Anion Analysis of Ionic Liquids and Ionic Liquid Purity Assessment by Ion Chromatography

Article abstract of DOI:10.1002/zaac.201600437

The simultaneous determination of halide impurities (fluoride, chloride, bromide, and iodide) and ionic liquid (IL) anions (tetrafluoroborate, hexafluorophosphate, and triflimide) using ion chromatography was developed with a basic, non-gradient ion chromatography system. The non-gradient method uses the eluent Na₂CO₃/NaHCO₃in water/acetonitrile (70:30 v:v) on the AS 22 column to enable a rapid and simultaneous analysis of different IL and halide anions within an acceptable run-time (22 min) and with good resolution R of larger than 2.4, a capacity k′ between 0.4 and 5.1, selectivities α between 1.3 and 2.1, and peak asymmetries A_sof less than 1.5. Halide impurities below 1 ppm (1 mg·L^–1of prepared sample solution) could be quantified. A range of ionic liquids with tetrafluoroborate [BF₄]^–, hexafluorophosphate [PF₆]^–, and bis(trifluoromethylsulfonyl)imide (triflimide) [NTf₂]^–anions combined with cations based on imidazole, pyridine, and tetrahydrothiophene could be analyzed for their anion purity. The IL-cations do not influence the chromatographic results. With the analysis of 18 ILs differing in their cation-anion combination we could prove the general applicability of the described method for the anion purity analysis of ionic liquids with respect to halide ions. The IL-anion purity of most ILs was above 98 wt %. The highest IL-anion purity was 99.8 wt %, implying anion impurities of only 0.2 wt %. The used halide anion from the synthesis route was the major anion impurity, yet with chloride also bromide and fluoride (potentially from hydrolysis of [BF₄]^–) were often detected. When iodide was used, at least chloride but sometimes also bromide and fluoride was present. However, even if the IL-anion content is above 99 wt %, it does not necessarily indicate an ionic liquid devoid of other impurities. From the IC analysis, one can also deduce a possible cation impurity if one takes into account the expected (calculated) IL-anion content. A matching experimental and theoretical IL-anion content excludes, a higher experimental content indicates the presence of residual KBF₄, NH₄PF₆, or LiNTf₂salt from the halide to IL-anion exchange.

Full text of DOI:10.1002/zaac.201600437

Journal of Inorganic and General Chemistry
DOI: 10.1002/zaac.201600437
ARTICLE
Zeitschrift für anorganische und allgemeine Chemie
Anion Analysis of Ionic Liquids and Ionic Liquid Purity
Assessment by Ion Chromatography
Christina Rutz,[a] Laura Schmolke,[a] Vasily Gvilava,[a] and Christoph Janiak*[a]
Dedicated to Professor Walter Frank on the Occasion of his 60th Birthday
Abstract. The simultaneous determination of halide impurities matographic results. With the analysis of 18 ILs differing in their cat-
(
(
fluoride, chloride, bromide, and iodide) and ionic liquid (IL) anions
tetrafluoroborate, hexafluorophosphate, and triflimide) using ion
ion-anion combination we could prove the general applicability of the
described method for the anion purity analysis of ionic liquids with
respect to halide ions. The IL-anion purity of most ILs was above
chromatography was developed with a basic, non-gradient ion
chromatography system. The non-gradient method uses the eluent
Na CO /NaHCO in water/acetonitrile (70:30 v:v) on the AS 22
2 3 3
column to enable a rapid and simultaneous analysis of different IL and
halide anions within an acceptable run-time (22 min) and with good
resolution R of larger than 2.4, a capacity kЈ between 0.4 and 5.1,
selectivities α between 1.3 and 2.1, and peak asymmetries A of less
s
than 1.5. Halide impurities below 1 ppm (1 mg·L–1 of prepared sample
solution) could be quantified. A range of ionic liquids with tetrafluoro-
borate [BF
98 wt%. The highest IL-anion purity was 99.8 wt%, implying anion
impurities of only 0.2 wt%. The used halide anion from the synthesis
route was the major anion impurity, yet with chloride also bromide
–
and fluoride (potentially from hydrolysis of [BF
4
] ) were often de-
tected. When iodide was used, at least chloride but sometimes also
bromide and fluoride was present. However, even if the IL-anion con-
tent is above 99 wt%, it does not necessarily indicate an ionic liquid
devoid of other impurities. From the IC analysis, one can also deduce
a possible cation impurity if one takes into account the expected (cal-
culated) IL-anion content. A matching experimental and theoretical IL-
anion content excludes, a higher experimental content indicates the
–
–
4
] , hexafluorophosphate [PF
6
] , and bis(trifluoromethyl-
–
sulfonyl)imide (triflimide) [NTf
2
]
anions combined with cations
based on imidazole, pyridine, and tetrahydrothiophene could be ana- presence of residual KBF , NH PF , or LiNTf salt from the halide to
4
4
6
2
lyzed for their anion purity. The IL-cations do not influence the chro-
IL-anion exchange.
Introduction
been reported. They are used in organic chemistry, catalysis or
electrochemistry.[
25–30]
Ionic liquids (ILs) are saline substances containing weakly
coordinating anions and cations.
point of these salts should be below 100 °C preferably even a
below room temperature. ILs are alternatives to aqueous or
There are generally two different routes to synthesize ionic
liquids. The common way is to use a two-step process. The
first step is the formation of an organic halide salt via alkyl-
ation using a haloalkane. After that a metathesis reaction intro-
duces the weakly coordinating anion to form the ionic liquid.
