Dual-host approach for liquid-liquid extraction of potassium fluoride/chloride via formation of an integrated 1-D polymeric complex

Article abstract of DOI:10.1039/c0cc03469j

Extraction of KF/KCl from aqueous solution by distinct cationic and anionic receptors is demonstrated. Structural studies of the extracted complexes showed 1D-polymers of the receptors containing respective ions in their cavities. The Royal Society of Che

Full text of DOI:10.1039/c0cc03469j

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 4721–4723
www.rsc.org/chemcomm
COMMUNICATION
Dual-host approach for liquid–liquid extraction of potassium fluoride/
chloride via formation of an integrated 1-D polymeric complexw
I. Ravikumar, Subrata Saha and Pradyut Ghosh*
Received 25th August 2010, Accepted 24th February 2011
DOI: 10.1039/c0cc03469j
Extraction of KF/KCl from aqueous solution by distinct cationic
and anionic receptors is demonstrated. Structural studies of the
extracted complexes showed 1D-polymers of the receptors
containing respective ions in their cavities.
and we isolated single crystals of [L2(F)(H2O)][N(Bu)4] (1) and
[L2(Cl)][N(Bu)4] (2) from water upon reaction of L2 with
n-Bu4NF and n-Bu4NCl respectively. Single crystal X-ray
studies of 1 showed recognition of mono hydrated Fꢀ inside
the C3v symmetric cleft of L2.z Fig. 1c shows that encapsulated
Fꢀ is bound strongly via three (N–Hꢁ ꢁ ꢁFꢀ) bonds and one
(O–Hꢁ ꢁ ꢁFꢀ) hydrogen bond with all three amide hydrogens of
L2 and encapsulated water molecule O4 respectively in a
distorted tetrahedral fashion (Fig. 7S and Table 5S, ESIw).
Further, the oxygen atom of the encapsulated water (O4) is in
strong contact with the all three -C6F5 moieties of L2 with
O4ꢁ ꢁ ꢁCg1, O4ꢁ ꢁ ꢁCg2 and O4ꢁ ꢁ ꢁCg3 (Cg1, Cg2 and Cg3 are the
centroids of three -C6F5 moieties) distances of 3.060, 3.161 and
3.169 A respectively (Fig. 1c). The strong H2Oꢁ ꢁ ꢁp contacts in
1 could be the consequence of ditopic recognition of Fꢀ and
H2O inside the small cavity of the receptor. The monotopic
recognition of Clꢀ inside the C3v symmetric cleft of L2 is
observed in case of 2 where Clꢀ has a strong hydrogen
bonding interaction with all three amide hydrogens of L2
(Fig. 10S, Table 6S, ESIw).
The extracted complexes of 3 and 4 were obtained by the
liquid–liquid extraction of an aqueous phase containing KF
(100 mM) or KCl (100 mM) with equimolar amounts of L1
and L2 (10 mM) in CHCl3. Details of extraction and crystall-
ization of 3 and 4 are given in the ESI.w Complex 3,
[L2(F)(H2O)][L1(K)] was subjected to various studies like
1H-NMR, 19F-NMR, IR, SEM-EDX, PXRD and Single
crystal X-ray diffraction.zSimilar experiments and character-
ization were carried out with the exception of 35Cl-NMR
instead of 19F-NMR studies for 4, [L2(Cl)(H2O)][L1(K)].z
Competitive extraction from aqueous solutions containing
mixtures of KF (100 mM) and KNO3 (100 mM) or KCl
Ion-pair recognition and extraction by synthetic hosts is an
area of general interest, with promising analytical, industrial
and environmental applications.1 In particular, the selective
liquid–liquid extraction of alkali metal salts is important
due to its applications in separation technology, production
of common living materials, medicine and maintenance of
osmotic balances in cells.2 Recently, Sessler et al. demon-
strated extraction of KF or KCl from water using polymeric
extractants with calixpyrroles and crown ether pendants.3 In
the literature, a dual-host (using separate anion and cation
hosts) approach has also been demonstrated for extraction of
different alkali salts.4 Fluoride recognition by synthetic
receptors mostly takes place in organic solvents, which greatly
limits the scope of their use,5 though fluoride recognition in
aqueous media is known in polyammonium cryptands.6
Herein we show a simple tripodal neutral receptor, L2 (which
has previously shown anion binding ability in organic
solvents),7 as a suitable receptor for encapsulation of Fꢀ/Clꢀ
in water that also acts as a partner in dual-host systems with
crown ether, L1 for extraction of KF/KCl from water. Further
we successfully demonstrate the single crystal X-ray structure
of the extracted species as integrated 1-D polymeric complexes
of KF/KCl. To the best of our knowledge this represents the
first dual-host system capable of liquid–liquid extraction of
KF from aqueous media and the first single crystal X-ray
structure of any extracted ion-pair in a dual host approach.
