4.5 Hz, 1H
4.37–4.49 ppm m, 1H
50.3 MHz, CDCl3
, 3.7 dd, Js11.6, Js8.1
. Ž
2
3
Ž
.
Hz, 1H
.
Ž
.
,
slowly added to the reaction mixture which is
stirred for 20 min until the evolution of HCl
stops. Samples can be taken at given intervals,
quenched at y208C with methanol and anal-
ysed. Propene is estimated as the difference
between the amounts of converted allyl chloride
and evolved 2-chloropropane. Each experiment
is duplicated and in some cases, triplicated.
GC analysis is performed with a 10% Car-
Ž
.
.
1
3
C NMR
Ž
.
: d 46.3
s134
Ž
J
13C – H
s154 Hz, –CH Cl
Hz, –CH2–
170.1 ppm
.
, 51.5 J
Ž
13C – H
s156 Hz, –CH–
2
.
, 53.5
CO
Ž
J
13
.
,
C – H
Ž
.
.
2.3.2. 4b: IR (neat, KBr): 1723 cmy (nCO
1
)
MS
Ž
CI, NH3
.
mrz: 192
Ž
2.8
.
, 190
Ž
20.4
.
.
,
q
bowax on Gas Chrom Q
Ž
80–100 mesh
.
column
188 29.4
Ž
.
wC H Cl CO MeNH x; 99
Ž
100
3
5
2
2
4
Y
q
Ž
2 m=1r8 ; gas vector: nitrogen at 14 ml
min ; injection temperature: 2708C; detection
wC H CO Me x.
3
4
2
y
1
MS
Ž
Ž
EI, 70 eV
.
mrz: 143
Ž
10
.
, 141
Ž
Ž
67
48
.
,
q
temperature: 2508C; temperature programma-
139 100
.
wC H Cl CO x; 137
Ž
15
.
, 135
.
3
5
2
y
1
tion: 608C during 0.5 min, 88C min
608C using ethyl crotonate as internal stan-
dard.
up to
wC H ClCO Mex; 115
Ž
2.6
.
C H Cl CO Me,
3
5
2
3
5
2
2
q
1
.
113
Ž
15.7
.
, 111
Ž
26.2
.
wC H Cl x; 99
Ž
100
.
3
5
2
q
q
wC H CO Me x; 77
Ž
32
.
, 75
Ž
95
.
wC H Cl x;
3
4
2
3
4
q
q
Hydrogen chloride could be replaced by the
appropriate amounts of chlorine or phosgene by
weighting the autoclave before and after addi-
tion of the reagent. Reactions in the presence of
additives are run under the same conditions, the
appropriate amount of additive being placed
591
Ž
38
.
wCO Me x; 39
Ž
93
.
wC H x.
2
3
3
H NMR
Ž
200 MHz, CDCl3
, 3.00
s, 3H , 3.75
, 3.85
, 4.39–4.52 ppm
50.3 MHz, CDCl3
, 47.2
.
: d 2.75
dd, Js16.5,
Ž
dd,
2
3
2
Js16.5, Js8.5 Hz, 1H
.
Ž
3
2
Js4.7 Hz, 1H
.
, 3.73
Ž
.
Ž
dd, Js
3
2
11.4, Js7.6 Hz, 1H
.
Ž
dd, Js11.4,
m, 1H
3
Js4.8 Hz, 1H
.
Ž
.
.
1
3
after PdCl in the Schlenk tube.
C NMR
Ž
.
: d 39.8
2
Chlorocarbonylation of ethylene is run under
the same conditions, the order of addition of the
Ž
J
13C – H
s131 Hz, –CH –
Hz, –CH Cl
.
Ž
J
13C – H
s152
2
.
, 51.9
Ž
J
13
s147 Hz, –OCH ,
.
2
C – H
3
gases being:
tial pressures12 bar
ide
Ž
1
.
HCl
Ž
9 bar
, and
.
,
Ž
Ž
3
.
.
2
.
.
ethylene
Ž
par-
52.6
CO
Ž
J
13C – H
s159 Hz, –CH– , 169.7 ppm
.
.
carbon monox-
Ž
.
.
Ž
partial pressures12 bar
2.3.3. 20a, c: IR (neat, KBr): 1800 cmy (nCO
MS CI, CH4 mrz: 223 6.9 , 221 15.5
219 10.5 26.0 , 185
100 , 183
25.1
1
)
.
,
2
2
1
.3. Product identification
Ž
.
Ž
.
Ž
.3.1. 4a: IR (neat, KBr): 1800 cmy (nCO
1
)
q
Ž
.
Ž
.
wC H BrCl CO x; 187
Ž
.
3
6
2
q
MS
Ž
CI, CH4
.
mrz: 179
Ž
6.4
.
, 177
9.5
wC H Cl CO x; 115
Ž
.
21.8
, 141
1.3
, 103
.
,
Ž
Ž
69.4
.
wC H BrClCO x; 159
Ž
Ž
Ž
6.1
1.5
3.7
.
.
.
,
,
3
5
q
q
75
Ž
22.2
, 139
8.9
wC H ClCO x.
.
wC H Cl COCl x; 143
Ž
157
141
111
.
, 155
, 139
wC H Cl x; 105
Ž
19.5
.
wC H BrCl x; 143
3
6
2
3
5
q
q
Ž
61.8
13
13.8
.
Ž
100
.
Ž
.
,
Ž
8.6
Ž
.
Ž
13.3
.
wC H Cl CO x; 113
,
.
3
5
2
3
5
2
q
q
1
Ž
.
, 111
Ž
13.7
.
wC H Cl x; 105
Ž
4.2
.
6.5
.
Ž
4.2
.
, 103
Ž
11.5
3
5
2
3
5
2
q
q
Ž
.
wC H ClCO x.
3
4
3
4

1
84
M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
.3.4. 21a, c: IR (neat, KBr): 1800 cmy (nCO
1
)
,
2
Evaporation of dichloromethane provides 1.579
18.8 mmol of 33 70% yield
Bps213–2148C. IR
MS
65 11.5
, 229
2.3 , 201
, 185 77.9
Ž
CI, CH4
, 263
100
6.1
.
mrz: 269
Ž
2.3
.
, 267
Ž
9.2
.
g
Ž
.
Ž
.
.
q
2
Ž
2
Ž
1
1
1
Ž
.
Ž
5.3
.
wC H Br ClCO x; 231
Ž
KBr, neat
. H NMR 200 MHz,
.
: 1785, 1750
Ž
3
6
2
q
y
1
1
48.1
03
19.1
.
