Reversible uptake of COS, CS2, and so2: Ionic liquids withO-alkylxanthate, O-alkylthiocarbonyl, and O-alkylsulfite anions

Article abstract of DOI:10.1002/chem.200802602

CO₂-binding organic liquids (CO₂BOLs) are mixtures of a base (typically an amidine or guanidine) and an alcohol, and have been shown to reversibly capture and release CO₂ with low reaction energies and high gravimetric CO<

Full text of DOI:10.1002/chem.200802602

FULL PAPER
DOI: 10.1002/chem.200802602
Reversible Uptake of COS, CS2, and SO2: Ionic Liquids with O-
Alkylxanthate, O-Alkylthiocarbonyl, and O-Alkylsulfite Anions
AHCTUNGTRENNUNG
David J. Heldebrant,*[a] Clement R. Yonker,[b] Philip G. Jessop,[c] and Lam Phan[c]
Abstract: CO2-binding organic liquids
(CO2BOLs) are mixtures of a base
(typically an amidine or guanidine) and
an alcohol, and have been shown to re-
versibly capture and release CO2 with
low reaction energies and high gravi-
metric CO2 capacity. We now report
the ability of such liquid blends to
chemically bind and release other acid
gases such as CS2, COS, and SO2 analo-
gously to CO2. These systems bind with
sulfur-containing acid gases to form
colored ionic liquids with new O-alkyl-
xanthate, O-alkylthiocarbonyl, and O-
alkylsulfite anions. The capture and
thermal stripping of each acid gas from
these systems and their applicability to-
wards flue gas desulfurization is dis-
cussed.
Keywords: acidity
·
flue gas
·
green chemistry · ionic liquids ·
reversibility
Introduction
ever, they are produced in other processes such as sulfur re-
covery, carbon black production, and aluminum produc-
tion.[4] Although the mixed oxides of sulfur in flue gas might
be waste products of coal combustion or any other process
that emits these gases, they still are marketable, hence, if a
feasible method is available to capture and deliver pure
gases (SO2, COS, CS2), then it could become economically
viable. There is a large market for SO2 in the wine industry
and the cement industry as gypsum (CaSO4). CS2 has been
used as an organic solvent and a vital component in the pro-
duction of rayon and cellophane.[6–8] COS can be potentially
used as a fumigant.[7,8] For a CO2 scrubber to be applicable
in post-combustion capture, a solvent system has to be com-
patible with these other acid gases. Current flue gas scrub-
bers are desulfurized with aqueous caustic soda and lime
solutions, which are highly corrosive and irreversible.[9–11]
These systems are damaging to the environment because
they generate large amounts of caustic waste that needs to
be disposed of. To stoichiometrically capture 1 ton of SO2
with caustic soda or lime, we calculate that 0.87 tons of CaO
and 0.62 tons of NaOH are needed. If we assume a 1m solu-
tion of either reagent, one would need 15.6 tons of water in
addition to the capture agent to keep the capture system a
liquid. We estimate that an SO2-binding organic liquid
(SO2BOL) described herein operating at full capacity would
require 1.2 tons of material to reversibly capture 1 ton of
SO2. The reversibility of SO2 capture is desirable because
the capture agent can be reused, unlike the irreversible CaO
and NaOH systems. Another advantage of a reversible pro-
cess is that the captured SO2 can be released and fed direct-
ly into the Klaus process rather than disposal. Desulfurizing
As the search for cleaner energy continues, the need for effi-
cient flue gas scrubbers intensifies. In our previous work, we
have described how CO2-binding organic liquids
(CO2BOLs) have great potential as energy-efficient organic
CO2 scrubbers.[1] CO2BOLs are mixtures of amidine or gua-
nidine bases with alcohols, which can reversibly bind with
CO2.[1–3] In industrial flue gas streams, CO2 is not the only
acid gas present; there are numerous acid gases that are
produced during the combustion of coal. Carbonyl sulfide
(COS, ꢀ4.9ꢀ10ꢁ3 gkgꢁ1 coal burned)[4], carbon disulfide
(CS2, ꢀ6.5ꢀ10ꢁ5 gkgꢁ1 coal burned),[4] and sulfur dioxide
(SO2 up to 3000 ppm)[5] can all be present in flue gas
streams, depending on the coal type and combustion pro-
cess. COS and CS2 are minor components of flue gas; how-
[a] Dr. D. J. Heldebrant
Materials Chemistry and Surface Research Group
Pacific Northwest National Laboratory
Richland, WA 99352 (USA)
Fax : (+01)509-375-2186
E-mail: david.heldebrant@pnl.gov
[b] Dr. C. R. Yonker
Molecular Interactions & Transformations Group
Pacific Northwest National Laboratory
[c] Dr. P. G. Jessop, L. Phan
Department of Chemistry
Queens University
90 Bader Lane, Kingston, ON, K7L 3N6 (Canada)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200802602.
Chem. Eur. J. 2009, 15, 7619 – 7627
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7619

flue gas with a recyclable organic system likely would be
more energy efficient, reduce corrosion associated with
aqueous acid formation, and eliminate the consumption of
tons of caustic soda and lime.
organic liquids (CS2BOLs). This manuscript describes the
reactivity of SO2, COS, and CS2 with amidine and guanidine
bases and alcohols and discusses potential applications of re-
versible scrubbing systems for postcombustion gas capture.
Lewis acid gases such as CO2, SO2, COS, and CS2 (CS2 is
a liquid at room temperature but is listed here as an “acid
gas” because it occurs as a gaseous species in industrial flue
gas) react with water to form their corresponding acids
H2CO3, H2SO3, H2CO2S, and H2COS2, respectively. Al-
though most of these acids are quite unstable with respect
to loss of the gas, they can be stabilized by the addition of a
base. For example, COS and CS2 have been shown to form
thiobicarbonic acid and dithiobicarbonic acid and make
their conjugate bases in alkaline aqueous solution.[12] There-
fore, the stabilizing effect of these bases should also work
for the reactions of these gases with alcohols. Alkylcarbonic
acids, formed by the reaction of CO2 with alcohol, are unsta-
ble in the absence of a base and are known to readily decar-
boxylate unless under CO2 pressure,[13–15] and the same insta-
bility was expected for COS and CS2 derivatives.[12] It has re-
cently been shown that alkylcarbonic acids can be reversibly
trapped as liquid alkylcarbonate salts when in the presence
of amidines and guanidines [this is the basis of the
CO2BOLs, Eq. (1)].[1–3,16] This reactivity should be similar
for COS and CS2, both of which are isoelectronic with CO2
and both of which have a central electrophilic carbon
atom.[12] CS2 has been shown to react with cellulose in the
presence of a strong base (NaOH) to form cellulose xan-
Results and Discussion
The reaction of COS with DBU and 1-hexanol: The reactiv-
ity of COS towards amidine/alcohol and guanidine/alcohol
blends was first monitored by changes in conductivity of a
solution of 1,8-diazabicycloACTHNUTRGENUGN[5.4.0]undec-7-ene (DBU) and 1-
hexanol in MeCN. As in our previous studies of reactions
with CO2,[1,3,22] guanidine and amidine bases paired with al-
cohols are nonconductive solutions until CO2 reacts and
forms the amidinium or guanidinium alkylcarbonate salt
[Eq. (1)]. We anticipated similar reactivity for COS with 1-
hexanol and DBU to form the corresponding carbonyl sul-
fide binding organic liquids [COSBOLs, Eq. (2)].
