Solvate Ionic Liquids as Reaction Media for Electrocyclic Transformations

Article abstract of DOI:10.1002/ejoc.201501614

Solvate ionic liquids (SILs) consisting of lithium bis(trifluoromethylsulfonyl)imide dissolved in tri-or tetraglyme have recently emerged as a novel class of ionic liquids. Herein, the first use of solvate ionic liquids as a replacement for molecular solvents in electrocyclization reactions is reported. The SILs promoted both Diels-Alder and [2+2] cycloaddition reactions, compared to an appropriate molecular solvent, and 5 M lithium perchlorate in diethyl ether. The Gutmann acceptor number (AN) of these solvate ionic liquids has also been determined by ³¹P NMR spectroscopy to be 26.5, thus being modest Lewis acids.

Full text of DOI:10.1002/ejoc.201501614

DOI: 10.1002/ejoc.201501614
Communication
Ionic Liquids
Solvate Ionic Liquids as Reaction Media for Electrocyclic
Transformations
Daniel J. Eyckens,[a] Megan E. Champion,[a] Bronwyn L. Fox,[a]
Prusothman Yoganantharajah,[b] Yann Gibert,[b] Tom Welton,[c] and Luke C. Henderson*[a]
Abstract: Solvate ionic liquids (SILs) consisting of lithium compared to an appropriate molecular solvent, and 5
M lithium
bis(trifluoromethylsulfonyl)imide dissolved in tri- or tetraglyme perchlorate in diethyl ether. The Gutmann acceptor number
have recently emerged as a novel class of ionic liquids. Herein, (AN) of these solvate ionic liquids has also been determined by
the first use of solvate ionic liquids as a replacement for molec- 31P NMR spectroscopy to be 26.5, thus being modest Lewis
ular solvents in electrocyclization reactions is reported. The SILs acids.
promoted both Diels–Alder and [2+2] cycloaddition reactions,
Introduction
Ionic liquids have permeated many areas of chemistry, biology,
materials science, and are prevalent in industrial applications.
They are commonly advocated for their negligible vapour pres-
sure, solubilising ability, and their excellent application to or-
ganic synthesis.[1] They have also found excellent utility in en-
ergy applications as electrolytes for batteries and as corrosion
inhibitors. Recently, Watanabe et al. have developed and elec-
trochemically, physically, and thermally characterised a range of
novel solvate ionic liquids (SILs) as potential electrolytes for lith-
ium ion batteries.[2] While these SILs have excellent prospects
in energy storage applications, their potential use as an alterna-
tive to both traditional organic solvents and commonly em-
Figure 1. Solvate ILs (G3TFSI and G4TFSI) used in this study and 5M LPDE
ployed ionic liquids for organic transformations remains com-
pletely unknown.
(structure is simplified for comparison).
The preparation of these solvate ILs attracts no synthetic
challenges, simple dissolution of Li[NTf2] in either tri- or tetra-
glyme yields the ionic liquids, G3TFSI or G4TFSI, respectively,
within a few hours (Figure 1). Additionally, they possess various
attractive attributes for organic chemists: (i) they are aprotic;
(ii) they remain as liquids at low temperatures (e.g. 0 °C), while
maintaining their highly polar nature. They also possess a hard
cation, which has been “softened” due to complexation with
the polyether moieties and thus have great potential to stabilise
polar intermediates or, potentially, participate in Lewis acid acti-
vation.
This latter point is reminiscent of the 5
M lithium perchlorate
in diethyl ether (5M LPDE) pseudo-ionic liquid, reported by
Grieco et al.[3] which saw popularity in the late 1990s and early
2000s.[4]
Despite its excellent potential, the use of 5M LPDE has fallen
out of favour over the past decade. This has most likely arisen
due to the potential explosive nature of perchlorates, in addi-
tion to tedious drying procedures, and the high maintenance
nature of this solvent.
Thus, the lithium-derived solvate ionic liquids examined here
are a potential alternative to 5M LPDE, which obviate these
limitations as both the G3TFSI and G4TFSI solvents can be
stored over various drying agents (molecular sieves, MgSO4,
etc.) or heated to remove water, without any detrimental ef-
fects, nor the potential of explosion.
