Fructose-derived ionic liquids: Recyclable homogeneous supports

Article abstract of DOI:10.1016/j.tetlet.2003.09.121

A new homogeneous support has been developed that is based on a fructose-derived ionic liquid. It has been applied to the preparation of a series of supported Diels-Alder adducts. The cleavage of these adducts from the ionic liquid has resulted in some interesting observations regarding the base stability of imidazolium cations. Using a transesterification cleavage, the ionic support can be recovered and recycled.

Full text of DOI:10.1016/j.tetlet.2003.09.121

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 8399–8402
Fructose-derived ionic liquids: recyclable homogeneous supports
Scott T. Handy* and Maurice Okello
Department of Chemistry, State University of New York at Binghamton, Binghamton, NY 13902, USA
Received 29 July 2003; revised 12 September 2003; accepted 15 September 2003
Abstract—A new homogeneous support has been developed that is based on a fructose-derived ionic liquid. It has been applied
to the preparation of a series of supported Diels–Alder adducts. The cleavage of these adducts from the ionic liquid has resulted
in some interesting observations regarding the base stability of imidazolium cations. Using a transesterification cleavage, the ionic
support can be recovered and recycled.
©
2003 Published by Elsevier Ltd.
5
In the context of a project directed at the synthesis of
room temperature ionic liquids (RTILs) from biorenew-
able sources, we have reported the preparation of a new
Janda). Since the supports are homogeneous in a
variety of conventional organic solvents, reactions can
be performed under conventional solution-phase condi-
tions. At the same time, by changing the polarity of the
solvent (most frequently by the addition of methanol),
the support and supported molecule will precipitate,
resulting in facile separation by filtration.
1
family of RTILs based on fructose. While these new
materials can be used as solvents in conventional
2
organic reactions, their protic nature and the presence
of a hydroxyl group indicated that there could be some
new applications of these materials beyond those typi-
3
cally encountered with ‘non-functionalized’ RTILs. In
While this is a major step forward, there are still
limitations to the current soluble supports. First, using
simple PEG supports, the maximum loading that can
be achieved is even lower than that of polystyrene
this communication, we report our preliminary studies
utilizing the hydroxymethylene group as a tether for
supported organic synthesis.
6
supports (<1 mmol/g). Second, the highly oxygenated
Supported synthesis is a widely employed technique
that has greatly facilitated the synthesis of many com-
pounds and is the critical element behind the explosion
nature of the PEG supports can interfere in certain
types of reactions or even be degraded under certain
reaction conditions.
4
in combinatorial synthesis. Traditionally, supported
synthesis has employed a heterogeneous material such
as cross-linked polystyrene to support one of the reac-
tants. The primary advantage of such a choice is that
the supported material, being heterogeneous, can be
readily separated by simple filtration from the reaction
medium and by-products. At the same time, this hetero-
geneity limits the types of reactions and reaction condi-
tions that can be employed. Further, using simple
polystyrene supports, which are typically functionalized
to <10%, the maximum loading is <2 mmol/g.
With this in mind, our recently reported fructose-
derived ionic liquid 1 appeared to hold several advan-
tages. First, the molecular weight of the ionic liquid is
relatively low and each molecule can be used as a
support, meaning that reasonably high effective load-
ings of ionic support 1 (>2 mmol/g) could be achieved.
Second, it is well established that the solubility proper-
ties of ionic liquids in general can be tuned based upon
7
the anion and alkyl chain length in the cation. As a
result, supports could be developed that would be
immiscible in either polar solvents or non-polar sol-
vents and this immiscibility used to separate reaction
These limitations have led more recently to the develop-
ment of a variety of ‘soluble’ supports (e.g. the
polyethyleneglycol (PEG) supports popularized by
8
by-products from the supported products.
This tunable miscibility was also observed with the
fructose-derived ionic liquids. Thus, ionic liquid sup-
port 1 was found to be immiscible with very low
polarity solvents (hexane and cyclohexane) and with
Keywords: Diels–Alder reaction; ionic liquid; soluble support; car-
bene; transesterification.
*
Corresponding author.
0
040-4039/$ - see front matter © 2003 Published by Elsevier Ltd.
doi:10.1016/j.tetlet.2003.09.121

