Influence of structural variation in room-temperature ionic liquids on the selectivity and efficiency of competitive alkali metal salt extraction by a crown ether

Article abstract of DOI:10.1021/ac010061v

An improved method for the preparation of 1-alkyl-3-methylimidazolium hexafluorophosphates provides a series of room-temperature ionic liquids (RTILs) in which the 1-alkyl group is varied systematically from butyl to nonyl. For competitive solvent extraction of aqueous solutions of alkali metal chlorides with solutions of dicyclohexano-18-crown-6 (DC18C6) in these RTILs, the extraction efficiency generally diminished as the length of the 1-alkyl group was increased. Under the same conditions, extraction of alkali metal chlorides into solutions of DC18C6 in chloroform, nitrobenzene, and 1-octanol was undetectable. The extraction selectivity order for DC18C6 in the RTILs was K⁺ > Rb⁺ > Cs⁺ > Na⁺ > Li⁺. As the alkyl group in the RTIL was elongated, the K⁺/Rb⁺ and K⁺/Cs⁺ selectivities exhibited general increases with the larger enhancement for the latter. For DC18C6 in 1-octyl-3-methylimidazolium hexafluorophosphate, the alkali metal cation extraction selectivity and efficiency were unaffected by variation of the aqueous-phase anion from chloride to nitrate to sulfate.

Full text of DOI:10.1021/ac010061v

Anal. Chem. 2001, 73, 3737-3741
In flu e n c e o f S t ru c t u ra l Va ria t io n in
Ro o m -Te m p e ra t u re Io n ic Liq u id s o n t h e S e le c t ivit y
a n d Effic ie n c y o f Co m p e t it ive Alk a li Me t a l S a lt
Ex t ra c t io n b y a Cro w n Et h e r
Sa ngki Chun, Se rge i V. Dzyuba , a nd Ric ha rd A. Ba rts c h*
Department of Chemistry and Biochemistry, Texas Tech University, Lubbock, Texas 79409-1061
fonyl]amides by dicyclohexano-18-crown-6 (DC18C6) (Figure 1).
Subsequently, Rogers and co-workers8 reported the extraction of
sodium, cesium, and strontium nitrates from aqueous solutions
into 1-butyl-, 1-hexyl-, and 1-octyl-3-methylimidazolium hexafluo-
rophosphates by 18-crown-6, DC18C6, and 4,4′(5′)-di(tert-butyl-
cyclohexano)-18-crown-6. Very recently, Rogers and co-workers9
described “task specific” RTILs containing metal ion chelating
units and their use in the solvent extraction of cadmium(II) and
mercury(II) chlorides.
We now report an improved preparation of 1-alkyl-3-methylimi-
dazolium hexafluorophosphates with alkyl ) butyl f nonyl
(Figure 1), their physical characterization, and an investigation
of these RTILs as diluents for competitive solvent extractions of
alkali metal salts by DC18C6.
An improved method for the preparation of 1 -alkyl-3 -
methylimidazolium hexafluorophosphates provides a se-
ries of room-temperature ionic liquids (RTILs) in which
the 1 -alkyl group is varied systematically from butyl to
nonyl. For competitive solvent extraction of aqueous
solutions of alkali metal chlorides with solutions of
dicyclohexano-1 8 -crown-6 (DC1 8 C6 ) in these RTILs, the
extraction efficiency generally diminished as the length
of the 1 -alkyl group was increased. Under the same
conditions, extraction of alkali metal chlorides into solu-
tions of DC1 8 C6 in chloroform, nitrobenzene, and 1 -oc-
tanol was undetectable. The extraction selectivity order
for DC1 8 C6 in the RTILs was K+ > Rb+ > Cs+ > Na+
>
Li+. As the alkyl group in the RTIL was elongated, the K+/
Rb+ and K+/ Cs+ selectivities exhibited general increases
with the larger enhancement for the latter. For DC1 8 C6
in 1 -octyl-3 -methylimidazolium hexafluorophosphate, the
alkali metal cation extraction selectivity and efficiency
were unaffected by variation of the aqueous-phase anion
from chloride to nitrate to sulfate.
