Coordination and extraction of mercury(ii) with an ionic liquid-based thione extractant

Article abstract of DOI:10.1039/c3dt50410g

A neutral thione extractant, 1,3-diethylimidazole-2-thione (C ₂C₂ImT), was prepared from an ionic liquid (IL), 1,3-diethylimidazolium acetate, and used within a hydrophobic ionic liquid, 1-ethyl-3-methylimidazolium bis(trifluoromethane)sulfonamide ([C ₂mim][NTf₂]), to extract Hg(ii) from aqueous solutions of HgCl₂ or Hg(OAc)₂. Investigations of the extraction mechanism, spectroscopic analyses of the extracted species, and crystallographic studies of the interactions of C₂C₂ImT with Hg(ii) are reported, including the first structurally characterized Hg-NTf₂ coordination compound, Hg(C₂C₂ImT)₂(NTf ₂)₂. Coordination complexes of the thione ligand with Hg(ii) show variability in coordination numbers and geometries with stoichiometry, suggesting that the extraction mechanism is dependent on the speciation of mercury in aqueous solution. HgCl₂ can form neutral, extractable complexes with the thione in aqueous solution. Hg(OAc)₂ dissociates on dissolution in water and Hg(ii) is extracted through a cation exchange mechanism involving [Hg(C₂C₂ImT) ₂]²⁺ ions. The precipitation of neutral mercury complexes from the IL following the extraction of excess mercury suggests a simple and unusual way to recycle the IL.

Full text of DOI:10.1039/c3dt50410g

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Coordination and extraction of mercury(II) with an ionic
liquid-based thione extractant†
Cite this: Dalton Trans., 2013, 42, 12908
a,b
a
a
a
Wenjuan Lu, Patrick S. Barber, Steven P. Kelley and Robin D. Rogers*
A neutral thione extractant, 1,3-diethylimidazole-2-thione (C
IL), 1,3-diethylimidazolium acetate, and used within a hydrophobic ionic liquid, 1-ethyl-3-methylimidazo-
lium bis(trifluoromethane)sulfonamide ([C mim][NTf ]), to extract Hg(II) from aqueous solutions of HgCl
or Hg(OAc) . Investigations of the extraction mechanism, spectroscopic analyses of the extracted species,
and crystallographic studies of the interactions of C ImT with Hg(II) are reported, including the first
structurally characterized Hg-NTf coordination compound, Hg(C ImT) (NTf . Coordination com-
2 2
C ImT), was prepared from an ionic liquid
(
2
2
2
2
2 2
C
2
2
C
2
2
2 2
)
plexes of the thione ligand with Hg(II) show variability in coordination numbers and geometries with
stoichiometry, suggesting that the extraction mechanism is dependent on the speciation of mercury in
Received 12th February 2013,
Accepted 14th April 2013
2
aqueous solution. HgCl can form neutral, extractable complexes with the thione in aqueous solution.
Hg(OAc)
2
dissociates on dissolution in water and Hg(II) is extracted through a cation exchange mechan-
DOI: 10.1039/c3dt50410g
2+
ism involving [Hg(C
2
C ImT)
2
2
]
ions. The precipitation of neutral mercury complexes from the IL following
www.rsc.org/dalton
the extraction of excess mercury suggests a simple and unusual way to recycle the IL.
reaction between dialkylimidazolium acetate ILs and elemen-
tal sulfur, where the IL acts as a reservoir of reactive car-
Introduction
1
8
Separations of metal ions using ionic liquids (ILs), generally benes. Our interest in separation and extraction of heavy
1
10
defined as salts that melt below 100 °C, have attracted interest metals and studies of metal–thione complexes (with d and
ever since the use of extractants to partition metal ions from first row transition metals) led us to investigate the ability of
2
,3
aqueous media to IL was reported in 1999. Further studies imidazole-2-thione to extract mercury from aqueous solution.
have shown that ion exchange is the most commonly reported The ability of heterocyclic thiones to form neutral 1 : 1 and
mechanism for the extraction of metal ions into ILs. In ion 1 : 2 metal-to-ligand complexes with mercury(II) halides in the
4
–8
−
−
−
exchange, the IL ions enter the aqueous phase to maintain form HgX
charge balance as the metal ions or charged metal-extractant has been known for years.
2 2 2
L and HgX L , where L = thione, X = Cl , Br or I )
1
9,20
However, the structural simi-
complexes are extracted into the IL phase. The transfer of the larity of the imidazole-2-thione and the imidazolium cations
IL cation or anion into the aqueous phase not only pollutes of many ILs should increase the solubility of this extractant
the aqueous phase but also results in the loss of the IL. Our and the extracted metal complex in imidazolium-based ILs,
group has been interested in overcoming these limitations by promoting a neutral extraction mechanism and decreasing or
choosing the appropriate task specific ILs and/or ligand– eliminating the ion exchange and loss of IL to the aqueous
9
,10
solvent combinations.
phase. Here we present our investigation of the extraction of
Many efficient sulfur-containing extractants have been pro- mercury(II) from aqueous solutions into a hydrophobic IL
1
1–17
posed for extraction and separation of soft metals.
using our easily synthesized IL-based thione extractant.
Recently, we reported a simple and efficient method for
obtaining access to imidazole-2-thiones from a solventless
Experimental
a
Center for Green Manufacturing and Department of Chemistry, The University
Reagents and metal salts were used as received from commer-
cial sources (Sigma-Aldrich, Milwaukee, WI) unless otherwise
noted. All solvents were ‘solvent grade’ and used as received
without additional purification. The ILs used were 1,3-diethyl-
imidazolium acetate ([C C Im][OAc]) (BASF, Florham Park, NJ)
2 2
with 90% purity, 1-ethyl-3-methylimidazolium acetate
of Alabama, Tuscaloosa, AL 35487, USA. E-mail: rdrogers@as.ua.edu;
Tel: +1 205-348-4323
b
Key Laboratory for Special Functional Aggregated Materials of Education Ministry,
School of Chemistry and Chemical Engineering, Shandong University, Jinan 250100,
China
†
Electronic supplementary information (ESI) available. CCDC 924172–924175.
