Donor-stabilized phosphenium adducts as new efficient and immobilizing ligands in palladium-catalyzed alkynylation and platinum-catalyzed hydrogenation in ionic liquids

Article abstract of DOI:10.1002/adsc.200900133

The straightforward synthesis of a new donor-stabilized phosphenium ligand 3d by addition of bromodifurylphosphine to 1,3-dimethylimidazolium-2-carboxylate 1 is described. The obtained ligand exhibits a very strong π-acceptor character, comparable to that

Full text of DOI:10.1002/adsc.200900133

FULL PAPERS
DOI: 10.1002/adsc.200900133
Donor-Stabilized Phosphenium Adducts as New Efficient and
Immobilizing Ligands in Palladium-Catalyzed Alkynylation and
Platinum-Catalyzed Hydrogenation in Ionic Liquids
Samer Saleh,a Elie Fayad,a Michꢀle Azouri,a Jean-Cyrille Hierso,a,*
Jacques Andrieu,a,* and Michel Picqueta
a
Institut de Chimie Molꢁculaire de l’Universitꢁ de Bourgogne (ICMUB) – UMR 5260 CNRS, 9 avenue Alain Savary,
21078 Dijon, France
Fax : (+33)-3-8039-6098; e-mail: Jean-Cyrille.Hierso@u-bourgogne.fr or Jacques.Andrieu@u-bourgogne.fr
Received: February 26, 2009; Revised: May 7, 2009; Published online: June 30, 2009
Abstract: The straightforward synthesis of a new Conversely, a multiple recycling of the metal/ligand
donor-stabilized phosphenium ligand 3d by addition system in non-acidic media was achieved from plati-
of bromodifurylphosphine to 1,3-dimethylimidazoli- num-catalyzed hydrogenation of m-chloronitroben-
um-2-carboxylate 1 is described. The obtained ligand zene. The catalytic results obtained by employing the
exhibits a very strong p-acceptor character, compara- complex of plati
ble to that of triphenyl phosphite [P(OPh)3] or of PtCl2(3a)2] in comparison with the non-ionic related
tris-halogenophosphines, with a nCO(A1) at 2087 cmꢀ1 trans-tris(triphenylphosphine)platinum
dichloride
for its nickel tricarbonyl complex. This ligand, as [trans-PtCl2A(PPh3)2] complex clearly indicate that the
well as the related 3a which was obtained from simultaneous existence of a strong p-acceptor char-
chlorodiphenylphosphine, were tested in palladium- acter and a positive charge within the ligand 3a sig-
ACHTUNGTRNEnNUNG um(II) chloride with 3a [trans-
AHCTUNGTRENNUNG
CTHUNGTRENNUNG
ACHTUNGTRENNUNG
catalyzed aryl alkynylation and in the platinum-cata- nificantly increases the life-time of the platinum cat-
lyzed selective hydrogenation of chloronitroben- alyst. The selectivity of the reaction is also improved
ꢀ
zenes, both in an ionic liquid phase. In C C bond by decreasing the undesirable formation of dehaloge-
cross-coupling we observed that the increase of the nation products. This cationic platinum complex
p-acceptor character in ligand 3d, due to the intro- trans-PtCl2(3a)2 is the first example of a highly selec-
duction of an additional electron-withdrawing group, tive catalyst for hydrogenation of chloronitroarenes
provides a very efficient catalyst in the alkynylation immobilized in an ionic liquid phase. The system was
reaction of aryl bromides with phenylacetylene, in- recycled six times without noticeable metal leaching
cluding the deactivated 4-bromoanisole or the steri- in the organic phase, and no loss of activity.
cally hindered 2-bromonaphthalene. The catalytic ac-
tivity decreases with recycling due to the sensitive- Keywords: catalysis; catalyst recycling; ionic liquids;
ness of ligands to protonation in the ionic phase. palladium; phosphenium salts; platinum
Introduction
phines was so far limited to a few examples because
of their multistep preparation.[1,7–9] In this context, we
In contrast to neutral phosphine ligands, imidazolium- have recently described a simple and straightforward
functionalized phosphines represent an interesting op- synthesis of imidazolium-2-phosphines 2 and 3 by ad-
portunity to strongly decrease the metal leaching in dition of monochlorophosphines to the dimethylimi-
homogeneous polyphasic catalysis, especially when dazolium-2-carboxylate 1 (see Scheme 1).[10] Besides,
using an ionic liquid phase.[1–4] Among these cationic this synthesis provides access to new dicationic or
ligands, the imidazolium-2-phosphines not only lead mixed halide donor-stabilized phosphenium adducts
to very active rhodium catalysts in styrene hydrofor- by extension of the reactivity of 1 towards others
mylation[1] or to palladium complexes able to catalyze chlorophosphines.[11]
the Heck reaction[5] but also represent rare examples
of base donor-p-acceptor adducts involving an elec- stretching frequency of the related Ni(CO)
tron-rich P
(III) center as Lewis acceptor.[6] However, complexes has shown that the ligands 2 and 3 behave
the access to a large library of imidazolium-2-phos- as phosphite-like strong p-acceptors. This unexpected
Moreover, the measurement of the A1 symmetric
3A(3a–c)
CHTUNGTRENNUNG
ACHTUNGTRENNUNG
Adv. Synth. Catal. 2009, 351, 1621 – 1628
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1621

Samer Saleh et al.