An alkali metal salt like KBF or acids like HBF can be used
[
1,2]
By definition the melting
[3,4]
organic solvents,
and are regarded as a new liquid me-
dium[
5,6]
for the preparation of inorganic materials.
[7–14]
Ionic
4
4
liquids are stable at high temperatures, have no vapor pressure
to give, e. g., a tetrafluoroborate ionic liquid in the anion ex-
and a high ionic conductivity.[
15–17]
Contrary to common or-
[31–33]
change reaction (Scheme 1).
For halide free ionic liquids,
ganic solvents they are not flammable.[
18,19]
The density, vis-
direct alkylation of imidazole, pyridine or amine can be done
cosity, and hydrophilic properties of ILs can be systematically
varied through the choice of cation, anion, or cation-anion
combination.[ The miscibility of ionic liquids with water or
organic solvents varies with the side-chain lengths in the cation
and with the character of the anion. Because of their physical
[34–36]
using esters or anhydrides.
The problems about halide containing ionic liquids are well
known.[ Halides reduce the thermal stability and stability of
the inter-ionic network.[ The viscosity and the density also
depend on the purity of the ionic liquids. Furthermore, halide
20]
37]
38]
properties ionic liquids are regarded as designer solvents with impurities decrease the electrochemical window, i.e. oxidation
a potential wide range of applications in science and technol- and reduction potential limits.[ It requires a more positive
21]
ogy.[
21–24]
A wide range of applications for ionic liquids has potential to oxidize perfluorinated anions than chloride, brom-
ide, and iodide. A significant decrease of the anodic potential
can be observed with halide impurities.[
21]
*
Prof. Dr. C. Janiak
E-Mail: janiak@uni-duesseldorf.de
Although these problems are known, just a few analytical
methods have been reported for the quantification of halide
impurities in ILs. Common methods are reversed-phase liquid
[
a] Institut für Anorganische Chemie und Strukturchemie
Universität Düsseldorf
Universitätsstr. 1
40225 Düsseldorf, Germany
[39–43]
[44–46]
chromatography,
ion pair chromatography,
hydro-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/zaac.201600437 or from the au-
thor.
[
47,48]
philic interaction chromatography,
capillary electro-
phoresis,[
49,50]
and ion chromatography.
[49,51]
Z. Anorg. Allg. Chem. 2017, 643, 130–135
130
© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Journal of Inorganic and General Chemistry
ARTICLE
Zeitschrift für anorganische und allgemeine Chemie
tively. The instrument was controlled by Chromeleon® software
version 7.1.0.898). The injection volume was 25 μL. Milli-Q 18 MΩ
(
water was used for all eluent, sample, and standards preparation. Ana-
lytical grade sodium carbonate and sodium hydrogen carbonate (both
Applichem) were used to prepare the eluent. HPLC grade acetonitrile
(
4
VWR) was added to the eluent. The standard eluent used was a
.5 mmol·L Na CO and 1.0 mmol·L NaHCO mixture with an ad-
2 3 3
–
1
–1
dition of 30 vol% acetonitrile. For the method development we used
different eluent mixtures. The eluent was prepared daily and degassed
by bubbling nitrogen through the solution for 20 min.
The calibration standards were prepared from sodium fluoride
(
(
Applichem), sodium chloride (Fisher Chemical), potassium bromide
Gruessing), potassium iodide (Applichem), potassium tetrafluoro-
borate (Acros Organics), ammonium hexafluorophosphate (Alfa
Aesar), and bis(trifluoromethane)sulfonimide lithium salt (lithium tri-
flimide, TCI) by dissolving a defined amount in water with further
–
1
Scheme 1. Synthesis scheme of 1,3-dialkylimidazolium, 1-alkyl-
pyridinium, and 1-alkylthiophenium ionic liquids via N-alkylation and ionic liquid anions: 5–100 mg·L ). Exact masses and volumes for the
dilution to different precise concentrations (halides: 0.1–5 mg·L ,
–
1
anion exchange.
standard solutions can be found in the Supporting Information (Tables
S1–S9, Supporting Information). A calibration curve was established
by measuring seven samples of each anion as an external standard at
different concentrations. The concentration range of fluoride, bromide
Ion chromatography (IC) is a relatively easy, fast, and stan-
dard method for the simultaneous analysis of anions. IC is used and iodide was 0.1 to 5.0 mg·L , chloride 0.1 to 20.0 mg·L , and
routinely for the determination of halides in aqueous media.
–
1
–1
[52]
–1
the IL anions were calibrated from 5.0 to 100.0 mg·L (Table S13,
IC should also be applicable for the analysis of anions of ionic Supporting Information). Every determination was done threefold and
the averaged results including standard deviation were used. The esti-
liquids and anion impurities like halides but the utilization of
mated errors were calculated and added to the standard deviation. The
IC for the analysis of ionic liquids is rarely reported. The main
calibration was checked by analyzing control samples with known con-
challenge here is to achieve a good peak resolution. Hardacre
centrations regularly.
