Enhanced membrane transport of pharmaceutically active protic ionic liquids

Article abstract of DOI:10.1039/c1cc14314j

We show that pharmaceutically active protic ionic liquids can be designed to rapidly transport through model membranes as neutral hydrogen bonded clusters.

Full text of DOI:10.1039/c1cc14314j

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 11429–11431
www.rsc.org/chemcomm
COMMUNICATION
Enhanced membrane transport of pharmaceutically active protic ionic
liquidsw
Jelena Stoimenovski* and Douglas R. MacFarlane
Received 18th July 2011, Accepted 17th August 2011
DOI: 10.1039/c1cc14314j
We show that pharmaceutically active protic ionic liquids can be
designed to rapidly transport through model membranes as
neutral hydrogen bonded clusters.
are based on actives that are acids and bases (or simple salts
thereof) and hence we have also recently reported on pharma-
ceutically active protic ILs (PILs), where we reacted pharma-
ceutically active acids with pharmaceutical/biocompatible
bases to produce salts with dual activity, e.g. an analgesic/
nasal decongestant dual activity PIL.9
Ionic liquids (ILs) are currently being investigated for use in
a number of areas of science.1 Some years ago, Rogers and
co-workers proposed ILs based on active pharmaceutical
ingredients (APIs), whereby the synthesised ILs not only retain
pharmaceutical activity but also have the added benefits of
existing in a liquid form.2 Some of the benefits of the liquid
state include improved solubility, stability, bioavailability and
alternative delivery modes of the drug. Being amorphous,
pharmaceutical ILs could be a possible solution to one of
the more significant problems in the pharmaceutical industry –
polymorphic inter-conversion, whereby one polymorph of a
drug converts into another over time.3 Given that the counter-
ion chosen for the formation of the IL can have a significant
influence on the pharmacokinetics of the drug candidate, the
new salt form of the registered drug would need to be treated
as a new chemical entity, meaning that it can be separately
patented.4 Hence, the formation of ILs of existing APIs could
assist in the development of new forms of the API.
However, it is a well-established fact that ionic drugs do not
readily cross biological membranes and skin due to their lack
of sufficient lipophilicity.4 Stratum corneum is the main barrier
for permeation of most drugs through the skin and only a small
number of drug molecules have the optimal physicochemical
characteristics to penetrate it sufficiently to exert their therapeutic
effect. A number of strategies have been proposed to overcome
skin impermeability. Wotton et al. suggest that an increase in
the diffusion coefficient (D) or the partition coefficient (K) of the
drug, or both, can result in a multiplicative effect on the flux of
the permeant.10 Penetration enhancement is most commonly
achieved by incorporating solvents that promote transfer
through the stratum corneum.11
Here we describe a series of attenuated total reflectance–
Fourier transform infrared (ATR-FTIR) experiments designed
to investigate the permeation of different types of PILs
through a model membrane. We show that some of the PILs
exhibit strongly enhanced permeation properties, probably
due to the formation of hydrogen bonded complexes in the
membrane. The results of this study will assist in the design of
the most effective pharmaceutically active ILs.
Since their emergence, pharmaceutically active ILs have
been studied by a number of groups. Hough et al. and our
group have reported on pharmaceutically active ILs containing
cations/anions of various actives.5 One of the compounds,
lidocaine docusate, was shown to have enhanced analgesic
effect compared to the parent lidocaine chloride salt.5a This is just
one of the examples where combining two pharmaceutical/
biocompatible ions results in compounds with characteristics
superior to the parent drugs. Bica and co-workers have reported
ILs based on analgesic, anti-pyretic and anti-inflammatory
compounds, acetylsalicylic and salicylic acids,6 while Cybulski
et al. have synthesised and characterised ILs based on
antiseptic/disinfectant cations and enzyme/amino acid anions
and the resulting ILs were found to be very effective against
bacteria and fungi.7,8 A large number of drugs on the market
The use of the ATR-FTIR spectroscopic technique to study the
diffusion of permeants is well established in the pharmaceutical
field.12 The model membrane used for this study is a silicone
membrane. Although the stratum corneum is a more complex
molecular structure, the trends observed are thought to be
similar to those seen in the silicone membrane and it is thus
accepted as a useful model in this context.12a The synthetic
procedures and characterisation data for the compounds
investigated and the detailed experimental procedures are
given in the supporting information. Following the work of
Tantishaiyakul et al.,12a propylene glycol (PG) was used as the
solvent for the control experiments. PG was found not to
change the properties or diffusion barrier of the membrane.