FULL PAPERS
Scheme 1.
Scheme 2.
result is the consequence of the C!P bond in imida-
zolium-2-phosphines which displaces the formal posi-
tive charge from the imidazolium ring to the phospho- tion of a CO group in [Ni(CO)4]. Although less toxic
rus center. Since the combination of this electronic compounds than Ni(CO)4 (prepared in situ from
property with an ionic nature renders such ligands [NiACTHNUTRGNENUG(COD)2]) are now more commonly used (such as
very promising in the development of new continu- [Rh(CO)2Cl]2), we opted to prepare the nickel car-
ous-flow polyphasic catalytic processes, we describe in bonyl adduct in order to compare easily the parame-
this paper, firstly, the extension of our method to a ters of the new imidazolium-2-phosphane complexes
stronger p-acceptor ligand than 3a–c by addition of directly with the large database tabulated by
bis(5-methyl-2-furyl)bromophosphine to the imidazo- Tolman[12] and with our other nickel tricarbonyl phos-
lium 2-carboxylate 1 and, secondly, the catalytic prop- phenium complexes [Ni(CO)3ACTHNUTRGNEUNG
(3a–c)].[10] Thus, after
erties of some obtained complexes in palladium and addition of 3d to an equimolar amount of Ni(CO)4 in
platinum catalysis in an ionic liquid phase with a par- a toluene/CH2Cl2 solution, the related nCO(A1) stretch-
ticular attention on the recycling opportunities for the ing wave number was observed at 2087 cmꢀ1 in the in-
future development of continuous flow catalytic pro- frared spectrum. This higher value categorizes 3d as
cesses.
being much less electron-donating than the common
aryl tertiary phosphines [nCO(A1)=2069 cmꢀ1 with
PPh3] and more precisely in the range of the tris-halo-
genophosphines, or of the weaker electron-donating
Results and Discussion
phosphites
like
P
(OPh)3
(2085 cmꢀ1)
and
By the use of the procedure previously described for P(OCH2CH2CN)3 (2088 cmꢀ1).[12] This result is due to
2a–c, the bromodifurylphosphine slowly reacts with a the shift of positive charge from the imidazolium ring
slight excess of imidazolium 2-carboxylate 1 in to the phosphorus, which is consistent with the de-
CH2Cl2 at room temperature (see Scheme 1). The scription of 3a–d as having a donor-stabilized phos-
mixture was stirred two days and a light-brown solid phenium structure instead of the imidazolium-2-phos-
was obtained after removal of solvent (Scheme 2).
phine. Nevertheless, their coordination properties
The phosphorus NMR spectrum of this crude prod- remain identical to those reported with phosphine or
uct showed the presence of residual bis(5-methyl-2- phosphite ligands in Pt, Pd or Rh chemistry. This new
furyl)bromophosphine, which was easily eliminated class of ligands is therefore of fundamental and ap-
by simple washing with diethyl ether. Compound 3d plied interest in molecular catalysis with late transi-
was identified by the intense resonance detected at tion metals.[10,13,14] Herein, we have further investigat-
ꢀ78.9 ppm. The formation of the phosphorus-carbon ed their properties in two different catalytic reactions:
bond between the N-heterocycle and the phosphorus namely the palladium-catalyzed aryl alkynylation of
atom was confirmed by the presence of a doublet at aryl bromides and the platinum-catalyzed hydrogena-
141.39 ppm in the 13C NMR spectrum, with a typical tion of nitrobenzenes.
1
coupling constant JP, C of 47 Hz.[10,11] This observation,
combined with a 2:1 ratio for furyl versus imidazolium
groups in its proton NMR spectrum, unambiguously Palladium-Catalyzed Aryl Alkynylation in an Ionic
proved the formation of the expected ligand 3d.
Liquid: Immobilization and Recycling
The electron donor properties of the phosphorus
atom of a tertiary phosphine are easily evaluated by The palladium-catalyzed cross-coupling reaction be-
infrared measurement of the A1 symmetric stretching tween sp2-hybridized carbon aryl halides and sp-hy-
frequency detected in the corresponding nickel car- bridized carbon terminal acetylenes is a very impor-
bonyl complex prepared in situ by phosphine substitu- tant development in the field of alkyne chemistry
1622
asc.wiley-vch.de
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 1621 – 1628

Donor-Stabilized Phosphenium Adducts as New Efficient and Immobilizing Ligands
FULL PAPERS
since conjugated alkynes are recurring building blocks
in bioactive natural products, and in a wide range of
industrial intermediaries (pharmaceuticals, agrochem-
icals), or in molecular materials for optics and elec-
tronics.[15] This kind of cross-coupling reaction is effi-
ciently conducted in the combined presence of palla-
dium and a copper salt (mostly CuI) as co-catalyst
(Sonogashira reaction). Under certain conditions, the
reaction can be conducted with palladium in the ab-
sence of copper (Heck alkynylation); this alternative
is clearly valuable from a sustainability and economi-
cal point of view. Even more interesting is the devel-
opment of immobilized catalytic systems which opens
the way to recycling or continuous-flow chemistry. As
a consequence, the potential of the catalytic systems
combining [palladium/3a (or 3d)/base] in [BMIM]
[BF4] was explored for aryl alkynylation. It has been
previously shown that aryl bromides (substrates
Figure 1. Effect of the stronger p-acceptor ligand on palladi-
um alkynylation of different aryl bromides in [BMIM]
[BF4].