In this work, G3TFSI, G4TFSI and G4ClO4 are evaluated as
organic solvents and compared to 5M LPDE, and a suitable
molecular solvent as determined by the reaction. Also pre-
sented is NMR spectroscopic data relating to the Lewis acidity
[a] Deakin University, Waurn Ponds Campus,
Locked Bag 20000, Geelong, VIC 3220, Australia
E-mail: luke.henderson@deakin.edu.au
www.deakin.edu.au/research/ifm
[b] Faculty of Natural Sciences, Imperial College London
London SW7 2AZ, UK
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/ejoc.201501614.
Eur. J. Org. Chem. 2016, 913–917
913
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
of these ILs, from which the Gutmann acceptor number (AN) is
When the reaction was carried out in G3TFSI and G4TFSI
1
determined. It is the goal of this work to demonstrate prelimi- (Table 1, Entries 3 and 4, respectively), the H NMR spectra of
nary organic transformations, which can be carried out in these the crude mixtures showed complete consumption of aldehyde
solvents, with more in-depth examination of each reaction to 3, and the isolated yields were much higher at 56 % and 41 %,
be carried out in the future.
respectively. Note that the crude extract, when using either
solvate ionic liquid, typically contains residual glyme. This mate-
rial is easily removed by flash chromatography, and its presence
in the crude extract can be minimized by using successive brine
washes during workup. Finally, the use of G4ClO4 in this reac-
tion (Table 1, Entry 5) gave some conversion to the desired
product, though the isolated yield was quite low (15 %).
Examination of the crude reaction mixture for all examples
suggested a mixture of aldehyde 1, intermediate 3, and the
desired diene 4. To encourage higher yields we carried out the
reaction in the presence of 4 Å molecular sieves (Table 1, En-
try 6) to remove adventitious water. Finally, repeating the reac-
tion in G3TFSI accompanied by heating of the solution (80 °C)
for the final hour, gave 4 in a yield (Table 1, Entry 7, 70 %)
similar to those of the previous entries.
It is worth noting that heating of perchlorates is extremely
ill-advised due to their explosive potential. Thus, this highlights
a major advantage of G3TFSI and G4TFSI, which can be heated
to improve reaction yield or as part of any optimization process.
These simple demonstrations show that both G3TFSI and
G3TFSI can serve as viable solvents for organic transformations.
Highlighted in the original report by Cossio et al.[5] was that
Results and Discussion
The suite of solvents employed in this work includes G3TFSI,
G4TFSI, G4ClO4, 5M LPDE, and a suitable molecular solvent de-
pending on the chosen reaction. This suite gives a cross-section
of relevant solvents with consistency of ionic portions to give a
solid foundation of solvent comparison. Note that G3ClO4 was
omitted as this combination gives a solid salt.
Examination of the literature surrounding the use of
5M LPDE as a reaction solvent revealed a diverse collection of
organic reactions, which are enhanced in reaction rate, yield, or
selectivity by being conducted in that solvent. To that a [2+2]
reaction between dimethylketene (1) (generated in situ from
isobutyryl chloride) and (E)-cinnamaldehyde (2) was chosen,
giving alkene 4, after carbon dioxide extrusion.[5] This reaction
was reported not to proceed in the absence of LiClO4 and has
been reported in 5M LPDE; therefore, it was an obvious choice
for this study. Carrying out this reaction in chloroform (Table 1,
Entry 1) showed that the product was formed in trace amounts the aforementioned cycloaddition, when carried out with 4-
(<10 %), as expected. When repeating the literature condi- nitrobenzaldehyde does not eliminate CO2 but stops at the 2-
tions,[5] the high yield (95 %) reported originally in 5M LPDE oxetanone 7, thus necessitating a comparative reaction to de-
was unable to be reproduced.
termine if this outcome was observed in the solvate ionic liq-
uids.
Repeating this reaction under literature conditions in G3TFSI
gave an excellent yield of 94 % (Table 2, Entry 1); indeed, halv-
ing the reaction time had little effect on the isolated yield
(90 %, Table 2, Entry 2). As such, applying these optimized reac-
tion conditions to G4TFSI also gave a good yield of 75 %
(Table 2, Entry 3). Once again, the solvate ionic liquids proved
to be superior to both 5M LPDE and the examined molecular
solvent (chloroform), which gave modest and poor yields, re-
spectively. Due to the limitations of the perchlorate anion and
the poor results seen previously, this investigation continued
without the use of G4ClO4.