EXPERIMENTAL SECTION
Reagents. Inorganic and organic compounds were reagent-
grade commercial products and were used as received unless
noted otherwise. Dicyclohexano-18-crown-6 (a mixture of the cis-
syn-cis and cis-anti-cis isomers) was obtained from Aldrich.
Deionized water was prepared by passing distilled water through
three Barnstead D8922 combination cartidges in series. Before
use in solvent extraction, chloroform was washed with deionized
water to remove the ethanol stabilizer and saturate the chloroform
with water. Nitrobenzene was distilled under reduced pressure
and washed with deionized water. The 1-octanol was washed with
distilled water.
Apparatus. The 1H NMR spectra were taken in acetone-d6
(∼0.1 M solutions) with a Bruker AF-200 NMR spectrometer.
Proton chemical shifts are reported downfield from TMS. IR
spectra were recorded as films between NaCl plates with a Perkin-
Elmer model 1600 FT-IR spectrophotometer. UV-visible spectra
were measured with a Shimadzu UV-visible spectrophotometer.
Melting points and glass transition temperatures were determined
with a Shimadzu DSC-50 differential scanning calorimeter and a
LTC-50 low-temperature assembly. Alkali metal cation concentra-
tions in aqueous solutions were determined with a Dionex DX-
120 ion chromatograph using a Dionex CS12A column. For the
solvent extraction studies, a Glas-Col multipulse vortex mixer and
a Clay Adams Compact II centrifuge were used to mix and
separate the organic and aqueous solutions, respectively. The
extractions were performed in 15-mL, metal-free, polypropylene
centrifuge tubes (Elkay), and samples were removed with Hamil-
Room-temperature ionic liquids (RTILs) are attracting ever-
increasing attention as solvents and catalysts for a wide variety
of organic reactions.1,2 Applications of RTILs in separation pro-
cesses have received much less attention.3-9 When employed in
solvent extraction processes, the negligible vapor pressure and
low flammability of a RTIL are important advantages over
conventional organic diluents. For metal ion separations, Dai and
co-workers6 found large distribution coefficient values for extrac-
tion of strontium nitrate from aqueous solutions into disubstituted
imidazolium hexafluorophosphates and bis[(trifluoromethyl)sul-
(1) Seddon, K. R. J. Chem. Technol. Biotechnol. 1 9 9 7 , 68, 351.
(2) Welton, T. Chem. Rev. 1 9 9 9 , 99, 2071.
(3) Huddleston, J. G.; Willauer, H. D.; Swatloski, R. P.; Visser, A. E.; Rogers, R.
D. J. Chem. Soc., Chem. Commun. 1 9 9 8 , 1765,
(4) Blanchard, L. A.; Hancu, D.; Beckman, E. J.; Brennecke, J. F. Nature 1 9 9 9 ,
399, 28.
(5) Armstrong, D. W.; He, L.; Liu, Y.-S. Anal. Chem. 1 9 9 9 , 71, 3873.
(6) Dai, S.; Ju, Y. H.; Barnes, C. E. J. Chem. Soc., Dalton Trans. 1 9 9 9 , 1201.
(7) Visser, A. E.; Swatloski, R. P.; Rogers, R. D. Green Chem. 2 0 0 0 , 1, 1.
(8) Visser, A. E.; Swatloski, R. P.; Reichert, W. M.; Griffin, S. T.; Rogers, R. D.
Ind. Eng. Chem. Res. 2 0 0 0 , 39, 3596.
(9) Visser, A. E.; Swatloski, R. P.; Reichert, W. M.; Mayton, R.; Sheff, S.;
Wierzbicki, A.; Davis, J. H., Jr.; Rogers, R. D. J. Chem. Soc., Chem. Commun.
2 0 0 1 , 135.