cheaper and more accessible than iodides but more
ꢀ
challenging due to the stronger C X bonding) can be
efficiently used in aryl alkynylation in an ionic
liquid.[16,17] The system [Pd/3PPh3] (1 mol%) in
[BMIM]
ACHTUNGTRENNUNG[BF4] is efficient for the coupling of a variety 2-bromonaphthalene. The results are summarized in
of activated and deactivated aryl bromides to phenyl- Figure 1.
acetylene; however the recycling was limited by detri-
Under these conditions, the palladium precursor
mental leaching of the monophosphine ligand combined with 3d was found to be a superior system,
throughout the extraction process with organic sol- in all cases, than the analogous palladium complex
vents. We thus decided to explore the potential of the containing 3a, in terms of activity and reproducibility
less lipophilic charged donor-stabilized phosphenium of results. For instance, the coupling of 4-bromoaceto-
adducts in aryl alkynylation reactions. We employed phenone and 2-bromonaphthalene to phenylacetylene
the catalytic systems combining palladium and the was problematic with 3a, while [Pd/3d] gave good to
catalysis auxiliaries 3a and 3d in the presence of pyr- high yields of coupling products (67–100%) (see
rolidine as an organic base (Scheme 3). After prelimi- Figure 1). A less clear cut result was found with the
nary investigation of general reaction conditions electronically deactivated substrates 4-bromoanisole
(temperatures varied from 80 to 1308C, and experi- and 2-bromotoluene since for both ligands only 24 to
ment time between 2 and 40 h) we determined suita- 50% conversions were obtained.
ble conditions to be around 1208C for 20 h. The inor-
ganic carbonate K2CO3 was found to be inefficient in investigations and optimization were focused on the
[BMIM][BF4]. system incorporating 3d. Table 1 summarizes the
Based on these preliminary indications, the catalytic
ACHTUNGTRENNUNG
Under these conditions, comparative studies were scope and limitation of the [Pd/3d/pyrrolidine] cata-
conducted on four aryl bromides with different steric lytic system for phenylacetylene arylation in [BMIM]
and electronic features: the activated electron-poor 4- [BF4], together with the recycling runs conducted
bromoacetophenone, the deactivated electron-rich 4- from the best conversions. Under the optimized con-
bromoanisole, the ortho-substituted electronically de- ditions, and in the noticeable absence of CuI co-cata-
activated 2-bromotoluene and the sterically hindered lyst, complete arylation of phenylacetylene were per-
formed from bromobenzene, 3-bromotoluene and 4-
bromoacetophenone (quantitative yields, entries 1–3).
Good to moderate coupling was also obtained from 3-
bromoanisole (82% yield, entry 8), 4-bromoanisole
(58% yield, entry 9) and 2-methylbromobenzene
(50% yield, entry 4), indicating that either strongly
unfavorable electronic or steric effects slightly
hamper the performances of the system. However,
the catalyst activity was clearly limited for the conver-
sion of the trisubstituted 2-bromomesitylene (both
electronically and sterically deactivated) even over a
Scheme 3.
long reaction time (entry 5, 26% yield of coupling
Adv. Synth. Catal. 2009, 351, 1621 – 1628
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1623

Samer Saleh et al.
FULL PAPERS
Table 1. Palladium-catalyzed alkynylation with various aryl bromides in [BMIM]ACHTUNRGTNEUNG[BF4].
Entry
Aryl Bromide
Conditions Tp [8C]/t [h]
Run 1
1st Recycling
Yield [%][a]
2nd Recycling
Yield [%][a]
Yield [%][a]
99 (90)
1
2
3
4
125/40
125/40
90/40
99
60
75
60
40
99 (95)
99 (90)
50
125/40
5
125/48
26
6
7
125/40
125/40
95 (85)
95 (85)
70[b]
40[b]
8
125/40
125/40
82
58
32[b]
9
[a]
Yields based on aryl bromides (average of two runs), determined by GC (external standard); isolated yields obtained
after chromatography on silica column reach 90 to 95% of these values; isolated yields are in brackets.
In these recycling experiments NaNH2 was employed to regenerate the pyrrolidine base.