Table 1. Comparison of solvents in the [2+2] cascade formation of dienes.
Entry
Solvent
t [h]
T [°C]
Yield [%][a]
1
2
3
4
5
CHCl3
5M LPDE
G3TFSI
G4TFSI
G4ClO4
G3TFSI
G3TFSI
6
6
6
6
6
r.t.
r.t.
r.t.
r.t.
r.t.
<5
ca. 20
56
41
15
Table 2. Formation of oxetanone 7.
6[b]
7
6
r.t.
80 °C
60
70
6[c]
[a] Isolated yield post chromatography. [b] Activated molecular sieves
(100 mg) were used throughout the reaction. [c] The reaction mixture was
heated for the final hour of the specified time.
Attempts at optimising this reaction by freshly preparing the
5M LPDE, extensive drying of both LiClO4 and diethyl ether,
gave highly variable results at best. The yield of 4 in 5M LPDE
varied between 5–40 %, with <20 % being typical for repeated
attempts. Despite the complete consumption of the aldehyde
2, a complex mixture of products was typically observed in the
1H NMR spectra of the crude mixture.
Entry
Solvent
t [h]
Yield [%][a]
1
2
3
4
5
G3TFSI
G3TFSI
G4TFSI
5M LPDE
CHCl3
6
3
3
3
3
94
90
75
45[b]
<5
[a] Isolated yield post chromatography. [b] Reported yield of 90 % in 6 h.[3]
Eur. J. Org. Chem. 2016, 913–917
www.eurjoc.org
914 © 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Keeping with the theme of electrocyclisation reactions, a ate (9) and isoprene required the use of extremely high temper-
Diels–Alder reaction between isoprene and dimethyl acetylene- atures to obtain good yields (cf. 195 °C and 66 %, respec-
dicarboxylate (6) was examined. It is well known that Diels– tively).[8] The difficult nature of this reaction was reinforced by
Alder chemistry is compatible with ionic liquids,[6] salt addi- the poor result obtained when using chloroform as the solvent
tives,[7] and is another class of reaction, reported to proceed (Table 3, Entry 7) giving only traces of the desired product 12.
well in 5M LPDE (94 % in 7 h, Table 3, Entry 1).[3b]
The use of G3TFSI and G4TFSI (Table 3, Entries 8 and 9) gave
more promising, albeit low, isolated yields of 73 % and 21 %,
respectively, although it was found that the isolated product
was 13 possessing the trans configuration of the esters and
not the expected cis configuration of 12. We presume that this
epimerization has occurred due to the high temperature em-
ployed in this reaction. The reason for the drastic reduction in
yield in G4TFSI is unknown but may be due to a slightly differ-
ent solubility of isoprene in this IL compared to the G3TFSI
variation.
Table 3. Diels–Alder reactions in solvate ionic liquids.
Therefore, we repeated this reaction with 9 at room tempera-
ture in G3TFSI overnight (Table 3, Entry 10) to minimize possible
epimerization. Unfortunately, no reaction was observed in the
crude material, and only starting material 9 was isolated.
When employing dimethyl fumarate (10) as the dienophile,
a much less pronounced improvement in yield was noted in the
solvate ionic liquids, when compared to the molecular solvent
(Table 2, Entry 11 vs. Table 2, Entries 12 and 13). Nevertheless,
this was still greatly improved from previous reports of this
same transformation requiring a substantially longer reaction
time (cf. 36 h, 70 °C).[9]
Entry
Reactant
Solvent
t [h]
T [°C][a]
Yield [%][b]
1
2
3
4
5
6
7
8
9
10
11
12
13
8
8
8
8
8
8
9
9
9
5M LPDE
[bmim][PF6]
G3TFSI
G4TFSI
G3TFSI
G4TFSI
CHCl3
G3TFSI
G4TFSI
G3TFSI
CHCl3
7 h
2 h
r.t.
80[d]
100
100
100
100
100
100
100
r.t.