10.1021/ac010061v CCC: $20.00 © 2001 American Chemical Society
Published on Web 06/21/2001
Analytical Chemistry, Vol. 73, No. 15, August 1, 2001 3737

Hz), 1.89 (2H, pen, J ) 7.6 Hz), 3.98 (3H, s), 4.28 (2H, t, J ) 4.0
Hz), 7.58 (1H, t, J ) 1.8 Hz), 7.63 (1H, t, J ) 1.8 Hz), 8.76 (1H, s);
IR 3160, 2965, 1467, 836 cm-1
.
3 -Methyl-1 -pentylimidazolium hexafluorophosphate: 66%
1
yield; H NMR δ 0.89 (3H, m), 1.37 (4H, m), 1.96 (2H, m), 4.04
(3H, s), 4.34 (2H, t, J ) 7.2 Hz), 7.68 (1H, t, J ) 1.8 Hz), 7.74 (1H,
t, J ) 1.8 Hz), 8.94 (1H, s); IR 3170, 2961, 1576, 1458, 835 cm-1
.
Anal. Calcd for C9H17F6N2P: C, 36.25; H, 5.75; N, 9.39. Found: C,
36.24; H, 6.04; N, 9.48.
1 -Hexyl-3 -methylimidazolium hexafluorophosphate: 70%
1
yield; H NMR δ 0.84 (3H, m), 1.39 (6H, m), 1.96 (2H, m), 4.02
Fig u re 1 . Structures of dicyclohexano-18-crown-6 and 1-alkyl-3-
(3H, s), 4.32 (2H, t, J ) 7.2 Hz), 7.64 (1H, t, J ) 1.8 Hz), 7.70 (1H,
methylimidazolium hexafluorophosphates.
t, J ) 1.8 Hz), 8.87 (1H, s); IR 3169, 2933, 1576, 1458, 836 cm-1
.
ton GasTight syringes. Elemental analysis was performed by
Desert Analytics Laboratory of Tucson, AZ.
Anal. Calcd for C10H19F6N2P: C, 38.46; H, 6.13; N, 8.97. Found:
C, 38.60; H, 6.28; N, 8.97.
General P rocedure for the P reparation of 1 -Alkyl-3 -
methylimidazolium Bromides.10 A flask containing 1-methylimi-
dazole (60 mmol) and a primary alkyl bromide (60 mmol) was
placed in a silicone oil bath. With magnetic stirring, the bath was
heated to 140 °C during a 10-min period. In the latter stages of
this heating, an exothermic reaction took place forming an
emulsion that disappeared after a few minutes to form a transpar-
ent, golden, slightly viscous solution. When the emulsion disap-
peared, the flask was removed from the oil bath and the contents
were allowed to stir and cool in the air for 10 min. The flask was
returned to the 140 °C oil bath for 10-15 min. After cooling to
room temperature, the hygroscopic 1-alkyl-3-methylimidazolium
bromide was obtained in nearly quantitative yield and of sufficient
purity for use directly in the preparation of the corresponding
hexafluorophosphate salt.
1-Heptyl-3-methylimidazolium hexafluorophosphate: 78%
1
yield; H NMR δ 0.87 (3H, m), 1.35 (8H, m), 1.97 (2H, m), 4.03
(3H, s), 4.33 (2H, t, J ) 7.2 Hz), 7.65 (1H, t, J ) 1.8 Hz), 7.72 (1H,
t, J ) 1.8 Hz), 8.90 (1H, s); IR 3166, 2929, 1574, 1463, 834 cm-1
.
Anal. Calcd for C11H21F6N2P: C, 40.49; H, 6.49; N, 8.59. Found:
C, 40.63; H, 6.63; N, 8.59.