8
400
S. T. Handy, M. Okello / Tetrahedron Letters 44 (2003) 8399–8402
polar solvents (water and methanol). On the other
hand, the same ionic liquid with a dicyanimide anion
was completely miscible with polar solvents such as
water and methanol, but immiscible with lower polarity
solvents such as hexane, toluene, and diethyl ether.
These complimentary behaviors could be of consider-
able use in future applications of these ionic liquid
supports. For the present research, efforts focussed on
the use of the triflimide support 1.
Scheme 1. Acylation of ionic support 1.
In order to explore the possible application of fructose-
derived ionic liquids as a new class of soluble supports,
our first focus was on the preparation of some simple
acrylate-derived Diels–Alder adducts. The first stage in
these studies was therefore to prepare acrylate 2 by the
reaction of acryloyl chloride with alcohol 1 (Scheme 1).
Immediately, problems were encountered. Attempts
using standard pyridine or triethylamine mediated con-
ditions afforded a dark, tarry mixture from which only
modest amounts of the desired acrylate could be sepa-
rated. A report by Aggarwal did shed some light as to
the potential problem, since he had noted that the use
of imidazolium-based ionic liquids as the solvent for the
Baylis-Hillman reaction resulted in extensive deproto-
nation of the C2-position of the imidazolium ring,
thereby greatly complicating the resulting reaction
mixture.
9
In an effort to determine if a similar situation was
responsible for our observations, an attempt was made
to acylate alcohol 1 in the absence of base. Under neat
reaction conditions, acrylate 2 was obtained along with
hydrogen chloride addition product 3 in a 3:1 ratio.
Since product 3 did not interfere in subsequent reac-
tions, this mixture could be used as is. Further opti-
mization lead to the discovery that acrylate 2 could be
cleanly prepared by acylation at low temperature
(−20°C) in the presence of triethylamine without the
1
0
concomitant formation of any of product 3.
Table 1. Diels–Alder reactions and support cleavage reactions using acrylate 2
Entry
R1
R2
R3
R4
R5
R6
R
Conditionsa
Yield (%)b
endo/exoc
Recovered 1
1
2
3
4
5
6
7
H
H
CH3
H
H
H
H
H
H
H
CH3
H
CH3
H
H
CH3
CH3
CH3
H
H
H
H
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
-CH2-
H
H
H
H
CH3
CH3
CH3
CH3
CH3
CH3
Bn
A,C
B,D
B,D
B,E
B,F
A,F
B,F
B,F
B,F
B,F
B,G
B,H
65
66
29
74
60
60
63
57
46
66
91
77
3:1
NA
6:1
5:1
all endo
3:1
NA
NA
5:1
H
H
H
H
H
H
d
-(CH ) -
2
2
H
-CH2-
97
92
91
90
98
90
92
CH3
CH3
H
H
H
H
H
H
H
H
H
e
8
9
10
11
12
H
H
H
-(CH ) -
all endo
3:1
3:1
2
2
-CH2-
-CH2-
H
Et
a
A=Toluene, reflux, 4 h; B=neat, hydroquinone, 120°C, 12 h; C=LiOH, EtOH/H O, 12 h; D=LiOH, MeOH/H O, 12 h; E=2.5 N NaOH,
2
2
reflux, 2 h; F=KCN, MeOH, 12 h; G=KCN, BnOH, 12 h; H=KCN, EtOH, 12 h.
b
c
Isolated yield over three steps from ionic support 1.
Determined by 1H NMR.
d
e
Ortho/meta ratio determined by 1H NMR.
Using ionic support recycled from entry 7.