[b]
after 48 h), under similar conditions [Pd/3PPh3] gave of the immobilized catalytic systems. From extraction
a complete conversion.[17] after experiments using the ligand 3d, the 31P NMR
After ethereal extraction of the coupling product signature of bis(5-methyl-2-furyl)bromophosphine
under argon, the resulting ionic liquid phase was and its corresponding phosphine oxide was unambigu-
reused without any pre-treatment and refilled with ously detected (singlets at 1.7 and 15.9 ppm in CDCl3,
the coupling reagents and the pyrrolidine base (1st re- respectively). Scheme 4 depicts a reasonable pathway
cycling, entries 1–3). While the recycling of bromo- to explain the formation of these phosphine products
benzene gave excellent results (99% yield), function- and the presence of inactive palladium(0).
alized aryl bromides such as 3-bromotoluene and 4-
Due to the specific features of the donor-stabilized
bromoacetophenone showed a drop of activity (60 phosphenium adducts[10] the accumulation of a large
and 75% yield, respectively). A second recycling con- excess of acidic pyrrolidinium bromide would be con-
firmed this tendency (see Table 1). 31P NMR experi- ducive to the C!P bond cleavage through the proto-
ments on the extracted phases did not point out any nation of the carbenic 1,3-dimethylimidazol-2-ylidene
traces of ligand leak into the organic phase, in con- of the ligand at the relatively high temperatures em-
trast to experiments where triphenylphosphine was ployed. The concomitantly formed phosphine halide
employed as ligand.[17] However, careful 31P NMR ex- being oxygen-sensitive, a mixture of the phosphine
periments conducted over long periods of time on and its phosphine oxide is detected in the NMR spec-
concentrated ethereal phases after recycling were in- tra. We thus believe that the large excess of pyrrolidi-
formative to understand the origin of the deactivation nium bromide could be at the origin of the systemꢃs
1624
asc.wiley-vch.de
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 1621 – 1628

Donor-Stabilized Phosphenium Adducts as New Efficient and Immobilizing Ligands
FULL PAPERS
tions.[20–24] For instance, with d8-transition metals,
Pd/C was found to be an efficient and selective cata-
lyst,[20] while characterization of such catalyst at the
molecular state remains troublesome. For sustainable
chemistry development, ionic liquids are attracting
much interest in catalytic transformations as they can
have a profound effect on activity and selectivity.
Indeed a carbon-supported platinum system in ionic
liquids has been shown to be more active than the
analogous palladium system or Raney nickel in the
hydrogenation of halonitrobenzenes and performs
Scheme 4. Acidic deactivation pathway proposed for [Pd/
3a,d] catalytic systems.
better in the ionic liquid [BMIM]
ACHTUNGTREN[UNNG BF4] than in metha-
nol. Besides, this platinum catalyst affords not only a
deactivation. As a consequence we looked for an ele- higher selectivity but can be then recycled more than
gant method to overcome these troubles upon recy- five times without any apparent decrease of activity
cling. A possible approach was investigated, founded and selectivity.[25] It therefore represents, to the best
on the successful coupling of bromobenzene, 3-bro- of our knowledge, one of the most relevant catalytic
motoluene and 3-bromoanisole (run 1, entries 6–8). systems for this selective hydrogenation reaction with-
The strategy aimed at the regeneration of the pyrroli- out any equivalent in homogeneous catalysis. Due to
dine base by addition of NaNH2 with simultaneous the good performance of this heterogeneous Pt/C
elimination of the acidic protons in the form of the system in ionic liquids, we have thus examined the
volatile ammonia.[29] This method eventually allowed catalytic properties of our air-stable cationic donor-
a second run as the expected coupling products were stabilized phosphenium Pt(II) complex in the homo-
obtained in 30 to 70% yield (1st recycling, entries 6– geneous catalyzed hydrogenation of chloronitroben-
8). However, the course of the reaction was seriously zenes (CNB) to chloroanilines (CA) in an ionic
hampered by the formation of residual NaBr that pro- liquid, anticipating that the presence of the cationic
duces a very viscous mixture which is certainly at the imidazolium moiety should inhibit an eventual metal
origin of mass-transfer problems.
leaching into the extracting non-polar organic phase.
A first catalytic experiment was carried out with
The present catalytic investigations have thus
proved the potential of imidazolium phosphines as li- the platinum complex 4 trans-PtCl2[3a]2 in the pres-
gands in the copper-free palladium-catalyzed alkyny- ence of 25 equivalents of m-CNB in 10 mL of
lation of aryl bromides. From a general point of view [BMIM]
ACHTUNGTERN[NUNG BF4] under 20 bars of hydrogen at 1108C
the selectivity towards the expected coupling product during 48 h (See Scheme 5).
is good to excellent; nevertheless room of progress is
After ethereal extraction of the organic products,
possible in terms of activity of the immobilized the related GC analysis exhibited a complete conver-
system and its stability towards acidic attack. In order sion of m-CNB to m-CA and only traces of aniline
to avoid this problem of incompatibility of such li-
ACHTUNGTREN(NUNG <0.5%). We then performed a kinetic monitoring of
gands in catalyses performed in acidic media, we the reaction and found that the conversion increased
transferred our interest to another important catalytic with time from 55% after 4 h to 63% and 100%, re-
reaction which does not generate acidic protons, such spectively after 18 and 48 h. This diminution of the re-
as the platinum-catalyzed hydrogenation of nitroben- action rate after 4 h could be a significant indication
zenes with ligands 3.
of a poisoning of the catalyst by the m-CA formed.