94[c]
96[e]
48
78
73
10 min
10 min
30 min
30 min
20 min
20 min
20 min
16 h
81
<5
Another reaction for which 5M LPDE was commonly used
was the rapid pseudo-pinacol rearrangement of epoxides, such
as that reported with styrene oxide (14).[4f] When this same
rearrangement was attempted, in both G3TFSI and G4TFSI
(Scheme 1) none of the rearranged product 15 was observed,
73[e]
21[e]
0
9
10
10
10
20 min
20 min
20 min
100
100
100
45
74
65
G3TFSI
G4TFSI
1
with only starting material being present in the H NMR spec-
trum of the crude product. Extension of reaction time and tem-
perature of this process gave the same result.
[a] Microwave irradiation. [b] Isolated yield. [c] Reported yield.[3b] [d] Reported
yield.[6]
Also, [bmim][PF6] at 80 °C gave a 98 % yield of the Diels–
Alder product 11 in 2 h. It is interesting to note that, while the
reported conditions for this Diels–Alder reaction in ionic liquids
use 5 equiv. of the dienophile, this number of equivalents was
used in this study to remain consistent with the literature After
10 min at 100 °C in G3TFSI, diene 11 was obtained in a moder-
ate 48 % isolated yield (Table 3, Entry 3). Interestingly, this reac-
tion proceeded better in G4TFSI giving 11 in a 79 % isolated
yield over 10 min (Table 3, Entry 4). Extension of the reaction
time to 30 min (Table 3, Entries 5 and 6) gave improved yields
for both G3TFSI and G4TFSI (73 % and 81 %, respectively).
Again, this highlights the benefit of being able to heat these
ionic liquids without limitation as part of an optimization proc-
ess. We noted that isoprene is only sparingly soluble in both
G3TFSI and G4TFSI, forming two layers when added to the same
reaction vessel. The high yields observed suggest that isoprene
is miscible only at higher temperatures and that the reaction
proceeds rapidly at a much lower effective concentration of the
dienophile, compared to previous reports.
Scheme 1. Attempted rearrangement of styrene oxide (14) with solvate ionic
liquids.
This outcome for G3TFSI and G4TFSI was attributed to the
lithium ion having reduced “availability” to act as a Lewis acid.
This was thought to be due to the chelation-coordination of
the glymes being stronger than that of the diethyl ether coordi-
nation to Li+ in solution for 5M LPDE. The structure of solvate
ionic liquids was recently investigated by Watanabe[2e] using
molecular dynamics simulation and is consistent with this
hypothesis.
The Diels–Alder reaction between isoprene and dimethyl
To test the hypothesis that the lithium cation is less Lewis-
maleate (9) or dimethyl fumarate (10) was also investigated. acidic in these solvate ILs when compared to the naked salt or
Previous reports of a Diels–Alder reaction with dimethyl male- in 5M LPDE, their relative Lewis acidities were determined by
Eur. J. Org. Chem. 2016, 913–917
www.eurjoc.org
915
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
NMR spectroscopy, as reported by Britovsek et al.[10] using a liquids are easy to prepare, store, and can be used without prior
modified Gutmann methodology.[11] This technique monitors specialist training. Their ability to be stored for extensive peri-
the shift of 31P atom of phosphine oxides in the presence of ods of time, kept anhydrous, and their stability over a broad
Lewis acids (Table 4).
range of temperatures enables them to be used where other
ionic liquids (including 5M LPDE) may not be the optimal
choice. Additionally, they exhibit mild Lewis-acidic characteris-
tics, both possessing Gutmann Acceptor Numbers of 26.5 as
determined by 31P NMR spectroscopy. Thus, they can poten-
tially be applied to any reaction in which Lewis acid catalysis
has been successful.
Table 4. Determination of the Gutmann Acceptor Number (AN) for solvate
ionic liquids.