For ESI and crystallographic data in CIF or other electronic format see DOI:
0.1039/c3dt50410g
1
([C mim][OAc]) with 95% purity from IoLiTec Ionic Liquids
2
1
2908 | Dalton Trans., 2013, 42, 12908–12916
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Technologies Inc. (Tuscaloosa, AL), and 1-ethyl-3-methylimida- 10 mL acetonitrile. Slow evaporation of the acetonitrile at room
zolium bis(trifluoromethane)sulfonamide ([C mim][NTf ]) temperature produced yellow crystals within two days.
2
2
2
1
which was synthesized according to a literature procedure.
mim][NTf ] was prepared through the metathesis reaction
of [C mim]Cl with LiNTf in water. Once the reaction was com-
Synthesis of bis(1,3-diethylimidazole-2-thione)mercury(II)
chloride (Hg(C ImT) Cl
plete, the product was washed with water to remove NaCl until 1,3-Diethylimidazole-2-thione (0.2343 g, 1.5 mmol) was com-
[C
2
2
2
C
2
2
2
)
2
2
−
a AgNO
3 2
test confirmed no residual Cl . The IL was then dried pletely dissolved in 10 mL methanol. A solution of HgCl
using a rotary evaporator and stored in a sealed flask until (0.1375 g, 0.5 mmol) in 10 mL methanol was added dropwise
1
used. The H NMR spectra were recorded using a Bruker to the stirring solution of C
AV-360 (Karlsruhe, Germany) spectrometer operating at 2 h. Slow evaporation of the methanol in a dessicator produced
60 MHz. The mercury(II) concentrations in the aqueous colorless crystals within two weeks.
phases were determined with a Perkin-Elmer Analyst 400
atomic absorption spectrometer (Waltham, MA). The C ImT
concentrations in the aqueous phase were measured with a
Lambda XLS absorption spectrophotometer (Perkin Elmer Ltd, UV-visible spectroscopy was used to monitor the partitioning
Beaconsfield, HP9 2FX, UK). Powder X-ray diffraction (PXRD) of the ligand, C ImT, to the aqueous phase from the IL
data were collected on a Bruker D2 Phaser powder X-ray diffr- phase. 1 mL of a deionized water and 1 mL of 0.01 M
actometer (Madison, WI) with Ni-filtered Cu Kα radiation. ImT/[C mim][NTf ] solution were contacted by mixing
2 2
C ImT. The mixture was stirred for
3
Partitioning studies with 1,3-diethylimidazole-2-thione
ImT)
2 2
C
2 2
(C C
2 2
C
C
2
C
2
2
2
Infrared (IR) analysis was performed using the attenuated total both in a glass vial with 5 min of vortexing followed by 2 min
reflectance method on the neat samples with a Bruker Alpha of centrifugation at 2000 g to ensure that the phases were fully
FT-IR instrument, Bruker Optics, Inc. (Billerica, MA), equipped mixed and separated. The concentration of C
2
C
2
ImT in the
with a diamond ATR cell.
aqueous phases was determined by diluting the aqueous solu-
tion to the appropriate concentration range for absorbance
measurement at 254 nm. A baseline was obtained for DI water
using a quartz cuvette for each experiment before recording
any spectra. The associated Beer’s Law plot was determined
Synthesis of 1,3-diethylimidazole-2-thione (C C ImT)
2
2
The extractant, C C ImT, was synthesized according to a litera-
2
8
2
1
ture procedure. 11.00 g of [C C im][OAc] (59.40 mmol) and
2
2
−5
−4
from 10 to 10 M C
out a known amount of C
The C C ImT concentrations in the IL phases were calculated
2
C
2
ImT solutions prepared by weighing
1
.87 g sulfur (7.29 mmol, equivalent to 58.30 mmol of sulfur
2 2
C
ImT and diluting with DI water.
atoms) were placed in a round bottom flask. The flask was par-
tially immersed in a silicone oil bath with a controlled temp-
erature of 50 °C. The mixture was stirred and was observed to
darken from yellow to brown within a few minutes. After stir-
ring for 24 h, 30 mL of acetonitrile was added to the reaction
medium. The resulting mixture was stirred for an additional
2
2
by mass balance. The partition coefficient of C
IL phase was calculated as the concentration of C
2
C
2
ImT to the
ImT in
2 2
C
the IL phase divided by its concentration in the aqueous
phase.
1
0 min to ensure dissolution of the formed thione and then
Extraction studies
filtered to remove any unreacted sulfur. The acetonitrile was
removed under reduced pressure and the product mixture was
redissolved in 50 mL dichloromethane. The dichloromethane
solution was washed with water (3 × 30 mL) and 5% (w/w)
sodium bicarbonate solution (3 × 30 mL) to remove the
residual [C C Im][OAc] and acetic acid. The dichloromethane
was removed under vacuum to give 4.37 g of a light brown
colored product. Yield = 47.6%. The product was further
purified by recrystallization from water to give the light yellow
Prior to conducting extraction experiments, the IL was pre-
equilibrated with water. 20 mL of IL were mixed with 5 mL
deionized water with 2 h of stirring and 2 min of centrifu-
gation at 2000 g to separate the phases. All extractions were
performed using 1 mL of a deionized water solution as the
2
2
upper aqueous phase and 1 mL of pre-equilibrated [C mim]-
2
2
[NTf ] as the lower nonaqueous phase. The two phases were
contacted by mixing both in a glass vial with 5 min of vortex-
ing followed by 2 min of centrifugation at 2000 g to ensure
that the phases were fully mixed and separated. The initial
mercury(II) concentration in the aqueous phase was kept at
1
final product. H NMR (360 MHz, DMSO-d ) δ 7.18 (s, 1H),
6
1
3
3
6
.97 (q, J = 7.2 Hz, 2H), 1.24 (t, J = 7.2 Hz, 3H). FT-IR (cm−
079, 2974, 1566, 1410, 1272, 1255, 1211, 1156, 946, 789, 713,
79, 513.
)
0
.01 M in all experiments, while the initial concentrations of
+
−
+
−
+
H , Cl , Na , OAc , and C
varied.