Thus, while our donor-stabilized phosphenium plati-
num complex led to a high selectivity, which main-
tains the dehalogenation reaction to a lower level, the
activity remained lower than that of the Pt/C hetero-
geneous catalyst.[25] However, care should be taken in
Platinum-Catalyzed Hydrogenation in an Ionic
Liquid: Immobilization and Recycling
The catalytic hydrogenation of halonitrobenzenes to
the corresponding haloanilines without detrimental
dehalogenation is an exciting challenge due to their
several applications in the manufacture of herbicides,
pesticides or chemical fertilizers.[18] Since the discov-
ery of a ruthenium phosphine complex as catalyst for
these hydrogenation reactions,[19] different transition
metal complexes have been examined with the view
to get haloaminobenzenes without by-products result-
ing from dehalogenation or cross-coupling reac- Scheme 5.
Adv. Synth. Catal. 2009, 351, 1621 – 1628
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1625

Samer Saleh et al.
FULL PAPERS
this comparison because of the different dilution con- dehalogenation level, the amount of aniline starting
ditions used in this study, particularly regarding the only to be detectable from the fourth recycling and
substrate which is forty times more diluted in the reaching a maximum of only 2.3% at the seventh re-
present molecular system (see Experimental Section). cycling (see Figure 2).
Additional catalytic experiments were performed
under conditions identical to the ones described been compared to those found with its neutral triphe-
above with variation of the nature of CNB and the nylphosphine platinum analogue 5 trans-PtCl2A(PPh3)2
These excellent catalytic performances have then
CHTUNGTRENNUNG
ligand. The donor-stabilized phosphenium platinum and are given in Figure 2. In contrast to complex 4,
complex remained totally inactive in hydrogenation only a partial conversion of 85% was observed with a
of o-CNB while only little difference has been point- rapid decrease down to 32% at the second recycling,
ed out with Pd/C[20] or Pt/C.[25] Conversely, a partial that is, a global TON=42 for complex 5 after three
conversion of 85% was observed with the analogous runs whereas complex 4 performed a TON=73.5 at
bis-(triphenylphosphine)dichloroplatinum complex 5. the same stage. Additionally, this neutral complex
This latter observation shows the very strong influ- showed from the first catalytic run a poor selectivity
ence of p-acceptor character in phosphorus ligands on with the formation of 5% of dehalogenated aniline
the platinum catalytic activity.
and the presence of a significant amount of insoluble
Since our donor-stabilized phosphenium platinum zero-valent metal which is the consequence of rapid
catalyst exhibited an unprecedented selectivity in con- catalyst deactivation. We believe that such differences
version of m-CNB to m-CA in homogeneous phase, in activity and selectivity between the platinum cata-
we evaluated its potential in recycling processes. lysts 4 and 5 are directly linked to the cationic nature
After 48 h, the organic compounds were extracted and structure of the coordinated phosphorus ligand
from [BMIM]ACHTUNGTRENNUNG[BF4] several times with diethyl ether 3a. Indeed, in contrast to the triphenylphosphine
under an inert atmosphere. The ionic liquid phase ligand, the positive charge present in ligand 3a makes
was then reloaded with m-CNB (substrate to catalyst it insoluble in diethyl ether and no ligand leaching
ratio=25) before being reintroduced into the auto- from the ionic liquid phase occurs during the extrac-
clave. The catalytic conditions described for the first tion process. The catalyst that remains in the ionic
catalytic run were then applied to the new catalytic liquid solution is therefore stabilized while it rapidly
mixture. After the recycling, the ionic liquid was re- decomposes if the ligand is extracted such as in the
covered and reused for a further catalytic experiment case of PPh3. This is consistent with the absence of in-
using the above procedure until the sixth recycling. soluble zero-valent metal in the case of complex 4.
The conversion of m-CNB to m-CA and the amount The outstanding life-time of this platinum complex,
of undesirable dehalogenation product for each cata- without significant loss of activity after a total of 14
lytic run are summarized in Figure 2.
days at 1108C, also shows unambiguously the robust-
It is noteworthy that the activity remained con- ness of the donor-stabilized phosphenium ligands,
stantly high (between 94 and 100%) even after seven provided no acidic protons are used or formed during
catalytic runs, corresponding to a cumulative TON of the reaction as previously shown in the palladium al-
167. The selectivity was also remarkable with a weak kynylation reaction. Moreover, it should be pointed
out that the recycling experiments performed here
prove also the robustness of ligand 3a towards water
which is co-produced during the hydrogenation cata-
lytic cycle.