The 18-Crown-6, L1 and tripodal amide, L2 (Fig. 1a and b)
is our choice of cation and anion receptors respectively
because of high selectivity of L1 toward potassium ion,8 and
solubility of L2 in most of the common organic solvents even
in water. Water solubility of L2 gave us the opportunity to
attempt complexation of Fꢀ and Clꢀ in an aqueous medium
Department of Inorganic Chemistry, Indian Association for the
Cultivation of Science, 2A & 2B Raja S. C. Mullick Road,
Kolkata 700 023, India. E-mail: icpg@iacs.res.in
w Electronic supplementary information (ESI) available: Experimental
details. CCDC 701763, 710731, 732760 and 804542. For ESI and
crystallographic data in CIF or other electronic format see DOI:
10.1039/c0cc03469j
Fig. 1 (a) L1, (b) L2 and (c) single crystal structure of encapsulated
Fꢀ in the cavity of L2.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 4721–4723 4721

View Article Online
show the necessity of dual receptors towards the extraction of
the salt from aqueous media. To determine the effect of pH
towards KF/KCl extraction by L1 and L2, experiments were
carried out at three different pH, 4.5, 6.3 and 9.3. The
maximum extraction of KF (48%) and KCl (44%) is observed
at pH B6.0 whereas no extraction of KF and KCl is observed
at pH B9.0 (Table 10S, ESIw).
Energy-Dispersion X-ray (EDX) analyses were carried out
on the extracted complexes 3 and 4 (Fig. 18S, ESIw). In cases
of 3 and 4, the EDX spectrum shows peaks in addition to C, N
and O, that can be assigned to K (3.312 KeV), F (0.677 KeV)
and K (3.312 KeV), Cl (2.621 KeV), respectively indicating the
presence of KF/KCl which are absent in the case of L2. Since,
L2 contains pentafluorophenyl moieties, a separate peak for
fluoride is not observed in case of 3.
We have carried out a detailed IR studies for L2 and
complexes 1–4 (Fig. 21S, ESIw). The peak at 1660 cmꢀ1 in
L2 may be attributed to the -CQO (amide I) stretching
frequency. In 1 and 2 -CQO stretching frequency remains
unaltered whereas, 3 and 4 showed existence of two different
carbonyl stretching frequencies. In the case of 3, two different
-CQO stretching frequencies appear at 1683 and 1652 cmꢀ1
whereas in 4 corresponding frequencies appear at 1681 and
1655 cmꢀ1. This suggests that in these two complexes the
carbonyl groups are in two different environments.
Fig. 2 Partial 1H-NMR spectra of free L2 and complexes 1, 2, 3 and 4
in CDCl3 at 298 K. a and b denote the existence of n-Bu4N+ and K+
ꢁ
L1 cations in complexes 1, 2 and 3, 4, respectively.
(100 mM) and KNO3 (100 mM) has also resulted in 3 and 4
respectively. Fig. 2 shows 1H-NMR studies on L2 and
complexes 1–4. In 1 and 2 a large downfield chemical shift
of the -NH signal (Dd = 4.52 and 2.49 ppm respectively) w.r.t.
