Ionophilic phosphines: Versatile ligands for ionic liquid biphasic catalysis

Article abstract of DOI:10.1021/ol702664a

(Chemical Equation Presented) Phosphine ligands bearing an imidazolium fragment were easily prepared by one-step radical chain addition of secondary phosphines to allyl or vinyl imidazolium salts. These ligands were used to prepare new ionophilic second g

Full text of DOI:10.1021/ol702664a

ORGANIC
LETTERS
2008
Vol. 10, No. 2
237-240
Ionophilic Phosphines: Versatile
Ligands for Ionic Liquid Biphasic
Catalysis
Crestina S. Consorti,* Guilherme L. P. Aydos, Gu
1nter Ebeling, and
Ja1ırton Dupont*
Laboratory of Molecular Catalysis, Institute of Chemistry - UFRGS, AVenida Bento
Gonc¸alVes, 9500 P.O. Box 15003, 91501-970, Porto Alegre, RS Brazil
consorti@iq.ufrgs.br; dupont@iq.ufrgs.br
Received November 2, 2007
ABSTRACT
Phosphine ligands bearing an imidazolium fragment were easily prepared by one-step radical chain addition of secondary phosphines to allyl
or vinyl imidazolium salts. These ligands were used to prepare new ionophilic second generation Grubbs-type catalysts. The catalyst immobilized
in 1-butyl-3-methyl imidazolium ILs shows good catalytic activity in RCM reactions of several substrates and, depending on the media employed,
is stable up to eight cycles.
Multiphase organometallic catalysis, in particular, liquid-
liquid biphasic catalysis involving two immiscible phases,
may offer the possibility of circumventing the problems
associated with the homogeneous process such as product
separation, catalyst recycling, and the use of organic solvents.
Nowadays, there are several alternatives under investigation
as fluids for multiphase catalysis for ring-closing metathesis
reactions including the resurgence of water,1-3 perfluorinated
hydrocarbons,4-6 and supercritical fluids, in particular CO2.7,8
Indeed, the advent of water-soluble organometallic com-
plexes, especially those based on sulfonated phosphorus-
containing ligands, has enabled various biphasic catalytic
reactions to be conducted on an industrial scale. However,
the use of water as a catalyst immobilizing phase has its
limitations: (i) it is a highly polar and coordinating protic
solvent, (ii) it is reactive toward many organometallic
complexes and substrates, (iii) from an environmental
perspective, trace amounts of organic compounds in water
are notoriously difficult to remove, and (iv) the synthesis of
(4) (a) da Costa, R. C.; Gladysz, J. A. Chem. Commun. 2006, 2619-
2621. (b) da Costa, R. C.; Gladysz, J. A. AdV. Synth. Catal. 2007, 349,
243-254.
(5) Matsugi, M.; Curran, D. P. J. Org. Chem. 2005, 70, 1636-1642.
(6) Yao, Q. W.; Zhang, Y. L. J. Am. Chem. Soc. 2004, 126, 74-75.
(7) Michalek, F.; Madge, D.; Ruhe, J.; Bannwarth, W. Eur. J. Org. Chem.
2006, 577-581.
(8) Furstner, A.; Ackermann, L.; Beck, K.; Hori, H.; Koch, D.;
Langemann, K.; Liebl, M.; Six, C.; Leitner, W. J. Am. Chem. Soc. 2001,
123, 9000-9006.
(1) Mohr, B.; Lynn, D. M.; Grubbs, R. H. Organometallics 1996, 15,
4317-4325.
(2) Jordan, J. P.; Grubbs, R. H. Angew. Chem., Int. Ed. 2007, 46, 5152-
5155.
(3) Zarka, M. T.; Nuyken, O.; Weberskirch, R. F. Macromol. Rapid
Commun. 2004, 25, 858-862.
10.1021/ol702664a CCC: $40.75
© 2008 American Chemical Society
Published on Web 12/22/2007

specially designed water-soluble ligands/organometallic com-
plexes is essential for its use.