The two types of PILs that were chosen for study were (i)
those that have been shown to be proton transferred and fully
dissociated13 due a sufficiently large difference in pKa between
acid and base and (ii) those that are thought to be proton
School of Chemistry, Monash University, Clayton, Victoria 3800,
Australia. E-mail: Jelena.Stoimenovski@monash.edu;
Fax: +61 3 9905 4535; Tel: +61 3 9905 4540
w Electronic supplementary information (ESI) available: synthetic
procedures, diffusion profiles of the studied compounds and an
example FTIR membrane experiment. See DOI: 10.1039/c1cc14314j
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 11429–11431 11429

View Online
present as a PG solution a very different behaviour is
observed; it permeates slowly through the membrane, achieving
only 15% of saturation in 20 h. The two pure salt starting
materials for this PIL, sodium salicylate and NTH3+ sulphate,
when dissolved in PG, permeate o 10% in a 20 h period. Thus
the IL form is significantly more permeable than either of the
inorganic salt forms of the actives.
Salicylic acid (SalH) in PG was also examined in order to
determine how the permeation of a neutral compound would
compare to the ionic species. The salicylic acid was found to
permeate about as rapidly as the salicylate-containing IL.
Since the two starting materials, which are simple salts, do
not cross the membrane, whereas the pure acid form does, it
appears that neutral species are the most rapidly permeable in
this type of membrane. This is in accord with the general
observation that ionic compounds do not cross the membrane
as readily as neutral compounds. The behaviour of the IL may
therefore appear paradoxical. However, our previous studies9
have suggested, on the basis of crystallographic analysis
of analogous compounds and gas phase ab initio structure
calculations, that NTH3+ Salꢀ may exist as a cyclic, two ion-pair
complex.9 Thus, as this neutral complex, the PIL can cross the
membrane as rapidly as unionised neutral species. Further-
more, it is likely that the cyclic complex ceases to exist in the
PG solution, dissociating into individual ions and hence limiting
the permeability. The fact that some level of permeation
is observed, as shown in Fig. 2, may relate to the presence of
a small equilibrium fraction of neutral species being present,
or formed at the membrane interface. As these are removed by
permeation from the membrane’s outer surface, their concen-
tration is maintained by the equilibrium process in the liquid.
Thus, it appears that the permeation of salt forms of actives
can be enhanced by formation of a PIL in which there is
significant association present in the form of neutral species. In
Fig. 1 The types of protic ILs investigated: (a) proton transferred
and fully dissociated;13 (b) proton transferred but form hydrogen
bonded clusters.9
transferred, but form hydrogen bonded clusters9 (Fig. 1). The
PILs in (ii) are dual active PILs, with both the cation and
anion pharmaceutically active. This is the first report of the
pharmaceutical IL bromohexinium ibuprofenate; we have
previously described the preparation and properties of the
other compounds.9,13 In that previous work, the IR spectroscopy
and Walden plot analysis of the transport properties were used
to assess the extent of proton transfer; proton transfer was found
to be high in all cases. However, as discussed in more detail below,
the existence of ion pairs and clusters was identified, supported by
a crystal structure analysis of analogous compounds.9
Fig. 2 shows the relative absorbances of the species crossing
the membrane over time for NTH3+ Salꢀ (as a pure liquid and
as a PG solution) and its starting materials (in PG). ATR-
FTIR spectroscopy allows the concurrent measurement of the
diffusional parameters of both the primary permeant and its
formulation solvent, if both species diffuse into the membrane
under study (and their absorbance bands are distinct from
each other).12b In the case of the salicylate compounds we
monitored the carboxylate asymmetrical stretching band at
1570 cmꢀ1, whereas in salicylic acid the CQO stretching at
1665 cmꢀ1 was analysed. Internal hydrogen bonding present in
salicylic acid reduces the frequency of the carbonyl stretching
the case of NTH3 Salꢀ, the permeation of both ions is
+
strongly enhanced. Fig. 3 summarises the extent of permeation
over time for two types of PILs. Like NTH3+ Salꢀ, Bro+ Ibuꢀ
shows evidence of ion pairing; the IR data suggests full proton
transfer, but it sits well below the ideal line on the Walden
absorption, which is normally observed at 1760 cm .
ꢀ1 20 Fig. 2
+
plot. It shows the same permeation trends as NTH3 Salꢀ,
+
clearly illustrates that NTH3 Salꢀ permeates the model
membrane very rapidly, achieving saturation (100% relative
absorbance) in B20 h. However, when this compound is
achieving saturation in B20 h.