COS absorption by DBU and 1-hexanol was rapid, similar
to that of CO2. The uptake of COS was complete within
thirty seconds and was mildly exothermic, causing the solu-
tion temperature to jump from 24 to 268C. The conductivity
of the MeCN solution was initially 25 mS prior to COS addi-
tion, likely due to small impurities in the solution. The con-
ductivity of the solution rose to 2000 mS after COS addition,
indicating the formation of a salt, which we propose is the
O-alkylthiocarbonate salt [Eq. (2)]. Mixtures of COS/DBU
and COS/1-hexanol in MeCN were nonconductive, showing
that the formation of the charged product from COS is only
possible in the presence of both base and alcohol.
ꢁ
thates (ROCS2 Na+) in the production of viscose rayon.[6]
There are, however, no published examples of organic rever-
sible solvent systems with COS and CS2.
SO2, while chemically different from CO2, COS, and CS2,
should show similar binding to alcohols due to its electro-
philic sulfur atom.
The elemental analysis of the isolated product matched
that expected for the formulation [DBUH+][ꢁOSCOC6H13].
Spectroscopic characterization of the DBU/1-hexanol/
COS ionic liquid was performed by using both 1H and
Pyramidal SO2 complexes have been shown to be able to
thermally release SO2 at less than 2008C.[17] SO2 has been
shown to react with alkaline water to form bisulfite salts,
which is the basis of current SO2 scrubbing technology.[5,9–12]
However, we are unaware of any published accounts of re-
actions of alcohols and bases with SO2 gas. There are litera-
ture precedents for an organic reversible SO2 binding and
release system; Eckert and Liotta et al.[18,19] were the first to
show a nonionic switchable solvent system with SO2 as the
miscibility trigger based on the early work of Drake
et al.[19,20] Although there are literature precedents for alkyl
sulfite salts of amidine and guanidines,[21] there are not, to
our knowledge, any reports of the use of such mixtures as
switchable solvents or as SO2 scrubbing agents.
1
13C NMR and IR spectroscopy (Table 1). The H NMR spec-
trum of 1:1:1 mixtures of DBU, 1-hexanol, and COS (Fig-
ure S1 in the Supporting Information) showed no evidence
ꢁ
of free alcohol O H and showed the protonated DBU
cation (Scheme 1) at d=7.6 ppm ([D3]MeCN) for COS,
which is slightly downfield of the protonated DBU cation
Table 1. Key bands in the IR spectrum of the product from DBU/1-hexa-
nol/COS, where the proposed structure is [DBUH+][C6H13OCSOꢁ].
[a]
Assignment Product [cmꢁ1
]
A
ACHTUNGTRENNUNG
A
]
A
]
C=N
C=O
1693
1573
1107
1060
1649
–
1580
1100–1120
1055
–
–
–
ꢁ
ꢁ
Analogously to our CO2BOL blends, herein we report
novel reversible SO2 binding organic liquids (SO2BOLs),
COS binding organic liquids (COSBOLs), and CS2 binding
C O C
ꢁ
C S
[a] This work.
7620
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7619 – 7627

CO2-Binding Organic Liquid Systems
FULL PAPER
species with CS2 is only possible in the presence of both
base and alcohol.
Scheme 1. Bases in their protonated form (from left to right): DBU,
TMG, Hꢃnigꢄs. * denotes 13C listed in the tables.
observed in our CO2 studies at d=7.0 ppm ([D3]MeCN).
The H NMR spectrum (Figure S1 in the Supporting Infor-
mation) also showed the downfield shifting of the terminal
The elemental analysis of the isolated product matched
that expected for the formulation [DBUH+][ꢁS2COC6H13].
The product of the reaction of CS2 with DBU and 1-hexa-
1
1
alcohol R CH2 O from d=3.5 to 3.8 ppm ([D3]MeCN),
comparable to the d=4.1 ppm of the same methylene in di-
hexylcarbonate,[23] which indicates that the COS was binding
to the oxygen of the alcohol and not the imine nitrogen of
the base. The 13C NMR spectrum showed that the bridge-
head carbon in DBU (noted in Scheme 1) shifted downfield,
indicative of protonation. DBU shifted from d=160.2 to
164.9 ppm ([D3]MeCN) with COS, which is similar to the
DBUH+ in the CO2BOL system at d=164.9 ppm (CDCl3).
The thiocarbonyl carbon of the ROC(O)Sꢁ anion was ob-
served at d=185.5 ppm ([D3]MeCN), which is 31 ppm
downfield of free COS (d=154 ppm in the solid state and in
nol was characterized with H, 13C NMR, and IR spectrosco-
ꢁ
ꢁ
1
py (Table 2). The H NMR spectrum (Figure S5 in the Sup-
Table 2. Key bands in the IR spectrum of the product of DBU/1-hexanol/
CS2, where the proposed structure is [DBUH+][C6H13OCS2ꢁ].
Assignment Product
G
R
G
[ꢁS2COBu][31]
ACHTUNGTRENNUNG
]
[a]
N
]
[cmꢁ1
G
]
R
C=N
1644
1150
1107
1060
982
1649
–
–
1153
1107
1065
962
ꢁ
C O
–
–
–
–
1143
–
1055
–
ꢁ
ꢁ
C O C
C=S
ꢁ
CS2
SO2 at ꢁ658C).[24,25] Although we were unable to find pub-
[a] This work.