et al. published a method about the determination of chloride
impurities in tetrafluoroborate ionic liquids.[ A good resolu-
53]
All ionic liquids were prepared following standard preparative pro-
tion for the halide impurities in an adequate time was reached, cedures.[21] Details can be found in the Supporting Information. Six
but the tetrafluoroborate peak exhibited a strong tailing and its different tetrafluoroborate-based ILs were analyzed. These ILs were
symmetry should be improved. A good peak symmetry could 1-allyl-tetrahydro-1H-thiophenium-tetrafluoroborate
([ATh][BF
([BMIm][BF
4
]),
1
-butyl-3-methyl-imidazolium-tetrafluoroborate
-butyl-pyridinium-tetrafluoroborate ([BPy][BF
4
]),
be obtained by using an eluent gradient system, which allows
1
4
]), 4-butyronitrile-
]), 1-lauryl-pyridinium-
]), and 1-octyl-tetrahydro-1H-thiophen-
ium-tetrafluoroborate ([OTh][BF ]). For [BMIm][BF ] three different
samples were analyzed, one obtained from anion exchange with KBF
and two different batches from anion exchange with HBF . The two
hexafluorophosphate-based ILs were 1-allyl-3-methyl-imidazolium-
variation of the eluent concentration during the measure-
ment.[54,55]
pyridinium-tetrafluoroborate ([NCC
tetrafluoroborate ([LPy][BF
3 4
Py][BF
Nesterenko et al. reported a method for the determi-
4
nation of the hydrolysis products from alkali metal hexafluoro-
phosphate salts by using a reagent-free eluent generator in
4
4
4
,
2
012. The separation of the hydrolysis products fluoride
4
–
–
(F ), hydrogenmonofluorophosphate (HPO F ), hydrogen-
3
2
–
–
phosphate (HPO4 ), and difluorophosphate (PO F ) anions hexafluorophosphate ([AMIm][PF ]) and 4-butyronitrile-pyridinium-
2
2
6
could be achieved with a good resolution.[ With a gradient hexafluorophosphate ([NCC
56]
3
6
Py][PF ]). Eight triflimide-based ILs
system a simultaneous analysis of different IL anions in combi- 1-allyl-3-methyl-imidazolium-triflimide ([AMIm][NTf
2
]), 1-allyl-
([ATh][NTf ]), 1-butyl-3-
nation with halide impurities has been reported[
54,55]
but a gra-
tetrahydro-1H-thiophenium-triflimide
2
2
methyl-imidazolium-triflimide ([BMIm][NTf ]), 1-butyl-pyridinium-
dient system is more expensive and may not be readily avail-
able with every IC device. Herein we describe a method for
the simultaneous determination of halide anions in ionic li-
quids with the anions tetrafluoroborate, hexafluorophosphate,
or triflimide using a basic, standard ion chromatography sys-
triflimide ([BPy][NTf
flimide ([BTh][NTf ]), 1-lauryl-pyridinium-triflimide ([LPy][NTf
-octyl-3-methyl-imidazolium-triflimide ([OMIm][NTf ]), and
-octyl-pyridinium-triflimide ([OPy][NTf ]) were analyzed.
2
]), 1-butyl-tetrahydro-1H-thiophenium-tri-
2
2
]),
1
1
2
2
tem without a gradient pump or a reagent-free eluent generator. IC samples were prepared by dissolving a defined mass of IL in a
defined volume of either deionized water (if soluble) or in the prepared
eluent solution to achieve an IL-anion concentration between 50–
–
1
Experimental Section
100 mg·L . Exact masses and volumes for the IL sample solutions
can be found in the Supporting Information.
All ion chromatographic measurements were carried out with a Dionex
ICS 1100 instrument with suppressed conductivity detection. The
Supporting Information (see footnote on the first page of this article):
suppressor (AERS 500, Dionex) was regenerated with an external Synthesis equations for ILs, concentrations of solutions for calibration,
water module. The system was equipped with the analytical column
IonPac AS 14 or IonPac AS 22 from Dionex (4ϫ250 mm) with the
corresponding guard column AG 14 or AG 22 (4ϫ50 mm), respec-
concentrations for sample solutions, chromatograms from column
AS 14, chromatograms without column, chromatograms with column
AS 22 for method development, chromatograms with AS 22 of IL sam-
Z. Anorg. Allg. Chem. 2017, 130–135
www.zaac.wiley-vch.de
131
© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Journal of Inorganic and General Chemistry
ARTICLE
Zeitschrift für anorganische und allgemeine Chemie
ples, cake diagrams of anion fractions in ILs, data tables, formulae for the AS 14 column material for the separation of the IL anions
data analysis.
and to exclude other instrumental artefacts, measurements
without column were performed. These measurements showed
the suitability of the ICS 1100 system for the analysis of ionic
liquid and halide anions. Without an analytical column sharp
peaks of tetrafluoroborate, hexafluorophosphate, and triflimide
were detected after a few seconds using eluents at different
concentrations containing sodium carbonate/sodium hydrogen
carbonate mixture or sodium hydroxide (Figures S10–S21,
Supporting Information). With these results the detection fail-
Results and Discussion
Method Development and Optimization of IC Separation
The selection of the ion-exchange column for the separation
of ILs and halides is not trivial. Fluoride can react with silica,
so silica-based anion-exchange columns are not applicable.[56]
Our initial separation studies were done with an AS 14 col-
umn. The substrates of the AS 14 column are highly cross-
linked macroporous resin pellets with a diameter of 9 μm.
They consist of styrene, which is cross-linked with divin-
ylbenzene and is functionalized with quaternary alkyl ammo-
nium groups. As the AS 14 column is routinely used for the
analysis of halide anions in aqueous solutions, we tried to de-
velop a method for the analysis of the IL anions together with
the halide anions. This application for the AS 14 column is,
so far, unknown in literature. We investigated the use of dif-
ferent eluent solutions of sodium carbonate/sodium hydrogen
carbonate mixture, sodium borate or sodium carbonate at dif-
ferent concentrations to try to reach a good peak resolution.