L2 is observed, indicative of halide encapsulation inside the
cleft of L2 and its strong interactions with the -NH protons. In
the cases of 3 and 4, a similar downfield shift of -NH signals
are observed as in the cases of 1 and 2 respectively, indicating
the presence of Fꢀ and Clꢀ encapsulated L2 in the extracted
products. Though, 1 and 3 showed similar shifts in the -NH
resonance of L2, a major noticeable upfield shift of 0.17 and
The single-crystal X-ray structure of 3z revealed the
formation of a 1-D coordination polymer composed of L1,
L2 and KF (Fig. 3) where K+ and Fꢀ are encapsulated inside
L1 and L2 respectively. In the case of 4 a similar 1-D
coordination polymer is obtained (Fig. 17S, ESIw). Structural
analyses showed that 3 and 4 are isostructural. Interestingly,
two out of the three amide carbonyl oxygen atoms of Fꢀ
encapsulated L2 form two coordinate covalent bonds with the
encapsulated K+ of L1, in the case of 3, which upon propagation
form the 1-D polymeric chain. In this metal–organic polymeric
complex, K+ is bound to all six oxygen atoms of L1 with
K+ꢁ ꢁ ꢁO distances B2.8 A with two more oxygen atoms
coordinating from the carbonyl groups of Fꢀ encapsulated
L2 with K+ꢁ ꢁ ꢁO distances 2.73 and 2.77 A, resulting in eight
coordinate covalent bonds. The presence of carbonyl groups in
different environments is also evident from the IR analysis.
The low energy band in IR analyses of 3 or 4 could be assigned
a
b
0.20 ppm of -CH2 and -CH2 resonances respectively in the
aliphatic region of L2 are observed in case of 3 whereas in case
of 1 -CH2 resonances are matched with L2. Similar upfield
shifts in the aliphatic region (0.16 and 0.21 ppm) are also
observed in 4 whereas in 2, no such shifts are present. The
-CH2 chemical shifts observed in 3 and 4 could be attributed to
the strong interaction of Fꢀ/Clꢀ encapsulated L2 with cationic
counter part of complexes. In 3 and 4, a new peak at 3.6 ppm
clearly indicated the presence of a crown-ether moiety in the
extracted complexes. Detailed analyses of 1H-NMR spectra of
3 and 4 upon calculating -CH2a or -CH2b integration of L2 to
the b protons of L1 indicated a 1 : 1 involvement of L1 and L2
(Fig. 12S and 14S, ESIw).
In the 19F-NMR and 35Cl-NMR studies (Fig. 16S, ESIw)
hydrated Fꢀ and Clꢀ resonances for KF and KCl in D2O are
observed at ꢀ125.08 and 0.00 ppm respectively. Complexes 1
and 3 both showed a sharp resonance with a downfield
chemical shift of 14.86 ppm that represents encapsulation of
Fꢀ in the cavity of L2 in both the cases. In 2 and 4 a 14.84 ppm
downfield shift of Clꢀ resonance is observed where Clꢀ signals
showed broadening, indicative of Clꢀ encapsulation and
complexation. As control experiments, extractions were also
performed in an analogous manner using L1 and L2,
separately. No quantifiable fluorine or chlorine signal in
19F- and 35Cl-NMR was observed in the organic phase when
L1 was used as an extractant. There is no observable change in
the -NH resonance in 1H-NMR of L2 obtained from the
organic phase when L2 is used alone as an extractant for KF
or KCl (Fig. 22S, 23S and 27S, ESIw). These studies indeed
Fig. 3 Single crystal X-ray structure of (a) monomeric L1ꢁL2ꢁKF
(complex 3) with partial atom-labeling and (b) a 1-D polymer of
complex 3. Non acidic hydrogen atoms and disordered crown ether
moieties are omitted for clarity.
c
4722 Chem. Commun., 2011, 47, 4721–4723
This journal is The Royal Society of Chemistry 2011

View Article Online
studies have been performed to establish a suitable mechanism
for the liquid–liquid extraction of KF/KCl.
PG thanks the Department of Science and Technology for
financial support through a Swarnajayanti Fellowship. SS
acknowledges CSIR, India for JRF. Single crystal X-ray
structures of complexes 1–4 were performed in the DST
funded National Single Crystal X-ray Facility at IACS.