13
ꢁ
lished C NMR spectroscopy chemical shift data for RO
(O)Sꢁ anions, there are reports of RO-C(O)-SR carbonyl
esters at d=162–173 ppm (CDCl3) and RO-C(S)-OR thio-
porting Information) of the 1:1:1 mixture of DBU, 1-hexa-
nol, and CS2 showed protonated DBU at d=8.4 ppm
([D3]MeCN), slightly downfield of the CO2 and COS deriva-
carbonyl esters at d=203–216 ppm (CDCl3).[26] The RO
ꢁ
C(O)Sꢁ anions are less shielded than a corresponding ester
and would appear slightly more downfield.
ꢁ
tives. There was no evidence of free alcohol O H, but there
ꢁ
ꢁ
O
The IR spectrum of DBU/1-hexanol/COS showed evi-
was downfield shifting of the alcohol methylene R CH2
ꢁ1
ꢁ
dence of N H bands at 3223 and 3084 cm and no evidence
from d=3.5 ppm ([D3]MeCN) in the free alcohol to d=
4.3 ppm in the product. The 13C NMR spectrum showed that
the bridgehead carbon in DBU shifted downfield to d=
166.4 ppm ([D3]MeCN and neat; Figure S6 in the Support-
of the broad band of the free OH in 1-hexanol at 3328 cmꢁ1,
thereby confirming that DBU is protonated and the alcohol
is deprotonated.[22] Most importantly, key bands found in the
IR spectra of salts of the structure [K+][ꢁOSCOalkyl] are
also found in the spectrum of our product. We assign the ob-
ing Information), indicative of protonated DBU. Similarly,
ꢁ
the [DBUH+]
[HSO4 ] salt is observed at d=166.3 ppm in
served bands by analogy to the bands of [DBUH+]
[Clꢁ] and
CDCl3.[22] The bound dithiocarbonyl of the CS2BOL was
visible downfield at d=230.0 ppm (neat), which is 32 ppm
downfield of free CS2 in 1-hexanol (d=193 ppm) and near
[K+][ꢁOSCOalkyl].
Attempts to crystallize COSBOL derivatives were unsuc-
cessful because the DBU-based COSBOLs are all liquids. It
is expected that other bases untested at this time could yield
a solid COSBOL for crystallographic analysis.
ACTHNUTRGNEUNG
the reported value for [K+][ꢁS2COMe] in D2O at d=
235.6 ppm.[27] The IR spectrum (Table 2; Figure S7 in the
Supporting Information) of the DBU/1-hexanol CS2BOL
ꢁ
showed no evidence of the free O H in 1-hexanol. Key
The reaction of CS2 with DBU and 1-hexanol: CS2 is expect-
ed to bind to amidines and guanidines and alcohols to form
the proposed O-alkylxanthate salt [Eq. (3)]. CS2 was very
slow to react with DBU and 1-hexanol in MeCN to form
the proposed CS2BOL salt. A solution of CS2, DBU, and 1-
hexanol (0.1m in each) in MeCN at 248C took nearly 2 h to
react, with the conductivity of the solution rising from 54 to
5400 mS after CS2 binding. The small conductance of the so-
lution prior to CS2 addition is likely due to small impurities
in the solution. After CS2 addition, the reaction temperature
spontaneously rose from 24 to 278C over two hours. Equi-
molar mixtures of CS2/DBU and CS2/1-hexanol showed no
conductance, demonstrating that the formation of a charged
bands found in the IR spectra of salts of the structure [K+]
[ꢁS2COalkyl][28–31] are also found in the spectrum of our
product.
Although the product of the reaction of DBU and CS2
with 1-hexanol was a liquid, the product from benzyl alcohol
was a solid salt and was characterized crystallographically.
Colorless block-shaped crystals of the salt (Figure 1) exhibit
the xanthate sp2-hybridized CS2 carbon bound through the
oxygen of the reagent alcohol. The sulfur atoms are equiva-
lent and consistent with a delocalized xanthate anion with
ꢁ
ꢁ
bond lengths of S(1) C
(xanthate) 1.681 ꢂ. The S(1)-C
126.58.
ACHTUNGTRENNUNG
G
ACHTUNGTRENNUNG
Chem. Eur. J. 2009, 15, 7619 – 7627
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7621

D. J. Heldenbrant et al.
ꢁ
ACTHNUTRGNEUNG
[DBUH+][Cl3CCO2 ] (d=5.5 ppm, CDCl3).[22] The pKa of
these sulfur analogues of the alkyl carbonic acids are un-
known at this time, however, the SO2 derivative is closer in
ACTHNUTRGNEUNG
acidity to that of the HCl ([DBUH+][Clꢁ]) salt (d=
1
10.2 ppm, [D3]MeCN). The H NMR showed no evidence of
ꢁ
free alcohol O H, suggesting the base was the protonated
species in solution. The 13C NMR spectrum showed down-
field shifting of the bridgehead carbon in DBU, indicative of
protonation, with the DBU shifting from d=160.2 to
166.5 ppm ([D3]MeCN), which appears to be closer in acidi-
ty to the sulfuric acid salt of DBU (d=166.3 ppm in
CDCl3).
Figure 1. Molecular structure of [DBUH+][ꢁS2COCH2Ph].
Other SO2BOLs were synthesized using TMG and
1
The reaction of SO2 with amine, amidine, and guanidine
bases and 1-hexanol: The reaction of SO2 with amidine or
guanidine and alcohol blends to form the proposed
SO2BOL salt [Eq. (4)] were first monitored by conductivity.
SO2 absorption by 0.1m solutions of DBU and 1-hexanol in
MeCN at 278C was rapid, with most of the reactions com-
pleted within twenty seconds. The SO2 uptake was exother-
mic as noted by a spike in the temperature from 27 to 328C.
The conductivity of the solution rose from 5 to 2900 mS after
SO2 uptake. The increase in conductance suggested SO2 was
being chemically bound as a charged species and not just
being physically absorbed by DBU and 1-hexanol. The
charged species is believed to be the O-alkylsulfite salt
shown in Equation (4). There was a minor increase in con-
ductivity (25 to 200 mS) when SO2 was added to 0.1m DBU
in MeCN, which is attributed to trace amounts of water re-
acting to form the DBU bisulfite salt. There was no conduc-
tance seen for a solution of SO2 in 0.1m 1-hexanol in MeCN.