3 -Methyl-1 -octylimidazolium hexafluorophosphate: 85%
yield; 1H NMR δ 0.87 (3H, m), 1.33 (10H, m), 1.95 (2H, m), 4.04
(3H, s), 4.35 (2H, t, J ) 7.4 Hz), 7.69 (1H, t, J ) 1.8 Hz), 7.75 (1H,
t, J ) 1.8 Hz), 8.99 (1H, s); IR 3168, 2928, 1575, 1460, 835 cm-1
.
Anal. Calcd for C12H23F6N2P: C, 42.35; H, 6.81; N, 8.23. Found:
C, 42.59; H, 6.57; N, 8.26.
3 -Methyl-1 -nonylimidazolium hexafluorophosphate: 87%
yield; 1H NMR δ 0.87 (3H, m), 1.36 (12H, m), 1.99 (2H, m), 4.05
(3H, s), 4.36 (2H, t, J ) 7.2 Hz), 7.70 (1H, t, J ) 1.8 Hz), 7.76 (1H,
General P rocedure for the P reparation of 1 -Alkyl-3 -
methylimidazolium Hexafluorophosphates. HPF6 (22.2 mL of
60% solution in water, 0.15 mol) was added slowly to a magnetically
stirred solution of the 1-alkyl-3-methylimidazolium bromide (0.15
mol) in 30 mL of water in a plastic bottle cooled in an ice bath.
When the addition was completed, the mixture was stirred for 10
min in the ice bath and then for 10 min at room temperature.
The mixture was poured into a separatory funnel containing Et3N
(20.9 mL, 0.15 mol) in 200 mL of deionized water. The mixture
was shaken repeatedly, and the layers were separated. The organic
layer was washed with 200 mL of warm, deionized water. (As
determined by ion chromatographic analysis of the contacting
aqueous phase, the 1-alkyl-3-methylimidazolium hexafluorophos-
phate was usually contaminated at this stage by significant levels
of alkali metal salts from the HPF6.) The organic layer was
dissolved in dichloromethane (100 mL), and deionized water (200
mL) was added. After vigorous, magnetic stirring of the mixture
for 30 min, the water layer was replaced with 200 mL of fresh
deionized water and the mixture was vigorously stirred for 30 min.
The organic layer was separated, and the dichloromethane was
evaporated in vacuo. Residual water in the 1-alkyl-3-methylimida-
zolium hexafluorophosphate was removed by azeotropic distilla-
tion with benzene using a Dean-Stark trap. The benzene was
evaporated in vacuo and the oil was dried in vacuo (oil pump) at
110-130 °C for 4-8 h.
t, J ) 1.8 Hz), 8.98 (1H, s); IR 3170, 2926, 1575, 1466, 832 cm-1
.
Anal. Calcd for C13H25F6N2P: C, 44.07; H, 7.11; N, 7.91. Found:
C, 44.38; H, 7.32; N, 7.94.
P hase-Transition Temperature Determination. A weighed
amount of the 1-alkyl-3-methylimidazolium hexafluorophosphate
(2-16 mg) was sealed under air in an aluminum crucible. The
temperature in the DSC instrument was lowered to -130 to -140
°C by pouring liquid nitrogen into the low-temperature assembly.
After all of the liquid nitrogen had evaporated from the cooling
unit and the temperature started to rise, heating at 30 °C/ min
was initiated.
Density Measurements. A 1.0-mL volumetric flask was filled
to the mark with the 1-alkyl-3-methylimidazolium hexafluorophos-
phate at room temperature and weighed. A heat gun was used to
remove air bubbles in the sample.
Solubility in Water. In a 15-mL polypropylene centrifuge tube,
1.00 g of the RTIL and 2.5 mL of deionized water were shaken on
a vortex mixer for 30 min and centrifuged for 10 min. A 10-µL
aliquot of the aqueous phase was removed with a microsyringe
and diluted to 2.5 mL with deionized water. The absorbance of
this solution at 211 nm was measured and compared with that
obtained from dissolving a weighed amount (1-9 mg) of the RTIL
in 2.5 mL of deionized water.