Conclusions
The results reported herein clearly demonstrate the
utility of donor-stabilized phosphenium adducts in
late transition metal catalysis. The introduction of an
additional electron-withdrawing group like furyl in
the cationic ligand allowed us to decrease even more
its already weak basicity and led to a very efficient
palladium catalyst in the alkynylation reaction of vari-
ous aryl bromides, including weakly reactive sub-
strates. Although the catalyst remained insoluble in
non-polar organic solvents and was thus successfully
immobilized in the IL phase during recycling at-
tempts, its activity in [BMIM]ACTHNUTRGNEUNG[BF4] decreased from
Figure 2. Pt-catalyzed hydrogenation results of m-CNB with
the recovered ionic liquid phase. % conversion of m-CNB
to m-CA with platinum complexes 4 (*in the background)
and 5 (* in the foreground) and % aniline formed with
platinum complex 4 (&). Yields were determined by GC.
1626
asc.wiley-vch.de
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 1621 – 1628

Donor-Stabilized Phosphenium Adducts as New Efficient and Immobilizing Ligands
FULL PAPERS
Avance 300 spectrometer. All chemical shifts are relative to
SiMe4 (1H and 13C NMR) and 85% H3PO4 (31P NMR) and
are given in ppm. The elemental analyses were performed in
the first recycling. The deactivation was found to be
due to a protonation reaction of the C!P bond of
the ligand in the presence of the large quantity of pyr-
rolidium salt formed upon the successive recycling.
This observation suggests a certain incompatibility of
such ligands with catalytic reactions performed in
acidic media, or which generate easily transferred
protons, especially when no treatments are done on
the reused phase (water washing, etc.). Conversely,
the catalytic results obtained with the trans-PtCl2(3a)2
complex compared to its related neutral complex
our Institute on Fisons EA 1108 apparatus. The
a
[Pd(allyl)Cl]2, PtCl2ACHTGUNTERN(NUG PPh3)2 5, bromoaryles, phenylacetylene,
R
m-CNB and o-CNB were commercial products and were
used as received. The bis(5-methyl-2-furyl)bromophosphine,
1,3-dimethylimidazolium 2-carboxylate 1, donor-stabilized
phosphenium 3a, PtCl2(3a)2 4, and ionic liquid [BMIM]ACHTUNGTRENNUNG[BF4]
were prepared according to the literature.[10,26–28] The solids
(catalyst and reagents) and the ionic liquid were degassed
under vacuum before use.
trans-PtCl
(PPh3)2 in the hydrogenation of chloro-
[BF4] confirmed the recy-
A
Synthesis of Donor-Stabilized Phosphenium Adduct
3d
cling potential of donor-stabilized phosphenium li-
gands. The simultaneous existence of a strong p-ac-
ceptor character and a positive charge in the phos-
phorus ligand substantially improved the life-time of
the platinum catalyst. The chemical selectivity of the
reaction was also eventually increased by suppressing
the detrimental formation of the dehalogenation
product. Therefore, this cationic platinum complex
represents, to the best of our knowledge, the first ex-
ample of highly selective homogeneous catalyst for
the reduction of nitroarenes in an ionic liquid phase.
The platinum catalytic system was successfully recy-
cled seven times without noticeable metal leaching in
the organic phase and any significant loss of activity.
To a suspension of 1,3-dimethylimidazolium 2-carboxylate 1
(0.216 g, 1.54 mmol) in CH2Cl2 (5 mL) was very slowly
added the bis(5-methyl-2-furyl)bromophosphine (0.488 g,
1.79 mmol). A purple colour appeared immediately which
slowly changed to light brown after 18 h of stirring. The mix-
ture was stirred for two days at room temperature. The sol-
vent was removed under vacuum to afford a light brown
solid which was washed twice with 5 mL of degassed pen-
tane and ether, and dried under vacuum; yield: 0.539 g
1
(95%); mp 1318C. H NMR (300 MHz, CDCl3): d=8.28 (s,
2H, N-CH=CH-N), 6.88 (s, 2H, H3 furyl), 6.09 (m, 2H,
H4 furyl), 3.94 (s, 6H, NCH3), 2.32 (s, 6H, CH3 furyl); 13C{1H}
NMR (75 MHz, CDCl3): d=160.81 (d, JP, C =2.3 Hz, 2C,
C2 furyl), 141.39 (d, JP,C =47.0 Hz, 1C, NCN), 138.24 (s, 2C,
C5 furyl), 127.60 (s, 1C, N-CH=), 127.13 (s, 1C, N-CH=),
3
123.39 (d, 2C, C4 furyl), 108.57 (d, JP,C =8.3 Hz, 2C, C3 furyl),
3
4
37.46 (d, JP, C =8.3 Hz, 2C, NCH3), 13.98 (d, JP, C =19.6 Hz,
2C, CH3 furyl); 31P{1H} NMR (121 MHz, CDCl3): d=ꢀ78.9
(s); anal. calcd. for C15H18N2O2PBr (368.02): C 48.90,
H 4.93, N 7.61; found: C 48.37, H 4.91, N 7.91.