Details for these eluent mixtures can be found in the figure
captions of the chromatograms in Figures S4–S9 (Supporting
Information). Test measurements were carried out with potas-
sium tetrafluoroborate, ammonium hexafluorophosphate, and
–
–
ure of [PF ] and [NTf ] could be traced to the AS 14 column.
6
2
Hence, for subsequent successful measurements we used the
analytical column AS 22 from Dionex. The substrate of the
AS 22 column is a highly crossed-linked modified styrene di-
vinylbenzene resin derivatized with alkyl-ethanol ammonium
functionalities. Its capacity is 210 μeq with a low hydrophobi-
–1
city. Using the analytical column AS 22 and a 4.5 mmol·L
–1
sodium carbonate and 1.4 mmol·L sodium hydrogen carb-
onate eluent, the analysis of tetrafluoroborate was achieved
within a retention time of 15 min with an acceptable baseline
separation and peak shape (Figure 2). A small peak tailing of
the tetrafluoroborate peak could be observed, but due to an
asymmetry factor of 1.7 (that is larger than 1.5) the shape
should still be improved (Figure 2).
lithium bis(trifluoromethylsulfonyl)imide (triflimide). Only the
–
[
BF4] peak but neither hexafluorophosphate nor triflimide
could be detected with the AS 14 column (Figures S4–S9, Sup-
–
porting Information). Detection of [BF ] with the AS 14 col-
4
umn suffered from its asymmetric peak (Figure 1) which failed
to become symmetrical in the tested eluents.
Figure 2. Chromatogram of potassium tetrafluoroborate; peak
–
–
–
assignment: 2 F , 3 Cl , 5 [BF
4
] . Column: AS 22; flowrate:
–
1
–1
1
1
.00 mL·min ; eluent: 4.5 mmol·L
sodium carbonate and
–
1
.4 mmol·L sodium hydrogen carbonate; run-time: max. 60 min.
Still, the simple aqueous sodium carbonate/sodium
hydrogen carbonate eluent was not applicable for the analysis
of hexafluorophosphate and triflimide salts or ILs. Hexafluoro-
phosphate was only detected at a long retention time of 55 min
showing a distinctive tailing (asymmetry factor of 2.2) (Figure
S28, Supporting Information), whereas triflimide could not be
Figure 1. Example chromatogram of potassium tetrafluoroborate
–
showing the typical asymmetric [BF
4
] peak on the AS 14 column;
–
–
–
–1
peak assignment: 2 F , 3 Cl , 5 [BF
4
] . Flowrate: 1.20 mL·min ; elu- detected at all (Figure S29, Supporting Information).
ent: 2.5 mmol·L–1 sodium carbonate and 2.5 mmol·L
–1
sodium
The retention time for anions depends of the charge, size,
and polarizability of the anion. The addition of organic sol-
vents to the eluent can be used to change the retention time
hydrogen carbonate; run-time: max. 60 min.
The retention time of the tetrafluoroborate peak correlates of anions. Hu et al. could show a simultaneous analysis of
with the elution potency of the eluent. With a higher concen- hexafluorophosphate with halide ions like fluoride, chloride,
trated eluent, the retention time could be shortened (Figure S6, and bromide with acceptable peak shape and baseline separa-
tion within 25 min using an AS 22 column.[ Consequently,
57]
Supporting Information).
Hydrophobic ions, such as hexafluorophosphate or triflimide in our runs a separation of fluoride, chloride, bromide, iodide,
may be strongly retained by styrenedivinylbenzene polymeric and the IL anions tetrafluoroborate, hexafluorophosphate, and
anion-exchangers and this will affect both the runtime and re- triflimide was achieved within an acceptable run-time of
–1
producibility of separations. Due to the hydrophobic nature of 22 min using an eluent mixture of 4.5 mmol·L sodium carb-
ions like hexafluorophosphate or triflimide, it is necessary to onate und 1.4 mmol·L–1 sodium hydrogen carbonate with
minimize the non-specific adsorption of the analytes by using 30 vol% acetonitrile added. All peaks show a baseline separa-
[56]
a hydrophilic anion exchange.
To prove the unsuitability of tion and a good resolution (Figure 3).
Z. Anorg. Allg. Chem. 2017, 130–135
www.zaac.wiley-vch.de
132
© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Journal of Inorganic and General Chemistry
ARTICLE
Zeitschrift für anorganische und allgemeine Chemie
should be between 0.9 to 1.1, but values up to 1.5 for A are
S
accepted as well.[
58–63]
The determined quality parameters
confirm a successful simultaneous chromatographic separation
of the IL-anion and halide impurities.
With the aforementioned method for the AS 22 column we
measured a series of synthesized ionic liquids (see Supporting
Information for chromatograms, Figures S30–S43). All chro-
matograms showed the same good peak resolution, which we
could achieve in the above test measurements (cf. Figure 3 and
–
1
Figure 3. Chromatogram of sodium fluoride (0.30 mg·L ), sodium
–1
–1
chloride (0.50 mg·L ), potassium bromide (0.20 mg·L ), potassium
–1
–1
tetrafluoroborate (10.00 mg·L ), sodium iodide (0.50 mg·L ), ammo- Table 1). The anion purity of most of the ILs was above
nium hexafluorophosphate (10.00 mg·L ), and lithium triflimide 98 wt%. Detailed values are given in the Supporting Infor-
–1
–
1
–
–
–
–
–
(
7
4
10.00 mg·L ); peak assignment: 2 F , 3 Cl , 4 Br , 5 [BF
4
] , 6 I ,
mation. We were able to determine the halide impurities of
the ILs below 1 ppm (below 1 mg·L–1 in the prepared sample
solution) (Tables S14–S16, Supporting Information).
It is not trivial, however, to judge the overall purity of an
IL. Even if the IL-anion content is above 99 wt%, it does not
necessarily indicate an ionic liquid devoid of impurities. The
–
–
–1
[PF
6 2
] , 8 [NTf ] . Column: AS 22; flowrate: 0.90 mL·min ; eluent:
–1
–1
.5 mmol·L sodium carbonate, 1.4 mmol·L sodium hydrogen carb-
onate and 30% acetonitrile; run-time: set to max. 22 min. For chroma-
tograms of the actual investigated ILs with column AS 22 see Figures
S30–S43 (Supporting Information).
In order to judge the quality of the chromatographic separa- examples of [BMIm][BF ] and [BPy][NTf ] illustrate how the
4
2
tion we calculated the factors for resolution R, capacity kЈ, IC anion analysis can give additional information on the IL
selectivity α, and asymmetry AS for every adjacent peak purity and, hence, for a possible optimization of the anion ex-
(Table 1) (formulae are given in Supporting Information).
change and washing step in the IL synthesis. With more than
The quality parameter resolution R should be larger than 0.8, 99 wt% IL-anion fraction the ILs [BMIm][BF ] (from KBF4
4
the capacity kЈ should give a value between 1 and 5, the selec- or from HBF ), [BMIm][NTf ] and [BPy][NTf ] (from
4
2
2
tivity α values lie between 1.05 and 3, and the asymmetry AS LiNTf ) seem to be quite pure (Table 2 and Figure 4). Yet,
2
Table 1. Quality parameters for the chromatographic separation of halide and IL anions.a)
c)
c)
hc) /μs
Rc)
kЈc)
αc)
c)
Peak
t
R
/min
w
b
/min
/
A
S
H
2
O
3.05
4.14
5.31
6.70
8.25
10.98
16.62
18.49
0.29
0.34
0.22
0.19
0.33
0.54
0.78
0.87
0.73
0.15
0.70
0.07
2.76
0.07
0.65
0.27
3.46
4.18
6.78
5.96
6.28
8.55
2.39
n.a.
0.00
0.35
0.74
1.20
1.70
2.60
3.88
5.06
n.a.
1.55
1.45
1.40
1.21
1.31
1.51
1.44
1.13
b)
b)
Fluoride
Chloride
Bromide
2.11
1.62
1.42
1.53
1.49
1.30
n.a.
b)
–
b)
[BF
4
]
b)
Iodide
–
b)
[PF
6
]
[NTf
2
]– b)
a) Based on chromatogram in Figure 3; column: AS 22; flowrate: 0.90 mL·min ; eluent: 4.5 mmol·L sodium carbonate, 1.4 mmol·L–1 sodium
–
1
–1
hydrogen carbonate and 30% acetonitrile; run-time: 22 min. b) Anion source: sodium fluoride, sodium chloride, potassium bromide, potassium
R b
tetrafluoroborate, sodium iodide, ammonium hexafluorophosphate, and lithium triflimide. c) Retention time t , peak width w , peak height h,
resolution R, capacity factor kЈ, selectivity α, and asymmetry A ; see Supporting Information for details of formulae.
S
Table 2. Mean value X¯ and measurement uncertainty σ
of halide and IL anions, and the calculated theoretical content of IL anion.
g
a)
¯
b)
–
c)
Ionic liquid
Sample conc.
Mean value X and measurement uncertainty σ
g
of anion concentrations
[BF [NTf
]– ]–
[IL-anion] theo
/mg·L
Deviation from
[IL-anion] theo /%
–1
–1
–1
–
/mg·L
/mg·L (wt%)
F–
Cl–
Br
–
4
2
[
(
[
(
[
(
[
(
[
(
BMIm][BF
from KBF
BMIm][BF
4
]
)
636.0
294.0
136.0
138.6
170.0
0.5 Ϯ0.1 (0.2)
0.4 Ϯ0.2 (0.15)
0.0
258 Ϯ7 (99.7)
–
–
–
245 Ϯ1
113 Ϯ1
52 Ϯ1
+5
4
4
]_1
0.1 Ϯ0.1 (0.1)
0.2 Ϯ0.2 (0.4)
0.0
0.1 Ϯ0.1 (0.1)
0.4 Ϯ0.4 (0.8)
0.2 Ϯ0.2 (0.2)
0.0
112 Ϯ2 (99.8)
within exp. error
from HBF
BMIm][BF
4
)
]_2
4
0.2 Ϯ0.1
(0.4)
0.0
52 Ϯ1 (98.5)
none
+4
from HBF
BMIm][NTf
from LiNTf
BPy][NTf
from LiNTf
4
)
2
]
–
–
97 Ϯ1 (99.8) 93 Ϯ1
2
)
2
]
0.1 Ϯ0.1 (Ͻ0.1) 0.2 Ϯ0.1 (0.15)
0.1 Ϯ0.1
(Ͻ0.1)
125 Ϯ3
(99.7)
118 Ϯ1
+6
2
)
a) IC samples were prepared by dissolving a defined mass of IL in a defined volume of either deionized water (if soluble) or in the prepared
–
1
eluent solution to achieve an IL-anion concentration between 50–100 mg·L . Exact masses and volumes for the IL sample solutions can be
found in the Supporting Information. b) Every determination was done in triplicate and the averaged results including standard deviations were
g
used. The estimated errors were calculated and added to the standard deviation to give the measurement uncertainty σ . c) The standard deviation
for the expected theoretical [IL-anion]–theo value stems from the weighing and volume measurement uncertainty when preparing the solution
from the neat IL.