Competitive Solvent Extraction of Alkali Metal Salts from
Water by DC1 8 C6 in Chloroform, Nitrobenzene, and 1 -Oc-
tanol. A mixture of 2.0 mL of 20 mM DC18C6 in chloroform,
nitrobenzene, or 1-octanol and 5.0 mL of a 2.0 mM (in each)
aqueous solution of five alkali metal chlorides in a 15-mL
1 -Butyl-3 -methylimidazolium hexafluorophosphate:3 64%
yield; 1H NMR δ 0.93 (3H, t, J ) 7.2 Hz), 1.36 (2H, sept, J ) 7.5
(10) Dzyuba, S. V.; Bartsch, R. A. J. Heterocycl. Chem. 2 0 0 1 , 38, 265.
3738 Analytical Chemistry, Vol. 73, No. 15, August 1, 2001

polypropylene centrifuge tube was shaken on a vortex mixer for
5 min (10 min with nitrobenzene). The tube was centrifuged for
5 min (10 min with nitrobenzene) to promote phase separation.
A 1.5-mL portion of the organic phase was removed with a syringe
and added to 3.0 mL of deionized water in a new centrifuge tube.
The tube was shaken on a vortex mixer for 5 min (10 min with
nitrobenzene) and centrifuged for 5 min (10 min with nitroben-
zene). A 1.0-mL portion of the aqueous phase was removed with
a syringe and transferred to a 10-mL volumetric flask. The flask
was filled to the mark with deionized water, and the alkali metal
cation concentrations were determined by ion chromatography.
Competitive Solvent Extraction of Alkali Metal Salts from
Water into RTIL Solutions of DC1 8 C6 . For each RTIL, the
competitive solvent extraction experiment was performed in
triplicate. A mixture of 1.00 g of a 20 mM solution of DC18C6 in
the RTIL and 2.5 mL of a 2.0 mM (in each) aqueous solution of
the alkali metal salts in a 15-mL polypropylene centrifuge tube
was shaken on a vortex mixer for 30 min, centrifuged for 10 min,
shaken on a vortex mixer for 30 min, and centrifuged for 10 min.
A 1.0-mL portion of the aqueous phase was removed with a syringe
and transferred to a 10-mL volumetric flask. The flask was filled
to the mark with deionized water, and the alkali metal cation
concentrations were determined by ion chromatography. Control
experiments were conducted in the absence of DC18C6 to correct
for alkali metal salt extraction by the RTIL itself.
Ta b le 1 . De n s it ie s , S o lu b ilit ie s in Wa t e r, a n d P h a s e
Tra n s it io n s o f 1 -Alk yl-3 -m e t h ylim id a zo liu m
He x a flu o ro p h o s p h a t e s
1-alkyl
group
density
(g/ mL)
water solub
(g/ 100 mL)
midpoint of glass
∆Cp
(mJ/ s‚K)
transition (°C)a
butyl
pentyl
hexyl
heptyl
octyl
1.363
1.333
1.307
1.274
1.237
1.202
1.88
1.23
0.75
-77
-80
-80
-84
-71
-66
14b
0.18
0.16
0.13
0.23
0.07
0.02
5.8c
0.37
∼0.20
nonyl
∼0.15
a
Heating rate of 30 °C/ min. b Melting point. c Units are ∆H (kJ/ mol).
zolium hexafluorophosphates based upon two-step reactions from
1-methylimidazole were 64-87% and increased as the alkyl group
of the 1-alkyl group was elongated.
P hysical P roperties of 1-Alkyl-3-methylimidazolium Hexa-
fluorophosphates. The data in Table 1 show that the densities
and water solubilities of the 1-alkyl-3-methylimidazolium hexafluo-
rophosphates decrease monotonically as the 1-alkyl group is
lengthened.