2
mim in the aqueous phase were
Synthesis of (1,3-diethylimidazole-2-thione)mercury(II)
The extractant was dissolved in a water-immiscible IL,
mim][NTf ], to form the nonaqueous phase with initial
ImT concentrations varying from 0 to 0.04 M. The 0.01 M
methanol. A solution of C C ImT (0.0781 g, 0.5 mmol) in 5 mL Hg(II) working solutions were prepared by dissolving mercury(II)
2
chloride ([Hg(C C ImT)-κCl-μ Cl] )
2
2
2
[C
2
2
HgCl
2
(0.1375 g, 0.5 mmol) was completely dissolved in 10 mL
2 2
C C
2
2
methanol was added dropwise to the stirring solution of chloride or mercury(II) acetate in deionized water. HCl and
+
−
HgCl
2
. A white precipitate formed as the mixture stirred for NaCl were used for adjusting H and Cl concentration in the
1
h. The solid was collected by filtration and dissolved in HgCl2 aqueous phases from 0.1–1 M, while 1-ethyl-3-methyl-
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 12908–12916 | 12909

View Article Online
Paper
imidazolium acetate [C
Dalton Transactions
2
mim][OAc] and NaOAc were used for were solved by direct methods, and the crystal structures of
adjusting C mim and OAc− concentration in the Hg(OAc)2 [Hg(C C ImT)-κCl-μ Cl] and Hg(C C ImT)(NTf ) were solved
+
2
2
2
2
2
2
2
2 2
aqueous phases from 0–2 M.
by locating the Hg atom by Patterson methods. Non-hydrogen
The distribution ratios (D) for mercury ions were deter- atoms in all structures were located from the difference map
mined as the ratios of Hg(II) concentration in the lower (non- and refined anisotropically through least squares refinement
2
aqueous) phase to the upper (aqueous) phase. The Hg(II) against F . In all structures, hydrogen atoms were placed in
concentrations in the aqueous phases were determined by calculated positions and allowed to ride on the carrier atom.
diluting each aqueous phase to the appropriate concentration Hydrogen atoms on methyl groups were refined using a riding
2
+
2
range for atomic absorption analysis. Hg was atomized and rotating model. In the solution of [Hg(C
2
C
2
2
ImT)-κCl-μ Cl] , the
detected at 253.65 nm by operating in the flame mode using presence of large difference map peaks near the Hg atom indi-
Hg-specific hollow cathode lamps. A calibration curve was con- cated problems with the data. All reflections at 2θ > 50.5°
−
1
structed by using 10, 20, 40, 60, 80, and 100 mg L aqueous which deviated by more than three standard deviations from
solutions of HgCl prior to every analysis to ensure accurate the calculated structure factors were eliminated from the
measurements. The metal ion concentrations in the nonaqu- refinement using the OMIT command (although they were still
2
eous phases were calculated by mass balance.
included in the calculation of R factors). The data to parameter
ratio following the removal of erroneous reflections was
approximately 18.6, and the residual difference map peaks
were greatly reduced in intensity. A correction for extinction
was also applied.
CCDC 924172, 924173, 924174, and 924175 contain the
supplementary crystallographic data for the structures
described in this paper.
X-ray crystallographic characterization
Single crystal X-ray diffraction data were collected on a Bruker
diffractometer (Madison, WI) with a Platform 3-circle gonio-
meter and an Apex II CCD area detector using graphite mono-
chromated Mo-Kα radiation (λ = 0.71073 Å). Crystals were
cooled during collection under a cold nitrogen stream using
an N-Helix cryostat (Oxford Cryosystems). A hemisphere of
data was measured using a strategy of omega scans of 0.5° per
frame. Empirical absorption corrections were applied. Data
collection, integration, and absorption corrections were per-
Results and discussions
Coordination chemistry
2
2
formed using the APEX2 software suite from Bruker. Struc- We had hypothesized that using an IL-based thione which was
ture solution and refinement was conducted using the likely to form neutral complexes with Hg(II), as well as show
2
3
SHELXTL software package from Bruker. Packing diagrams high solubility with the IL, would promote the efficient extrac-
were created using Mercury available from the Cambridge tion of neutral species. Prior to conducting extraction studies,
2
4
Crystallographic Data Center.
we explored the coordination chemistry of the thione ligand
The crystallographic details and parameters of the struc- with Hg(II) through single crystal X-ray diffraction studies in
tures described in this paper can be found in Table 1. The order to understand the nature of the Hg–thione interactions.
2
crystal structures of
C
2
C
2
ImT, [Hg(C
2
C
2
ImT)-κCl-μ Cl]
2
,
2 2
The thione extractant, 1,3-diethylimidazole-2-thione (C C ImT),
Hg(C C ImT) Cl , and Hg(C C ImT)(NTf ) were determined at was prepared from a solventless reaction of 1,3-diethylimidazo-
2
2
2
2
2
2
2 2
1
00 K. The crystal structures of C C ImT and Hg(C C ImT) Cl
lium acetate IL and elemental sulfur, as previously reported
2
2
2
2
2
2
Table 1 Crystallographic details
2
2
C C
2
ImT
[Hg(C
2
C
2
ImT)-κCl-μ Cl]
2
2
Hg(C C
ImT) Cl
2 2 2
Hg(C C ImT) (NTf )
2 2 2 2 2
CCDC number
Formula
Formula weight (g mol
Crystal system
Space group
Temperature (°C)
Cell dimensions
a (Å)
924172
924173
924174
924175
C H F N O S Hg
18 24 12 6 8 6
7 12
C H N
2
S
14
C H24Cl
4 4
N S
2
Hg
2
C
14 24
H N
Cl
4 2
2
S Hg
−
1
)
156.25
Monoclinic
P2 /n
835.48
Monoclinic
P2 /c
583.98
Monoclinic
P2 /c
1073.38
Monoclinic
P2 /n
1
1
1
1
−173.0
−173.0
−173.0
−173.0
6.495(2)
14.431(5)
8.949(3)
90
9.258(5)
9.365(5)
14.749(9)
90
12.9363(6)
11.8914(6)
13.1361(6)
90
8.4298(5)
17.833(1)
11.8196(7)
90
b (Å)
c (Å)
α (°)
β (°)
γ (°)
Cell volume (Å )
93.467(3)
90
837.3(5)
4
103.015(8)
90
1245(1)
2
92.657(2)
90
2018.6(2)
4
99.958(2)
90
1750.0(2)
2
3
Z
Total/unique reflections
21 407/3228
0.0382
0.0348
0.0913
6212/2182
0.0659
0.0573
0.1463
48 754/7616
0.0444
0.0274
27 092/4028
0.0471
0.0212
R
R
int
1
[I > 2σ(I)]
[I > 2σ(I)]
wR
2
0.0670
0.0429
1
2910 | Dalton Trans., 2013, 42, 12908–12916
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
center of inversion in the space group P2
1
/c. Each mercury(II)
center is coordinated by a single monodentate thione, a non-
bridging chloride, and two bridging chlorides in a distorted
tetrahedral geometry.