S. T. Handy, M. Okello / Tetrahedron Letters 44 (2003) 8399–8402
8401
With the desired acrylate in hand, the Diels–Alder
ionic support was recovered in both the first and second
use. Further the isolated yield of the final ester product
is very similar between the two runs, indicating that the
support remains intact.
reactions themselves could be investigated (Table 1). In
11
all cases, standard thermal conditions were employed.
For cyclopentadiene, a solution of the acrylate and the
diene was heated to reflux in toluene until starting
material was consumed by TLC. For the other dienes,
the reaction was performed neat in the presence of
hydroquinone to avoid extensive oligomerization of the
In conclusion, we have developed a new type of soluble
support—a hydroxylic ionic liquid based on fructose.
This support affords higher potential loadings than
other simple homogeneous or heterogeneous supports
and can be readily recovered and recycled. Product/by-
product separation can be performed by either extrac-
tion or removal of the volatiles in vacuo. The key
present challenge in using this support is the limited
base stability of the imidazolium core, since it can be
readily deprotonated. Efforts are currently underway to
develop new ionic supports that avoid this limitation
and will be reported in due course.
1
2
diene. In all cases, the Diels–Alder products 4 could
be readily isolated by removal of the volatiles in vacuo.
The final step in this short sequence was cleavage of the
Diels–Alder adducts from the ionic liquid support. Ini-
tial attempts focused on simple basic saponification of
the ester linkage (Table 1, entries 1–4). Although this
proved to be an efficient method for obtaining the
carboxylic acid products in reasonable overall yield,
recovery of the ionic liquid support was not as straight-
forward. The recovered material was a dark brown,
1
viscous liquid that was impure by H NMR. Again, we
Acknowledgements
suspected that the basic conditions were deprotonating
the C2 position of the imidazolium core and leading to
a variety of side reactions. This was supported by the
observation that basic transesterification with sodium
methoxide in methanol also resulted in a dark brown,
viscous liquid.
The authors thank the State University of New York,
the Research Foundation, and the New York State
Energy Research Development Agency (NYSERDA)
for funding this research.
In an effort to avoid this problem, attention was turned
to milder methods for cleaving the Diels–Alder product
from the ionic liquid support. After exploring a few sets
of conditions, the one that proved to be most effective
References
1
3
was a cyanide-mediated transesterification. Using a
catalytic amount of potassium cyanide in methanol, the
methyl esters 5 were obtained in generally good overall
1. (a) Handy, S. T.; Okello, M.; Dickenson, G. Org. Lett.
2003, 5, 2513–2515; (b) Handy, S. T. Chem. Eur. J. 2003,
9, 2938–2944.
2. Handy, S. T.; Okello, M. Tetrahedron Lett. 2003, 44,
8395.
1
4,15
yield (Table 1, entries 5–10).
ionic liquid support could be recovered in greater than
0% yield. After some experimentation, it was deter-
More importantly, the
9
3. For recent reviews on ionic liquids, see: Welton, T. Chem.
Rev. 1999, 99, 2071–2083; Wasserscheid, P.; Keim, W.
Angew. Chem., Int. Ed. 2000, 39, 3772–3789; Dupont, J.;
de Souza, R. F.; Suarez, P. A. Z. Chem. Rev. 2002, 102,
3667–3692; Ionic Liquids in Synthesis; Wassercheid, P.;
Welton, T., Eds.; Wiley-VCH: Weinheim, 2003.
4. For a review, see: Fruechtel, J. S.; Jung, G. Angew.
Chem., Int. Ed. Engl. 1996, 35, 17–42.
5. For reviews, see: Spanka, C.; Wentworth, P., Jr.; Janda,
K. D. Combinatorial Chemistry & High Throughput
Screening 2002, 5, 233–240; Dickerson, T. J.; Reed, N.
N.; Janda, K. D. Chem. Rev. 2002, 102, 3325–3343.
6. For a report focusing on methods to increase the loading
of these supports, see: Sunder, A.; Heinemann, J.; Frey,
H. Chem. Eur. J. 2000, 6, 2499–2506.
mined that this recovery could be best accomplished by
taking up the residual mixture of potassium cyanide
and ionic support 1 in 4N sodium hydroxide, followed
by extraction with methylene chloride and ethyl acetate.
Addition of basic alumina to remove coloration, fol-
lowed by filtration and removal of the volatile solvents
then afforded nearly colorless ionic support 1. This
support was pure and free of solvent and/or Diels–
1
Alder adducts as determined by H NMR.
This transesterification method for cleavage from the
support has one additional interesting feature—the
ability to prepare a number of different esters from the
same supported carboxylate. This attribute was briefly
examined by treatment of cyclopentadiene Diels–Alder
adduct 4 with potassium cyanide and either ethanol or
benzyl alcohol (Table 1, entries 11–12). In both cases,
the anticipated ester products were obtained in good
yield and the ionic support 1 could be recovered in
nearly quantitative yield.
7. For selected studies, see: Huddleston, J. G; Visser, A. E.;
Reichert, W. M.; Willauer, H. D.; Broker, G. A.; Rogers,
R. D. Green Chem. 2001, 3, 156–164; Golding, J.;
Forsyth, S.; MacFarlane, D. R.; Forsyth, M.; Deason, G.
B. Green Chem. 2002, 4, 223–229; MacFarlane, D. R.;
Forsyth, S. A.; Golding, J.; Deacon, G. B. Green Chem.
2
002, 4, 444–448; Holbrey, J. D.; Reichert, W. M.; Swat-
Finally, the fact that the ionic support 1 can be recov-
ered so readily means that it can be recycled and used
in further applications of supported synthesis. The
efficiency of this recycling can be seen in entries 7 and
loski, R. P.; Broker, G. A.; Pitner, W. R.; Seddon, K. R.;
Rogers, R. D. Green Chem. 2002, 4, 407–413.
8. There has already been a single report exploring the use
of a RTIL as a soluble support, see: Fraga-Dubreuil, J.;
Bazureau, J. P. Tetrahedron Lett. 2001, 42, 6097–6100.
8
of Table 1 in which it can be seen that >90% of the