Ž
.
, 227
Ž
50.4
.
wC H Br CO x;
Ž
n
.
, 1600 cm
Ž
n
.
3
5
2
CO
C
5C
q
3
4
Ž
.
Ž
.
, 199
77.8
, 155
Ž
6.1
.
wC H Br x; 187
CDCl3
Hz , 6.17
ppm 1H, dt Js5.8 Hz, Js1.65 Hz .
.
.
: d 4.89
Ž
2H, dd, Js2.2, Js1.65
3
5
2
q
3 3
.
Ž
.
, 183
34.4
6.1
Ž
.
Ž
wC H BrClCO x;
.
Ž
1H, dt, Js2.2, Js5.8 Hz , 7.58
.
3
5
q
3
4
59
49
19
Ž
7.6
6.1
8.4
.
.
, 157
Ž
Ž
.
29
.
wC H BrCl x;
Ž
3
5
q
Ž
Ž
, 147
.
wC H BrCO x; 121
Ž
Ž
6.1
.
,
3
4
q
.
wC H Br x; 105
Ž
8.4
.
, 103
6.9
.
3
4
q
wC H ClCO x.
3. Results and discussion
3
4
y
1
2
.3.5. 23b: IR (neat KBr): 1720 cm (nCO
)
3.1. Chlorocarbonylation of allyl chloride with
homogeneous palladium catalysts
q
MS
pea1k , 63, 59, 55, 27.
H NMR
200 MHz, CDCl3.: d 2.77
Js6.7 Hz
3H, s
Ž
EI
.
mrz: 124
Ž
M
.
, 122, 91, 87
Ž
base
.
Ž
Ž
2H, t,
.
, 3.7 ppm
Reaction of allyl chloride with hydrogen
chloride and carbon monoxide under moderate
pressures is catalysed by molecular Pd com-
2
2
.
, 3.73
Ž
2H, t, Js6.7 Hz
Ž
.
.
1
3
C NMR
.
8.7, 51.7, 170.5 ppm CO .
Ž
50.32 MHz, CDCl3 : d 37.1,
.
pounds as well as 10% PdrC
Scheme 2 .
Ž .
Compounds 1–6 are characterised by GC,
3
Ž
1
13
GCrMS, IR, and H and C NMR. Trace
amounts of hydrogen and phosgene are detected
in the gas phase. Running the reaction with high
2
.4. 3,4-Dichlorobutanoic acid
A mixture of 4a
carbonate 2 g, 14.5 mmol
heated in dichloromethane
brown solution is cooled down to RT and dis-
Ž
2.5 g, 14 mmol
and water
10 ml for 4 h. The
.
, potassium
purity carbon monoxide Air Liquide N60 also
Ž .
leads to the formation of hydrogen, probably via
the water–gas shift reaction w25x involving
Ž
.
Ž
Ž
5 ml is
.
.
residual water present in the reagents used.
Blank experiments indicate that the formation of
2-dichloropropane 5 and propene 6 requires the
presence of both CO and HCl; the absence of
CO gives mostly rise to the formation of the
adduct of HCl on allyl chloride. The expected
product, 3,4-dichlorobutanoyl chloride, 4a,
which is only cited by Tsuji et al. w14x has been
tilled under reduced pressure
the acid 34 59% yield
Fs62–658C. Anal.: found
40.8 , H 4.28 4.17% . IR
nOH , 1750, 1710 cm
MHz, CDCl3 : d 4.18
Ž
1 Torr
.
to provide
Ž
.
.
Ž
calc.
.
: C 41.6
br,
200
Ž
.
Ž
.
Ž
nujol
.
: 3500
. H NMR
Ž
4
Ž
y
1
1
.
Ž
nCO
.
3
.
Ž
2H, dd, Js5.9, Js
3
4
1
7
.7 Hz
.08 ppm
50.32 MHz, CDCl
.
, 6.12
1H, dt, Js15.3, Js5.9 Hz
: d 42.1 t, J
Ž .
1H, dt, Js15.3, Js1.7 Hz ,
3 3 13
Ž
.
.
C
s
fully characterised see Section 2 . Systematic
Ž .
NMR
65.9 Hz
Ž
.
Ž
13C – H
studies have been carried out with further
methanolysis of the reaction mixture in order to
simplify GC analyses.
3
1
.
, 123.1
Ž
d, J
13C – H
s151.4 Hz , 144.3
.
Ž
d, J
13C – H
s162.3 Hz , 170.8 ppm.
.
The best solvents for the reaction are halo-
2
.5. 2(H)-5-Furanone
genated hydrocarbons
ŽCH Cl , C H Cl2., but
2 2 2 4
A mixture of 4a
cium carbonate 2.808 g, 28 mmol
in water 20 ml for 5 h. The clear yellow
solution is cooled down to RT and the buteno-
lide 33 is extracted with dichloromethane 10=
ml . The organic layer is dried over MgSO4.
Ž
4.7 g, 27 mmol
.
and cal-
is refluxed
Ž
.
Ž
.
Ž
5
.
Scheme 2.

M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
185
the reaction proceeds also in toluene or without
solvent, although lower selectivity and yield in
reaction temperature Table 1, entries 2, 6–8 ;
Ž .
however, selectivity for this compound de-
creases sharply above 1008C, owing to an in-
crease in the formation of 1, as already ob-
served by Tsuji et al. w14x. The conjugated ester
2b is always detected as traces, even at higher
temperatures.
The amount of 4 produced is stable after 8 h
reaction as the build-up of 5 which follows a
parallel evolution Table 1, entries 9–15 . Nev-
Ž .
4
are observed Table 1, entries 1–5 . The
Ž .
overlapping of the peaks of toluene and 4b
avoids in this case the determination of yield
and selectivity for this compound. The reaction
is prevented when protic reagents
methanol, water are added to the system, even
in small quantities. Although small amounts of
b are still produced in addition to 5, the main
compound obtained under these conditions is
-methoxypropene. The reaction starts only
Ž
e.g.,
.
1
ertheless, the selectivity for 4 decreases with the
reaction time due to steady formation of 1 and 3
up to 23 h. Monitoring by GC of the first stages
3
above 608C. The yield of 4 increases with the
Table 1
Chlorocarbonylation of allyl chloride: influence of the reaction parameters on yields and product selectivity
Entry
Solvent
Temperature
Ž8C.
Time
Žh.
CO
Žbar.
HCl
Žbar.
N2
Žbar.
Yield Ž%.
Selectivity
1
arb
3arb
4arb
5
6
4 Ž%.