Experimental Section
General Procedures
All reactions were performed in Schlenk-type flasks under
an argon atmosphere except hydrogenation experiments
which were performed in a Parr 300-mL stainless steel auto-
clave equipped with a magnetic drive and an internal glass
vessel. The temperature was controlled by a rigid heating
mantle and by a single loop cooling coil. Solvents were puri-
fied and dried by conventional methods and distilled under
argon. Chromatography was performed on silica gel (230–
400 mesh, Merck Kieselgel 60). GC and GC-MS analyses
were performed in our Laboratories on Shimadzu GC-2014
with a Supelco Equity-5 capillary column (30 mꢄ0.25 mmꢄ
0.25 m), Hewlett–Packard HP-6890 series with a HP-5 capil-
lary column (30 mꢄ0.25 mmꢄ0.25 m) and ThermoFinnigan
Representative Catalytic Experiment for the Heck
Alkynylation Reaction in [BMIM]ACHTUNRGTNEUNG[BF4]
The solid mixture of [PdACTHNUTRGNE(UGN allyl)Cl]2 (6.3 mg, 0.0348 mmol of
Pd), donor-stabilized phosphenium (3a: 32 mg, 0.102 mmol
or 3d: 38 mg, 0.102 mmol) and bromobenzene (0.358 mL,
536 mg, 3.416 mmol) was degassed for 15 min in a 20-mL
Schlenk tube equipped with a magnetic stirrer bar. Under
argon were added 3 mL of [BMIM]ACHTNUTRGNEUNG[BF4] previously de-
gassed under reduced pressure for 10 min. The Schlenk tube
was heated in an oil bath at 908C to give a red-orange solu-
tion. To the ionic liquid solution was added, out of the oil
bath, 0.35 mL pyrrolidine (1.6 equiv., 390 mg, 5.46 mmol,
d=0.87) then 0.60 mL phenylacetylene (1.6 equiv., 558 mg,
5.46 mmol, d=0.93). The resulting mixture was heated at
1258C for 40 h under argon. The product was extracted
from the ionic liquid phase by the addition of diethyl ether
(two times 5 or 10 mL) and decanting off the ether from the
ionic liquid phase (GC yield: 99%). After evaporation the
residue was purified by silica gel chromatography (diethyl
ether/hexane: 1/9) to give (2-phenylethynyl)benzene; isolat-
ed yield: 540 mg (90%).
TRACE GC with
a RTX-5 capillary column (5 mꢄ
0.25 mmꢄ0.25 m) and an UltraFast module. GC yields are
based upon external standard calibration from pure starting
substrates and reaction products. The program used in GC-
MS and GC for experiments was the following: carrier gas
He, gas flow 1.3 mLminꢀ1; injector 3008C, oven temperature
program: 1 min at 608C, 208Cminꢀ1 till 2808C, 5, 10 or
15 min at 2808C, detector FID 3008C or MSD; the program
used in UltraFast GC for experiments was the following:
carrier gas H2, gas flow 0.3 mLminꢀ1; injector 2508C,
column program: 6 sec at 508C, 2008Cminꢀ1 till 2008C, iso-
therm at 2008C, detector FID 2508C. 1H, 31P{1H} and
13C{1H} NMR spectra were recorded at 298 K on a Bruker
The products obtained were analysed by proton NMR,
mass spectrum analysis and elemental analysis, and were
found identical to commercial products.
Adv. Synth. Catal. 2009, 351, 1621 – 1628
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1627

Samer Saleh et al.
FULL PAPERS
[6] B. D. Ellis, P. J. Ragogna, C. L. B. Macdonald, Inorg.
Chem. 2004, 43, 7857–7867.
Recycling Experiment for the Heck Alkynylation
Reaction in [BMIM]ACTHNUTRGNEUNG[BF4]
[7] A. A. Tolmachev, A. A. Yurchenko, A. S. Merculov,
M. G. Semenova, E. V. Zarudnitskii, V. V. Ivanov,
A. M. Pinchuk, Heteroat. Chem. 1999, 10, 585–597.
[8] I. V. Komarov, M. Y. Kornilov, Tetrahedron 1995, 51,
11271–11280.
[9] N. Kuhn, J. Fahl, D. Blꢅser, R. Boese, Z. Anorg. Allg.
Chem. 1999, 625, 729–734.
[10] M. Azouri, J. Andrieu, M. Picquet, P. Richard, B. Han-
quet, I. Tkatchenko, Eur. J. Inorg. Chem. 2007, 31,
4877–4883.
[11] M. Azouri, J. Andrieu, M. Picquet, H. Cattey, Inorg.
Chem. 2009, 48, 1236–1242.
[12] C. A. Tolman, Chem. Rev. 1977, 77, 313–348.
[13] N. Kuhn, M. Goꢆhner, G. Henkel, Z. Anorg. Allg.
Chem. 1999, 625, 1415–1416.