The synthesis of the imidazolium phosphine ligands is
shown in Scheme 1. In the presence of catalytic amounts of
Classical transition-metal catalyst precursors are, in many
cases, “soluble” in imidazolium ILs and are not removed
from the ionic solution by the majority of organic com-
pounds. Indeed various catalytic processes can be “directly”
transposed in ionic liquids, such as those based on homo-
geneous transition-metal catalyst precursors and colloids,
with great advantages compared with those performed in
organic solvents or in water.9,10 This is one of the great
advantages of ILs in organometallic catalysis, that is, they
allow the direct transposition of well-known homogeneous
processes for liquid-liquid biphasic conditions without the
use of specially designed ligands/complexes that are neces-
sary for aqueous, perfluorinated, or supercritical fluid-based
catalytic processes. However, in some cases classical metal
catalysts are removed from the ionic liquid by the products
formed, and ionic modified ligands such as phosphines
should be used to minimize catalyst leaching.11 This is the
case in the hydroformylation reactions for which mono- and
diphosphine ligands containing the imidazolium moiety have
been developed12 or for the hydrogen transfer reactions for
which imidazolium containing η6-arene ligands have been
used.13 Ring-closing metathesis reactions14-16 can be per-
formed by simple dissolution of Grubbs first or second
generation catalysts,17,18 allenylidene ruthenium precursors19
or Hoveyda type catalysts bearing imidazolium fragments20
dissolved in ionic liquids. However, no simple ionophilic
ruthenium-phosphine catalysts precursors for ring-closing
metathesis in ILs have been reported so far.
Scheme 1. Synthesis of Phosphine Ligands
the appropriate radical initiator, the radical chain addition
of secondary phosphines to allyl or vinyl imidazolium salts
was easily performed.21 The addition of HPPh2 to the allyl
imidazolium salt 122 and vinyl imidazolium salt 222 is
straightforward, and almost quantitative conversions are
attained with the mild radical initiator azobis(isobutyronitrile)
(AIBN). Ligands 4 and 6 were isolated as colorless oils in
94% and 88% yields. Radical addition of the bulkier HPCy2
fails with AIBN, and with the high-temperature radical
initiator dicumyl peroxide, a complex mixture of products
was obtained. The target ligand 3 was successfully synthe-
sized in 87% yield as a white solid by using the radical
initiator azobis(cyclohexanecarbonitrile) (ABCN). The reac-
tion of the vinylimidazolium salt 2 with HPCy2 promoted
by ABCN gives the target ligand 5 together with byproducts
which could not be separated. Despite the low reactivity of
the allyl group toward secondary aliphatic phosphine addi-
tion, the method described offers a simple, single-step
strategy for the synthesis of phosphine ligands bearing an
ionic fragment. This methodology was succinctly mentioned
but not exploited in earlier literature.23
We report herein the synthesis of a new family of
phosphine ligands that bear an imidazolium fragment and
consist of a versatile ligand class for IL biphasic catalysis.
We also report the syntheses of a second generation Grubbs
type compound with high IL affinity and with potential for
RCM reaction in ionic media.
(9) Dupont, J.; de Souza, R. F.; Suarez, P. A. Z. Chem. ReV. 2002, 102,
3667-3691.
(10) Parvulescu, V. I.; Hardacre, C. Chem. ReV. 2007, 107, 2615-2665.
(11) Wasserscheid, P.; Waffenschmidt, H. ACS Symp. Ser. 2002, 818,
373-386.
(12) (a) Bronger, R. P. J.; Silva, S. M.; Kamer, P. C. J.; van Leeuwen,
P. Chem. Commun. 2002, 3044-3045. (b) Favre, F.; Olivier-Bourbigou,
H.; Commereuc, D.; Saussine, L. Chem. Commun. 2001, 1360-1361. (c)
Dupont, J.; Silva, S. M.; de Souza, R. F. Catal. Lett. 2001, 77, 131-133.
(13) Geldbach, T. J.; Dyson, P. J. J. Am. Chem. Soc. 2004, 126, 8114-
8115.
In particular this method has several advantages over
others reported so far such as those based on base-catalyzed
addition of primary or secondary phosphines to 1-vinylimi-
dazole,24 alkylation of imidazoles with bromoalkyl(diphe-
nylphosphine) oxides folowed by phosphine reduction,25,26
or nucleophilic substitution using bromo-alkyl-1-methylimi-
dazolium salts.27 However, the phosphines are prepared in
four steps in the case of base-catalyzed method and the
(14) (a) Astruc, D. Actual. Chim. 2004, 3-11. (b) Astruc, D. New J.
Chem. 2005, 29, 42-56.
(15) Deshmukh, P. H.; Blechert, S. Dalton Trans. 2007, 2479-2491.
(16) Handbook of Metathesis; Grubbs, R. H., Ed.; Wiley-VCH: Wein-
heim, Germany, 2003.
(20) Semeril, D.; Olivier-Bourbigou, H.; Bruneau, C.; Dixneuf, P. H.
Chem. Commun. 2002, 146-147.
(21) Rauhut, M. M.; Semsel, A. M.; Wystrach, V. P.; Currier, H. A. J.
Org. Chem. 1961, 26, 5138-5139.
(17) Buijsman, R. C.; van Vuuren, E.; Sterrenburg, J. G. Org. Lett. 2001,
3, 3785-3787.
(22) See Supporting Information.
(18) Stark, A.; Ajam, M.; Green, M.; Raubenheimer, H. G.; Ranwell,
A.; Ondruschka, B. AdV. Synth. Catal. 2006, 348, 1934-1941.
(19) (a) Audic, N.; Clavier, H.; Mauduit, M.; Guillemin, J. C. J. Am.
Chem. Soc. 2003, 125, 9248-9249. (b) Clavier, H.; Audic, N.; Mauduit,
M.; Guillemin, J. C. Chem. Commun. 2004, 2282-2283. (c) Clavier, H.;
Audic, N.; Guillemin, J. C.; Mauduit, M. J. Organomet. Chem. 2005, 690,
3585-3599. (d) Rix, D.; Clavier, H.; Coutard, Y.; Gulajski, L.; Grela, K.;
Maudint, M. J. Organomet. Chem. 2006, 691, 5397-5405. (e) Rix, D.;
Caijo, F.; Laurent, I.; Gulajski, L.; Grela, K.; Mauduit, M. Chem. Commun.
2007, 3771-3773. (f) Thurier, C.; Fischmeister, C.; Bruneau, C.; Olivier-
Bourbigou, H.; Dixneuf, P. H. J. Mol. Catal. A-Chem. 2007, 268, 127-
133. (g) Yao, Q. W.; Zhang, Y. L. Angew. Chem., Int. Ed. 2003, 42, 3395-
3398. (h) Yao, Q. W.; Sheets, M. J. Organomet. Chem. 2005, 690, 3577-
3584.
(23) Danopoulos, A. A.; Winston, S.; Gelbrich, T.; Hursthouse, M. B.;
Tooze, R. P. Chem. Commun. 2002, 482-483.
(24) Kottsieper, K. W.; Stelzer, O.; Wasserscheid, P. J. Mol. Catal.
A-Chem. 2001, 175, 285-288.
(25) Tsoureas, N.; Danopoulos, A. A.; Tulloch, A. A. D.; Light, M. E.
Organometallics 2003, 22, 4750-4758.
(26) Herrmann, W. A.; Kocher, C.; Goossen, L. J.; Artus, G. R. J. Chem.-
Eur. J. 1996, 2, 1627-1636.
(27) (a) Yang, C. L.; Lee, H. M.; Nolan, S. P. Org. Lett. 2001, 3, 1511-
1514. (b) Lee, S. G.; Zhang, Y. J.; Piao, J. Y.; Yoon, H.; Song, C. E.;
Choi, J. H.; Hong, J. Chem. Commun. 2003, 2624-2625. (c) Lee, H. M.;
Zeng, J. Y.; Hu, C. H.; Lee, M. T. Inorg. Chem. 2004, 43, 6822-6829. (d)
Field, L. D.; Messerle, B. A.; Vuong, K. Q.; Turner, P. Organometallics
2005, 24, 4241-4250.