1
2
3
4
5
6
7
8
9
r
100
100
100
100
100
60
4
4
4
4
4
4
4
4
2
24
24
24
24
24
24
24
24
48
48
48
48
48
48
48
12
18
36
48
48
48
48
24
24
24
12
16
18
12
18
24
24
24
36
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
r
12
11
6
4
3
3
–
3
1
1
1
2
23
26
30
24
0
15
23
19
16
52
7
33
22
15
–
34
8
82
87
91
–
CH Cl
2
2
C H Cl
2
C H Me
MeOH
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
CH Cl
2
4
6
5
traces
traces
traces
2
4
5
8
11
13
13
12
traces
1
2
3
2
0
traces
1
12
2
3
2
2
6
4
6
3
2
4
7
4
3
3
10
12
36
6
3
5
3
6
5
7
8
89
92
74
80
76
75
70
61
64
64
86
87
83
77
70
71
76
74
70
45
81
86
85
83
83
84
84
81
83
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
80
12
41
24
25
30
31
30
30
32
18
20
29
34
16
20
25
32
30
29
29
25
33
38
34
36
41
42
39
9
12
20
16
19
20
18
16
16
15
24
19
25
18
8
11
16
22
24
13
27
28
26
18
14
15
12
12
14
2
130
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
23
22
22
24
27
26
27
28
12
14
17
23
11
17
22
18
17
21
23
19
22
27
25
26
27
28
27
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
4
8
2
2
16
23
32
65
4
4
4
4
4
4
4
4
4
4
4
4
4
2
2
2
2
2
2
9
9
3
6
2
2
2
4
5
1
1
2
9
2
14
18
24
9
6
9
9
9
9
2
2
traces
2
1
1
1
5
1
2
1
1
5
2
2
2
4
4
4
4
4
4
27
21
15
10
6
2
2
2
14
18
9
2
9
3
2
3
2
Reaction conditions: wPdCl xs2 mmol, wC H Clxs100 mmol, solvents10 ml.
2
3
5

1
86
M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
of the reaction indicates this selectivity is the
highest after 40 min Fig. 1 . The initial selec-
tivity is 70% suggesting that 4 is the kinetic
product of the chlorocarbonylation reaction. This
is confirmed by the steady formation of 1 with
variation of these values at the same initial total
pressure by introducing a sufficient amount of
Ž
.
nitrogen. Inspection of Table 1
for a total pressure of 33 bar, entries 29–34 for
a total pressure of 48 bar indicates that the
Ž
entries 26–28
.
time
Ž
Fig. 1 and Table 1
.
, hence suggesting that
selectivity for 4 remains practically constant;
this compound is either a decomposition prod-
uct of 4 or the product of a side reaction
involving another, less active, palladium cata-
lyst.
The increase in carbon monoxide pressure
from 12 to 48 bar leads to the expected increase
the lowest ones are observed for the lowest
COrHCl pressure ratio
i.e., 1.33 , which corre-
Ž .
sponds to the optimal ratio for the formation of
1 in both total pressure conditions. The highest
yield of 4 is observed at COrHCls2 for a
total pressure of 33 bar, but at lower values
in the yields of any carbonylation product
Ž
Ta-
Ž
i.e., 1.3–1.7
to low amounts of 3 produced
has been at least duplicated and provides results
within "3% , the trends for its formation with
.
for a total pressure of 48 bar. Due
ble 1, entries 2, 16–19 . Again, the increase in
.
Ž
each experiment
the formation of 5 follows the increase in the
production of 4. However, at higher pressures,
the formation of 1 and 3 occurs more rapidly,
therefore leading to a slight decrease in selectiv-
ity for 4. The effect of hydrogen chloride pres-
sure is more contrasted and varies with the
partial pressure of carbon monoxide. When pCO
s48 bar, a linear increase in the formation of 3
.
respect to partial and total pressures are less
clear.
The chlorocarbonylation reaction proceeds
with different types of palladium complexes
Ž
Table 2, entries 2, 35–40 . There is no influ-
.
ence of the initial oxidation state for neutral
complexes; moreover, PdrC is operative, lead-
ing to similar a yield of 4, albeit with a lower
and 4 is observed
is further observed up to 14 bars
where a plateau is reached Table 1, entries 2,
3–25 . However, at that stage, the amount of 1
Ž
Table 1, entries 20–22
.
. This
Ž
pCO s24 bar
.
Ž
selectivity Table 2, entry 41 ; further results
Ž .
2
.
with this precursor will be detailed below. How-
ever, the catalyst activity seems to be sensitive
to the nature of the ligands present on the
catalyst precursors. Ligands which are easily
protonated such as h -allyl w26x and acac w27x,
or labile like PhCN induce better activities; the
lower activity of the dicationic complex
increases more rapidly, hence lowering the se-
lectivity for the dichloro derivative. A delicate
balance dictates the highest yields and selectivi-
ties for 4. Since these values can depend on the
total pressure in the autoclave, we examine the
3
wPdŽMeCN.4 xŽBF4.2 may result from the
stronger retention of the acetonitrile ligands in
the coordination sphere of the palladium centre
during the formation of the active species.
Noteworthy is the total inhibition by triph-
enylphosphine entries 42, 43 which prevents
Ž .
the formation of the catalyst or poisoned it; both
reasons may take place. In fact, addition to
PdCl2 of 1 equivalent of triphenylphosphine
strongly reduce the catalytic activity and prod-
Fig. 1. Evolution of allyl chloride conversion and yield of 4b in
the first stages of the chlorocarbonylation of allyl chloride Žreac-
tion conditions: wPdCl xs 2 mmol, wC H Clxs100 mmol,
uct yields Table 2, entry 44 . Addition of ureas
Ž .
which are reported to facilitate the carbonyla-
tion of allyl chlorides w28x is also depressing the
catalyst performances Table 2, entries 45–47 .
Ž .
2
3
5
CH Cl s10 ml, p s24 bar, pHCl s9 bar, T s1008C, ts4
2
2
CO
h..

M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
187
Table 2
Chlorocarbonylation of allyl chloride: influence of the reaction parameters on yields and product selectivity
Entry
ComplexrAdditive
SrC
Conversion Ž%.
Yield Ž%.
Selectivity
1
arb
3arb
4arb
5
6
4 Ž%.