Lewis acid
Et3P=O
AN[a]
δ(31P)
Δδ(31P)
None
n-Heptane
G3TFSI
46.88
46.66
57.96
57.96
46.76
46.75
64.48
0.22
0.00
11.30
11.30
0.10
0.09
17.82
5.07, 1.30
–
0.00
26.5
26.5
0.23
0.21
41.8
11.9, 3.10
G4TFSI
Triglyme
Tetraglyme
LiTFSI
Acknowledgments
The authors would like to thank the Institute for Frontier Materi-
als (IFM), the Strategic Research Centre (SRC) for Chemistry and
Biotechnology, and the Facuilty of Science, Engineering, and
Built Environment for their continued funding. We additionally
thank IFM for their postgraduate scholarship to D. E.
[bmim][TFSI]
51.68, 47.94
[a] Acceptor number (AN) was calculated according to AN
2.348 × Δδ[31P(Et3P=O)].[11]
=
In this study, triethylphosphine oxide was used in a 1:3 ratio
of phosphine oxide/Lewis acid in C6D6. The 31P resonances of
triethylphosphine oxide were standardised using the addition
of n-heptane (AN = 0). Interestingly, the change in δ(31P) for
both G3TFSI and G4TFSI was the same, indicating a very similar
level of Li+ accessibility within the ILs, giving AN values of 26.5.
To confirm that the AN observed was due to the IL and not the
ether constituent, the AN of each glyme was also measured and
found to be negligible (AN < 0.3), as expected.
Evaluating non-glyme-solvated LiTFSI gave an acceptor num-
ber of 41.8, markedly higher than that of both G3TFSI and
G4TFSI. This is expected as the C6D6–Li+ interaction would be
relatively weak, compared to the complexing glyme, and thus
the Li+ ion should be more accessible (i.e. a stronger Lewis acid)
than the Li+ ion present in both G3TFSI and G4TFSI. Unfortu-
nately, an AN was unable to be determined for 5M LPDE under
the same conditions. This was due to the instant precipitation
of LiClO4 once added to the NMR solvent, through removal of
the ether O–Li+ stabilization. Attempts to measure the 31P NMR
signals in neat 5M LPDE [with a trace of deuterium (CDCl3) for
signal lock] also failed to give a usable signal.
Finally, to put the above AN values into perspective, the AN
of [bmim][TFSI] was also determined by comparing the Lewis
acidity of the imidazolium and lithium cations. Two 31P NMR
signals were observed in this instance, giving AN values of 11.9
and 3.10. Presumably, the observed Lewis acidity of
[bmim][TFSI] is dominated by hydrogen-bonding effects be-
tween the phosphine oxide and the acidic hydrogen atom at
C2, and the pair at C4/C5 on the imidazolium ring. Considering
that the Lewis acidity of the C2 hydrogen atom present in imid-
azolium-derived ionic liquids has major ramifications for cataly-
sis in ionic liquids,[12] this bodes very well for the use of the
solvate ionic liquids in similar reactions.
Keywords: Ionic liquids · Lithium · Solvates · Lewis acids ·
Acceptor number
[1] a) J. P. Hallett, T. Welton, Chem. Rev. 2011, 111, 3508–3576; b) M. A.
Martins, C. P. Frizzo, A. Z. Tier, D. N. Moreira, N. Zanatta, H. G. Bonacorso,
Chem. Rev. 2014, 114, 1–70; c) H. Olivier-Bourbigou, L. Magna, D. Morvan,
Appl. Catal. A 2010, 373, 1–56; d) N. V. Plechkova, K. R. Seddon, Chem.
Soc. Rev. 2008, 37, 123–150; e) H. Weingartner, Angew. Chem. Int. Ed.
2008, 47, 654–670; Angew. Chem. 2008, 120, 664; f) Q. Zhang, S. Zhang,
Y. Deng, Green Chem. 2011, 13, 2619.
[2] a) Y. Kitazawa, K. Iwata, S. Imaizumi, H. Ahn, S. Y. Kim, K. Ueno, M. J.
Park, M. Watanabe, Macromolecules 2014, 47, 6009–6016; b) T. Mandai,
K. Yoshida, K. Ueno, K. Dokko, M. Watanabe, Phys. Chem. Chem. Phys.
2014, 16, 8761–8772; c) H. Moon, R. Tatara, T. Mandai, K. Ueno, K. Yo-
shida, N. Tachikawa, T. Yasuda, K. Dokko, M. Watanabe, J. Phys. Chem. C
2014, 118, 20246–20256; d) S. Terada, T. Mandai, R. Nozawa, K. Yoshida,
K. Ueno, S. Tsuzuki, K. Dokko, M. Watanabe, Phys. Chem. Chem. Phys.
2014, 16, 11737–11746; e) S. Tsuzuki, W. Shinoda, M. Matsugami, Y. Ume-
bayashi, K. Ueno, T. Mandai, S. Seki, K. Dokko, M. Watanabe, Phys. Chem.