8
402
S. T. Handy, M. Okello / Tetrahedron Letters 44 (2003) 8399–8402
9. Aggarwal, V. K.; Emme, I.; Mereu, A. Chem. Commun.
0.5 mL (5 mmol) of 2,3-dimethylbutadiene. The mixture
2
002, 1612–1613.
was purged with argon, then sealed and heated in an oil
bath to 120°C. After 12 h, the tube was cooled to room
temperature and the excess diene and any volatile by-
products removed by concentration in vacuo. The crude
product was the triturated with cyclohexane (2×10 mL)
and any residual cyclohexane was removed from the
supported product in vacuo to afford the Diels–Alder
adduct ready for cleavage from the support.
1
0. General procedure for acylation: To 0.2443 g (0.5487
mmol) of ionic support 1 was added 660 mL of
dichloromethane followed by 116 mL (0.8339 mmol) of
triethylamine. The solution was chilled to −20°C and a
solution of 46 mL (0.5487 mmol) of acryloyl chloride in
1
10 mL of dichloromethane was added dropwise. The
reaction was stirred for 30 min and then quenched with
00 mL of 5N sodium hydroxide. The resulting mixture
2
13. Herzig, J.; Nudelman, A.; Gottlieb, H. E.; Fischer, B. J.
Org. Chem. 1986, 51, 727–730.
was diluted with 5 mL of dichloromethane and 2 mL of
water. The aqueous layer was extracted with dichloro-
methane (3×5 mL) and the combined organic extracts
were dried with magnesium sulfate, filtered, and concen-
trated to afford 0.2604 g (96% by H NMR) of acrylate
2
important, since unless all of the amine hydrochloride
by-product was neutralized prior to extraction and con-
centration, significant amounts (up to 20%) of the hydro-
gen chloride addition product 3 were obtained.
14. General procedure for transesterification: the supported
Diels–Alder adduct with 2,3-dimethylbutadiene (crude,
0.5339 mmol) was dissolved in 5 mL of methanol. To this
solution was added 17 mg (0.2668 mmol) of potassium
cyanide and the reaction was stirred 12 h. The methanol
was removed in vacuo and the remaining residue was
triturated with cyclohexane (3×10 mL). The combined
cyclohexane washes were concentrated in vacuo to afford
0.60 g (63% isolated over three steps) of methyl 1,2-
dimethylcyclohexene-4-carboxylate.
1
. The method of quenching this reaction was quite
1
1. All crude Diels–Alder adducts displayed spectral proper-
ties consistent with those reported in the literature.
2. General procedure for the Diels–Alder reaction: In a
pressure tube with 0.2688 g (0.5339 mmol) of acrylate 2
was added 1.0 mg (0.0053 mmol) of hydroquinone and
15. The modest yields of the Diels–Alder adducts over this
three step sequence are most likely due to the volatility of
the methyl esters, since the yield for the less volatile
benzyl ester is much higher.
1