Physically absorbed SO2 cannot demonstrate an increase in
conductance because there would be no charged species
present in solution. The observed increase in conductance of
the DBU/SO2/1-hexanol mixture in MeCN suggests that
chemical binding of SO2 to form the charged species is only
possible when both alcohol and base are present. SO2 and 1-
hexanol were found to chemically react with other bases
such as 1,1,3,3-tetramethylguanidine (TMG) and ethyldiiso-
propylamine (Hꢃnigꢄs base, Scheme 1) to form other
SO2BOLs.
Hꢃnigꢄs base paired with 1-hexanol. The H NMR spectrum
showed the acidic NH+ of TMGH+ at d=10.5 ppm
([D3]MeCN) and the protonated Hꢃnigꢄs base at d=
9.2 ppm ([D3]MeCN) and no free alcohol O H, which sug-
gests that the base was the protonated species in solution.
The 13C NMR spectrum of these other SO2BOLs showed
TMGH+ at d=166.8 ppm ([D3]MeCN), while there was a
small 5 ppm downfield shift for Hꢃnigꢄs base at d=55.2 ppm
([D3]MeCN). No free alcohol O H was observed with
either TMG/1-hexanol or Hꢃnigꢄs base/1-hexanol after reac-
tion with SO2. There was no downfield shifting in the
H NMR spectrum of the alcohol methylene R CH2
upon addition of SO2 with any of the bases studied. The
13C NMR spectrum shows a slight upfield shift of the meth-
ylene from d=62 to 59 ppm ([D3]MeCN) when SO2 is pres-
ent, in comparison with d=64 ppm in the analogous
CO2BOL salt.
ꢁ
ꢁ
1
ꢁ
ꢁ
O
The SO2BOLs were also characterized by IR spectroscopy
(Table 3). TMG/1-hexanol (Figure S10 in the Supporting In-
+
ꢁ
formation) showed N H stretching of TMGH at 3084 and
3232 cmꢁ1, and Hꢃnigꢄs base (Figure S11 in the Supporting
ꢁ1
ꢁ
Information) had N H stretching at 3551 and 3373 cm .
Similar bands were observed with DBU/1-hexanol (Fig-
ure S9 in the Supporting Information). The N H stretches
confirm protonation of the bases, consistent with the H and
ꢁ
1
13C NMR spectroscopy data. Key IR bands in the salts of
the structure [K+][ꢁO2SOalkyl] were also observed in the
spectra of our products. Furthermore, the elemental analysis
of the DBU/1-hexanol/SO2 ionic liquid matched the theoret-
ical weight percent values of the proposed SO2BOL salt
(Table S1 in the Supporting Information).
We were unable to grow crystals of any SO2BOLs with
linear or branched alcohols or benzyl alcohol because they
all are liquids at room temperature. The analogous TMG bi-
The DBU/1-hexanol/SO2 system was spectroscopically
characterized by 1H and 13C NMR spectroscopy. The
1H NMR spectrum (Figure S8 in the Supporting Informa-
tion) showed the greatest downfield shift of the protonated
DBU amidinium nitrogen atom at d=10.6 ppm
([D3]MeCN) in this study. The [DBUH+] chemical shift ap-
pears to be linearly correlated to the acidity of the protonat-
ing acids (weak to strong: CO2 <COS<CS2 <SO2). We had
shown that the CO2 derivative (d=5.5 ppm, CDCl3) appears
to be close in acidity to the trichloroacetic acid salt of DBU,
sulfite was grown with TMG and SO2 and water. Block-
[HSO3 ] (Figure 2) show
the central sulfur bound to one double-bonded oxygen with
ꢁ
shaped yellow crystals of [TMGH+]
ꢁ
a bond length of S(1) O(1) 1.365 ꢂ, and two equivalent
ꢁ
single-bonded oxygen atoms with bond lengths of S(1) O(2)
ꢁ
1.474 ꢂ and S(1) O(3) 1.475 ꢂ. In bisulfite anions, the hy-
drogen atom can be bound to the sulfur (HSO3 ) or to an
oxygen atom (HOSO2 ). The position of the hydrogen atom
ꢁ
ꢁ
was not determined in the structure.
7622
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7619 – 7627

CO2-Binding Organic Liquid Systems
FULL PAPER
Table 3. Key bands in the IR spectra of the products of the reaction of SO2 with base/hexanol mixtures, where
the proposed structure is [BaseH+][C6H13OSO2ꢁ].
color change to orange-red, and
the largest increase in viscosity
seen in this study (Figure 3).
[DBUH+]
E
ACHTUNGTRENNUNG ACHTUGNTRENNUNG ACHTUGTNRENNUG AHTCUNGTRENUNNG AHCTUNRTGENNUNG ACTHUNGTRENNNUG
[TMGH+][ꢁO2SOR] [NEtiPr2H+][ꢁO2SOR] [Na+][ꢁO2SOCH3][32] [K+][ꢁO2SOCH3][32]
[a]
[a]
[cmꢁ1
]
[cmꢁ1
]
[cmꢁ1
]
[cmꢁ1
]
Three SO2BOLs were pro-
duced with TMG/1-hexanol,
DBU/1-hexanol, and Hꢃnigꢄs
base/1-hexanol reacted with
SO2. 1:1 mixtures of TMG/1-
hexanol and Hꢃnigꢄs base/1-
hexanol absorbed 3.0 molequiv
of SO2, whereas DBU/1-hexa-
nol absorbed 2.3 equiv. We at-
tribute the absorption of three
equivalents of SO2 to one
equivalent chemically binding
1622
1600
1467
1456
1324
1207
1158
–
1687
1600
1467
1405
1319
1239
1184
1121
1031
–
–
–
–
–
–
–
1467
1395
1319
1239
1184
1110, 1144
1036
982
1465
1439
–
1462
–
–
–
–
1177
1119
1039
–
1157
1105
–
–
645
1037
1001
671
671
673
670
[a] This work.
Figure 3. DBU/1-hexanol 1:1 mixtures after treatment with a) CO2,
b) COS, c) CS2, or d) SO2
ACTHNUTRGNEUNG
Figure 2. Molecular structure of [TMGH+][HSO3ꢁ].
to the alcohol and base and then two equivalents of SO2
physically dissolving in the generated SO2BOL ionic liquid.