2
−
The dimer is analogous to the well-known Hg Cl
2
6
2
6
anion, as well as other neutral, chloride-bridged pnicto-
2 2 2 2 8
Scheme 1 Synthesis of C C ImT from the IL [C C Im][OAc] and S .
2
7,28
29–31
32
gen,
chalcogen,
and organomercury molecules. The
Hg–S bond distance of 2.418(3) Å is slightly shorter than in the
1
8
3
3
(
Scheme 1). The thione was purified by recrystallization from
structure of Hg(ethylimidazole-2-thione)
2
Cl
2
(2.49(1) Å)
1
water and fully characterized using H NMR, infrared spec-
troscopy, and single-crystal X-ray diffraction (XRD).
X-ray quality crystals were obtained by slow evaporation
from water, and the asymmetric unit from the crystal structure
2 2
is shown in Fig. 1. The C C ImT molecules are linked in pairs
where the coordination geometry of the cation is distorted tet-
rahedral, and it is shorter than the sum of covalent radii of S
and tetrahedral Hg (ca. 2.53 Å). This bond length is longer
than the Hg–S bonds observed in linear Hg complexes with
3
4,35
sulfur donors (ca. 2.34 Å).
The dimers are supported by
by weak hydrogen bonding interactions between the thione
sulfur atoms and the hydrogens on the ethyl groups, as well as
neighboring imidazole ring hydrogen atoms and (2.921 and
intramolecular ionic interactions between the imidazolium
rings and the chloride ions whereas the imidazolium rings are
oriented such that they overlap with the bridging chloride
ligands.
2
.845 Å, respectively) as shown in Fig. S1 (ESI†). The bond
lengths and angles in the molecule are similar to those
2 2 2
Reaction of solutions of C C ImT and HgCl in methanol
2
5
reported for 1,3-dimethylimidazole-2-thione (C
In order to explore the coordinating ability of C
reacted it with mercury halides, since these are known to form
1
C
1
ImT).
under identical conditions as noted above, but in a 1 : 3 metal-
to-ligand ratio did not result in any precipitate. Crystals from
this reaction suitable for single crystal X-ray diffraction could
not be obtained through slow evaporation under ambient con-
ditions but were obtained through slow evaporation in a
dessicator.
2 2
C
ImT, we
2
0
neutral complexes with similar thiones. Two coordination
2
compounds, [Hg(C
2
C
2
ImT)-κCl-μ Cl]
2 2 2 2 2
and Hg(C C ImT) Cl ,
were obtained by reacting solutions of C C ImT and HgCl in
2
2
2
methanol at different metal-to-extractant ratios.
The complex resulting from the 1 : 3 reaction, Hg(C C ImT) Cl
2
2
2
2
2
[
Hg(C C ImT)-κCl-μ Cl] was obtained by reacting HgCl2
2
2
2
(Fig. 3), is a 1 : 2 complex of HgCl and the thione with a dis-
2
and the thione in a 1 : 1 ratio and redissolving the immediate
white precipitate in acetonitrile followed by slow evaporation.
Crystallographic analysis of this compound revealed that the
centrosymmetric dimer (Fig. 2) resides on a crystallographic
torted tetrahedral geometry. This compound also crystallizes
in the space group P2 /c with Z′ = 1. The Hg–Cl bonds are
1
asymmetric (Hg–Cl1 = 2.4772(7) Å, Hg–Cl2 = 2.5470(7) Å), as are
the Hg–S bonds (Hg–S1A = 2.5092(7) Å, Hg–S1B = 2.5198(6) Å).
The Hg–S bond distances are slightly longer than in
2
[
Hg(C
2
C
2
ImT)-κCl-μ Cl]
2
but close to the Hg–S bond lengths
3
3
in Hg(ethylimidazole-thione)
2 2
Cl (2.49 Å, 2.50 Å). The distor-
tion of Hg from ideal tetrahedral geometry is much less pro-
2
nounced than for [Hg(C
2
C
2
2
ImT)-κCl-μ Cl] . The Hg-centered
bond angles vary from 100.85(2)° for S1A–Hg–Cl2 to 115.65(3)°
for Cl1–Hg–S1A.
The complex could hypothetically possess up to C2v sym-
metry but is twisted into a peculiar lower symmetry arrange-
ment which appears to maximize intramolecular interactions.
Fig. 1 Thermal ellipsoid plot (50% probability) showing the formula unit of
2 2
C C ImT.
2 2
The C C ImT ligands are rotated away from this conformation
Fig. 2 Thermal ellipsoid plot (50% probability) showing the formula unit of
Fig. 3 Thermal ellipsoid plot (50% probability) showing the formula unit of
Hg(C ImT) Cl
2
[Hg(C
2
C
2
ImT)-κCl-μ Cl]
2
.