Notes and references
Fig. 4 Proposed mechanism of KF/KCl extraction.
z The crystals of complexes 2, 3 and 4 are very weakly diffracting at
higher Bragg angles. This resulted in poor quality of the data set.
However, this is the best possible data set for these complexes.
to the carbonyl bound to potassium of L1. On the other hand,
the encapsulated fluoride is hydrogen bonded to all three
amide -NH of L2 with Nꢁ ꢁ ꢁFꢀ distances ranging from
2.66–2.73 A and to a molecule of water as observed in the
case of 1. The extracted complexes of bulk 1-D polymeric
complexes 3 and 4 are obtained in 48% and 44% yield
respectively when 0.1 mmol of each L1 and L2 are used for
extraction, see ESI.w
Experimental powder X-ray patterns of bulk 3 and 4
showed similar powder X-ray patterns simulated from the
corresponding single crystal X-ray data of these complexes
(Fig. 19S, ESIw), suggesting high purity of the bulk extracted
complexes.
Crystallographic data for complex
1
(CCDC 701763):
C43H52F16N5O4, M = 1006.88, monoclinic, space group P21/n, a =
16.3394(11), b = 13.3936(9), c = 23.0429(15) A, b = 104.689(2)1,
V = 4878.0(6) A3, Dc = 1.368 g cmꢀ3, Z = 4, l = 0.71073 A, T =
100(2) K, 28 988 reflections, 6895 independent (Rint = 0.0293), and
4200 observed reflections [I > 2s(I)], 617 refined parameters, R =
0.0665, wR2
=
0.1883. Complex
2
(CCDC 710731):
ꢀ
C43H51ClF15N5O3, M = 1006.34, triclinic, space group P1, a =
11.7018(18), b = 14.141(2), c = 15.737(2) A, a = 88.438(6),
b = 77.895(5), g = 74.960(5)1, V = 2458.0(7) A3, Dc = 1.360 g cmꢀ3
,
Z = 2, l = 0.71073 A, T = 298(2) K, 14 575 reflections, 4642
independent (Rint 0.0379), and 3128 observed reflections
=
[I > 2s(I)], 601 refined parameters, R = 0.0624, wR2 = 0.1437.
Complex 3 (CCDC 732760): C36H41F16KN4O10, M = 1032.83, mono-
clinic, space group P21/n, a = 15.591(7), b = 19.049(8), c = 16.196(7) A,
b = 93.954(6)1, V = 4799(4) A3, Dc = 1.430 g cmꢀ3, Z = 4, l =
The 2D-DOSY has been used as an independent method to
propose a suitable mechanism for the extraction of KF/KCl
from water (Fig. 20S, ESIw) through the measurement of
the diffusion coefficients. The 1H-NMR spectra of all the
combinations are shown in the horizontal projection. All
signals are well resolved and can be classified according to
their self-diffusion coefficients. The DOSY spectrum of an
acetonitrile solution of L1 + L2 (10 mM) shows two distinct
sets of resonances with different diffusion coefficients of 2.58 ꢂ
10ꢀ9 m2 sꢀ1 and 5.0 ꢂ 10ꢀ9 m2 sꢀ1. It is clearly evident from
the spectra, that the addition of KPF6 (10 mM) to this solution
showed a single set of resonances with a diffusion coefficient of
0.71073 A, T = 100(2) K, 26 265 reflections, 4111 independent (Rint
=
0.1175), and 2206 observed reflections [I > 2s(I)], 574 refined para-
meters, R = 0.0832, wR2 = 0.2168. Complex 4 (CCDC 804542):
C39H41ClF15KN4O10, M = 1085.29, monoclinic, space group P21/c,
a = 10.676(10), b = 24.66(2), c = 18.522(18) A, b = 98.094(11)1, V =
4827(8) A3, Dc = 1.491 g cmꢀ3, Z = 4, l = 0.71073 A, T = 298(2) K,
21 302 reflections, 3379 independent (Rint = 0.0777), and 2565
observed reflections [I > 2s(I)], 631 refined parameters, R =
0.0738, wR2 = 0.1931.