Melting points and/ or glass transition temperatures for the
1-alkyl-3-methylimidazolium hexafluorophosphates with 1-alkyl
groups of butyl f nonyl were determined by DSC and are
presented in Table 1. Although a glass transition temperature of
6 °C has been reported for 1-butyl-3-methylimidazolium hexafluo-
rophosphate,11 we were unable to observe any phase transition
in the region of 0-10 °C for several measurements made with
different samples of this RTIL.
RESULTS AND DISCUSSION
Synthesis of 1 -Alkyl-3 -methylimidazolium Hexafluoro-
phosphates. Although hexafluorophosphate-based RTILs have
been utilized as solvents in many processes,1,2 no convenient
method has been reported for the preparation of such solvents in
large quantities and with sufficient purity for analytical studies.
Preparation of 1-butyl-3-methylimidazolium hexafluorophos-
phate by treatment of 1-butyl-3-methylimidazolium chloride with
HPF6 using a published procedure3 followed by washing several
times with deionized water to remove contaminating HCl afforded
only 35-40% yields of the pure RTIL (free from alkali metal cation
contaminants by ion chromatography). Since this RTIL has some
water solubility, the low yield is attributed to losses during the
washing steps. Replacing the HPF6 with KPF6 gave the crude RTIL
in 90-95% yield. However, repeated washings with deionized water
to remove contaminating potassium chloride reduced the yield
of the pure RTIL to approximately 50%. In view of such difficulties,
a new procedure was developed for the synthesis of 1-alkyl-3-
methylimidazolium hexafluorophosphates.
An aqueous solution of the 1-alkyl-3-methylimidazolium bro-
mide, formed by the neat reaction of 1-methylimidazole with
equimolar 1-bromoalkane at 140 °C,10 in a plastic bottle was stirred
magnetically in an ice bath and 1 equiv of HPF6 (60% aqueous
solution) was added slowly. After stirring in the ice bath and then
at room temperature, the contents were transferred to a separatory
funnel containing water and 1 equiv of Et3N. The 1-alkyl-3-
methylimidazolium hexafluorophosphate was separated, washed
with water, and dissolved in dichloromethane. The solution was
washed with water, and the dichloromethane was evaporated in
vacuo. Traces of water were removed from the residue by
azeotropic distillation with benzene using a Dean-Stark trap. The
benzene was evaporated in vacuo, and the resultant oil was dried
at elevated temperature in vacuo. Yields of 1-alkyl-3-methylimida-
Competitive Solvent Extraction of Alkali Metal Chlorides
by DC18C6 in Chloroform, Nitrobenzene, and 1-Octanol. For
comparison with solvent extraction results obtained with the
RTILs, competitive solvent extractions of alkali metal chlorides
by DC18C6 were performed in the freqently encountered organic
diluents of chloroform, nitrobenzene, and 1-octanol. From contact
of 2.0 mM (in each) aqueous solutions of the five alkali metal
chlorides with a 20 mM solutions of DC18C6 in chloroform,
nitrobenzene, and 1-octanol, transfer of alkali metal chlorides into
the organic phase was undetectable. This low extraction efficiency
is attributed to the high hydration energy of chloride ion.12
Competitive Solvent Extraction of Alkali Metal Chlorides
by DC1 8 C6 in 1 -Alkyl-3 -methylimidazolium Hexafluoro-
phosphates. To assess the levels of alkali metal chloride extrac-
tions by the RTIL alone, solvent extractions of aqueous solutions
containing 2.0 mM concentrations of each of the five alkali metal
chlorides were conducted with no DC18C6 in the organic phase.
For ease of comparison, the alkali metal cation concentrations in
the RTILs shown in Figure 2 are given as percent metal salt
loadings of the crown ether if it had been present in the diluent.
As can be seen, the levels of alkali metal chloride transfer into
the RTIL are appreciable for 1-butyl-3-methylimidazolium hexaflu-
orophosphate and diminish rapidly as the 1-alkyl group in the RTIL
is elongated. For all six RTILs, the extraction selectivity is Cs+
>
Rb+ > K+ > Li+ ∼ Na+. This is consistent with the relative
hydrophobicities of the alkali metal cations.
(11) Carmichael, A. J.; Hardacre, C.; Holbrey, J. D.; Seddon, K. R.; Nieuwen-
huyzen, M. Proc. Electrochem. Soc. 99-41, Molten Salts XII 2 0 0 0 , 209.
(12) Smith, D. W. J. Chem. Educ. 1 9 7 7 , 54, 540.
Analytical Chemistry, Vol. 73, No. 15, August 1, 2001 3739

Fig u re 2 . Influence of 1-alkyl group variation on efficiency of
competitive alkali metal cation extraction from aqueous solutions into
1-alkyl-3-methylimidazolium hexafluorophosphates in the absence of
dicyclohexano-18-crown-6.
Fig u re 4 . Influence of the 1-alkyl group variation on the K+/Cs+
and K+/Rb+ selectivies in competitive alkali metal cation extraction
from aqueous solutions by dicyclohexano-18-crown-6 in 1-alkyl-3-
methylimidazolium hexafluorophosphates.
1-hexyl-3-methylimidazolium hexafluorophosphate as a solvent for
the competitive solvent extraction of alkali metal cations by
DC18C8 is not apparent at this stage. (Unusual behavior of 1-hexyl-
3-methylimidazolium ion has been noted earlier by others.14)
The influence of varying the 1-alkyl group of the 1-alkyl-3-
methylimidazolium hexafluorophosphate diluent on the K+/ Cs+
and K+/ Rb+ selectivities for competitive alkali metal cation
extractions by DC18C6 is depicted in Figure 4. With the exception
of 1-hexyl-3-methylimidazolium hexafluorophosphate, the K+/ Cs+
selectivity and, to a lesser extent, the K+/ Rb+ selectivity increase
as the 1-alkyl group of the RTIL is elongated. Thus, the decrease
in extraction efficiency (Figure 3) as the 1-alkyl group of the
1-alkyl-3-methylimidazolium hexafluorophosphate is lengthened
is coupled with an increase in extraction selectivity.
Fig u re 3 . Influence of 1-alkyl group variation on efficiency of
competitive alkali metal cation extraction from aqueous solutions by
dicyclohexano-18-crown-6 in 1-alkyl-3-methylimidazolium hexafluo-
rophosphates.
Competitive Solvent Extraction of Alkali Metal Salts by
DC1 8 C6 in 1 -Octyl-3 -methylimidazolium Hexafluorophos-
phate. For competitive solvent extractions of alkali metal salts
from aqueous solution into common organic solvents by crown
ethers, the identity of the anion influences both the extraction
efficiency and selectivity.15-17 To probe the counteranion influence
in extractions involving RTIL diluents, competitive extractions of
alkali metal nitrates and alkali metal sulfates from aqueous
solutions by DC18C6 in 1-octyl-3-methylimidizolium hexafluoro-
phosphate were performed. A comparison of the percent crown
ether loadings for the five alkali metal cations as the anion was
varied from chloride to nitrate to sulfate is presented in Table 2.
It is immediately apparent that the efficiencies and selectivities
of alkali metal cation extraction are unaffected by this variation
in the aqueous-phase anion. This differs markedly from results
reported for competitive alkali metal salt extractions by DC18C6
in chloroform and 1-octanol.16,17 For these molecular solvents, the
metal ion loading was strongly affected by anion variation. This
Results from competitive solvent extraction of 2.0 mM (in each)
aqueous solutions of the alkali metal chlorides with 20 mM
solutions of DC18C6 in the six RTILs are shown in Figure 3. (The
data are corrected for the levels of alkali metal cations that are
extracted by the RTIL alone.) Immediately apparent is the change
in extraction selectivity from that observed in the absence of
DC18C6 (Figure 2). The observed extraction selectivity of K+
>
Rb+ > Cs+ > Na+ g Li+ reflects the relative complexing abilities
of 18-crown-6 ligands for the alkali metal cations.13
With the exception of 1-hexyl-3-methylimidazolium hexafluo-
rophosphate, there is a smooth decrease in the extraction
efficiency for each alkali metal cation species as the 1-alkyl group
in the RTIL is elongated (Figure 3). Thus, increasing the
lipophilicity of the RTIL is found to diminish the propensity for
alkali metal cation extraction by DC18C6. The somewhat greater
alkali metal extraction efficiencies with 1-hexyl-3-methylimidazo-
lium hexafluorophosphate as the diluent than those expected from
the correlation established by the other 1-alkyl-3-methylimidazo-
lium hexafluorophosphates were reproducible for different batches
of this RTIL. The reason for the slightly anomalous behavior of
(14) Carmichael, A. J.; Seddon, K. R. J. Phys. Org. Chem. 2 0 0 0 , 13, 591.