2
PdCl2
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
40
66
100
200
500
50
50
75
61
75
67
75
85
54
88
5
3
2
3
3
3
4
3
9
2
3
1
1
2
1
1
2
1
3
26
22
31
23
29
30
12
30
0
0
2
10
1
1
30
22
25
21
40
42
42
traces
16
traces
24
29
29
29
25
12
20
19
23
9
22
22
22
22
22
28
27
17
1
2
12
3
12
9
19
28
12
11
11
8
8
9
6
15
9
87
88
86
85
88
83
75
71
0
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
6
6
6
6
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
PdŽdba.2
PdCl ŽPhCN.
22
18
20
21
11
29
2
1
3
10
3
2
24
24
18
16
30
28
28
6
2
2
PdCl2 Žcod.
ŽC H PdCl.
3
5
2
PdŽacac.2
wPdŽMeCN. xŽBF .
4 2
4
PdrC
PdŽPPh .
traces
traces
traces
traces
traces
traces
1
traces
1
1
1
1
1
3
4
PdCl ŽPPh .
6
0
2
3 2
PdCl qPPh
17
25
17
13
82
93
59
54
86
83
85
18
38
24
57
72
78
69
73
47
63
54
traces
2
1
1
8
19
3
5
4
4
5
3
3
2
5
3
3
3
100
83
50
50
77
54
86
78
89
89
89
100
84
100
80
88
88
88
86
80
77
83
2
3
PdCl qTMU
2
a
PdCl qTMU
2
PdCl qTMTU
2
PdCl qSnCl
2
2
b
2
PdCl qSnCl
2
PdCl qLiCl
2
b
PdCl qLiCl
2
PdCl qCuCl
2
2
c
2
b
2
PdCl qCuCl
2
PdCl qCuCl
2
a
PdCl qPPNCl
traces
traces
traces
2
PdCl qMe NHPHCl
13
7
2
2
b
PdCl qMe NHPHCl
2
2
Bu NCl
1
1
1
1
2
1
2
1
18
18
23
22
21
16
15
15
4
PdCl2
PdCl2
PdCl2
PdCl2
PdCl2
wPdx, soluble
wPdx, insoluble
21
22
23
19
28
22
26
2
2
4
3
d
d
a
2
5
2
equivrPd.
b
c
d
equivrPd.
.5 equivrPd.
Recycling experiment: see text.
Reaction conditions: wPdxs2 mmol, wadditivexs1 equiv with respect to palladium, if not stated, wC H Clxs100 mmol, if not stated,
3
5
CH Cl s10 ml, p s24 bar, p s9 bar, Ts1008C, ts4 h.
2
2
CO
HCl
On the other hand, chlorides may have a strong
effect on activity and yields Table 2, entries
8–58 . Tin II chloride increases both conver-
sion and production of chlorocarbonyl products.
However, the selectivity for 4 is lower, due to
the higher amounts of 1 formed, especially in
the presence of an excess of the additive. Ionic
chlorides like LiCl are not improving the yields,
though the selectivity for the expected 4 is
results, but still with the production of large
amounts of 5, even if the chloride is in excess
with respect to palladium chloride. Onium salts
Ž
4
.
Ž .
are much less effective Table 2, entries 55–58 .
Ž .
Finally, variation of the substrate to catalyst
ratio, SrC, has no effect on the catalyst perfor-
mances up to SrCs200. For SrCs500, con-
version and yield of 4 decrease, which also
leads to a lower selectivity. The lack of re-
sponse on catalyst concentration may suggest
higher. Copper II chloride provides the best
Ž .

1
88
M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
that under the reaction conditions used, part of
the active species precipitates from the solution.
This is suggested by the observation at the end
of the reaction of a precipitate which corre-
presence of phosgene, carbon monoxide, hydro-
gen chloride and carbon dioxide, indicating some
phosgene hydrolysis. The presence of water in
small quantities is inherent to the introduction
of phosgene.
sponds to ca. 50% of the PdCl engaged. It is
2
noteworthy that changing PdCl to Pd
Ž . Ž
PdCl2 cod or PdCl2
Ž
dba.2
PhCN.2 and running the
chlorocarbonylation for two other SrC ratios
100 and 500 does not lead to strong variations
,
At the end of the reaction, the mixture is
made up of two phases: one red solution and
one green precipitate. Each phase contains dif-
2
Ž
.
ferent palladium compounds vide infra , which
Ž .
of the catalytic performances.
have been isolated and re-used for the carbony-
lation of allyl chloride. Both isolated systems
are active and give identical product distribu-
Ž .
tions Table 2, entries 64 and 65 . However, the
yield of 3,4-dichlorobutanoyl chloride is some-
This new synthesis of 3,4-dichlorobutanoyl
chloride represents a considerable improvement
compared to the only method described so far
w14x. The nature of the halogen source during
the catalytic process could be chlorine, phos-
gene, chloride ion or palladium chloride. The
reaction performed under a pressure of chlorine
what smaller.
3.2. Chlorocarbonylation of allyl chloride with
and carbon monoxide
tively , in the presence of palladium chloride, at
008C, for 4 h only gives 1,2,3-trichloropro-
Ž
5 and 24 bar, respec-
PdrC
.
1
The use of supported palladium catalysts
could provide an easier work-up of the reaction
mixture and the possibility of recycling the
catalyst. Since palladium on charcoal has been
pane. This reaction proceeds easily, even at
room temperature and without any metal as
catalyst. Carbonylation of 1,2,3-trichloropro-
pane to 4 does not proceed under the reaction
conditions used. The reaction run with phosgene
does not occur in the absence of carbon monox-
ide. In fact, 4 can be obtained under CO pres-
found to be active Table 1, entry 41 , we have
Ž .
examined these opportunities. Under the stan-
dard conditions used above, the catalytic system
is more sensitive to the amount of substrate
sure, but in poor yield
Ž
maximum 3%
.
. More-
involved Table 3, entries 41, 66–68 , a maxi-
Ž .
mum being observed for a substrate to catalyst
over, the infrared spectrum of gases shows the
Table 3
Chlorocarbonylation of allyl chloride: use of 10% PdrC
Entry
Complex
SrC
Conversion
Ž%.
Yield Ž%.
Selectivity
1
arb
3arb
4arb
5
6
4 Ž%.
2
PdCl2
50
33
50
100
200
100
100
100
100
100
75
78
88
58
35
14
43
48
57
48
3
6
9
5
2
1
2
2
2
2
1
2
3
1
26
28
30
18
9
traces
7
13
15
8
23
27
29
17
9
1
8
9
11
16
22
15
17
17
15
12
26
24
29
22
87
78
71
75
82
–
78
81
88
80
6
4
6
6
6
7
7
7
7
6
1
7
8
9
0
1
2
3
PdrC
PdrC
PdrC
PdrC
traces
traces
traces
traces
traces
traces
a
PdrC, recycled
PdrC, recycled
b
b
wPdx, soluble
PdrC, 5 CuCl
Pd–CurC, recycled
2
b
a
With methanolysis, see text.
b
Without methanolysis, see text.