Chem. Phys. 2015, 17, 126–129; f) K. Ueno, R. Tatara, S. Tsuzuki, S. Saito,
H. Doi, K. Yoshida, T. Mandai, M. Matsugami, Y. Umebayashi, K. Dokko,
M. Watanabe, Phys. Chem. Chem. Phys. 2015, 17, 8248–8257; g) K. Ueno,
K. Yoshida, M. Tsuchiya, N. Tachikawa, K. Dokko, M. Watanabe, J. Phys.
Chem. B 2012, 116, 11323–11331; h) C. Zhang, K. Ueno, A. Yamazaki, K.
Yoshida, H. Moon, T. Mandai, Y. Umebayashi, K. Dokko, M. Watanabe, J.
Phys. Chem. B 2014, 118, 5144–5153; i) C. Zhang, A. Yamazaki, J. Murai,
J.-W. Park, T. Mandai, K. Ueno, K. Dokko, M. Watanabe, J. Phys. Chem. C
2014, 118, 17362–17373.
[3] a) P. A. Grieco, J. D. Clark, C. T. Jagoe, J. Am. Chem. Soc. 1991, 113, 5488–
5489; b) P. A. Grieco, J. J. Nunes, M. D. Gaul, J. Am. Chem. Soc. 1990, 112,
4595–4596.
[4] a) A. Heydari, Tetrahedron 2002, 58, 6777–6793; b) S. Sankararaman, J. E.
Nesakumar, Eur. J. Org. Chem. 2000, 2003–2011; c) B. S. Babu, K. K. Bala-
subramanian, Tetrahedron Lett. 1998, 39, 9287–9288; d) B. P. Bandgar,
S. S. Makone, Synth. Commun. 2004, 34, 743–750; e) W. J. Kinart, C. M.
Kinart, J. Organomet. Chem. 2006, 691, 1441–1451; f) R. Sudha, K. Mal-
ola Narasimhan, V. Geetha Saraswathy, S. Sankararaman, J. Org. Chem.
1996, 61, 1877–1879.
[5] I. Arrastia, F. P. Cossio, Tetrahedron Lett. 1996, 37, 7143–7146.
[6] a) M. J. Earle, P. B. M. K. R. Seddon, Green Chem. 1999, 1, 23–25; b) I.
Meracz, T. Oh, Tetrahedron Lett. 2003, 44, 6465–6468.
[7] A. Kumar, Chem. Rev. 2001, 101, 1–17.
[8] F. X. Woolard, D. J. Mcphee (Amyris Inc., USA), WO2012103156, USA,
2012.
Conclusions
The use of solvate ionic liquids as solvents for organic transfor-
mations has been demonstrated for the first time. These ionic
Eur. J. Org. Chem. 2016, 913–917
www.eurjoc.org
916
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
[9]
Welch, L. Cabrera, P. A. Chase, E. Hollink, J. D. Masuda, P. Wei, D. W.
Stephan, Dalton Trans. 2007, 3407–3414.
J. H. Lee, W. H. Kim, S. J. Danishefsky, Tetrahedron Lett. 2009, 50, 5482–
5484.
[12] A. K. Chakraborti, S. R. Roy, J. Am. Chem. Soc. 2009, 131, 6902–6903.
[10] G. J. P. Britovsek, J. Ugolotti, A. J. P. White, Organometallics 2005, 24,
1685–1691.
[11] a) V. Gutmann, Coord. Chem. Rev. 1976, 18, 225–255; b) U. Mayer, V.
Gutmann, W. Gerger, Monatsh. Chem. 1975, 106, 1235–1257; c) G. C.
Received: December 28, 2015
Published Online: January 25, 2016
Eur. J. Org. Chem. 2016, 913–917
www.eurjoc.org
917
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim