926
A. Hubbard, T. Okazaki, and K. K. Laali
the computed energetics and the ability to produce three moles
of ArCl from one mole of TCICA.
0.133 mol (1 equiv); seeTables 1 and 2) and the Schlenk tube was
heated at ∼50◦C for the specified period, during which ICA by-
product appeared as a white solid (see recycling). The reaction
was worked up by addition of diethyl ether (3 × 0.5 mL), which
formed a separate layer from the IL. The biphasic mixture was
stirred vigorously and allowed to separate out before removing
the ether phase. The ether extract was neutralized with sodium
bicarbonate and filtered before analysis by GC or GCMS.
Conclusions
The utility of imidazolium-based Brønsted-acidic IL
[BMIM(SO3H)][OTf] as catalyst and solvent for electrophilic
chlorination of representative aromatic compounds withTCICA
has been demonstrated via a survey study. The reactions were
performed under mild conditions in good to excellent yields,
dependingonthesubstrates.Theabilitytosynthesizethreemoles
of ArCl from one mole of TCICA, the simple product isola-
tion step, facile recovery and recycling of the IL, and the use of
TCICA as a readily available, cheap chlorinating agent, are some
of the positive attributes of this method. The reaction generates
ICA as by-product, which is easily recovered from the IL. The
method avoids the use of chlorine gas and strong/corrosive acids.
Unlike the currently available methods, it does not produce large
quantities of acid by-products that need to be neutralized and dis-
posed of. Chemoselectivity towards mono- and dichlorination
can be controlled by adjusting the reaction conditions. Based
on substrate selectivity measurements, and a model DFT study,
a protosolvated trication is suggested as a key intermediate,
undergoing sequential transfer-chlorination/N-protonation.
Recycling of [BMIM(SO3H)][OTf] IL
Previously used IL containing ICA by-product was extracted
with methanol (3 × 1.0 mL) and transferred into a clean Schlenk
tube via syringe. Methanol was removed from the IL under vac-
uum and the acidity of the IL was checked with litmus paper.
Recycled ILs of pH 3 or lower were reused. Control experi-
ments showed no noticeable decrease in the conversions with
the recycled IL.
Yield Determination by GC
The reaction mixtures were directly analyzed by GC. In each
case, the GC conversions were determined by taking an aver-
age of three injections. Response factors were determined by
using 1:1 equimolar mixtures of authentic samples of products,
formed in various reactions. On this basis, a correction factor
of 3% was applied to the dichloroarene products. The response
factors determined for all other compounds were found to be
good within 1% and did not require correction.
Experimental
General
All reagents employed in the present study were high-purity
commercial samples that were used without further purification.
Reactions were carried out in small Schlenk tubes under a nitro-
gen or argon atmosphere. Diethyl ether used for extraction was
distilled over magnesium sulfate or from benzophenone/sodium
before use.
Gas chromatography analyses were performed on an HP
5890 instrument (using a capillary OV101 column). Gas
chromatography–mass spectrometry (GCMS) analyses were
performed on a Thermoquest Trace gas chromatograph cou-
pled to a Finnigan Polaris mass spectrometer. NMR spectra
were recorded on a 400 MHz Bruker instrument. Infrared (IR)
spectra were obtained on a Perkin Elmer Fourier-transform IR
instrument.
Product Confirmation
Isocyanuricacid:[21] δH ([D6]DMSO)11.16(s, 3H). Chloroethyl-
benzene: GCMS m/z 140.0, 142.0 (M+), 125.1 (M+ − CH3),
105.2 (M+ − Cl). Dichloroanisole: GCMS m/z 176.0, 178.0,
180.0 (M+). Trichloroanisole: GCMS m/z 209.9, 211.9,
213.9 (M+). Chloro-p-xylene: GCMS m/z 140.0, 142.0
(M+), 105.1 (M+ − Cl). Dichloro-p-xylene: GCMS m/z
174.0, 175.9, 178.0 (M+), 139.1, 141.0 (M+ − Cl), 103.1
(M+ − HCl2). Chloro-p-fluoroanisole: GCMS m/z 160.1, 162.1
(M+). Dichloro-p-fluoranisole: GCMS m/z 193.9, 196.0,
198.0 (M+). Chloro-p-diethylbenzene: GCMS m/z 168.1,
170.1 (M+), 133.1 (M+ − Cl). Dichloro-p-diethylbenzene:
GCMS m/z 202.1, 204.1, 206.0 (M+), 187.0, 189.0 (M+ − CH3),
167.1, 169.1 (M+ − Cl). Chlorodurene: GCMS m/z 168.0,
170.0 (M+), 133.1 (M+ − Cl). Dichlorodurene: GCMS m/z
202.0, 204.0, 206.0 (M+), 167.1, 169.1 (M+ − Cl). Trichloro-
durene: GCMS m/z 236.0, 238.0, 240.0 (M+), 201.0,
203.0, 204.0 (M+ − Cl). Chloromesitylene: GCMS m/z 154.0,
156.0 (M+), 119.1 (M+ − Cl). Dichloromesitylene: GCMS
m/z 188.0, 190.0, 192.1 (M+), 153.3, 155.1 (M+ − Cl),
115.2 (M+ − H3Cl2). Trichloromesitylene: GCMS m/z 222.0,
224.0/226.0(M+), 187.0, 189.0(M+ − Cl), 151.0(M+ − HCl2),
115.1 (M+ − H2Cl3).
Synthesis of Brønsted-Acidic IL [BMIM(SO3H)][OTf][7]
N-Methylimidazole (98.2 mmol, 7.8 mL) was stirred (with-
out any added solvent) with 1,4-butane sultone (98.2 mmol,
10.0 mL) at room temperature for 2 days, whereby a solid
mass was formed. It was washed repeatedly with diethyl ether
and with toluene, and dried under vacuum to give the corre-
sponding zwitterion in 94.7% yield. A stoichiometric amount
of triflic acid (92.9 mmol, 8.2 mL) was added and the reac-
tion mixture was stirred at 40◦C for 2 days during which time
the zwitterion dissolved/liquefied, resulting in the formation of
[BMIM(SO3H)][OTf] in near-quantitative yield. The IL was
washed repeatedly with toluene and with ether and dried under
Identity of all other products was confirmed by GC co-
injection with authentic samples. Isomer distributions were
confirmed by GC coinjection with authentic samples.
1
vacuum. The purity of the IL was assessed by H NMR spec-
troscopy. Its NMR data were in agreement with the reported
values.[7b,10b]
Computational Protocols
Structures were optimized by the DFT method at B3LYP/
6–31+G(d,p) level using the Gaussian 03 package.[22,23] All
computed geometries were verified by frequency calculations
to have no imaginary frequencies. Global minima were located
by changing the conformation (by varying the directions of the
OH groups), and by comparing the energies of the optimized
TCICA Chlorination Procedure
The IL (3.33 mmol, 1.23 g) was charged into a Schlenk tube
equipped with a magnetic stirring bar under a nitrogen atmos-
phere and TCICA (0.133 mmol, 31.0 mg) was added. The aro-
matic substrate was then introduced (0.40 mmol (3 equiv) or