Reaction conditions: wPdxs2 mmol, wC H Clxs100 mmol, if not stated, CH Cl s10 ml, p s24 bar, pHCl s9 bar, Ts1008C,
3
5
2
2
CO
ts4 h.

M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
189
chloride and cinnamyl chloride are converted
with HCl into the addition products 1,2-dichlo-
robutane and 1,3-dichloropropylbenzene, re-
spectively.
The chlorocarbonylation of crotyl chloride 7
leads, in addition to the isomeric 3-chloro-but-
1
-ene 8, to different types of compounds
Scheme 3.
Ž
Scheme 3, Table 4 which are mostly halo-
.
genated hydrocarbons arising from addition of
HCl to the double bond, 9, reduction of the
double bond, 10, dimerisation, 11, and
oligomerisation, 12, of crotyl chloride. Simi-
ratio of 50. The solid recovered after the reac-
tion with SrCs100
Table 3, entry 67 per-
Ž .
formed with methanolysis of the reaction prod-
larly, in the case of allyl chloride, butenes
1-ene and cis- and trans-but-2-enes are present
Ž
but-
ucts, when engaged in a new reaction shows no
.
more activity for chlorocarbonylation
entry 69 . This solid contains palladium, but in
an amount smaller than that of the starting
catalyst 5.8 instead of 7.98% , and chlorine
11.86 instead of 1.06% . After the attempted
reaction, the amounts of palladium and chlorine
have decreased again 2.74 and 9.22, respec-
tively . Leaching of palladium from carbon sup-
Ž
Table 3,
in significant amounts. Mono- and dichloroacyl
chlorides, 14 and 15, are minor components
which account for 6% of the reaction mixture
.
Ž
.
Table 4 ; no unsaturated acyl chlorides are
Ž .
Ž
.
detected by GCrMS. The structure of 15 is
ascertained on the basis of the presence of the
fragment mrzs74 in the mass spectrum, re-
sulting from a McLafferty rearrangement
Ž
.
ports is known w29x and is favoured here by the
presence of carbon monoxide and hydrogen
chloride. The lack of activity of the recycled
solid may originate from poisoning during the
methanolysis process. In fact, if the methanoly-
sis of the reaction mixture is not carried out, the
solid recovered by centrifugation from a stan-
dard reaction with SrCs100 is active, but less
Ž
Scheme 4
GC corresponding to the expected diastereomers
ratio of 45:55
Chlorocarbonylation of 3-chloro-but-1-ene 8
gives the same products Table 4 . The conver-
.
and the occurrence of two peaks in
Ž
.
.
Ž
.
sion of 8 is higher, but the product distribution
is very similar, therefore suggesting that the
reactions involved during this process occur via
than the fresh catalyst
sides, the recovered red solution freed from the
reaction products is also active Table 3, entry
, so suggesting the soluble nature of the
active species. The effect of additives as re-
ported above is limited to copper II chloride
Table 3, entry 72 . This effect is lower than in
the case of PdCl and is not further observed
Table 3, entry 70 . Be-
Ž .
Ž
71
.
Table 4
Chlorocarbonylation of crotyl chloride and 3-chloro-1-butene:
comparison of the product distributions
Ž
.
Entry Compound Conversion Yield Ž%.
Ž%.
Ž
.
7
8
9
10 11 12 13 14
15
2
7
7
7
4
5
6
7
8
7q
8
78
91
–
22 9 29 9 4
24 9 25 10 2
28 11 22 7 3
1
1
2
15
19
1
1
5
6
for the recycled solid Table 3, entry 73 .
Ž .
18 traces 4
Ž1:1.
3
.3. Chlorocarbonylation of substituted allyl
a
7
7
Buta-1,
–
33 13 13 7 1
1
n.d. 1
4
chlorides
3
-diene
a
Not determined.
The reaction could be extended to other allyl
chlorides with very different results depending
on the nature of the substituents. Methallyl
Reaction conditions: wPdCl xs2 mmol, wallyl chlorideŽs.xs100
2
mmol, CH Cl s10 ml, p s24 bar, p s9 bar, T s1008C,
2
2
CO
HCl
ts4 h.

1
90
M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
tained with allyl chloride. Moreover, in the
absence of hydrogen chloride, 16 mostly leads
to oligomeric materials: only small amounts of
1
7 and 19c and the isomeric but-2-enoyl bro-
mide 22c are detected 4, 1 and 1%, respec-
tively . This could be explained by the occur-
Scheme 4.
Ž
.
the same intermediates and pathways. Further,
chlorocarbonylation of a 1:1 mixture of these
two allyl chlorides provides the same product
distribution.
rence of radical processes. In fact, chlorocar-
bonylation of 16 is not inhibited by radical
scavengers like galvinoxyl, but no more
oligomers are formed under these conditions.
Finally, it is worthy to note that buta-1,3-di-
ene is converted into the same reaction mixture
3.5. Chlorocarbonylation of ethylene
under the chlorocarbonylation used Table 4 . In
Ž .
fact, addition of HCl to buta-1,3-diene leads to a
mixture of 7 and 8 which undergoes the chloro-
carbonylation. However, selectivities are poor
and this reaction has not been optimised further.
In its basic work on nickel-catalysed reac-
tions of allyl halides, Chiusoli reported the for-
mation of hexenoic esters by co-reaction of allyl
chloride with ethylene, carbon monoxide and
methanol w31x. In the present case, the reaction
3
.4. Extension to allyl bromide
undergoes a completely different pathway, re-
sulting in the formation of 3-chloropropionic
The easier reaction of organic bromides with
derivatives 23, and 5 and 6 Eq. 3 .
Ž Ž ..
transition metal complexes than organic chlo-
rides w30x prompts us to examine the chlorocar-
bonylation of allyl bromide with the anticipation
of preparing the corresponding dichloroacyl
bromide. Under the standard conditions used,
3
Ž .
allyl bromide is converted
pattern similar to that of allyl chloride
. Structures 17 and 18 are assigned on the
Ž
71%
.
into a product
Ž
Scheme
The chlorocarbonylation of ethylene has been
reported as a stepwise process in the presence of
stoichiometric amounts of palladium dichloride
w18x. The formation of 23a is also achieved by
5
.