After extraction with diethyl ether under argon the resulting
dark coloured ionic liquid was reused without any specific
treatment, but was degassed under reduced pressure for
15 min. The aryl halide was added and the mixture degassed
for 15 min. The reaction was carried out after addition of
pyrrolidine and phenylacetylene and work-up was done
under the same conditions as employed for the first run.
Catalytic Hydrogenation Experiments with Platinum
Complexes
The autoclave was purged three times under vacuum/argon
before introduction of the catalytic solution. The reactor
was charged with
10ꢀ2 mmol] or
10ꢀ2 mmol], m- or o-chloronitrobenzene (71 mg, 0.45 mmol)
and 10 mL of [BMIM][BF4] and then heated at 1108C.
4
[trans-PtCl2(3a)2: 21 mg, 1.857ꢄ
5
[trans-PtCl2A(PPh3)2: 15 mg, 1.857ꢄ
CHTUNGTRENNUNG
[14] J. Andrieu, M. Azouri, P. Richard, Inorg. Chem.
Commun. 2008, 11, 1401–1404.
AHCTUNGTRENNUNG
[15] H. Doucet, J.-C. Hierso, Angew. Chem. 2007, 119, 850–
888; Angew. Chem. Int. Ed. 2007, 46, 834–871.
[16] J.-C. Hierso, M. Picquet, H. Cattey, P. Meunier, Synlett
2006, 3005–3008.
When the desired temperature was reached, the reactor was
filled with 20 bars of hydrogen and the mixture was stirred
for 48 h. The reactor was cooled until room temperature
and the organic products were extracted with diethyl ether
and then analyzed by Ultrafast GC.
[17] J.-C. Hierso, J. Boudon, M. Picquet, P. Meunier, Eur. J.
The identification of the organic compounds was per-
formed by comparison with pure commercial compounds.
Org. Chem. 2007, 4, 583–587.
[18] P. J. Chenier, Survey of Industrial Chemistry, 2nd re-
vised edn., VCH, Weinheim, 1992.
[19] J. F. Knifton, J. Org. Chem. 1976, 41, 1200–1206.
[20] V. Kratky, M. Kralik, M. Mecarova, M. Stolcova, L. Za-
libera, M. Hronec, Appl. Catal. A 2002, 235, 225–231.
[21] A. Corma, P. Sema, P. Concepciꢇn, J. J. Calvino, J. Am.
Chem. Soc. 2008, 130, 8748–8753.
Acknowledgements
We wish to acknowledge the Ministꢀre de l’Enseignement Su-
pꢁrieur et de la Recherche for support and for PhD fellow-
ship to S. S. and M. A. We also thank the Universitꢁ de Bour-
gogne, the Centre National de la Recherche Scientifique and
the “Conseil Rꢁgional de Bourgogne” for financial support
of this work.
[22] P. Maity, S. Basu, S. Bhaduri, G. K. Lahiri, Adv. Synth.
Catal. 2007, 349, 1955–1962.
[23] N. Maity, P. R. Rajamohanan, S. Ganapathy, C. S. Gopi-
nath, S. Bhaduri, G. K. Lahiri, J. Phys. Chem. C 2008,
112, 9428–9433.
[24] F. Cꢈrdenas-Lizana, S. Gꢇmez-Quero, M. A. Keane,
Appl. Catal. A 2008, 334, 199–206.
[25] D.-Q. Xu, Z.-Y. Hu, W.-W. Li, S.-P. Luo, Z.-Y. Xu, J.
References
Mol. Catal. A 2005, 235, 137–142.
[1] D. J. Brauer, K. W. Kottsieper, C. Liek, O. Stelzer, H.
Waffenschmidt, P. Wasserscheid, J. Organomet. Chem.
2001, 630, 177–184.
[2] S.-G. Lee, Y. J. Zhang, J. Y. Piao, C. E. Song, J. H. Choi,
J. Hong, Chem. Commun. 2003, 2624–2625.
[3] R. P. J. Bronger, S. M. Silva, P. C. J. Kamer, P. W. N. M.
van Leeuwen, Dalton Trans. 2004, 1590–1596.
[4] J. Andrieu, M. Azouri, Inorg. Chim. Acta 2007, 360,
131–135.
[26] J.-C. Hierso, A. Fihri, R. Amardeil, P. Meunier, H.
Doucet, M. Santelli, Tetrahedron 2005, 61, 9759–9766.
[27] M. Picquet, D. Poinsot, S. Stutzmann, I. Tkatchenko, I.
Tommasi, P. Wasserscheid, J. Zimmermann, Top. Catal.
2004, 29, 139–143.
[28] P. A. Z. Suarez, J. E. L. Dullius, S. Einloft, R. De Souza,
J. Dupont, Polyhedron 1996, 15, 1217–1219.
[29] While this work was in progress, Wan et al. reported a
similar strategy using aqueous NaOH treatement to re-
cover the free amine in a Heck reaction, see ref.[5]
[5] Q.-X. Wan, Y. Liu, Y. Lu, M. Li, H.-H. Wu, Catal. Lett.
2008, 121, 331–336.
1628
asc.wiley-vch.de
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 1621 – 1628