Z. Anorg. Allg. Chem. 2017, 130–135
www.zaac.wiley-vch.de
133
© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Journal of Inorganic and General Chemistry
ARTICLE
Zeitschrift für anorganische und allgemeine Chemie
–
–
when comparing the expected theoretical [BF ] or [NTf ]
procedure, the IL was dried in vacuo to remove residual water
4
2
content with the experimentally found amount it can become and other solvents like dichloromethane. Any remains of HBF4
apparent that the overall IL-purity can be less due to remaining would be removed further within this step due to the HBF4
alkali metal salt of the IL-anion from the exchange process volatility.
(
Table 2). Thus, it can be problematic when determining and
Therefore, we were not only able to analyze the IL for its
halide content; we could also draw conclusions about the qual-
ity of the anion exchange and purification process. Even when
the IL-anion percentage is above 99 wt% (Figure 4), it does
not necessarily indicate an ionic liquid devoid of impurities.
When judging the anion purity of an IL it is not only necessary
to quantify possible halide impurities but also to compare the
experimental and expected (theoretical) IL-anion content from
the examined sample mass (for more examples see Tables
S14–S16, Supporting Information).
ascertaining IL purity only by halide and IL-anion analysis.
Conclusions
A new method for anion purity analysis of ionic liquids by
ion chromatography is described using a simple, non-gradient
chromatographic system. The method is based on the AS 22
–
Figure 4. Weight percentage of anions in different ILs. Blue: [BF
4
] ,
] . The compound for the
halide to IL-anion exchange is given in parentheses. For more graphi-
cal presentations of the determined anion weight percentages see Fig- nitrile. The method allowed the simultaneous determination of
–1
column with the eluent 4.5 mmol·L sodium carbonate,
–
–
–
–
black: F , red: Cl , green: Br , orange: [NTf
2
1
.4 mmol·L–1 sodium hydrogen carbonate, and 30 vol% aceto-
ures S44–S50 (Supporting Information).
the IL anions tetrafluoroborate, hexafluorophosphate, and tri-
flimide and the halide anions fluoride, chloride, bromide, and
iodide with resolutions R larger than 2.4, a capacity kЈ between
The ionic liquid [BMIm][BF ] was synthesized in two dif-
4
ferent ways. Following the alkylation of imidazole the anion
exchange was carried out using either KBF or HBF . The
0.4 and 5.1, selectivities α between 1.3 and 2.1 and peaks
4
4
asymmetries A of less than 1.5. Halide impurities below
S
triflimide-based ionic liquids were synthesized with lithium tri-
flimide salt for the anion exchange. We were able to trace the
different anion exchange in the ion chromatographic analysis
of the product by comparing the measured concentration to the
expected concentration of the IL anion, here tetrafluoroborate
and triflimide, which is calculated from the sample mass and
volume of solvent (Table 2).
1
ppm (below 1 mg·L–1 of the prepared sample solution) could
be quantified. With the analysis of 18 ILs differing in their
cation-anion combination, the general applicability of the de-
scribed method for the IL-anion purity was demonstrated. The
IL cations were based on imidazole, pyridine and tetra-
hydrothiophene. The cations do not appear to affect the chro-
matographic process. Further, we show that caution has to be
exerted to base the IL purity on the weight fraction of the IL
anion and the halide impurities only. An IL-anion content
above 99 wt% does not necessarily indicate an ionic liquid
devoid of other impurities. The experimentally found IL-anion
concentration must also be compared to the calculated ex-
pected IL-anion content from the sample concentration. A
matching experimental and calculated IL-anion content ex-
cludes, a higher experimental content indicates the presence of
residual KBF , NH PF , or LiNTf salt from the halide to IL-
The sample of [BMIm][BF ], which was synthesized using
4
KBF for the anion exchange contained a higher amount of
4
tetrafluoroborate in comparison to the expected (theoretical)
content.
Similar
for
[BMIm][NTf ] (LiNTf )
and
2
2
[
BPy][NTf ] (LiNTf ) a higher than expected anion content
2
2
–
was found. Whereas, the experimental [BF ] content of two
4
charges of the [BMIm][BF ] IL synthesized by anion exchange
4
with HBF was almost identical to the expected concentration.
4
The main reason for a higher than theoretical anion content
seems to be the solubility of the KBF , NH PF , or LiNTf
salt used for anion exchange. The purification consisted of
washing with water to remove the alkali halide (here potassium
or lithium chloride) as well as the excess KBF , NH PF , or
4
4
6
2
4
4
6
2
anion exchange.
4
4
6
Acknowledgements
LiNTf salt from the anion exchange reaction (see Supporting
2
Information for details). Due to the low water solubility of We thank Mrs. Susann Wegner for the synthesis of the ionic liquids.