238
Org. Lett., Vol. 10, No. 2, 2008

Table 1. Optimization in the 1,7-Octadiene RCM with Catalyst Precursor 8a
cycle/yield (%)b (time (min))
entry
solvent
1
2
3
4
5
6
7
8
1
2
3
4
5
BMI.NTf2/toluene
BMI.PF6/toluene
BMI.FAPc/toluene
BMI.PF6/toluened
BMI.NTf2/toluened
98(80)
98(70)
-
99(10)
99(10)
95(90)
98(50)
96(90)
99(40)
81(120)
99(40)
-
92(60)
98(80)
94(90)
88(90)
99(25)
99(30)
86(90)
90(60)
95(180)
89(180)
-
-
a Reaction conditions: 1,7-octadiene (2.0 mmol), toluene (4 mL), IL (1.5 g), catalyst (0.0050 mmol), 45 °C. b GC yield. c FAP ) tris(perfluoroethyl)-
trifluorophosphate. d Reaction performed with the second-generation Grubbs catalyst.
phosphorus center must be protected and deprotected. In the
case of the nucleophilic substitution method only ionophilic
Table 2. RCM of Various Dienes Promoted by 8a
ligands containing the diphenyl phosphine moiety have been
reported so far. The radical chain addition method reported
herein allows the one-step generation of diphenyl and
dicyclohexyl ionophilic phosphine ligands.
Next, 728 and the phosphine ligand 3 were reacted in
toluene, according to known methodology, affording the
Grubbs catalyst analogue 8 as a pink powder in 75% yield
(Scheme 2). For toluene-IL biphasic systems, compound 8
Scheme 2. Synthesis of Grubbs Catalyst Analogue 8
a Reaction conditions: substrate (2.0 mmol), toluene (4 mL), BMI.PF6
(1.5 g), 45 °C. b R ) CO2Et. c GC yields. d No IL, homogeneous reaction
presents a high IL phase affinity with undetectable Ru
amounts (measured by atomic absorption for BMI.NTf2/tol-
uene biphasic system at room temperature) in toluene. It is
worth noting that for the same biphasic system, the conven-
tional second-generation Grubbs catalyst29 has a very low
IL phase affinity and is mainly present in the toluene phase.
The catalytic activity of the catalyst precursor 8 was first
evaluated in the ring closing metathesis reaction of 1,7-
octadiene yielding cyclohexene in IL/toluene biphasic media
(Table 1). Three different hydrophobic ILs (BMI.NTf2,
BMI.PF6, and BMI.FAP (FAP ) (tris(perfluoroethyl)tri-
fluorophosphate))) in the presence of toluene as cosolvent
were employed for this study. We found that the concentra-
tion of compound 8 could be lowered from 5 to 0.25 mol %
without a significant reduction in catalytic activity. This may
be ascribed to the catalytic activity being mainly controlled
in toluene. e Homogeneous reaction in toluene performed using the second
generation Grubbs catalyst. f BMI.PF6/H2O biphasic system.
by diffusion of the substrate into the ionic phase on this scale.
Although excellent RCM yields are obtained for the first
three cycles in BMI.NTf2 (entry 1, Table 1), a noticeable IL
phase color change from pinkish to brown is observed after
the first cycle. Solutions of compound 8 in BMI.FAP (entry
3, Table 1) turn immediately colorless with a complete lack
of catalytic activity. The use of BMI.PF6 (entry 2, Table 1)
improves both catalytic activity and recyclability of com-
pound 8. In BMI.PF6/toluene media, compound 8 can be
recycled up to eight times with virtually no loss of catalytic
activity. Atomic absorption analysis of the toluene phase after
each recycling revealed a ruthenium content below the
detection limits of the technique (<2 ppm, i.e., less than 1.5%
of the initial Ru content).
(28) Sanford, M. S.; Love, J. A.; Grubbs, R. H. Organometallics 2001,
20, 5314-5318.
(29) (a) Huang, J.; Stevens, E. D.; Nolan, S. P.; Petersen, J. L. J. Am.
Chem. Soc. 1999, 121, 2647. (b) Scholl, M.; Trnka, T. M.; Morgan, J. P.;
Grubbs R. H. Tetrahedron Lett. 1999, 40, 2247. (c) Ackermann, L.; Fu¨rstner,
A.; Weskamp, T.; Kohl, F. J.; Herrmann W. A. Tetrahedron Lett. 1999,
40, 4787.
Comparative examples with the second-generation Grubbs
catalyst were performed in both BMI.PF6/toluene and
BMI.NTf2/toluene catalytic systems (entries 4 and 5, Table
1). Although high 1,7-octadiene conversion over a short
Org. Lett., Vol. 10, No. 2, 2008
239

reaction time was obtained in the first cycle, it decreases
abruptly as can be seen by the longer reaction time required
to achieve 90% conversions in subsequent cycles. This fact
is not unexpected since most of the Grubbs catalyst is
removed dissolved in the toluene phase, and by the fourth
cycle virtually no Ru is present in the ionic phase.
Next, we decided to compare the catalytic activity of
compound 8 for the RCM of several diene substrates to better
assess the scope of the reaction. As summarized in Table 2,
8 proved to be effective for obtaining di- and trisubstituted
olefins (entries 1 and 2, Table 2). Additionally, recycling of
the ionic phase was attempted for the more reactive substrate
diallyl diethylmalonate, showing a small drop in the catalytic
activity after the second cycle (97% and 81% yield (1.5 h)
for second and third cycles, respectively). The reaction profile
for this substrate is presented in Figure 1, showing a sigmoid
change to orange is accompanied by copious ethene evolu-
tion. The performance of compound 8 for the RCM of the
poorly reactive di(methallyl diethylmalonate) substrate gives
only 10% yield of the tetra-substituted olefin product (entry
3, Table 2). Homogeneous experiments performed in toluene
comparing compound 8 and the Grubbs second generation
catalyst showed a slight difference in the catalytic activity
(7% and 18% yields, respectively) (entries 4 and 5, Table
2). The catalyst precursor 8 was found to be completely
inactive for the RCM reaction of the water-soluble substrate
homopropargylamine hydrochloride in a biphasic BMI.PF6/
H2O system (entry 6, Table 2).
In summary, we have designed a new class of ionophilic
phosphine ligands to operate in ionic liquid media, easily
attained in a one-step procedure. The proper choice of starting
materials provides a versatile method to access a wide range
of phosphine ligands containing imidazolium fragments. The
viability of these ligands was demonstrated by employing
the dicyclohexyl derivative 3 to synthesize the Grubbs type
compound 8. The catalyst precursor presents good activity
for the RCM reactions performed in biphasic BMI.PF6/
toluene and good recycling properties. In particular, in this
new catalytic system, the ruthenium concentration can be
significantly lowered from 5 to 0.25 mol % without
significant reduction of catalytic activity, as compared to
other Ru-catalytic reported so far. Further work is in progress
to access Grubbs-type compounds with less hindered N-
heterocyclic carbene ligands in order to address the problem
of less reactive substrates.
Acknowledgment. The authors are grateful to CT-
PETRO-CNPq and CAPES for partial financial support. We
also thank Prof. EÄ der C. Lima (IQ-UFRGS) and Ju´lio C. P.
Vaghetti (IQ-UFRGS) for atomic absorption analysis.
Figure 1. Conversion rate for the RCM of 2-allyl-2-methallylma-
lonate (reaction conditions of entry 2, Table 2).
Supporting Information Available: Full experimental
details and compounds characterization. This material is
available free of charge via the Internet at http://pubs.acs.org.
profile typical for catalyst precursor preactivation. In fact,
after 10 to 15 min of the reaction start, an IL phase color
OL702664A
240
Org. Lett., Vol. 10, No. 2, 2008