2
C
2
2
2
.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 12908–12916 | 12911

View Article Online
Paper
Dalton Transactions
2 2 2 2 2 2
Fig. 4 Mercury(II) distribution ratios in [C mim][NTf ] from 0.01 M HgCl (black) and Hg(OAc) (red) aqueous solutions as a function of C C ImT concentration in
the IL (left) and slope analysis (right).
about the Hg–S bond axes (S1A–Hg–S2A–C2A and S2A–Hg–
S1B–C2B torsion angles are 168.95(9)° and 167.05(8)°, respect-
The extraction experiments were conducted by dissolving
2 2
C C ImT in the IL in various concentrations and contacting
ively, vs. 180° for C2v symmetry) and C–S bond axes (imidazo- these solutions with aqueous solutions of 0.01 M HgX (X = Cl,
2
lium A and B ring planes form angles to the S–Hg–S plane of OAc). Mercury concentrations in the aqueous phase were deter-
7
9.28° and 82.59°, respectively, vs. 90° for C2v symmetry), mined by atomic absorption, and the concentration in the IL
bringing the positively-charged rings closer to one of the chlor- phase was determined by mass balance. The distribution ratio
2
+
ide ligands. Cl1, which forms a longer bond to Hg than Cl2, of Hg between aqueous and IL phases was calculated as the
2
+
forms a shorter contact to C
2 2
C ImT B than does Cl2 to concentration of Hg in the IL phase divided by its concen-
C C ImT B. Also, one of the ethyl groups on ring A is oriented tration in the aqueous phase. 1 mL of a deionized water solu-
2
2
nearly in plane with the ring, a feature not seen in any of the tion containing Hg(II) was mixed with 1 mL of pre-equilibrated
other crystal structures containing C ImT, which allows it to [C mim][NTf ] and centrifuged for phase separation. The
make an intramolecular close contact to Cl1 (see Fig. 3). initial mercury(II) concentration in the aqueous phase was kept
The intermolecular interactions include short hydrogen at 0.01 M in all experiments, while the initial C ImT concen-
2
C
2
2
2
2 2
C
contacts with the imidazolium rings serving as donors and the trations varying from 0 to 0.04 M. The results are depicted in
chloride and sulfur atoms serving as acceptors, and charge- Fig. 4.
ordered stacking of chloride ions and imidazolium rings. Pos-
2 2
The distribution ratios for Hg(II) into [C mim][NTf ] with
2
+
sibly due to the fact that positive charge of the Hg ion is split no extractant are quite low with Hg(II) preferring the aqueous
over two thione ligands rather than one, this crystal structure phase (DHgCl2 = 0.74; DHg(OAc)2 = 0.18), indicating the IL itself is
has short contacts between overlapping imidazolium rings. a very poor extractant as expected. Hg(II) can be extracted into
These π–π stacking interactions occur in dialkylimidazolium the IL phase with the addition of C
2
C
2
ImT with distribution
3
6,37
ILs
and are known to be important in their interactions ratios increasing with extractant concentration, reaching 181
3
8
with conjugated systems.
2 2 2
for the 0.01 M HgCl aqueous phase/0.04 M C C ImT in IL.
The observed distribution ratios are comparable to results pre-
viously reported in IL extraction systems containing mercury-
Extraction studies
39
specific extractants either dissolved in the IL or incorporated
9
,40
Having confirmed the coordinating ability of the C C ImT as a task specific functional group.
2
2
ligand, we moved on to study the liquid–liquid extraction of
mercury(II) from water using this ligand as an extractant. higher for HgCl
C C ImT was dissolved in the hydrophobic, room temperature mercury speciation might affect the extraction efficiency. The
The D values for Hg(II) at higher thione concentrations are
2
than Hg(OAc) , suggesting that aqueous
2
2
2
IL 1-ethyl-3-methylimidazolium bis(trifluoromethane)sulfon- predominate mercury species in 0.01 M aqueous HgCl
solu-
while Hg(OAc)
to extracting mercury(II), the distribution ratio of the extractant does not form appreciable amounts of Hg-acetate complexes
2
4
1
amide, [C
2
mim][NTf
2
], and used as the extraction phase. Prior tions is presumed to be the neutral HgCl
2
,
2
4
2
2 2
C C
ImT between the IL and water, which could influence the when dissolved in water. To study this further, we proceeded
extraction efficacy, was measured by contacting 0.01 M to explore the extraction mechanism for these two mercury
C C ImT in IL with water. The partition coefficient of C C ImT salts.
2
2
2 2
between IL and aqueous is 2.75, indicating it is much more
The extraction stoichiometry was studied by slope analysis
soluble in the IL, most likely due to the structural similarity of of the distribution ratios as a function of the concentration of
the imidazole-2-thione and the imidazolium cations of the IL.
C
2
C
2
ImT in the IL phase. As shown in Fig. 4 (right), both
1
2912 | Dalton Trans., 2013, 42, 12908–12916
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 5 Mercury distribution ratios at room temperature between [C
2
mim][NTf
2
] and various aqueous phases including HgCl
2
(left) and variable NaCl (●) and HCl
(
○) concentrations, and Hg(OAc)
correspond to systems without extractant (overlapping). The solid triangular symbols denote the distribution ratio of Hg between the IL and deionized water with
▲) or without (▼) C
ImT. The concentration of C ImT in the IL for all cases was 0.02 M.
2 2 2 2
(right) and variable [C C ImT; dashed lines
mim][OAc] (●) and NaOAc (○) concentrations. Solid lines correspond to systems with C
2
+
(
2
C
2
2 2
C
−
, HgCl −
2−
systems show a linear relationship between log D and log species in 1 M Cl , HgCl
ImT]. The slopes for the aqueous HgCl (2.18) and Hg- mately equal concentrations in 0.1 M Cl , and HgCl
OAc)2 (1.96) systems indicate the extracted species in both dominant species in 0.01 M. This equilibrium indicates that
cases are 2 : 1 thione : Hg(II). for the 0.01 M solutions used in our extraction studies, neutral
In order to determine if neutral or ion exchange mecha- HgCl is likely the dominant Hg species and that this species
2
3
, and HgCl
4
are in approxi-
−
[
(
C
2
C
2
2
2
is the pre-
4
4
2
nisms were occurring, the effects of ion concentration and pH can easily coordinate the thione. The trend in distribution
on the extractions were studied. Fig. 5 shows the distribution ratios suggests that upon doubling the chloride concentration
values as a function of aqueous ion concentrations for the from 0.01 to 0.02 M, the concentration of neutral HgCl
HgCl2 (left) and Hg(OAc)2 (right) systems. The effect of increases, but order-of-magnitude increases in the chloride
additional ions on the extraction of HgCl and Hg(OAc) was concentration convert the HgCl into anionic mercury chloride
studied for [C mim][NTf ] with no extractant and with 0.02 M complexes, decreasing the amount of extractable mercury.
C C ImT. As shown in Fig. 5, the distribution ratios for both Taken together, the results suggest extraction of HgCl into
2
2
2
2
2
2
2
2
2
2 2 2 2 2 2 2
HgCl and Hg(OAc) into [C mim][NTf ] with no extractant are the IL via the formation of the HgCl –2C C ImT neutral
less than 1 regardless of the concentration of acid or salt in complex depicted in Fig. 3. To confirm if this complex was
1
the aqueous phase, indicating the inability of the [C mim]- present in the IL after extraction, H NMR of the crystallogra-
2
[
NTf
2
] to extract either neutral HgCl
solution.
2 2
or cationic mercury in phically characterized HgCl thione complexes in DMSO were
1
Hg(OAc)
2
measured and compared to H NMR spectra of Hg-containing
In the C C ImT/[C mim][NTf ] extraction system, distri- IL solutions in DMSO (Fig. S5†). The NMR spectra of the IL
2
2
2
2
bution ratios for HgCl
2
showed the same trend for both HCl phase after extraction shows a downfield shift of the thione
and NaCl with a chloride ion dependence. In both cases the D C4/C5 proton peak which is similar to the shift observed for
values increase upon the addition of 0.01 M chloride and then Hg(C C ImT) Cl . Through the electron donation of the thione
2
2
2
2
decrease with further increases of chloride concentration. The into the positively charged Hg(II) ion, electron density is
similar distribution coefficients for equal concentrations of removed from the ligand, and the C4/C5 proton peaks become
2
HCl vs. NaCl indicate that the chloride ion itself, and not deshielded. Furthermore, [Hg(C C ImT)-κCl-μ Cl] is insoluble
2
2
2
−
+
43
1
other species such as OH or H
3
O , affect the extraction.
in the IL, and attempts to measure its H NMR in the IL were
The formation of different Hg(II) species in chloride con- unsuccessful, indicating that this complex does not have
taining aqueous solutions is dependent on the concentration sufficient solubility in the IL phase to be the extracted species.
4
4
of chloride anions as shown in eqn (1):
These results suggest that Hg(C
soluble species.
2
C
2
ImT)
2
Cl
2
is the extracted
in the IL led us
2
Hg2 Ð ½HgClꢀþ Ð HgCl2 Ð ½HgCl3ꢀꢁ Ð ½HgCl4ꢀ2ꢁ ð1Þ
þ
2
The insolubility of [Hg(C C ImT)-κCl-μ Cl]
2 2
to investigate the possibility of recycling the IL by forming and
The cationic mercuric species to the left of the equilibrium filtering the insoluble dimeric mercury complex. The extraction
−
2+
−
predominate at low Cl concentration ([Hg ] > [Cl ]), while experiment was conducted by contacting 1 mL IL and 1 mL
−
2
the anionic species predominate at high Cl concentration aqueous solution with 1 M thione and 2 M HgCl , respectively.
[Hg ] < [Cl ]). Distributions of aqueous mercury(II) chlorides As we expected, a precipitate forms following biphasic contact.
species are known to exist where [HgCl4] is the predominant After vortexing for 1 min, the precipitate was collected.
2
+
−
(
−
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 12908–12916 | 12913

View Article Online
Paper
Dalton Transactions
To test this as a recycling methodology, additional extrac-
tant was added to the IL and contacted with a new solution of
M HgCl . Another batch of precipitate was collected indicat-
2
2
ing the system may be able to operate under a continuous
process by operating with an excess concentration of
extractant.
To determine to composition of the precipitate, the com-
bined precipitates were dissolved in acetonitrile. Upon slow
solvent evaporation, X-ray quality crystals were obtained and
the unit cell was determined. The unit cell was identical to
that for the crystal structure shown in Fig. 2 of [Hg(C C ImT)-
2
2
2
κCl-μ Cl]
2
. These results indicate that the neutral Hg(II) species
is formed and has a low solubility limit in the IL. One could
imagine recycling the IL and thione within a continuous Fig. 6 Thermal ellipsoid plot (50% probability) showing the formula unit of Hg
2 2 2 2 2
(C C ImT) (NTf ) .
process by precipitating and collecting the Hg–thione complex
and stripping the Hg(II) from the thione with HCl solution.
Since Hg(OAc)2 does not form appreciable amounts of
acetate complexes when dissolved in water, and cationic Hg–O bond lengths (Hg–O2 = 2.964(2) Å, Hg–O4 = 2.803(3) Å),
mercury is the predominate mercury species in the aqueous especially compared to the short bond length of 2.3503(7) Å
4
2
phase, we hypothesized that cation exchange would be the for the Hg–S1 bond.
most probable extraction mechanism for the extraction of Hg(II) Further evaluation of the bond distances imply that [Hg-
by C ImT in this system. This mechanism is represented (C could be a stable linear dication in the IL. In Hg-
2
+
2
C
2
2 2 2
C ImT) ]
in eqn (2):
(C C ImT) (NTf ) , the linear Hg–S distances of 2.3503(7) Å are
2
2
2
2 2
much shorter than the Hg–S distances observed in the tetrahe-
Hg2þ
þ 2C2C2ImTðILÞ þ 2½C2mimꢀþ
þ 2½NTf2ꢀꢁ
ðaqÞ
2+
ðaqÞ
ðILÞ
ðILÞ
dral C C ImT complexes, suggesting a linear Hg-(thione)
2
2
2
!
HgðC2C2ImTÞ ðNTf2Þ
þ 2C2mimþ
ð2Þ
cation is not too dissimilar than the linear, covalently bound
2
2ðILÞ
4
6
molecule HgCl
2
.
The C–S distances within the coordinated
+
According to eqn (2), increasing C
2
mim concentration in thione of S1–C2 = 1.736(3) Å and S2–C2A = 1.826(3) Å are sig-
the aqueous phase would cause a decrease in mercury(II) nificantly longer than the 1.688(1) Å in the uncoordinated
+
extraction. The influence of the aqueous phase C
OAc concentrations on Hg(OAc) extraction by C C
2 2 2
2
mim and
ImT was much shorter and closer together than those in the uncoordi-
determined and the results are shown in Fig. 5 (right). The nated thione. The similarity of these bond distances to those
2 2
C C ImT, and the imidazole C2–N bond distances are both
−
+
47
C
2
mim concentration in the aqueous phase was adjusted observed in crystal structures of 1,3-diethylimidazolium salts
using [C mim][OAc], which is soluble in water. A decrease in D suggests that the imidazolium rings are now partially positive.
2
values for Hg(II) into [C mim][NTf ] was observed with increas- The ligated Hg(II) cation would have much greater solubility in
2
2
ing concentration of [C
aqueous phase. The significantly greater effect of [C
2
mim][OAc] but less so NaOAc in the the IL than uncoordinated or hydrated Hg(II) cations. The long
−
2
mim]- Hg–O bonds to the NTf
2
ion and the intramolecular hydrogen
[
OAc] on the D values compared to NaOAc supports the idea bonding interactions between the uncoordinated oxygen
+
−
that C
2
mim must enter the aqueous phase, while at the same atoms of the NTf
2
moiety and the neighboring hydrogen
¬
time suggesting the extraction mechanism is fairly indepen- atoms from the ethyl groups (Fig. S6†) suggest that NTf
dent of OAc concentration.
2
is
−
held in place by weak, charge-based interactions, in contrast to
2
+
The cation exchange mechanism for Hg(OAc)
2
was also sup- the linear [Hg(C
as a crystalline
We were able to find one other reported crystal structure of
solid following the extraction of a 0.02 M Hg(OAc) aqueous a linear, cationic Hg(II)-bis(imidazole-2-thione) complex, [Hg(1-
2
C
2
ImT)
2
]
core.
2 2 2 2 2
ported by the isolation Hg(C C ImT) (NTf )
2
4
8
solution with a solution of 0.02 M C
2
C
2
ImT in [C
2
mim][NTf
2
]. t-butyl-3-H-imidazole-2-thione)
2
][BF
4
]
2
4 2 4 2
(Hg(C ImT) (BF ) ).
Upon mixing and separation of the phases, suitable crystals The complex also features a weakly coordinating counterion
for single crystal X-ray diffraction formed spontaneously in the which interacts weakly with the Hg ion through three separate
[C mim][NTf ] phase after a few days. The complex, shown in anions. Hg–F bond lengths in the range 2.833(2)–2.960(3) Å
2 2
Fig. 6, crystallizes in the space group P2
1
/n with the Hg atom indicate that these anions are interacting in a similar manner
residing on a crystallographic center of inversion. Mercury is to the NTf2− anion. [Hg(1-t-butyl-3-H-imidazole-2-thione)2]-
−
coordinated by two C
2
C
2
ImT ligands and two chelating NTf
2
4 2
[BF ] also has long C–S distances within the thione (1.743(4),
anions to give a slightly distorted octahedral coordination geo- 1.742(3) Å), short Hg–S bond distances of 2.3454(9) and
2
+
−
metry. To our knowledge no other Hg -NTf coordination 2.3495(8) Å, and a nearly linear S–Hg–S bond angle of 172.86(3)°.
2
complexes have been structurally characterized. As a weak and The slightly bent geometry may be due to the asymmetric ligand
4
5
−
hard Lewis base, NTf
2
would be expected to interact only and presence of strong N–H⋯F hydrogen bonding between the
weakly with the Hg center which is supported by the very long thiones and BF4− both bound to the Hg ion. The common
1
2914 | Dalton Trans., 2013, 42, 12908–12916
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
features observed in these structures suggest that in the
absence of strongly coordinating anions, imidazole-2-thiones
may be adopting a zwitterionic resonance structure, forming a
8 M. L. Dietz and D. C. Stepinski, Talanta, 2008, 75, 598.
9 A. E. Visser, R. P. Swatloski, W. M. Reichert, R. Mayton,
S. Sheff, A. Wierzbicki, J. H. Davis Jr. and R. D. Rogers,
Environ. Sci. Technol., 2002, 36, 2523.
2
S–Hg–S core, similar to HgCl .
10 J. D. Holbrey, A. E. Visser, S. K. Spear, W. M. Reichert,
R. P. Swatloski, G. A. Broker and R. D. Rogers, Green Chem.,
2
003, 5, 129.
Conclusions
1
1 M. Mojski, Talanta, 1978, 25, 163.
The ligand 1,3-diethylimidazole-2-thione (C
synthesized in a simple, one pot process from reaction of an IL
with elemental sulfur, and can be used to extract HgCl and 13 R. Singh and S. N. Tandon, Talanta, 1997, 44, 843.
2
C
2
ImT) can be 12 Y. Baba, Y. Umezaki and K. Inoue, Solvent Extr. Ion Exch.,
1986, 4, 15.
2
Hg(OAc)
mim][NTf
Hg(II) were isolated and crystallographically characterized 15 I. Ishikawa and T. Sato, Nippon Kagaku Kaishi, 2000, 11,
showing that the coordination numbers and geometries can 779.
vary with Hg(II) counter ion and ligand stoichiometry. The extrac- 16 M. E. Núňez, E. R. de San Miguel, F. Mercader-Trejo,
tion mechanisms are dependent on the speciation of mercury in J. C. Aguilar and J. de Gyves, Sep. Purif. Technol., 2006, 51, 57.
the aqueous solution: HgCl can form Hg(C ImT) Cl com- 17 D. Pulak and B. Sukalyan, Int. J. ChemTech Res., 2011, 3,
plexes and be extracted as a neutral species, while Hg(OAc) 1349.
dissociates in aqueous solution and is extracted through a 18 H. Rodríguez, G. Gurau, J. D. Holbrey and R. D. Rogers,
2
from aqueous systems into the hydrophobic IL 14 J. S. Preston and A. C. Preez, Solvent Extr. Ion Exch., 1994,
[C
2
2
]. Coordination complexes of this ligand with 12, 667.
2
C
2 2
2
2
2
2+
cation exchange mechanism involving [Hg(C
2 2 2
C ImT) ]
ions.
Chem. Commun., 2011, 47, 3222.
The precipitation of neutral mercury complexes from the IL 19 R. Shunmugam and D. N. Sathyanarayana, J. Coord. Chem.,
following the extraction of excess HgCl suggests a simple and
1983, 12, 151.
2
unusual way to recycle the IL as well as the extractant.
20 Z. Popović, D. M. Calogović, Ž. Soldin, G. Pavlović,
N. Davidović and D. V. Topić, Inorg. Chim. Acta, 1999, 294, 35.
21 H. Luo, S. Dai, P. V. Bonnesen, A. C. Buchanan,
J. D. Holbrey, N. J. Bridges and R. D. Rogers, Anal. Chem.,
Acknowledgements
2004, 76, 3078.
WL would like to thank the China Scholarship Council for 22 APEX 2 AXScale and SAINT, version 2010, Bruker AXS, Inc.,
financial support and SPK would like to thank the U.S. Depart- Madison, WI.
ment of Energy Nuclear Energy University Programs for a 23 G. M. Sheldrick, SHELXTL, structure determination software
graduate research fellowship (DE-NE0000366). This work was suite, v.6.10, Bruker AXS Inc., Madison, WI, 2001.
supported by the U.S. DOE Office of Nuclear Energy’s Nuclear 24 C. F. Macrae, I. J. Bruno, J. A. Chisholm, P. R. Edgington,
Energy University Programs (Sub-Contract – #120427, Project –
3123).
P. McCabe, E. Pidcock, L. Rodriguez-Monge, R. Taylor,
J. van de Streek and P. A. Wood, J. Appl. Crystallogr., 2008,
#
47, 466.
2
2
5 G. B. Ansell, J. Chem. Soc., Perkin Trans. 2, 1972, 7, 841.
6 P. L. Goggin, P. King, D. M. McEwan, G. E. Taylor,
P. Woodward and M. Sandstrom, J. Chem. Soc., Dalton
Trans., 1982, 875.
References
1
2
3
V. A. Cocalia, K. E. Gutowski and R. D. Rogers, Coord.
Chem. Rev., 2006, 250, 755.
S. Dai, Y. H. Ju and C. E. Barnes, J. Chem. Soc., Dalton
Trans., 1999, 1201.
27 L. Liu, Q. Zhang and W. Leung, Acta Crystallogr., Sect. E:
Struct. Rep. Online, 2004, 60, 394.
28 N. A. Bell, M. Goldstein, T. Jones and I. W. Nowell, Inorg.
Chim. Acta, 1980, 43, 87.
Metal-Ion Separation and Preconcentration, Progress and
Opportunities, ed. M. L. Dietz, A. H. Bond and R. D. Rogers, 29 C. Stalhandske and F. Zintl, Acta Crystallogr., Sect. C: Cryst.
ACS Symposium Series 716, American Chemical Society,
Washington, DC, 1999, p. 418.
M. P. Jensen and A. H. Bond, J. Am. Chem. Soc., 2002, 124,
Struct. Commun., 1986, 42, 1449.
30 J. Choi, S. Y. Park, H. Y. Yang, H. J. Kim, K. Ihm, J. N. Nam,
J. R. Ahn and S. U. Son, Polym. Chem., 2011, 2, 2512.
31 N. Davidovic, D. Matkovic-Calogovic, Z. Popovic and
I. Vedrina-Dragojevic, Acta Crystallogr., Sect. C: Cryst. Struct.
Commun., 1998, 54, 574.
4
5
9
870.
A. E. Visser, M. P. Jensen, I. Laszak, K. L. Nash,
G. R. Choppin and R. D. Rogers, Inorg. Chem., 2003, 42,
2
197.
M. Baghdadi and F. Shemirani, Anal. Chim. Acta, 2009, 634,
86.
32 E. C. Spencer, M. B. Mariyatra, J. A. K. Howard,
A. M. Kenwright and K. Panchanatheswaran, J. Organomet.
Chem., 2007, 692, 1081.
6
7
1
X. Q. Sun, B. Peng, J. Chen, D. Q. Li and F. Luo, Talanta, 33 M. Cannas, F. A. Devillanova, G. Marongiu and G. Verani,
008, 74, 1071. J. Inorg. Nucl. Chem., 1981, 43, 2383.
2
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 12908–12916 | 12915

View Article Online
Paper
Dalton Transactions
3
3
4 D. Grdenic, Q. Rev. Chem. Soc., 1965, 19, 303.
5 L. Pauling, The Nature of the Chemical Bond, Cornell Univer-
sity Press, Ithaca, NY, 3rd edn, 1960.
6 A. R. Choudhury, A. S. Winterton, A. I. Cooper and
K. A. Johnson, J. Am. Chem. Soc., 2005, 48, 16792.
7 A. Mele, G. Romanό, M. Giannone, E. Ragg, G. Fronza, 44 F. A. Cotton and G. Wilkinson, Advanced Inorganic Chem-
G. Raos and V. Marcon, Angew. Chem., Int. Ed., 2006, 45, 1123. istry, Wiley-Interscience, New York, 5th edn, 1988, p. 612.
8 J. D. Holbrey, W. M. Reichert, M. Nieuwenhuyzen, 45 S. H. Strauss, Chem. Rev., 1993, 93, 927.
41 K. J. Powell, P. L. Brown, R. H. Byrne, T. Gajda, G. Hefter,
S. Sjoberg and H. Waner, Pure Appl. Chem., 2005, 77, 739.
42 A. P. Kreshkov, L. N. Balyatinskaya, Yu. F. Milyaev and
A. V. Filonova, Zh. Neorg. Khim., 1974, 19, 77.
3
3
3
43 S. Sjőberg, Acta Chem. Scand., Ser. A, 1977, 31, 705.
O. Sheppard, C. Hardacre and R. D. Rogers, Chem. 46 A. F. Wells, Structural Inorganic Chemistry, Clarendon Press,
Commun., 2003, 476.
Oxford, 3rd edn, 1962, p. 345.
3
4
9 A. E. Visser, R. P. Swatloski, S. T. Griffin, D. H. Hartman 47 G. Gurau, H. Rodríguez, S. P. Kelley, P. Janiczek,
and R. D. Rogers, Sep. Sci. Technol., 2001, 36, 785.
R. S. Kalb and R. D. Rogers, Angew. Chem., Int. Ed.,
2011, 50, 12024.
0 J. D. Holbrey, A. E. Visser, S. K. Spear, W. M. Reichert,
R. P. Swatloski, G. A. Broker and R. D. Rogers, Green Chem., 48 J. G. Melnick, K. Yurkerwich and G. Parkin, Inorg. Chem.,
003, 5, 129. 2009, 48, 6763.
2
1
2916 | Dalton Trans., 2013, 42, 12908–12916
This journal is © The Royal Society of Chemistry 2013