Ionic liquids (notably guanidinium cations) have previously
been shown to physically absorb up to 2 molequiv of
SO2.[33–36] In our SO2-uptake experiments, the TMG/1-hexa-
nol solution initially becomes very viscous and hot while re-
taining a colorless appearance, which we also saw in the ab-
sorption of CO2 in our CO2BOL systems. We feel this obser-
vation was due to the chemical binding of SO2 to form the
SO2BOL ionic liquid. As more SO2 is bubbled through the
SO2BOL, the color changes to a bright yellow-orange and
the viscosity begins to decrease. It has been reported earlier
that physical dissolution of SO2 into ionic liquids results in a
color change to dark yellow and a decrease in viscosity.[36] It
has been shown that ionic liquids can physically absorb two
molar equivalents of SO2.[33,35] However, SO2BOLs are supe-
rior because they are capable of absorbing three molar
equivalents. SO2BOL systems differ from the ionic liquids
studied by the groups of Brennecke,[33] Wang,[35] and Ri-
beiro[36] because the SO2BOL is not an ionic liquid until one
equivalent of SO2 is chemically bound.
Gravimetric uptake of acid gases: COS is gravimetrically ab-
sorbed by a DBU/1-hexanol mixture in analogy to the
uptake of CO2. DBU/1-hexanol was able to absorb
ACHTUNGTRENNUNG1.1 molequiv of COS gas, measured gravimetrically. 9 mmol
each of DBU and 1-hexanol (combined mass of 2.25 g) ab-
sorbed 10 mmol (0.60 g) of COS. The COS uptake was
mildly exothermic and generated a lime green viscous liquid
(Figure 3). The characteristic increase in viscosity as seen in
our CO2BOL system is consistent with the formation of an
ionic liquid and not just physical adsorption of the COS.
A DBU/1-hexanol mixture also reacts with 1 molequiv of
CS2 to yield the CS2BOL salt. CS2 is unlike CO2, COS, and
SO2 in that it is a liquid that boils at 468C. CS2 uptake by
DBU and 1-hexanol greater than 1 molequiv was not stud-
ied because liquid CS2 is soluble in the CS2BOL liquid. The
addition of 9 mmol of CS2 liquid to 9 mmol of DBU and 1-
hexanol was extremely exothermic, coinciding with a rapid
Chem. Eur. J. 2009, 15, 7619 – 7627
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7623

D. J. Heldenbrant et al.
For all acid gases used in this study, the DBU/1-hexanol
solvent pair always remained liquidlike when binding the
acid gases (Figure 3). Unlike the CO2 systems, COS and CS2
and SO2 caused noticeable color changes in DBU and 1-hex-
anol mixtures. The difference in color between the different
acid gases with DBU and 1-hexanol suggests the color is
specific to the anion. Ionic liquids with nitrosomethanide
anions have similarly been shown to exhibit vivid red and
blue colors.[37] Schultzꢄs group attributed the colors of those
ionic liquids to weak n!p* HOMO–LUMO electronic tran-
sitions.[37] Xanthate ions have been shown to have visible
colors, attributed to a lone pair on sulfur that donates elec-
trons to an antibonding p orbital.[38] The colors of the COS
derivatives are likely due to similar electronic transitions
from sulfur within the anion. The color in the SO2 deriva-
tives may be a combination of both the electronic transitions
from sulfur within the anion and the interactions of physi-
cally absorbed SO2 with the anion.[36] It is possible that if
these colors are unique to the O-alkylxanthate, O-alkylthio-
carbonate, and O-alkylsulfite anions they could be used as
potential chemical sensors for acid-gas detection.
hexanol, losing 1 molequiv in 30 sec, showing first-order ki-
netics. COS was the next fastest gas to strip from DBU and
1-hexanol, similarly showing first-order kinetics; however,
only 60% of the COS was removed from the COSBOL at
1408C after 3 min. The incomplete stripping of COS in our
burette system is due to an established thermodynamic equi-
librium between the gaseous COS and the chemically bound
COS in the COSBOL. Complete removal of COS from the
COSBOL can be achieved by sparging a gas through the bu-
rette at these temperatures or heating the COSBOL to
higher temperatures.
Two cycles of absorption and release of COS from DBU
and 1-hexanol were performed, with thermal stripping at
1008C under N2. Minimal solvent loss (<5%) was observed
when a reflux condenser was employed during thermal strip-
ping of the COS to temperatures as high as 1408C. IR (Fig-
ures S1–4 in Supporting Information) and 13C NMR spectra
(Figure 5) were recorded of the DBU and 1-hexanol show-
Gas release: Thermal stripping of the COSBOL, CS2BOL,
and SO2BOL systems was performed to demonstrate their
applicability to gas capture and release. As mentioned in
our previous work, organic CO2BOLs were predicted to
have a much lower specific heat than aqueous amine sys-
tems, making them much more energy efficient in stripping
CO2.[1] This same property is also expected to hold true for
the organic SO2BOLs, COSBOLs, and CS2BOLs compared
with aqueous systems. The thermal stripping of each acid
gas with the exception of CS2 from the DBU/1-hexanol sol-
vent pair was demonstrated on our automated gas burette
system (Figure 4).[39]
Figure 5. 13C NMR of a neat 1:1 mixture of DBU and 1-hexanol. From
bottom to top: DBU and 1-hexanol, contacted with COS, heated to
1408C, and recontacted with COS.
ing complete absorption and stripping of COS for two
cycles. The 13C NMR spectra (Figure 5) show, from bottom
to top, a neat mixture of DBU and 1-hexanol; the DBU and
1-hexanol contacted with COS; the same mixture after ther-
mal release of COS (stripped at 1408C); and then that mix-
ture contacted with COS again. The COS peak is visible at
d=185 ppm only when the COSBOL is formed. The IR
spectra (Figures S2–4 in the Supporting Information) also
show the complete binding and dissociation of COS from
DBU and 1-hexanol.
The release of CS2 from DBU and 1-hexanol was
AHCTUNTGRENaNGUN chieved by heating the CS2BOL to 130–1508C and con-
densing the liberated CS2 with a jacketed condenser. The
13C NMR spectrum of the clear and colorless distillate
showed a singlet at d=195 ppm ([D3]MeCN) confirming it
to be CS2. No formal recycling experiments were performed
with CS2. However, it is expected to react similarly to COS.
Extreme care should be taken in the case of heating the
CS2BOL in the presence of moisture; heating CS2 in the
presence of H2O liberates COS and H2S, and both are
known to be toxic gases.[7] The hydrolysis of CS2 into COS
Figure 4. Removal of COS from DBU and 1-hexanol (HexOH), SO2
from Hꢃnigꢄs base and 1-hexanol, SO2 from DBU and 1-hexanol, and
CO2 from DBU and 1-hexanol at 1408C. Heating began at 1 min.