1
13
basis of CI
Ž
methane
.
mass spectra, H and
C
NMR spectra. The acyl bromide 19 has been
identified by comparison with an authentic sam-
ple. The MS spectra of the dihaloacyl halides 20
the reaction of phosgene with ethylene, catal-
ysed by RhrC, PdrC or Pd
Ž .
PPh3 Cl under
2
and 21
and indicate the presence of both acyl chloride
see Section 2 are compared to those of
Ž .
high pressures of carbon monoxide w32x. This
reaction corresponds to the formal addition of
the elements of phosgene on ethylene. However,
4
and bromide, since peaks corresponding to re-
spective loss of HCl and HBr are observed.
However, a comparison of the relative intensi-
q
ties of the peaks at mrzs229
Ž
C H Br CO
.
,
.
3
5
2
q
q
1
85
Ž
C H BrClCO
.
and 157
Ž
C H BrCO
3
5
3
5
suggests that the acyl chloride predominates.
Moreover, the IR spectra of these two com-
pounds show a strong band attributed to n
COCl
y
1
centred at 1800 cm
which fits well with the
one attributed to 4. Hence, the product distribu-
tion is unexpectedly different from what is ob-
Scheme 5.

M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
191
no phosgene is detected in the reaction medium
and, furthermore, phosgene does not react with
ethylene under the reaction conditions used, even
allyl intermediate followed by reductive elimi-
nation w6,7x. A control experiment with a mix-
ture of these isomers under the standard reaction
conditions used gives 3 with only traces of 4.
The yellow-green solid which is recovered at
the end of all chlorocarbonylation reactions dis-
play the typical features reported for the
in the presence of PdCl . The other products
2
involving ethylene are obtained in low amounts
and were identified by GCrMS as C H Cl
7
14
2
and C H ClCO Me. They formally originate
6
12
2
from the coupling of
Ž
i
.
allyl chloride, two
ii three
the equivalent of phosgene and
, with evolution of hydrogen chloride.
oligomeric ill-characterised wPd
Ž
CO
which is best de-
scribed as an oligomer, 24, where Pd2
CO.2
units are bridged by chloro ligands w31–37x. The
elemental analysis of this product C: 7.9, H:
0.4, Cl: 22.9, Pd: 59.9% fits well with that of
.Clxn Ž
1975,
y
1
ethylene and hydrogen chloride, and
Ž
.
616, 444, 246, 217 cm
.
ethylene
methanol
Ž
Ž
.
Allyl chloride is converted into 2-chloropropane
and propene. Small amounts of the expected
chlorocarbonylation products of allyl chloride
Ž
.
the compound obtained by reacting PdCl with
2
1
b–4b are obtained
Ž
Table 5
.
. When the reac-
allyl chloride and carbon monoxide as reported
tion is performed without allyl chloride, practi-
cally no products are formed: the expected com-
pound 23b is detected in only small amounts
by Schnabel and Kobler
19.92, Pd: 60.24%
w38,39x.
Ž
C: 7.8, H: 0.3, Cl:
.
Ž
2% yield
The formation of this compound starts even
.
.
with small amounts of allyl chloride added to
the reaction medium. An optimal quantity of
allyl chloride is required for its production
ble 5 : the ratio of 2 is presumably associated
with a specific reaction pathway vide infra
Ž
Ta-
.
Ž
.
.
The supernatant deep-red solution consists of
a mixture of the organic reaction products and
of palladium complexes which cannot be puri-
fied. Due to the similar results obtained for both
the soluble and insoluble Pd-containing frac-
tions, we assume they are both precursors of the
active species which is formed in the presence
3
.6. Reaction mechanism
The formation of 4 and its analogues does
not result from chloration of 1 or 2 which are
reported to be formed via carbonylation of an
Table 5
Chlorocarbonylation of ethylene: influence of the allyl chloride to ethylene ratio on the yield of compound 23b
Entry
Allyl chloride
Žmmol.
Ratio
Conversion Ž%.
allyl chloride
Yield Ž%.
b–4ba
23bb
5a
1
76
77
78
79
80
81
82
83
84
0
7
12
25
52
100
117
148
222
2
10
26
30
53
68
68
57
16
0,11
0,19
0,40
0,90
1,61
2,17
2,74
3,96
83
92
99
90
81
68
56
35
traces
traces
71
74
70
47
42
31
22
16
2
3
6
5
5
3
a
Based on allyl chloride.
Based on ethylene.
b

1
92
M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
used: the order of reactivity is: ethylene)allyl
chloride)3-chloro-1-butene)crotyl chloride.
Methallyl chloride and cinnamyl chloride are
not converted into the expected products: the
adducts with HCl are exclusively formed in-
stead. Hence, the reaction occurs preferentially
on terminal double bonds. In the case of crotyl
chloride, the same product distribution as for
3-chloro-1-butene is explained by an intercon-
Scheme 6.
of carbon monoxide and hydrogen chloride. This
active species may result from first the oxida-
tive addition of HCl on the Pd
Ž
I
.
–Pd
I bond of
Ž .
2
4 leading to 25 which rearranges to the
version between both substrates which operates
3
chloro-bridged dimer 26
Ž
Scheme 6
.
.
via a
Ž
h -crotyl
.
palladium intermediate. It
In the presence of a substrate, this dimer can
be split into two different monomers, 27 and 28,
which may react further through migratory in-
sertions. However, migratory insertion within
should be pointed out that the two reaction
pathways starting from 28 take place with oppo-
site regioselectivities, though leading to the same
product. To our knowledge, there is no indica-
tion in the literature about the regioselectivity of
the chlorocarbonylation reaction. In the case of
alkoxycarbonylation, regioselectivity of the cou-
pling depends strongly on the nature of the
substrate and ligands attached to palladium w40x.
In the present case, the absence of bulky lig-
ands, the neutral character of the palladium
centre will favour 31 thanks to an intramolec-
ular stabilisation of the intermediate. Moreover,
the formation of the acyl chloride in the chloro-
carbonylation of allyl bromide speaks for this
pathway, since ClrBr exchange does certainly
not take place in the steps preceding 30.
2
7 will not provide the expected chloroacyl
chloride; in fact, 27 is certainly too unstable and
will decompose into palladium and hydrogen
chloride. On the other hand, 28 offers two
possibilities which depend on the timing of
migratory processes involving the substrate and
the ClrCO ligands. If the migratory
Ä
Cl, sub-
strate process is preferred, 29 is formed and
4
will evolve via CO migratory insertion, then
reductive elimination to the expected chloroacyl
chloride. Conversely, the formation of a chloro-
carbonyl complex 30 may occur prior to the
migratory insertion involving the substrate
which leads to 31
Ž
Scheme 7
.
.
Inspection of Table 1 clearly shows that the
amount of 4 formed corresponds grossly to the
amount of propene and 2-chloropropane
The chlorocarbonylation should be therefore
sensitive to the substitution on the substrate
Scheme 7.

M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
193
evolved, therefore suggesting that part of the
allyl chloride is sacrificed during the catalytic
process. Scheme 7 proposes the formation of
zerovalent palladium species which should be
converted back to PdCl2 and then to 24 for
further catalytic cycles. Although reoxidation of
active system. This is in full coherence with the
optimal ratio of 2 observed for the chlorocar-
bonylation of ethylene. Moreover, the involve-
ment of allyl chloride in the reoxidation process
of palladium limits the yields in 4 to 36%
Ž
taking into account the initial amount of PdCl2 ,
.
Pd
Ž
0
.
to Pd
Ž
II
.
is known to be mediated by
except in the cases where CuCl is used. This
2
CuCl in aqueous HCl w41x, under the reaction
addition may induce some reoxidation due to
the presence of residual water in the autoclave;
it is however limited, since no improvement is
2
conditions used, carbon monoxide and hydrogen
chloride are inoperative for this redox process.
We suggest that allyl chloride initiate this re-
observed with an excess of this salt
entries 53 and 54
Ž
Table 2,
generation which starts with the formation of
.
.
3
the
Ž
h -allyl
.
species 32
Ž
Scheme 8
.
. We sug-
gest that the next step corresponds to the attack
3
.7. Synthesis of 2(H)-5-furanone
of HCl on this intermediate affording PdCl and
2
propene which reacts further with HCl to give
Reaction of 4a with bases would provide
-5-furanone 33 Eq.
.. which is an or-
ganic intermediate of great potentialities for bio-
5
. Such a process has been recently reported for
2
Ž
H
.
Ž
Ž
4
allylmolybdenum complexes w26x. In fact, reac-
tion of allylchloropalladium dimer with an ex-
cess of dry HCl in dichloromethane gives quan-
logically active molecules e.g., chrysanthemic
Ž
acid, cerulenine and staganincine w42x, entero-
titative yields of PdCl and 5. The amounts of
2
lactones w43x, prostaglandins w44x, soluble poly-
1
–3 built during chlorocarbonylation arise from
mers w45x
.
. Several bases have been examined
for this reaction.
the trapping of 32 with carbon monoxide. Hence,
the higher is the CO partial pressure, the higher
is the yield in 1–3, as indicated in Table 1. The
beneficial effect of the HCl partial pressure for
the generation of 1 and, to a lesser extent, of 5
could be explained by the higher availability of
zerovalent palladium species obtained according
to Scheme 6.
4
Ž .
Treatment with potassium carbonate in di-
chloromethane–water mixture gives only 4-
chlorobut-2-enoic acid 34 in 59% isolated yield
Ž .., although it has been reported that
Eq. 5
reaction of 3,4-dichlorobutanoic acid with this
In agreement with Scheme 8, two equivalents
of allyl chloride are required to reoxidise the
zerovalent palladium built during the conversion
Ž
CO
of one wPdŽ .Clx unit into the catalytically
2
Scheme 8.

1
94
M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
base in dichloromethane provides the expected
are required for the regeneration of PdCl . They
2
butenolide w46x.
are limited to allyl chloride owing to the reac-
tion conditions which forbid the use of well-
known systems for zerovalent palladium oxida-
tion; this undoubtedly represents the major
drawback of the chlorocarbonylation process
described above.
5
Ž .
Examination of different bases
triethylamine, pyridine has led to poor results.
Only the use of calcium carbonate in refluxing
water gives a practically quantitative yield GC
Ž
NaOH, KOH,
.
Acknowledgements
Ž
.
This work was supported by operating grants
of 33. The compound is recovered by extraction
with dichloromethane due to its high solubility
from CNRS fellowship to NC and Soci e´ t e´
Ž .
Nationale des Poudres et Explosifs. We wish to
thank Dr. L e´ colier for his continuous interest
in water and the formation of an azeotrope
Section 2; 70% isolated yield
Ž
see
.
.
and SNPE’s Etablissement de Toulouse
Ž
Drs.
This procedure competes very favourably
with the more conventional routes starting from
Deweerdt and Raysse, Mr. Souiry
.
for technical
help for carbonylation reactions with chlorine
and phosgene.
Ž
i
.
bromation of g-butyrolactone, followed by
dehydrobromation of a-bromo-g-butyrolactone
with triethylamine w47x, or
ii partial sodium
Ž
.
borohydride reduction of the Diels–Alder adduct
of furane and maleic anhydride, followed by a
retro-Diels–Alder reaction w48x.
References
w1x S. Patai ŽEd.., The Organic Chemistry of Functional Groups,
The Chemistry of Acyl Halides, Interscience Publishers, New
York, 1972.
w2x R. Sustmann, H.G. Korth, in: J. Falbe ŽEd.., Methoden der
Organischen Chemie, Vol. E5, Georg Thieme Verlag,
Stuttgart, 1985, p. 587.
4
. Conclusions
w3x I. Tkatchenko, in: G. Wilkinson, F.G.A. Stone, E.W. Abel
ŽEds.., Comprehensive Organometallic Chemistry, Vol. 8,
Chlorocarbonylation of olefins by a COq
Pergamon, Oxford, 1982, p. 102.
HCl mixture is observed under mild conditions
for ethylene and allylic halides. Its extension to
other olefins is limited by the competitive addi-
tion of the hydracid on the double bond.
Chlorocarbonylation of allyl chloride gives
dichloro-3,4-butanoyl chloride in good yields
and the reaction has been scaled up for its
w4x D. Forster, Adv. Organometal. Chem. 17 Ž1979. 255–267.
w5x M.C. Bonnet, J. Combes, B. Manzano, D. Neibecker, I.
Tkatchenko, J. Mol. Catal. 52 1989 263–278.
Ž
.
w6x W.T. Dent, R. Long, G.H. Whitfield, J. Chem. Soc. Ž1964.
1
588–1594.
w7x D. Medema, R. van Helden, C.F. Kohll, Inorg. Chim. Acta 3
Ž1969. 255–265.
w8x G.P. Chiusoli, L. Cassar, Angew. Chem., Int. Ed. Engl. 6
Ž1967. 124–133.
conversion into the valuable 2
All the experimental data are compatible with a
CO
process involving wPd Clx as a catalyst pre-
H -5-furanone.
Ž .
w9x T. Alderson, V.A. Engelhardt ŽDuPont., US Pat. 3 065 242
Ž1962.; Chem. Abstr. 58 Ž1964. 89112.
Ž
.
w10x N. von Kutepow, K. Bittler, D. Neubauer, H. Reis ŽBASF.,
German Pat. 1 237 116 Ž1967.; Chem. 68 Ž1968. 21550.
w11x J.F. Knifton ŽTexaco., US Pat. 3 880 898 Ž1975.; Chem.
Abstr. 83 Ž1975. 113698.
n
cursor. We suggest that the active species for
this reaction arises from oxidative addition of
HCl on the Pd–Pd bond of this oligomer. How-
ever, further work is needed for demonstrating
this process and also for explaining the discrep-
ancies reported in the literature on the prepara-
w12x J. Tsuji, M. Morikawa, J. Kiji, Tetrahedron Lett.
1963
Ž .
1061–1064.
w13x J. Tsuji, M. Morikawa, J. Kiji, J. Am. Chem. Soc. 86 Ž1964.
851–4853.
4
w14x J. Tsuji, J. Kiji, S. Imamura, M. Morikawa, J. Am. Chem.
tion of wPdŽCO.xCl x species. Specific oxidants
Soc. 86 Ž1964. 4350–4353.
y n

M.C. Bonnet et al.rJournal of Molecular Catalysis A: Chemical 143 (1999) 181–195
195
w15x T. Susuki, J. Tsuji, Tetrahedron Lett. Ž1968. 913–915.
w16x T. Susuki, J. Tsuji, J. Org. Chem. 35 Ž1970. 2982–2986.
w17x J.A. Scheben, I.L. Mador, in: P.N. Rylander, H. Greenfield
ŽEds.., Catalysis in Organic Synthesis, Academic Press, New
w31x G.P. Chiusoli, Acc. Chem. Res. 6 Ž1973. 422–427.
w32x A.U. Blackman ŽNational Distillers and Chemical., US Pat. 3
119 861 Ž1964.; Chem. Abstr. 60 Ž1964. 9155.
w33x E. Fink, Compt. Rend. 126 Ž1898. 646–648.
York, 1976, p. 183.
w34x W. Manchot, Chem. Ber. 58 Ž1925. 2518–2521.
w35x E.-O. Fischer, A. Vogler, J. Organometal. Chem. 3 Ž1965.
w18x W.W. Pritchard ŽDuPont., US Pat. 3 681 449 Ž1972.; Chem.
Abstr. 77 Ž1972. 126017.
161–162.
w19x N. B e´ nard, M.C. Bonnet, S. L e´ colier, I. Tkatchenko, J.
Chem. Soc., Chem. Commun. Ž1993. 1448–1450.
w20x R.F. Heck, Palladium Reagents in Organic Syntheses, Aca-
w36x P.L. Goggin, J. Mink, J. Chem. Soc., Dalton Trans. Ž1974.
534.
w37x D. Belli dell’Amico, F. Calderazzo, C.A. Veracini, N. Zan-
demic Press, New York, 1985, p. 17.
dona, Inorg. Chem. 23 Ž1984. 137.
w21x W.T. Dent, R. Long, A.J. Wilkinson, J. Chem. Soc. Ž1964.
w38x A. Treiber, Tetrahedron Lett. Ž1966. 2831–2832.
w39x W. Schnabel, E. Kobler, J. Organometal. Chem. 19 Ž1969.
1
585.
w22x R.F. Heck, Palladium Reagents in Organic Syntheses, Aca-
455–457.
demic Press, New York, 1985, p. 18.
w23x D. Coulson, Inorg. Synth. 13 Ž1972. 112.
w24x M.C. Bonnet, L. Carmona, I. Tkatchenko, submitted.
w25x W.A. Herrmann, in: B. Cornils, W.A. Herrmann ŽEds..,
w40x M. Sperrle, G. Consiglio, Chem. Ber.rRecueil 130 Ž1997.
1557–1565.
w41x J. Smidt, W. Hafner, R. Jira, R. Sedlmeier, A. Sabel, Angew.
Chem., Int. Ed. Engl. 1 Ž1962. 80.
Applied Homogeneous Catalysis with Organometallic Com-
pounds, A Comprehensive Handbook, Vol. 2, VCH, Wein-
heim, 1996, p. 957.
w42x T.S. Brima ŽNational Distillers and Chemical., US Pat. 4 356
310 Ž1982.; Chem. Abstr. 98 Ž1983. 16569.
w43x D.N. Kirk, L.M. McLaughlin, A.M. Lawson, K.D.R. Setchell,
S.K. Patel, J. Chem. Soc., Perkin Trans. I Ž1985. 35–37.
w44x M.R. Binns, R.K. Haynes, D.E. Lambert, P.A. Shober, Tetra-
hedron Lett. 26 Ž1985. 3385–3388.
w45x H. Cherdron, H. Ohse, R. Palm, K. Friedhelm ŽShell., Ger-
man Pat. 1 228 065 Ž1966.; Chem. Abstr. 66 Ž1967. 55994.
w46x A.A. Gevorkyan, A.T. Manukyan, M.S. Sargsyan, Arm.
Khim. Zh. 43 Ž1990. 390–393; Chem. Abstr. 114 Ž1991.
w26x K.E. Oglieve, R.A. Henderson, J. Chem. Soc., Chem. Com-
mun. Ž1993. 474–478.
w27x I. Guibert, D. Neibecker, I. Tkatchenko, J. Chem. Soc.,
Chem. Commun. Ž1989. 1850–1853.
w28x H. Urata, H. Maekawa, S. Takahashi, T. Fuchikami, J. Org.
Chem. 56 Ž1991. 4320–4322.
w29x M.M. Barreto Rosa, M.C. Bonnet, I. Tkatchenko, Heteroge-
neous catalysis and fine chemicals II, Stud. Surf. Sci. Catal.
80993.
5
9 Ž1991. 263–268.
w47x C.C. Price, J.M. Judge, Organic Syntheses Coll., Vol. V,
w30x J.K. Kochi, Organometallic Mechanisms and Catalysis, Aca-
Wiley, New York, 1973, pp. 255–258.
demic Press, New York, 1978, p. 156.
w48x S. Takano, K. Ogasawara, Synthesis Ž1974. 42–43.