+
+
NBH3 Aceꢀ and NHH3 Aceꢀ are both highly ionised
PILs but, in contrast to those discussed above, they sit closer
to the ideal line on the Walden plot,13 indicating a lesser
tendency to form associated species. These PILs also permeate
through the model membrane, but significantly slower than
the associated PILs. As before, our hypothesis is that this
arises from the transport of neutral species through the
membrane, but in these less associated PILs this process is
slower because the concentration of the neutral species is
considerably lower. To put this on a more quantitative basis,
Table 1 summarises the time for these compounds to permeate
to 50% saturation (t50). There is a clear distinction between the
associated and less associated PILs.
A molecule will permeate the membrane well if it is highly
soluble in the membrane and if it diffuses rapidly within it.
Further analysis of these factors for the two types of PILs
studied assists in understanding their different permeation
behaviour. Experimental values of relative absorbance intensities
Fig. 2 Membrane transport of NTH3+ Salꢀ (pure and as a saturated
PG solution) and its starting materials as saturated PG solutions.
c
11430 Chem. Commun., 2011, 47, 11429–11431
This journal is The Royal Society of Chemistry 2011

View Online
low diffusion coefficient. An IL form of the actives does not
necessarily address the solubility factor per se, other than the
fact that the active ions are presented at very high concentra-
tions. However, in cases where neutral species are present in
the IL, the solubility in a membrane such as that used here
would be enhanced. Solubility measurements in model mem-
branes such as these would allow determination of the rela-
tionship between permeability and structure to be understood
more fully and thus enable IL design. However, such measure-
ments are difficult to achieve accurately.
In summary, we have shown that protic pharmaceutically
active ILs are, in some cases, rapidly transported through a
model membrane, most likely as hydrogen bonded complexes.
We hypothesize that these paired compounds behave more like
‘‘neutral’’ species and hence cross the membrane faster than
the more ‘‘ionic’’ drugs. It is interesting to note that hydrogen
bonded complexes of un-ionised acid–base mixtures have also
been found to display unusual properties recently.15 One of the
advantages of this protic IL approach is that active formula-
tions can be designed that will be more membrane diffusive in
a way that is not otherwise possible.
Fig. 3 Membrane diffusion of the various types of Protic ILs.
against time were fitted using a previously reported membrane
diffusive flux equation:12a,b
!
1
n
2
X
A
4
ðꢀ1Þ
ꢀDð2n þ 1Þ p2t
¼ 1 ꢀ
ꢁ exp
ð1Þ
A
1
p
n¼0 2n þ 1
4h2
where A is the absorbance intensity at time t, D the permeant
diffusion coefficient, h the film thickness and AN the intensity
of the permeant peak corresponding to the saturation of the
membrane. Table 1 summarises the diffusion values obtained
from a fit of eqn (1) for the compounds studied. A single
diffusion coefficient did not provide an adequate fit to the data
Notes and references
1 M. Freemantle, An Introduction to Ionic Liquids, RSC Publishing
2009, Cambridge, UK, 2010.
2 R. D. Rogers, D. T. Daly, R. P. Swatloski, W. L. Hough,
J. H. Davis, M. Smiglak, J. Pernak and S. K. Spear, Application:
WO 2007044693, The University of Alabama, USA, 2007, p. 199.
3 (a) J. Stoimenovski, D. R. MacFarlane, K. Bica and R. D. Rogers,
Pharm. Res., 2010, 27, 521–526; (b) R. Ferraz, L. C. Branco,
C. Prudencio, J. P. Noronha and Z. Petrovski, ChemMedChem,
2011, 6, 975–985.
4 P. Heinrich Stahl and C. G. Wermuth, Handbook of Pharmaceutical
Salts; Properties, Selection, and Use, VHCA and Wiley-VCH, 2008.
5 (a) W. L. Hough, M. Smiglak, H. Rodriguez, R. P. Swatloski,
S. K. Spear, D. T. Daly, J. Pernak, J. E. Grisel, R. D. Carliss,
M. D. Soutullo, J. J. H. Davis and R. D. Rogers, New J. Chem.,
2007, 31, 1429–1436; (b) W. L. Hough and R. D. Rogers, Bull.
Chem. Soc. Jpn., 2007, 80, 2262–2269; (c) P. M. Dean,
J. Turanjanin, M. Yoshizawa-Fujita, D. R. MacFarlane and
J. L. Scott, Cryst. Growth Des., 2009, 9, 1137–1145.
6 K. Bica, C. Rijksen, M. Nieuwenhuyzen and R. D. Rogers, Phys.
Chem. Chem. Phys., 2010, 12, 2011–2017.