All three acid gases were measured for gas stripping at
1408C (Figure 4). Each ionic liquid contained exactly
ACHTUNGTRENNUNG
ꢁ
and H2S proceeds easily because the C S bond in thiobicar-
7624
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7619 – 7627

CO2-Binding Organic Liquid Systems
FULL PAPER
ꢁ
bonate is replaced with a much stronger C O bond. The al-
coholysis of CS2 with alcohols does not proceed to make
COS and thiols in analogy to the hydrolysis reaction be-
ice bath used to chill the receiving flask. We were able to re-
cover 85% of the Hꢃnigꢄs base and 1-hexanol from the
vacuum distillation, with the remaining 15% of solvent re-
covered as the SO2BOL in the cold trap of the vacuum
(ꢁ1968C). Attempts to use a colder cooling bath for the
catch flask resulted in condensing the evolved SO2 (b.p.
ꢁ108C) in the catch flask, which then reacted with the
Hꢃnigꢄs base and 1-hexanol and subsequently re-formed the
SO2BOL. The 1-hexanol and Hꢃnigꢄs base collected from
the vacuum distillation readily absorbed SO2, indicating no
loss in activity. IR spectra (Figure S9–10 in the Supporting
Information) recorded during the stripping experiment con-
firmed the absorption and release of SO2. The NMR spectra
showed no significant changes with or without SO2 present.
At this time, the stripping of SO2 is unoptimized because of
the unique physical properties of SO2. We are investigating
the effect of higher boiling bases and alcohols to improve
separation of SO2 as well as conducting formal recycling ex-
periments of the SO2BOL system.
For SO2 capture and release in a SO2BOL system, strong
bases such as DBU and TMG will not be useful due to their
excessively strong binding of SO2. Less basic simple amines
such as Hꢃnigꢄs base are more practical choices. Also,
simple amines such as Hꢃnigꢄs base do not form a CO2BOL
or react with CO2 and H2O as discussed in our previous
studies.[1] The demonstrated selectivity to SO2 over CO2 by
simple amines make SO2BOLs attractive SO2 desulfuriza-
tion solvents for purifying a CO2 stream.
ꢁ
ꢁ
cause the R CH2 O bond in the alcohol is much less readily
ꢁ
cleaved than the H O bond in water.
The release of the two molar equivalents of physically ab-
sorbed SO2 from the SO2BOLs was rapid and first order
with respect to SO2. The physically absorbed SO2 was ther-
mally stripped from TMG/1-hexanol and DBU/1-hexanol at
varied temperatures (100 to 1508C) rapidly. Placing the cor-
responding SO2BOL under vacuum overnight at room tem-
perature also stripped the physically absorbed SO2. Once
the physically absorbed gas had been removed, none of the
SO2BOL liquids showed any further mass loss, even after
36 h under vacuum, thereby indicating that the one equiva-
lent of SO2 was chemically bound and that SO2BOLs have
no appreciable vapor pressure at room temperature.
The thermal release of chemically bound SO2 from
SO2BOLs was more difficult to accomplish than the removal
of the dissolved SO2. Neither DBU/1-hexanol nor TMG/1-
hexanol mixtures released chemically bound SO2 at temper-
atures as high as 1508C, indicating a much stronger chemical
binding of SO2 than CO2, COS, and CS2. The strength of
acid-gas binding in these systems appears to be linearly cor-
related to the Lewis acidity of the acid gas in decreasing
order: SO2 >CS2 >COS>CO2. Although amidines and gua-
nidines did not release chemically bound SO2 at tempera-
tures less than 1508C, a less basic tertiary amine, such as
Hꢃnigꢄs base when paired with 1-hexanol, did release the
chemically bound SO2 at 1408C (Figure 4). In the desulfoxy-
lation of the Hꢃnigꢄs base and 1-hexanol SO2BOL, a large
reflux condenser (flowing ethylene glycol at 58C) and a
second liquid trap were employed to capture any evaporated
Hꢃnigꢄs base and prevent vapors of the base from contribu-
ting to the volume of evolved gas measured in the burette.
40% of SO2 was stripped at 1408C, which is due to the ther-
modynamic equilibrium between gaseous SO2 and chemical-
ly bound SO2 in the SO2BOL.
In all, COS, CS2, and SO2 can be thermally released from
a base and alcohol blend analogous to CO2 in our CO2BOL
systems, making them applicable to all acid-gas capture and
release. We have observed minimal liquid loss during ther-
mal stripping of acid gases in this study. Liquid losses can be
reduced further by reducing the volatility of the base and al-
cohol components by either chemical modification or in-
creasing their molecular weight. Formal lifetime and ther-
modynamic measurements of the COSBOL, CS2BOL, and
SO2BOL systems are underway.
The Hꢃnigꢄs base binds SO2 more weakly than the DBU
and TMG SO2BOLs. The Hꢃnigꢄs base SO2BOL released
SO2 at temperatures above 908C when heated under an N2
atmosphere. The ionic liquid re-formed when the compo-
nents were cooled to room temperature. The rapid breaking
and re-forming of this SO2BOL made it appear to be “dis-
tilled” at approximately 1558C with no appreciable solvent
loss. DimCarb (comprised of dimethylamine and CO2) is the
only “distillable ionic liquid” currently known.[40–42] It is
likely that multiple SO2BOLs will be distillable because SO2
has a very high density (2.6 gmLꢁ1) and boils at ꢁ10 8C,
thus making it unlikely to leave a flask under conditions of
reflux. At temperatures above 1558C the Hꢃnigꢄs base/1-
hexanol SO2BOL begins to turn brown and degrade. The
degradation is accelerated by the presence of air, likely due
to oxidation.
Conclusion
CO2BOL systems have been shown to chemically react with
three sulfur-containing acid gases. COS, CS2, and SO2 react
in the same manner as CO2 towards amidine/alcohol and
guanidine/alcohol blends. These systems are the first to
show reversible binding and release of COS and CS2. The
capture of the acid gases with DBU and 1-hexanol results in
viscous, colored ionic liquids (green=COS, orange=SO2,
red=CS2) containing novel O-alkylxanthate, O-alkylthiocar-
bonyl, and O-alkylsulfite anions. These new colored ionic
liquids hold potential as chemical sensors for specific acid
gases. Each SO2BOL, COSBOL, and CS2BOL was charac-
terized by spectroscopic and conductivity measurements to
confirm the chemical binding of the acid gases to the alkox-
ide anion. Elemental analysis also confirmed the chemical
composition of DBU and 1-hexanol COSBOL, CS2BOL,
Complete removal of SO2 from Hꢃnigꢄs base and 1-hexa-
nol was accomplished by vacuum distillation. SO2 was dis-
tilled off at 608C using a short-path distilling head with an
Chem. Eur. J. 2009, 15, 7619 – 7627
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7625

D. J. Heldenbrant et al.
flask. The system was purged with N2 and then the components were ana-
lyzed gravimetrically and by IR and NMR spectroscopy.
and SO2BOL ionic liquids. Each acid gas was thermally
stripped from the solvent, showing the potential for reversi-
ble COS, CS2, and SO2 capture and release similar to our
CO2BOL system. Hꢃnigꢄs base and 1-hexanol was shown to
be selective for SO2 and not CO2, thus making it an attrac-
tive SO2 scrubber. The Hꢃnigꢄs base and 1-hexanol SO2BOL
was also found to be “distillable” similarly to DimCarb. Ul-
timately, base/alcohol mixtures show promise for the capture
and release of four different acid gases, showcasing this new
class of acid-gas-specific scrubbing systems.
Elemental analysis samples were synthesized by bubbling (COS or SO2)
or syringing (CS2) into a mixture of DBU and 1-hexanol and then placing
the mixtures under vacuum to remove any physically adsorbed gas. Ele-
mental analysis was performed in duplicate by Columbia Analytical Serv-
ices.
Crystals of TMG bisulfite were grown by slowly bubbling SO2 through a
1:1 mixture of TMG and H2O. Crystals of CS2 and DBU and benzyl alco-
hol were grown by sealing a jar containing equimolar mixtures of DBU
and benzyl alcohol and CS2.
Experimental Section
Acknowledgements
This work was funded by the Pacific Northwest National Laboratoryꢄs
(PNNL) Energy Conversion Initiative (ECI), Internal Laboratory Direct-
ed Research and Development (LDRD). The Pacific Northwest National
Laboratory is operated by Battelle for the US Department of Energy.
The authors would like to thank Dr. Ruiyao Wang at Queens University
and Dr. Bill Dougherty at Villanova University for the X-ray analysis.
P.G.J. thanks the Natural Sciences and Engineering Research Council of
Canada and the Canada Research Chairs program.
Liquid chemical reagents were purchased from Aldrich Chemical Com-
pany. All bases were distilled from CaH2 under an N2 atmosphere and
then stored over 4 ꢂ molecular sieves. Alcohols were distilled from Mg
activated with I2 under an N2 atmosphere and stored over 4 ꢂ molecular
sieves. All handling of reagents was performed under an N2 atmosphere
in a dry box. SO2, COS, and CS2 were also purchased from Aldrich
Chemical Company and used without purification. The moisture content
of the alcohols and bases was measured using a Mettler Toledo DL-37
Karl Fischer Titrator. All reagents had moisture content below d=
10 ppm. A Varian 300 MHz spectrometer was used to acquire 1H NMR
and 13C NMR spectra.
[1] D. J. Heldebrant, C. R. Yonker, P. G. Jessop, L. Phan, Energy Envi-
ron. Sci. 2008, 1, 487–493.
[2] P. G. Jessop, D. J. Heldebrant, X. W. Li, J. Lu, J. P. Hallett, R. S.
Jones, P. Pollet, C. A. Thomas, C. A. Eckert, C. L. Liotta, Abstr. Pap.
Am. Chem. Soc. 2005, 229, U971–U971.
[3] L. Phan, D. Chiu, D. J. Heldebrant, H. Huttenhower, E. John, X. W.
Li, P. Pollet, R. Y. Wang, C. A. Eckert, C. L. Liotta, P. G. Jessop,
Ind. Eng. Chem. Res. 2008, 47, 539–545.
[4] N. J. Blake, D. G. Streets, J.-H. Woo, I. J. Simpson, J. Green, S. Mei-
nardi, K. Kita, E. Atlas, H. E. Fuelberg, G. Sachse, M. A. Avery,
S. A. Vay, R. W. Talbot, J. E. Dibb, A. R. Bandy, D. C. Thornton,
F. S. Rowland, D. R. Blake, J. Geophys. Res. 2004, 109, D15S05.
[5] R. J. Gleason, M. Richman, P. E. Cooke, Office, U.P.a.T., Ed., Re-
search-Cottrell, Inc., USA, 1982.
[6] A. Turbak in Encyclopedia of Polymer Science and Engineering, 2nd
ed., Vol. 14 (Eds.: H. F. Mark, N. M. Bikales, C. G. Overberger, G.
Menges), Wiley Interscience, New York, 1988, pp. 45–72.
[7] M. D. S. Lay, M. W. Sauerhoff, D. R. Saunders in Ullmanꢀs Encyclo-
pedia of Industrial Chemistry, Wiley-VCH, Weinheim, 2005.
[8] R. D. Ashford, Dictionary of Industrial Chemicals, 2nd ed., Wave-
length Publications, London, 1994.
[9] A. B. M. Heesink, W. P. M. Vanswaaij, Chem. Eng. Sci. 1995, 50,
2983–2996.
[10] A. Bandyopadhyay, M. N. Biswas, Chem. Eng. Commun. 2006, 193,
1562–1580.
Conductivity measurements were performed in a dry box at room tem-
perature inside a stirred (250 rpm) Parr 160 mL stainless steel pressure
vessel modified to include an Omega CDCE-90-1 10–10000 mS conductiv-
ity probe and an Omega PX01CO-200 A5T 0–200 psi pressure transduc-
er. Acid-gas binding was measured in 0.1m solutions in acetonitrile
(MeCN) by determining increases in conductivity and decreases in pres-
sure. COS and SO2 were delivered to the pressure cell directly from lec-
ture bottles through stainless steel lines. CS2 was delivered to the pres-
sure cell with an internal solvent delivery system as described in our pre-
vious studies.[1]
Liquid reagents were syringed into oven-dried single-neck round-bot-
tomed flasks fitted with a rubber septa pierced with a syringe needle con-
nected to an N2 line. The flasks were cooled to 08C and then the liquid
components were sparged with each acid gas delivered by a regulator
connected to a syringe needle, which was pierced through the rubber
septa on the flasks. The flasks were sparged for 10 min to form the ionic
liquids. After the elapsed time had finished, the flasks were flushed with
N2 to displace any residual acid gas, then sealed and weighed.