1 (a) J. W. Steed and J. L. Atwood, Supramolecular Chemistry:
Second Edition, Wiley, Chichester, U.K., 2009; (b) B. A. Moyer
and R. P. Singh, Fundamentals and Applications of Anion
Separations, ed., Kluwer Dordrecht, The Netherlands, 2004;
(c) K. Gloe, H. Stephan and M. Grotjahn, Chem. Eng. Technol.,
2003, 26, 1107; (d) P. A. Gale, Coord. Chem. Rev., 2003, 240, 191;
(e) J.-M. Lehn, Supramolecular Chemistry: Concepts and
Perspectives, VCH, Weinheim, Germany, 1995.
2 (a) G. Sweeney and A. Klip, Cell. Mol. Biol., 2001, 47, 363;
(b) A. R. Mansour and K. J. Takrouri, Chem. Eng. Commun.,
2007, 194, 803; (c) K. H. Park, L. Hernandez, S.-Q. Cai, Y. Wang
and F. Sesti, J. Biol. Chem., 2005, 280, 21893.
1.03 ꢂ 10ꢀ9 m2
s
ꢀ1. This indicates the formation of an
integrated co-ordination species between potassium encapsulated
L1 and L2, whereas there is no change in the -NH resonance of
the receptor. The addition of n-Bu4NF or n-Bu4NCl to the
above solution showed two sets of resonances with different
diffusion co-efficients which is attributed to the existence of
Bu4N+ groups (D = 1.334 ꢂ 10ꢀ9 m2 sꢀ1) and the integrated
species of L1 and L2 (D = 1.03 ꢂ 10ꢀ9 m2 sꢀ1) in solution,
further a downfield shift of the -NH resonances of L2,
indicated the encapsulation of Fꢀ/Clꢀ in its cavity.
3 A. Abdullah, D. J. Coady, S. K. Kim, A. Akar, C. W. Bielawski,
M. Marquez and J. L. Sessler, Angew. Chem., Int. Ed., 2008, 47,
9648.
4 (a) K. Kavallieratos, R. A. Sachleben, G. J. V. Berkel and
B. A. Moyer, Chem. Commun., 2000, 187; (b) K. Kavallieratos,
A. Danby, G. J. V. Berkel, M. A. Kelly, R. A. Sachleben,
B. A. Moyer and K. Bowman-James, Anal. Chem., 2000, 72, 5258;
(c) M. M. Murad, T. Hayashita, K. Shigemori, S. Nishizawa and
N. Teramae, Anal. Sci., 1999, 15, 1185; (d) G. Cafeo, G. Gattuso,
F. H. Kohnke, A. Notti, S. Occhipinti, S. Pappalardo and
M. F. Parisi, Angew. Chem., Int. Ed., 2002, 41, 2122.
5 M. Cametti and K. Rissanen, Chem. Commun., 2009, 2809.
6 (a) B. Dietrich, J.-M. Lehn, J. Guilhem and C. Pascard, Tetrahedron
Lett., 1989, 30, 4125; (b) Md. A. Hossain, J. M. Llinares, S. Mason,
P. Morehouse, D. Powell and K. Bowman-James, Angew. Chem.,
Int. Ed., 2002, 41, 2335.
From the above studies, we propose a suitable mechanism
for extraction of KF/KCl from water as depicted in Fig. 4.
Even though Dual-Hosts are employed for the extraction of
these salts, the potassium cation first assists the formation of
an integrated species of the dual receptors and participates in
the extraction process of fluoride or chloride from water.
In conclusion, a simple tripodal amide acts as a selective
fluoride/chloride binding receptor in an aqueous medium in
combination with a well known potassium binding crown
ether, a dual hosts approach, for the liquid–liquid extraction
of potassium fluoride/chloride even in the presence of nitrate.
Different solution state studies along with 2D-DOSY-NMR
7 I. Stibor, D. S. M. Hafeed, P. Lhotak, J. Hodacova, J. Koca and
M. Cajan, Gazz. Chim. Ital., 1997, 127, 673.
8 (a) C. J. Pedersen, J. Am. Chem. Soc., 1967, 89, 7017;
(b) J. M. Mahoney, A. M. Beatty and B. D. Smith, J. Am. Chem.
Soc., 2001, 123, 5847.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 4721–4723 4723