Structure-based tuning of Tm in lipid-like ionic liquids. Insights from Tf2N- salts of gene transfection agents

Article abstract of DOI:10.1039/c2cc33609j

Ionic liquids of cations bearing two lipid-like aliphatic tails are shown to have values of T_m that can be very low, and that can be tuned up or down by the addition, deletion, or combination of dipolar modules and side-chain double bonds.

Full text of DOI:10.1039/c2cc33609j

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 7522–7524
www.rsc.org/chemcomm
COMMUNICATION
Structure-based tuning of Tm in lipid-like ionic liquids. Insights from
Tf2NÀ salts of gene transfection agentsw
Arsalan Mirjafari,a Samuel M. Murray,b Richard A. O’Brien,a Alexandra C. Stenson,a
Kevin N. West*b and James H. Davis Jr.*a
Received 18th May 2012, Accepted 11th June 2012
DOI: 10.1039/c2cc33609j
Ionic liquids of cations bearing two lipid-like aliphatic tails are
shown to have values of Tm that can be very low, and that can be
tuned up or down by the addition, deletion, or combination of
dipolar modules and side-chain double bonds.
Conceptually, this conundrum can be solved by creating
imidazolium cations bearing two long aliphatic substituents
(prospectively unsaturated) in neighbouring positions, e.g. at
N-1 and C-2. In principle, such ions are ‘make-able’, but
their synthesis and purification is likely to be rather involved.
So, it seemed sensible to seek a quick and easy approach that
would allow us to first establish whether the idea of maintaining
low Tm values in salts with double (and arguably parallel) tails –
regardless of their cation head group structure – was even
plausible. This proved to be quite straightforward, and by
doing so we gained insights into lipidic IL structure-property
(Tm) relationships that add to and deepen those uncovered by
our earlier work.
An old English maxim asserts that ‘two heads are better than
one’.1 But what about two tails? The question is important to
the design of new room-temperature ionic liquids, especially
ones capable of dissolving highly lipophilic substrates. To that
end, we recently demonstrated that low values of Tm can
be maintained in ionic liquids containing imidazolium ions
bearing a very long (lipophilic) N-bound alkyl tail, provided it
is unsaturated in character.2 By doing so we exploited a key
structure-property principle foundational to the natural
phenomenon of homeoviscous adaptation.3 In that process,
organisms modulate the fluidity of their cell membranes on a
temperature-dependent basis by varying the proportions of
saturated and unsaturated phospholipid components. However,
unlike naturally occurring zwitterionic phospholipids, our new
IL cations bear a single long aliphatic substituent, not two.
And, since we are interested in creating ILs with high degrees of
lipophilic (but not amphiphilic) character, the synthesis of ILs
with two long, unsaturated aliphatic substituents appended to
their cationic head groups seems logical to pursue. But, in
natural systems the two long aliphatic chains of phospholipids
‘point the same direction,’ a structural attribute that allows
them to self-assemble into their distinctly biphasic construct
with water. In the imidazolium cations characteristically used in
ILs, alkylation of the two ring nitrogens with long aliphatic
substituents is known to result in cations in which the latter
effectively ‘point in opposite directions.’ This impacts the
nature of their nanoscale assemblies relative to those which
are observed when imidazolium cations with a single long
(unidirectional) substituent assemble.4
By exchanging the chloride anions of off-the-shelf, commercially
available gene transfection agents (‘cationic lipids’)5 for the
bis(trifluoromethanesulfonyl)amide anion, Tf2NÀ, fully hydro-
phobic ILs – two with Tm values below normal room tempera-
ture – are readily prepared. Furthermore, each of these salts
contains a cationic head group to which two long, parallel
aliphatic chains are covalently tethered. Ionic liquids 1 and 2 are
a case in point.
As shown in Fig. 1, ILs 1 and 2 each contain a trimethyl-
ammonio- moiety tethered to a glyceryl-derived module. The
latter component, in turn, connects to two C18 aliphatic
appendages. In 1, these appendages are fully saturated, while in
2, each appendage is cis-unsaturated between chain carbon atoms
9 and 10. In turn, mono-tailed reference compounds 3 and 4 are
first-generation lipidic ILs featuring N-methylimidazolium cations
a Department of Chemistry, The University of South Alabama,
Mobile, Alabama, 36688, USA.
E-mail: jdavis@jaguar1.usouthal.edu; Fax: +1 251 460 7359;
Tel: +1 251 460 7427
b Department of Chemical & Biomolecular Engineering,
The University of South Alabama, Mobile, Alabama 36688, USA.
E-mail: kevinwest@southalabama.edu
w Electronic supplementary information (ESI) available: Synthetic
details, NMR and MS data. See DOI: 10.1039/c2cc33609j
Fig. 1 Structures of new lipidic ILs (1, 2 and 5) and reference lipidic
ILs (3, 4 and 6).
c
7522 Chem. Commun., 2012, 48, 7522–7524
This journal is The Royal Society of Chemistry 2012

View Article Online
bearing a single C18 appendage, saturated in the former and
unsaturated in the latter.
Net excision of the carbonyl moieties from the structure of 2
formally results in cation structure 5. The effect of doing so is
remarkable. While the Tm value of 2 is 14.6 1C, that of 5 is
À42.2 1C, a DTm of 56.8 1C. This is clearly the direct result of
suppressing the strong, carbonyl-induced dipolar contribution
to the inter-particulate cohesive forces acting in 2. Not only is
the Tm value of 5 very low in absolute terms, it is quite similar
to that of imidazolium IL 6, lowest-melting (À46.8 1C) member
of the first generation of lipidic ILs. A ‘map’ illustrating the
effect of these changes upon Tm is shown in Fig. 2.
Significantly, the Tm difference (Table 1) between 1 and its
imidazolium counterpart 3 is a scant 1.3 1C, despite the
doubled cation side-chain organic content of the former.
While acknowledging the comparison of 1 and 3 to be some-
thing of an ‘apples and oranges’ matter because of their
different head group types, the fact remains that the Tm of 1
is still low in absolute terms (well below the 100 1C benchmark
used as the upper Tm cut off for defining ionic liquids).
This, then, lends credence to the hypothesis that a putative
imidazolium analogue incorporating two saturated C18 chains
arranged in parallel might be reasonably anticipated to like-
wise manifest a sub-100 1C value of Tm.
A comparison of the Tm values of new ILs 1 and 2 with each
other is also informative. As noted, these differ only in the
nature of their lipidic side chains as either saturated (1) or
unsaturated (2). Here again, as with the lipidic imidazolium
ILs, the incorporation of a double bond in the side chain
results in a substantial decrease in Tm – that of 1 is 54.8 1C,
while that of 2 is 14.6 1C, a DTm of 40.2 1C. This demonstrates
that the Tm-lowering effect of side chain unsaturation is
relevant not just when an IL head group is an imidazolium
ring. Rather, it is clearly operative across both classes of ionic
liquids, with their dramatically different head group types and
total organic content. This buttresses our expectation that the
‘double bond effect’ is likely to prove very general, and
broadly applicable when designing ILs of various classes
whenever low melting points must be retained while still
imbuing the material with a high degree of hydrocarbon
side-chain content.
Significantly, both 5 and 6 have two side-chain double
bonds, although they are in the same chain in 6 but two
separate side chains in 5. With this in mind, and noting that
there is a substantial difference in Tm resulting from having
one (IL 4) versus two (IL 6) double bonds in a given cation, it
seems likely that the double bond effect is cumulative. At the
same time, it appears that the specific locations of the side-chain
double bonds may be of secondary importance; this comports
with observations we made in our earlier study vis-a-vis the Tm
differences between ILs as a function of the position of double
bonds in the lipidic side chain.2
It also bears noting that the Tm decrease (40.2 1C) brought
about by the formal introduction of two double bonds into the
cation structure of 1 to generate cation 2 is a substantial
proportion – about 70% – of the magnitude of the Tm increase
(56.8 1C) that is induced by the formal introduction into 5 of two
dipolar carbonyl groups, a change which likewise generates 2
(Fig. 2). This indicates that it is possible to take a basic cation
framework and temper the impact that one modification to it has
– pertaining here to its conceptual ‘baseline’ Tm – by making a
simultaneous modification of opposing effect to the same ion.
The overall features of the DSC curves (Fig. 3) from which
the Tm values of 1, 2 and 5 are extracted are highly evocative of
those observed with many natural lipids and polymers.8,9
Among these is the magnitude of the enthalpy change associated
with the melting events. In this regard it is notable that the
transitions of ILs 2 and 5 are very weak compared to that of 1.
This is indicative of solid states on the parts of 2 and 5 in which
side-chain interactions are not optimal due to the disorder brought
about by the double-bond induced side-chain kinking, a pheno-
menon clearly observed with saturated versus unsaturated natural
Although the salutary effect upon Tm of unsaturated versus
saturated side chains is again evident in these results, the
absolute 14.6 1C Tm value of double-tailed lipid IL 2 is
still considerably higher than that observed for benchmark
imidazolium IL 4. Of course, as would be the case for any IL,
its Tm value is the outworking of a number of factors, among
which are the number and strengths of all possible cohesive
forces. In typical (e.g., non-functionalized) imidazolium-type ILs,
two inter-particulate force types – Coulombic and London – are
thought to predominate.6 However, the inter-particulate cohesive
forces acting in 1 and 2 would clearly include strong dipolar
contributions from the two ester linkages present in the cation
head group area of each.7 Note that the imidazolium IL
benchmarks 3 and 4 are devoid of such components. So, to
probe the impact these ester groups have upon Tm, we
prepared IL 5 by anion exchange from a third commercially
available cationic lipid chloride salt.
Table 1 Tm values of lipidic ionic liquids. The values for compounds
3, 4, and 6 are taken from ref. 2
Ionic liquid
Tm (oC)
Æ
1
2
3
4
5
6
54.8
14.6
53.5
À20.9
À42.2
À46.8
0.5
0.5
1.0
1.0
0.5
1.0
Fig. 2 Structural changes and their Tm impacts.
Chem. Commun., 2012, 48, 7522–7524 7523
c
This journal is The Royal Society of Chemistry 2012

View Article Online
for pursuing the synthesis of new imidazolium ILs with long
but adjacent aliphatic appendages, but those with a myriad of
other cationic heads as well as functional groups10 which can
be used in conjunction with these appendages to create
‘designer’ melting points.
The corresponding authors would like to thank the National
Science Foundation (Grant No. CBET-1133101) for their support
of our research.
Notes and references
1 Informative remarks on the origin of the phrase may be found
at: http://www.phrases.org.uk/meanings/two-heads-are-better-than-
one.html. Accessed 05/05/2012.
Fig. 3 DSC traces obtained for (top to bottom) 1, 2 and 5. Those of 1
and 5 are referenced to the left y axis while that for 2 is referenced to
the right y axis. The traces are offset to distinguish them, but are not
rescaled.
2 (a) S. M. Murray, R. A. O’Brien, K. M. Mattson, C. Ceccarelli,
R. E. Sykora, K. N. West and J. H. Davis, Jr., Angew. Chem., Int. Ed.,
2010, 49, 2755–2758; (b) J. H. Davis, Jr, International Symposium on
Ionic Liquids and Life Science, Keio University, Yokohama, Japan,
2007, Session VI.A.
lipids.9 This further implies that in these ILs, the unsaturated
side chains pack less densely than those which are saturated, a
phenomenon likely to affect properties of practical importance
such as the solubility of various gases in the ILs.
3 M. Sinensky, Proc. Natl. Acad. Sci. U. S. A., 1974, 71, 522–525.
4 X. Wang, F. W. Heinemann, M. Yang, B. U. Melcher, M. Fekete,
A.-V. Mudring, P. Wasserscheid and K. Meyer, Chem. Commun.,
2009, 7405–7407.
5 P. L. Felgner, T. R. Gadek, M. Holm, R. Roman, H. W. Chan,
M. Wenz, J. P. Northrop, G. M. Ringold and M. Danielsen, Proc.
Natl. Acad. Sci. U. S. A., 1987, 84, 7413–7417.
Collectively, the foregoing results provide useful insights
into features and relationships that are consistent with achieving
low Tm values in salts with lipid-like cation appendages, features
that are not apparent from our earlier study. Specifically, cis
double bonds in the adjacent/parallel side chains of the new salts
are observed to be powerful downward drivers of Tm even
though the alkyl content in these cations is effectively doubled
relative to first generation salts. Further, the more highly charge-
localized nature of the quaternary ammonium head groups of the
present salts appears to have little intrinsic impact on Tm in
lipidic ILs. Note that the Tm of 5, with its quaternary ammonium
head group, is a scant 5 1C higher that that of 6, the lowest
melting of the first-generation lipidic ILs. In the final analysis, we
believe the present findings not only provide ample justification
6 See, for example: I. Lopez-Martin, E. Burello, P. N. Davey,
´
K. R. Seddon and G. Rotherberg, ChemPhysChem, 2007, 14, 690–695.
7 The capacity of functional groups to impart dipolar character to IL
cations has been previously documented and quantified. See:
(a) N. K. Sharma, M. D. Tickell, J. L. Anderson, J. Kaar,
V. Pino, B. F. Wicker, D. W. Armstrong, J. H. Davis, Jr. and
A. J. Russell, Chem. Commun., 2006, 646–648; (b) P. Twu,
Q. Zhao, W. R. Pitner, W. E. Acree Jr., G. A. Baker and
J. L. Anderson, J. Chromatogr., A, 2011, 1218, 5311–5318.
´
8 E. M. Antipov, V. G. Kulichikhin and N. A. Palte, Polym. Eng. Sci.,
1992, 32, 1188–1203.
9 R. N. A. H. Lewis, D. A. Mannock and R. N. McElhaney, in
Methods in Molecular Biology, ed. A. M. Dopico, Humana Press,
Totowa, NJ, 2007, vol. 400, pp. 171–195.
10 (a) J. H. Davis, Jr., Chem. Lett., 2004, 33, 1072–1077; (b) S. Tang,
G. A. Baker and H. Zhao, Chem. Soc. Rev., 2012, 41, 4030–4066.
c
7524 Chem. Commun., 2012, 48, 7522–7524
This journal is The Royal Society of Chemistry 2012