7 J. Cybulski, A. Wisniewska, A. Kulig-Adamiak, Z. Dabrowski,
T. Praczyk, A. Michalczyk, F. Walkiewicz, K. Materna and
J. Pernak, Tetrahedron Lett., 2011, 52, 1325–1328.
8 W. L. Hough-Troutman, M. Smiglak, S. Griffin, W. M. Reichert,
I. Mirska, J. Jodynis-Liebert, T. Adamska, J. Nawrot, M. Stasiewicz,
R. D. Rogers and J. Pernak, New J. Chem., 2009, 33, 26–33.
9 J. Stoimenovski, P. M. Dean, E. I. Izgorodina and D. R. MacFarlane,
Faraday Discuss., 2011, DOI: 10.1039/C1FD00071C.
10 P. K. Wotton, B. Moellgaard, J. Hadgraft and A. Hoelgaard, Int.
J. Pharm., 1985, 24, 19–26.
11 C. H. Purdon, C. G. Azzi, J. Zhang, E. W. Smith and H. I. Maibach,
Crit. Rev. Ther. Drug Carrier Syst., 2004, 21, 97–132.
12 (a) V. Tantishaiyakul, N. Phadoongsombut, W. Wongpuwarak,
J. Thungtiwachgul, D. Faroongsarng, K. Wiwattanawongsa and
Y. Rojanasakul, Int. J. Pharm., 2004, 283, 111–116;
(b) A. C. Watkinson, H. Joubin, D. M. Green, K. R. Brain and
J. Hadgraft, Int. J. Pharm., 1995, 121, 27–36; (c) W. J. McAuley,
K. T. Mader, J. Tetteh, M. E. Lane and J. Hadgraft, Eur. J. Pharm.
Sci., 2009, 38, 378–383.
+
(including up to an n = 3 fit), except for the NBH3 Aceꢀ
case. However, introducing a second component, involving a
second diffusion coefficient, provided a very good fit to the
data (see ESI), as also observed by Dias et al.14 It appears
therefore that there are two processes contributing to the
diffusion of these compounds (D1 and D2, Table 1), one
approximately an order of magnitude slower than the other.
In the work of Dias et al., it was suggested that the second
process arose from diffusion through a membrane already
swollen by the permeant. It is interesting to note that the
diffusion processes being observed here are relatively rapid, in
some cases only two orders of magnitude slower than typical
diffusion coefficients of ions in water.
From Fick’s equation the overall flux, F, of molecules
through the membrane is determined by F = P.CIL where P
is the permeation coefficient given by P = DK/h and CIL is the
concentration of the IL, K the partition coefficient of the
species between the liquid phase and the membrane and h is
the thickness of the membrane. The appearance of the parti-
tion coefficient in these relationships indicates the role of the
intrinsic solubility of the permeant in the membrane. It is
likely that the very low permeability of some species in Fig. 2
and 3 relate primarily to a very low solubility, rather than a
Table 1 Membrane diffusion coefficients and t50 values for PILs
D1
Fraction
D1 ꢁ10ꢀ8
D2 ꢁ 10ꢀ8 t50
(hours)
Ra
(cm2 sꢀ1
)
(cm2 sꢀ1
)
13 J. Stoimenovski, E. I. Izgorodina and D. R. MacFarlane, Phys.
Chem. Chem. Phys., 2010, 12, 10341–10347.
14 M. Dias, S. L. Raghavan and J. Hadgraft, Int. J. Pharm., 2001,
216, 51–59.
15 K. Bica, J. Shamshina, W. L. Hough, D. R. MacFarlane and
R. D. Rogers, Chem. Commun., 2011, 47, 2267–2269.
NTH3 Salꢀ 0.9982 0.69 ꢂ 0.03 1.3 ꢂ 0.1
10.1 ꢂ 1.7 B2
6.1 ꢂ 0.9 B2
1.5 ꢂ 0.2 B20
+
Bro+ Ibuꢀ
0.9933 0.50 ꢂ 0.02 0.5 ꢂ 0.1
NHH3 Aceꢀ 0.9977 0.74 ꢂ 0.03 0.17 ꢂ 0.01
+
NBH3 Aceꢀ 0.9981 —
0.097 ꢂ 0.003 —
B50
+
SalH in PG 0.9993 0.58 ꢂ 0.01 1.11 ꢂ 0.02
15.6 ꢂ 0.8 B2
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 11429–11431 11431