(15) Yakshin, V. V.; Abashkim, V. M.; Laskorin, B. N. Dokl. Acad. Nauk SSSR
1 9 8 0 , 252, 373; Chem. Abstr. 1 9 8 0 , 93, 102126h.
(16) Olsher, U.; Hankins, M. G.; Kim, Y. D.; Bartsch, R. A. J. Am. Chem. Soc.
1 9 9 3 , 115, 3370.
(17) Hankins, M. G.; Bartsch, R. A.; Olsher, U. Solv. Extr. Ion Exch. 1 9 9 5 , 13,
983.
(13) Izatt, R. M.; Pawlak, K.; Bradshaw, J. S. Chem. Rev. 1 9 9 1 , 91, 1721.
3740 Analytical Chemistry, Vol. 73, No. 15, August 1, 2001

negligible extraction with customary organic diluents of chloro-
form, nitrobenzene, and 1-octanol. As the lipophilicity of the RTIL
is enhanced by lengthening the 1-alkyl group, the extraction
efficiency decreases, but the extraction selectivity increases. In
contrast to customary organic diluents, the efficiency and selectiv-
ity of alkali metal salt extraction from aqueous solutions into RTIL
solutions of DC18C6 is unaffected by variation of the counteranion
from chloride to nitrate to sulfate. Further exploration of the
intriguing features of ionic liquids in separation processes is
underway in our laboratories.
Ta b le 2 . Effe c t o f An io n o n P e rc e n t Lo a d in g fo r
Co m p e t it ive S o lve n t Ex t ra c t io n o f Alk a li Me t a l S a lt s
fro m Aq u e o u s S o lu t io n s b y Dic yc lo h e x a n o -1 8 -c ro w n -6
in 1 -Oc t yl-3 -m e t h ylim id a zo liu m He x a flu o ro p h o s p h a t e
loading (%)a,b
Li+
Na+
K+
Rb+
Cs+
chloride
nitrate
sulfate
0.2
0.2
0.2
0.4
0.4
1.0
22.8
22.3
22.7
13.6
13.2
13.6
4.8
4.6
5.0
a
Each extraction was conducted in triplicate. b Standard deviation
from the average was (0.3%.
ACKNOWLEDGMENT
This research was supported by the Texas Higher Education
Coordinating Board Advanced Research Program. We thank Dr.
Dominick J. Casadonte, Jr., for use of the DSC instrument
(purchased with a grant from the Dreyfus Foundation) and Dr.
Sindee L. Simon for helpful discussions. Presented at Pacifichem
2000, Honolulu, HI, December 14-19, 2000.
further underscores the uniqueness of RTILs as diluents in solvent
extraction.
SUMMARY
A convenient method for the preparation of high-purity 1-alkyl-
3-methylimidazolium hexafluorophosphates in two steps using
commercially available reagents is reported. Solutions of DC18C6
in these RTILs provide efficient extractions of alkali metal
chlorides from aqueous solutions under conditions that gave
Received for review January 16, 2001. Accepted May 1,
2001.
AC010061V
Analytical Chemistry, Vol. 73, No. 15, August 1, 2001 3741