KBF and LiNTf , an excess of this salt may not have been Our work was supported by the Deutsche Forschungsgemeinschaft
4
2
fully removed. The positive deviation from the expected theo- (DFG) through grant Ja466/31–1 within the priority program
“
Materials synthesis at room temperature” (SPP 1708).
retical content confirmed this for [BMIm][BF ] (KBF ),
4
4
[
BMIm][NTf ] (LiNTf ), and [BPy][NTf ] (LiNTf ) and
2 2 2 2
showed a leftover of KBF and LiNTf exchange salt in the
4
2
Keywords: Ion chromatography; Ionic liquids; Halides;
Purity
IL. In contrast, HBF is readily soluble in water, so it could be
4
easily removed during the washing process. After the washing
Z. Anorg. Allg. Chem. 2017, 130–135
www.zaac.wiley-vch.de
134
© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Journal of Inorganic and General Chemistry
ARTICLE
Zeitschrift für anorganische und allgemeine Chemie
[
[
[
[
[
[
[
[
32] P. D. Vu, A. J. Boydston, C. W. Bielawski, Green Chem. 2007, 9,
References
1158–1159.
33] D. Zhao, Z. Fei, T. J. Geldbach, R. Scopelliti, P. J. Dyson, J. Am.
Chem. Soc. 2004, 126, 15876–15882.
34] D. J. Kim, K. H. Oh, J. K. Park, Green Chem. 2014, 16, 4098–
[1] P. Wasserscheid, W. Keim, Angew. Chem. 2000, 112, 3926–3945;
Angew. Chem. Int. Ed. 2000, 39, 3772–3789.
2] H. Weingärtner, Angew. Chem. 2008, 120, 664–682; Angew.
Chem. Int. Ed. 2008, 47, 654–670.
[
4101.
35] R. Arvai, F. Toulgoat, B. R. Langlois, J.-Y. Sanchez, M. Médebi-
elle, Tetrahedron 2009, 65, 5361–5368.
36] M. Javaherian, F. Kazemi, M. Ghaemi, Chin. Chem. Lett. 2014,
[
[
3] T. Welton, Chem. Rev. 1999, 99, 2071–2084.
4] A. Kuchenbach, R. Giernoth, ChemistryOpen 2015, 4, 677–681;
C. Feldmann, Z. Naturforsch. 2013, 68b, 1057. Introduction to
special issue on ionic liquids in chemical synthesis.
2
5, 1643–1647.
37] K. R. Seddon, A. Stark, M. J. Torres, Pure Appl. Chem. 2000, 72,
275–2287.
[5] J. P. Hallett, T. Welton, Chem. Rev. 2011, 111, 3508–3576.
2
[6] T. Torimoto, T. Tsuda, K. Okazaki, S. Kuwabata, Adv. Mater.
38] S. Watanabe, H. Seguchi, K. Yoshida, K. Kifune, T. Tadaki, H.
Shiozaki, Tetrahedron Lett. 2005, 46, 8827–8829.
39] P. Stepnowski, A. Muller, P. Behrend, J. Ranke, J. Hoffmann, B.
Jastorff, J. Chromatogr. A 2003, 993, 173–178.
2
010, 22, 1196–1221.
[
[
[
[
7] C. Janiak, Z. Naturforsch. 2013, 68b, 1059–1089; C. Janiak,
AIMS Mater. Sci. 2014, 1, 41–45.
8] P. S. Campbell, M. H. G. Prechtl, C. C. Santini, P.-H. Haumesser, [40] P. Stepnowski, W. Mrozik, J. Sep. Sci. 2005, 28, 149–154.
Curr. Org. Chem. 2013, 17, 414–429.
9] D. Freudenmann, S. Wolf, M. Wolff, C. Feldmann, Angew. Chem.
Int. Ed. 2011, 50, 11050–11060.
[41] L. M. Meng, H. Yu, X. Huang, Y. J. Ma, Chin. J. Anal. Chem.
2012, 40, 409–414.
[42] E. Daghir-Wojtkowiak, S. Studzinska, B. Buszewski, R.
10] E. Ahmed, J. Breternitz, M. F. Groh, M. Ruck, CrystEngComm
Kaliszanc, M. J. Markuszewski, Anal. Methods 2014, 6, 1189–
1196.
2012, 14, 4874–4885; E. Ahmed, M. Ruck, Dalton Trans. 2011,
4
0, 9347–9357; M. F. Groh, U. Müller, E. Ahmed, A. Rothen-
[43] H. Yu, Y. M. Sun, C. M. Zou, Chin. Chem. Lett. 2014, 25, 1371–
berger, M. Ruck, Z. Naturforsch. 2013, 68b, 1108–1122.
1374.
[
[
11] E. R. Parnham, R. E. Morris, Acc. Chem. Res. 2007, 40, 1005–
[44] M. Y. Wang, H. Yu, P. Li, J. Li, Y. F. Gao, Chin. J. Chromatogr.
2014, 32, 773–787.
[45] C. M. Zou, H. Yu, M. Y. Wang, Chin. Chem. Lett. 2014, 25, 201–
204.
[46] Y. J. Ma, Y. F. Gao, H. Yu, M. Li, J. Liq. Chromatogr. Relat.
Technol. 2014, 37, 73–87.
1013.
12] E. R. Cooper, C. D. Andrews, P. S. Wheatley, P. B. Webb, P. Wor-
mald, R. E. Morris, Nature 2004, 430, 1012–1016.
[
[
13] J. Dupont, J. D. Scholten, Chem. Soc. Rev. 2010, 39, 1780–1804.
14] K. Klauke, B. Hahn, K. Schütte, J. Barthel, C. Janiak, Nano-
Struct. Nano-Objects 2015, 1, 24–31; R. Marcos Esteban, K.