Stripping of acid gases from the ionic liquid solution was performed
using an automated burette system.[39] All acid-gas and liquid mixtures
were syringed into an oven-dried round-bottomed flask, which was then
sealed and weighed. Acid gases were sparged through the solution for
5 min at 08C. The flasks were then capped and weighed again for gas
uptake. For experiments testing the thermal stripping of the chemically
bound acid gases, the ionic liquids were placed under vacuum overnight
to remove any physically absorbed gases. The flasks were then weighed
again and attached to the burette system, opened, and the flasks were
then submerged to their neck in a preheated oil bath.
[11] E. R. Lindgren, D. W. Pershing, D. A. Kirchgessner, D. C. Drehmel,
Environ. Sci. Technol. 1992, 26, 1427–1433.
[12] G. W. I. Luther III, Aquat. Geochem. 2004, 10, 81–97.
[13] J. P. Hallett, P. Pollet, C. L. Liotta, C. A. Eckert, Acc. Chem. Res.
2008, 41, 458–467.
The vacuum “distillation” of the Hꢃnigꢄs base and 1-hexanol SO2BOL
was performed using a jointed 14/20 short path distillation kit. The
SO2BOL was formed as above at 08C and then pumped under vacuum to
remove any physically absorbed gases. The flask was backfilled with N2
and then weighed. The flask was then connected to the short-path distil-
lation head fitted with a 25 mL round-bottomed catch flask chilled to
08C. The SO2BOL was then placed under full vacuum, and the flask was
dipped into a preheated oil bath set to 608C. Room temperature water
was recirculated inside the short-path distillation head. After 5 min, the
SO2BOL had been completely distilled out of the distillation flask and
had been collected as a clear and colorless nonviscous liquid in the catch
[14] R. R. Weikel, J. P. Hallet, C. L. Liotta, C. A. Eckert, Top. Catal.
2006, 37, 75–80.
[15] X. F. Xie, C. A. Eckert, C. L. Liotta, Ind. Eng. Chem. Res. 2004, 43,
2605–2609.
[16] P. G. Jessop, D. J. Heldebrant, X. W. Li, C. A. Eckert, C. L. Liotta,
Nature 2005, 436, 1102–1102.
[17] A. Earnshaw; N. Greenwood, Chemistry of the Elements, 1st ed.,
Permagon, Oxford, 1994.
[18] D. Vinci, M. Donaldson, J. P. Hallett, E. A. John, P. Pollet, C. A.
Thomas, J. D. Grilly, P. G. Jessop, C. L. Liotta, C. A. Eckert, Chem.
Commun. 2007, 1427.
7626
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7619 – 7627

CO2-Binding Organic Liquid Systems
FULL PAPER
[19] N. Jiang, D. Vinci, C. L. Liotta, C. A. Eckert, A. J. Ragauskas, Ind.
Eng. Chem. Res. 2008, 47, 627–631.
[33] J. L. Anderson, J. K. Dixon, E. J. Maginn, J. F. Brennecke, J. Phys.
Chem. B 2006, 110, 15059–15062.
[20] L. R. Drake, S. C. Stowe, A. M. Partansky, J. Am. Chem. Soc. 1946,
68, 2521–2524.
[34] J. Huang, A. Riisager, R. W. Berg, R. Fehrmann, J. Mol. Catal. A
2008, 279, 170–176.
[21] S. Sato, K. Suzuki, K. Kawaguchi, 2008, JP 2008-182221.
[22] D. J. Heldebrant, P. G. Jessop, C. A. Thomas, C. A. Eckert, C. L.
Liotta, J. Org. Chem. 2005, 70, 5335–5338.
[23] S. Sakai, Y. Kobayashi, Y. Isii, J. Org. Chem. 1971, 36, 1176–1180.
[24] A. J. Beeler, A. M. Orendt, D. M. Grant, P. W. Cutts, J. Michl, K. W.
Zilm, J. W. Downing, J. C. Facelli, M. S. Schindler, W. Kutzelnigg, J.
Am. Chem. Soc. 1984, 106, 7672–7676.
[25] G. A. Olah, M. R. Bruce, F. L. Clouet, J. Org. Chem. 1981, 46, 438–
442.
[26] G. Barany, A. L. Schroll, A. W. Mott, D. A. Halsrud, J. Org. Chem.
1983, 48, 4750–4761.
[35] Y. Wang, H. Pan, H. Li, C. Wang, J. Phys. Chem. B 2007, 111,
10461–10467.
[36] L. J. A. Siqueira, R. A. Ando, F. F. C. Bazito, R. M. Torresi, P. S.
Santos, M. C. C. Ribeiro, J. Phys. Chem. B 2008, 112, 6430–6435.
[37] H. Brand, J. F. Liebman, A. Schultz, P. Mayer, A. Villinger, Eur. J.
Inorg. Chem. 2006, 4294–4308.
[38] K. Takahashi, J. Am. Chem. Soc. 1976, 98, 7488–7498.
[39] S. R. F. Zheng, D. J. Heldebrant, D. D. Caldwell, C. L. Aardahl, T. S.
Autrey, J. C. Linehan, Rev. Sci. Instrum. 2008, 79, 084103.
[40] A. I. Bhatt, A. M. Bond, D. R. MacFarlane, J. Zhang, J. L. Scott,
C. R. Strauss, P. I. Iotov, S. V. Kalcheva, Green Chem. 2006, 8, 161–
171.
[27] A. M. Hounslow, S. F. Lincoln, R. Edward, T. Tiekink, J. Chem. Soc.
Dalton Trans. 1989, 233–236.
[28] E. Kesters, S. Gillissen, F. Motmans, L. Lutsen, D. Vanderzande,
Macromolecules 2002, 35, 7902–7910.
[41] P. G. Jessop, Y. Hsiao, T. Ikariya, R. Noyori, J. Am. Chem. Soc.
1994, 116, 8851–8852.
[42] W. Schroth, H. D. Schaedler, A. Joerg, Z. Chem. 1989, 29, 56–57.
[29] M. L. Shankaranarayana, C. C. Patel, Can. J. Chem. 1961, 39, 1633.
[30] F. G. Pearson, R. B. Stasiak, Appl. Spectrosc. 1957, 112, 116–120.
[31] L. H. Little, G. W. Poling, J. Leja, Can. J. Chem. 1961, 39, 745–754.
[32] A. Simon, H. Kriegsmann, H. Dutz, Chem. Ber. 1956, 89, 2390–
2396.
Received: December 11, 2008
Revised: March 6, 2009
Published online: June 23, 2009
Chem. Eur. J. 2009, 15, 7619 – 7627
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7627