Schütte, D. Marquardt, J. Barthel, F. Beckert, R. Mülhaupt,
C. Janiak, Nano-Struct. Nano-Objects 2015, 2, 28–34; R.
Marcos Esteban, K. Schütte, P. Brandt, D. Marquardt, H. Meyer,
F. Beckert, R. Mülhaupt, H. Kölling, C. Janiak, Nano-Struct.
Nano-Objects 2015, 2, 11–18.
[
47] G. Le Rouzo, C. Lamouroux, C. Bresson, A. Guichard, P. Moisy,
G. Moutiers, J. Chromatogr. A 2007, 1164, 139–144.
[
48] C. Lamouroux, G. Foglia, G. Le Rouzo, J. Chromatogr. A 2011,
1218, 3022–3028.
[
[
[
49] A. Markowska, P. Stepnowski, Anal. Sci. 2008, 24, 1359–1361.
50] M. Shi, Q. Gao, J. Feng, J. Chromatogr. Sci. 2012, 50, 547–552.
51] R. Q. Zhang, H. Yu, Y. Z. Liu, Chin. J. Chromatogr. 2012, 30,
[
[
[
[
15] Y. Lin, S. Dehnen, Inorg. Chem. 2011, 50, 7913–7915.
16] T. Lodge, Science 2008, 321, 50–51.
17] R. E. Morris, Chem. Commun. 2009, 21, 2990–2998.
18] J. Dupont, R. F. de Souza, P. A. Z. Suarez, Chem. Rev. 2002, 102,
7
28–732.
52] European Standards EN ISO 10304–1: 1995; EN ISO 10304–1:
997; 1999 EN ISO 10304–4:1999.
53] C. Villagrán, M. Deetlefs, W. R. Pitner, C. Hardacre, Anal. Chem.
004, 76, 2118–2123.
[
[
[
[
1
3667–3692.
2
[
[
19] C. E. Song, Chem. Commun. 2004, 9, 1033–1043.
20] A. E. Visser, R. P. Swatloski, W. M. Reichert, J. H. Davis Jr.,
R. D. Rogers, R. Mayton, S. Sheff, A. Wierzbicki, Chem. Com-
mun. 2001, 1, 135–136.
54] Dionex Corporation, “IC-MS Determination of Anionic Ionic Li-
quids, Counterions, and Impurities”, 2009.
55] Dionex Corporation, “Determination of Tetrafluoroborate, Per-
chlorate, and Hexafluorophosphate in a Simulated Electrolyte
Sample from Lithium Ion Battery Production”, 2010.
[21] P. Wasserscheid, T. Welton, Ionic Liquids in Synthesis, 1st ed.
Wiley-VCH, Weinheim, 2003.
[
56] L. Terborg, S. Nowak, S. Passerini, M. Winter, U. Karst, P. R.
Haddad, P. N. Nesterenko, Anal. Chim. Acta 2012, 714, 121–126.
57] Z. Hu, G. Pan, M. Ye, Chin. J. Chromatogr. 2009, 3, 337–340.
[22] W. Sundermeyer, Angew. Chem. 1965, 77, 241–258; Angew.
Chem. Int. Ed. Engl. 1965, 4, 222–238.
[
[23] W. Sundermeyer, Chem. Unserer Zeit 1967, 1, 150–157.
[24] S. V. Volkov, Chem. Soc. Rev. 1990, 19, 21–28.
[25] S. A. Siddiqui, U. C. Narkhede, S. S. Palimkar, T. Daniel, R. J.
Lahoti, K. V. Srinivasan, Tetrahedron 2005, 61, 3539–3546.
[58] M. H. Gey, Instrumentelle Analytik und Bioanalytik, 2. Aufl.,
Springer, Heidelberg, 2008.
[59] K. Camman, Instrumentelle Analytische Chemie, 1. Aufl., Spek-
trum-Akademischer Verlag, Heidelberg, 2001.
[
[
26] S. S. Shinde, B. S. Lee, D. Y. Chi, Org. Lett. 2008, 10, 733–735. [60] V. R. Meyer, Praxis der Hochleistungs-Flüssigchromatographie,
27] M. C. Buzzeo, R. G. Evans, R. G. Compton, ChemPhysChem
004, 5, 1106–1120.
28] J. G. Huddleston, H. D. Willauer, R. P. Swatloski, A. E. Visser,
R. D. Rogers, Chem. Commun. 1998, 1765–1766.
10. Aufl., Wiley-VCH, Weinheim, 2009.
[61] S. Petrozzi, Instrumentelle Analytik, 1. Aufl., Wiley-VCH,
Weinheim, 2010.
[62] Chromatogramme richtig integrieren und bewerten. Ein Praxis-
handbuch für die HPLC und GC, (Eds.: S. Kromidas, H.-J. Kuss),
Wiley-VCH, Weinheim, 2008.
2
[
[
[
[
29] J. L. Anderson, D. W. Armstrong, Anal. Chem. 2003, 75, 4851–
4
858.
30] H. Qiu, S. Jiang, X. Liu, L. Zhao, J. Chromatogr. A 2006, 1116,
6–50.
31] A. Aupoix, B. Pégot, G. Vo-Thanh, Tetrahedron 2010, 66, 1352–
356.
[
63] J. W. Dolan, LC-GC Europe 2002, 15, 334–337.
4
Received: November 23, 2016
Published Online: January 12, 2017
1
Z. Anorg. Allg. Chem. 2017, 130–135
www.zaac.wiley-vch.de
135
© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim