The reaction of primary aromatic amines with alkylene carbonates for the selective synthesis of bis-N-(2-hydroxy)alkylanilines: The catalytic effect of phosphonium-based ionic liquids

Article abstract of DOI:10.1039/c0ob00105h

At T ≥ 140 °C, different primary aromatic amines (pX-C ₆H₄NH₂; X = H, OCH₃, CH₃, Cl) react with both ethylene- and propylene-carbonates to yield a chemoselective N-alkylation process: bis-N-(2-hydroxyalkyl)anilines [pX-C ₆H₄N(CH₂CH(R)OH)₂; R = H, CH ₃] are the major products and the competitive formation of carbamates is substantially ruled out. At 140 °C, under solventless conditions, the model reaction of aniline with ethylene carbonate goes to completion by simply mixing stoichiometric amounts of the reagents. However, a class of phosphonium ionic liquids (PILs) such as tetraalkylphosphonium halides and tosylates turn out to be active organocatalysts for both aniline and other primary aromatic amines. A kinetic analysis monitored by ¹³C NMR spectroscopy, shows that bromide exchanged PILs are the most efficient systems, able to impart a more than 8-fold acceleration to the reaction. The reactions of propylene carbonate take place at a higher temperature than those of ethylene carbonate, and only in the presence of PIL catalysts. A mechanism based on the Lewis acidity of tetraalkylphosphonium cations and the nucleophilicity of halide anions has been proposed to account for both the reaction chemoselectivity and the function of the catalysts.

Full text of DOI:10.1039/c0ob00105h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
The reaction of primary aromatic amines with alkylene carbonates for the
selective synthesis of bis-N-(2-hydroxy)alkylanilines: the catalytic effect of
phosphonium-based ionic liquids†
Maurizio Selva,* Massimo Fabris, Vittorio Lucchini, Alvise Perosa and Marco Noe`
Received 10th May 2010, Accepted 7th July 2010
DOI: 10.1039/c0ob00105h
At T ≥ 140 ◦C, different primary aromatic amines (pX–C6H4NH2; X = H, OCH3, CH3, Cl) react with
both ethylene- and propylene-carbonates to yield a chemoselective N-alkylation process:
bis-N-(2-hydroxyalkyl)anilines [pX–C6H4N(CH2CH(R)OH)2; R = H, CH3] are the major products and
the competitive formation of carbamates is substantially ruled out. At 140 ◦C, under solventless
conditions, the model reaction of aniline with ethylene carbonate goes to completion by simply mixing
stoichiometric amounts of the reagents. However, a class of phosphonium ionic liquids (PILs) such as
tetraalkylphosphonium halides and tosylates turn out to be active organocatalysts for both aniline and
other primary aromatic amines. A kinetic analysis monitored by 13C NMR spectroscopy, shows that
bromide exchanged PILs are the most efficient systems, able to impart a more than 8-fold acceleration
to the reaction. The reactions of propylene carbonate take place at a higher temperature than those of
ethylene carbonate, and only in the presence of PIL catalysts. A mechanism based on the Lewis acidity
of tetraalkylphosphonium cations and the nucleophilicity of halide anions has been proposed to
account for both the reaction chemoselectivity and the function of the catalysts.
and both products are not flammable; therefore, they can be
used as stoichiometric reagents without added solvents; iii) when
Introduction
The hydroxyalkylation of primary aromatic amines is extensively
EC or PC serve as hydroxyalkylating agents, CO2 is the only
used, especially in medicinal chemistry, for the synthesis of
by-product. Thus alkylene carbonates can be safely used even
aminoalcohols as intermediates.1 Despite the broad scope, the
for large scale preparations, as documented by the extensive
literature reported over the years.8 Only few papers however,
claim the application of EC and PC for the hydroxyalkylation
of primary aromatic amines (Scheme 1, path a).9 In the presence
of different catalysts, the efficiency of the reaction is limited by
the dual electrophilic character of the cyclic carbonates EC and
PC which may cause competitive acylation (possibly followed by
cyclization)6 (Scheme 1, path b).
reaction poses a major concern from both safety and environmen-
tal standpoints: conventional N-hydroxyalkylation methods are
based on ethylene and propylene oxides which are highly toxic and
carcinogenic compounds.2 Moreover, the extreme flammability
of these oxides means low reaction temperatures or sealed
reactors are necessary.1,3 Alternative reagents such as ethylene
and propylene halohydrins [X(CH2)2OH, XCH2CH(OH)CH3; X =
Cl, Br] do not reduce the overall impact of the procedure;1a,4
halohydrins are also very toxic and corrosive products and their
use generates stoichiometric amounts of halogenated salts which
must be disposed of.5
Our long standing interest on green synthetic methodologies
promoted by organic carbonates,6 prompted us to look at alkylene
carbonates (ethylene carbonate and propylene carbonate: EC and
PC, respectively) as a valuable option for new hydroxyalkylation
protocols of anilines. Although the catalytic insertion of CO2
on alkylene oxides still represents the most important route
for the synthesis of both ethylene and propylene carbonates,7
nonetheless the latter offers a number of practical benefits over
their parent oxides as well as over halohydrins. Among them:
i) EC and PC are classified as irritant but non-toxic products;
ii) EC is a low-melting solid (35 ◦C), while PC is a liquid,
Scheme 1 Competitive reactions of anilines with alkylene carbonates.
This behaviour is quite general for the reaction of anilines with
both linear and cyclic dialkyl carbonates, where mixtures of N-
alkyl derivatives and carbamates are often obtained.10 Effective
solutions to improve the selectivity toward alkylation have been
recently proposed by us through the use of two classes of catalysts:
i) zeolites such as alkali metal exchanged faujasites (FAU11) and
more recently, ii) phosphonium-based ionic liquids (PILs). FAU
and PILs however, thanks to their different acid–base properties
and steric requisites, promote different reaction outcomes. FAU
catalyze the formation of mono-N-alkyl amines, while the more
active PILs steer the reaction towards the formation of bis-N-
alkyl derivatives. Scheme 2 describes some relevant examples: in
Dipartimento di Scienze Ambientali dell’Universita` Ca’ Foscari Calle Larga,
S. Marta 2137 – 30123, Venezia, Italy. E-mail: selva@unive.it
† Electronic supplementary information (ESI) available: 1H and 13C NMR
spectra of PILs 2–13, compounds 3a–d and 7a–b, and a synopsis of major
MS signals is reported for all reaction products 2a–d, 3a–d, 4a–d, 5a–d,
6a–b, 7a–b, 8a–b and 9a–b. See DOI: 10.1039/c0ob00105h
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5187–5198 | 5187
©

View Article Online
tosylate (TosO), methyl carbonate (OCO2Me), chloride (Cl) and
iodide (I) anions, has been explored. Bromide exchanged PILs
have proven to be the most efficient catalysts; while the variation
of the cation structure presented a moderate effect, if any. Accurate
kinetic profiles have been determined for the reaction of aniline
with EC, catalysed by PILs: since solventless conditions have
been used, the high concentrations of both reagents and products
allowed to monitor the process via 13C NMR spectroscopy. A
mechanistic hypothesis has been formulated to rationalise the
performance of the examined catalysts.
Results
The bis-N-(2-hydroxy)ethylation of aniline.
Scheme 2 Selective N-alkylation reactions catalysed by FAU and PILs.
The reaction of aniline (1a) with EC was selected as a model
to begin the investigation. Our previously reported results13
suggested the use of tri-isobutylmethylphosphonium tosylate [(i-
Bu)3PMe]Tos (PIL1) as a catalyst for a preliminary screening: this
salt had a structure similar to PILs active for the N-methylation
of primary aromatic amines with dialkyl carbonates (Scheme 2,
eqn 3), it was stable at a high temperature, commercially available
(in high-purity) and inexpensive. Two sets (a and b) of experiments
were carried out by varying the loading of both the catalyst and
ethylene carbonate, as well as the reaction temperature. In the first
set (a), a mixture of aniline (1a, 0.80 g, 8.6 mmol), and ethylene
carbonate (EC, in a two molar excess over aniline), was set to
react at 150 ◦C, in the presence of different amounts of PIL1 (the
molar ratio Q = PIL1 : 1a was varied between 0.05 and 0.75). All
reactions were stopped after 2 h and analysed by GC-MS. In the
second set (b), a mixture of aniline (0.8 g, 8.6 mmol), and PIL1
(0.1 molar equiv. with respect to aniline, Q = 0.1), was set to react
with increasing amounts of ethylene carbonate (the molar ratio
EC : 1a was 2, 2.2, and 4), at two different temperatures of 150 and
170 ◦C. The reactions were monitored by GC-MS at time intervals
of 2, 6, and 8 h. For both sets (a) and (b), the major products
were the mono- and the bis-N-(hydroxy)ethyl derivatives of aniline
(Scheme 4: compounds 2a and 3a, respectively). Traces (~1%) of N-
phenylmorpholine (4a) and 3-phenyloxazolidin-2-one (5a) along
with other unidentified by-products (total amount 1–5%) were
also detected. Compound 3a was isolated and fully characterised
by 1H and 13C NMR, while the structures of 2a and 4a and 5a were
assigned from their MS spectra and by comparison to authentic
commercial samples. An experiment was also conducted in the
absence of the salt PIL1. The results of sets (a) and (b) are reported
in Fig. 1 and Table 1, respectively. In the figure, the light gray
profile refers to the overall N-alkyl selectivity (SN-alk) expressed as:
SN-alk = [(2a+3a)/Conversion]¥100 (left ordinate), while the dark
gray profile is the ratio of bis- to mono-N-alkyl products, i.e.
(3a/2a) (right ordinate).
the presence of FAU, eqn (1) and (2) show the mono-N-alkylation
of anilines with dimethyl- and 2-(2-methoxyethoxy)ethylmethyl-
carbonates,10 with EC and PC,9a and with glycerol carbonate;12
while, eqn (3) refers to the PIL-promoted N,N-dimethylation of
primary aromatic amines with different methylalkyl carbonates.13
Excellent alkylation selectivities (up to 98%, at complete substrate
conversion) are achieved in all cases, although, the reactions show
a high activation energy irrespective of the catalyst (FAU or PILs),
which usually implies a temperature greater than 140 ◦C.
A more in-depth analysis of this scenario indicates that the
potential of cyclic alkylene carbonates as alkylating agents of
anilines is still largely unexplored; in addition, the use of PILs
as catalysts for this process, has been hardly investigated. In this
paper, we report that at temperatures in the range of 140–170 ◦C,
phosphonium-based ionic liquids catalyze the reaction of different
primary aromatic amines (pX–C6H4NH2; X = H, OCH3, CH3, Cl)
with both EC and PC, towards the formation of the corresponding
bis-N-hydroxyalkyl derivatives [pX–C6H4N(CH2CH(R)OH)2; R =
H, CH3] with a high selectivity (up to 97%) at complete conversion
(Scheme 3).
Scheme 3 Selective bis-N-hydroxyalkylations of anilines by ethylene and
propylene carbonates in the presence of PILs catalysts.
Interestingly, at 140 ◦C, the model reaction of aniline with
ethylene carbonate proceeds to the almost quantitative formation
of the bis-N-hydroxyethyl derivative, even without the assistance
of any catalyst/promoter.
Up to 13 different phosphonium salts have been used. They
have been prepared in order to guarantee an accessible syn-
thesis with high purity and high thermal and chemical sta-
bility, or selected from commercial sources. In particular, the
combination of four cations such as tri-i-butyl methylphos-
phonium [(i-Bu)3PMe], tri-n-butylmethyl phosphonium [(n-
Bu)3PMe], tri-n-hexylmethylphosphonium [(n-Hexyl)3PMe], tri-
n-octylmethylphosphonium [(n-Octyl)3PMe], with bromide (Br),
Scheme 4 Products observed in the reaction of aniline with ethylene
carbonate catalysed by PIL1.
5188 | Org. Biomol. Chem., 2010, 8, 5187–5198
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Table 1 The reaction of aniline (1a) with EC (EC) in the presence of [MeP(i-Bu)3]Tosa
Products (%)c
#
EC : 1a (mol : mol)b
T/◦C
150
t/h
Conversion (%, by GC)c
2a
3a
4a
5a
1
Othersd
Yield (3a,%)
1
2
3
4
5
6
7
8
2
2
6
2
6
2
6
2
6
8
2
6
9
74
95
72
97
97
99
97
98
98
96
99
98
67
51
64
50
49
23
46
17
9
5
38
6
2
5
2
1
4
150
45
46
74
48
79
86
41
81
95
1
1
2
170
1
2
2
1
2.2
170
1
1
1
9
2
2
1
2
74
61
10
11
12
4
170
52
15
2
1
a All reactions were carried out using a mixture of aniline (1a, 0.8 g) and compound PIL-1 in a 1 : 0.1 molar ratio. b The molar ratio ethylene
carbonate : aniline. c The reaction conversion and the amounts of products were determined by GC-MS analyses (columns 5 and 6, respectively).
The conversion was referred to aniline (the limiting reagent). d Total amounts of unidentified by-products.
significantly, but it was still in favour of the product 2a (50–51%;
entries 2 and 4).
The reaction outcome was greatly improved at 170 ◦C. Aniline
was detected in trace amounts (3–4%) after only 2 h (entries 5, 7
and 10), and, as the reaction proceeded further, compound 2a was
efficiently transformed in the bis-N-alkyl product 3a with yields
of 86–95% (by GC, entries 9 and 12).
◦
◦
At both 150 C and 170 C, the increase of the EC : 1a molar
ratio (from 2 to 4) moderately favoured the conversion of 2a to
the desired final product 3a (compare entries 1–2 to 3–4, and
entries 5–6 to 7–8 and 10–11). Nonetheless, when the reaction
mixtures were purified by FCC (gradient elution: MeOH–diethyl
ether/petroleum ether; from 0 : 1 : 4 to 1 : 7.5 : 1.5 v/v, total of
650 mL), EC was partly co-eluted with compound 3a; hence,
the higher the amount of ethylene carbonate, the lower the
final isolated yields (entries 9 and 12: yields of 74% and 61%,
respectively).16
Fig. 1 The reaction of aniline with ethylene carbonate (150 ◦C, 2h): effects
of the amount of PIL1.
The synthesis of phosphonium based ionic liquids (PILs)
Set (a) (Fig. 1). At 150 ◦C, the reaction proceeded even without
catalyst (Q = 0): the conversion of aniline was of 48% and the
mono-N-alkyl derivative 2a was formed almost quantitatively
(46%).14 However, the reaction outcome was improved by the
ionic liquid: the conversion was significantly larger (up to 73%)
with the addition of even small quantities of PIL1 (Q = 0.05–
0.1) and, by further increasing the Q ratio from 0.1 to 0.75, the
conversion finally rose up to a quantitative value. With respect to
the uncatalysed process, the presence of the ionic liquid promoted
the formation of both the mono- and the bis-N-alkyl derivatives
of aniline (compounds 2a and 3a). Although the overall N-alkyl
selectivity was very high (SN-alk = [(2a+3a)/conversion]¥100 = 96–
98%), mixtures of products 2a and 3a were always obtained. At Q
≥ 0.1, the (3a/2a) ratios were in the range of 0.2–0.33 (dark gray
profile, right ordinate).
PILs were prepared according to a new green protocol recently
reported by us.17 Four alkylphosphines (R3P; R = n-octyl, n-hexyl,
i-butyl, and n-butyl) were set to react with dimethyl carbonate
(MeOCO2Me) as a green methylating agent, to obtain methyl-
trialkylphosphonium methylcarbonate salts PIL5–8 [Scheme 5,
(a)]. These compounds were anion-exchanged by reaction with
Brønsted acids (H–A, A = TsO, Br, and I) [(Scheme 5, (b)] to yield
the corresponding tosylate, bromide, and iodide salts (PIL2–4,
PIL9–12, and PIL13, respectively) along with methyl hydrogen
Set (b), Table 1. Experiments were continued at a constant Q
ratio (PIL1 : 1a) of 0.1,15 by varying the temperature and EC:1a
ratio. At 150 ◦C, going from 2 to 6 h reaction times, a substantially
quantitative conversion (95–97%) was observed with an excellent
N-alkyl selectivity (SN-alk: 95–98%; entries 1–2 and 3–4). The ratio
of bis-N- to mono-N-alkyl compounds (3a/2a) also increased
Scheme 5 The synthesis of PILs.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5187–5198 | 5189
©

View Article Online
Table 2 List of ionic liquids synthesized in this work
Table 3 The reaction of aniline with ethylene carbonate at 170 ◦C, in the
presence of different PILsa
Anion
Products (%)b
Conversion
Cation
[TosO]
[OCO2Me]
[Br]
[I]
#
PIL
t/h (%)b
2a 3a 4a 5a Others
a
[i-Bu3PMe]
[n-Bu3PMe]
[n-Hex3PMe]
[n-Oct3PMe]
PIL5
PIL6
PIL7
PIL8
PIL9
PIL2
PIL3
PIL4
PIL10
PIL11
PIL12
1
2
3
4
5
6
7
8
9
PIL1 MeP(i-Bu)3 TosO
PIL2 MeP(n-Bu)3 TosO
2
4
2
4
97
99
80
96
85
91
97
100
98
49 46
25 68
57 17
42 41
56 18
45 34
48 51
36 58
13 64
1
1
1
PIL13
5
5
8
7
9
1
1
a [i-Bu3PMe][TosO] was commercially available (Aldrich).
3
4
2
PIL3 MeP(n-Hex)3 TosO 2
4
1
PIL4 MeP(n-Oct)3 TosO
PIL9 MeP(i-Bu)3 Br
2
4
2
2
2
2
carbonate (i.e., the half-ester of carbonic acid). The latter was
1
9
5
7
12
3
◦
unstable above -36 C18 and immediately decomposed to form
4
3
methanol and CO2, thus providing its built-in removal.
10 PIL10 MeP(n-Bu)3 Br
11 PIL11 MeP(n-Hex)3 Br
12 PIL12 MeP(n-Oct)3 Br
98
9
61 12
100
100
12 80
18 77
5
4
Overall, the two-step procedure allowed to synthesize 13
different PILs whose structures are summarized in Table 2.
These products were obtained on a 2–3 g scale. They were fully
characterized by 1H and 13C NMR in their natural state, not
dissolved in any solvent, at the lowest monitoring temperature
that kept them as liquids (see experimental for further details).
1
a All reactions were carried out at 170 ◦C, using a mixture of aniline (1a,
0.8 g), ethylene carbonate, and PIL in a 1 : 2 : 0.1 molar ratio, respectively.
b The reaction conversion and the amounts of products were determined
by GC-MS analyses. Others were unidentified by-products.
and PIL10, an apparent drop of the N-alkyl selectivity (SN-alk) was
observed (entries 9–10). This was mainly due to the formation
of N-phenylmorpholine 4a (9–12%). Two additional experiments
explained this result: under the conditions of entry 1 (Table 3),
aniline was set to react with ethylene carbonate in the presence of
water or of aq. HBr (both H2O and HBr were used in 10% mol with
respect to aniline). While the addition of water had no appreciable
effects on the reaction outcome, a catalytic amount of aq. HBr
prompted the transformation of compound 3a into 4a in 70% GC
yield, after 18 h (Scheme 6).19
The bis-N-(2-hydroxy)ethylation of aniline in the presence of
different ionic liquids
The reaction of aniline with ethylene carbonate was investigated
in the presence of the different PILs of Table 2. Prelimi-
nary experiments carried out with strongly basic IL such as
[i-Bu3PMe][OCO2Me] and [n-Oct3PMe] [OCO2Me] (PIL5 and
PIL8),17 gave complex mixtures of products. For this reason, the
use of the methylcarbonate exchanged PILs (PIL5–8) was not
further examined.
A first screening on the behaviour of the other PILs as catalysts
was carried out, based on the results of Table 1. In particular,
tosylate and bromide-exchanged salts (PIL2–4 and PIL9–12,
respectively), were used. The complete combination of four cations
(from C4 to C8: [i-Bu3PMe], [n-Bu3PMe], [n-Hex3PMe], and [n-
Oct3PMe] was available for these anions (TosO and Br, Table 2). A
mixture of aniline (1a, 0.8 g, 8.6 mmol), ethylene carbonate (EC)
and a PIL (the molar ratio 1a : EC : PIL was 1 : 2 : 0.1, respectively)
was set to react at 170 ◦C and the progress of the reaction was
monitored by GC-MS at time intervals in the range of 2–4 h.
Analogous to the commercial PIL1, the major products were the
desired mono- and the bis-N-(hydroxy)ethyl derivatives of aniline
(compounds 2a and 3a of Scheme 4). N-phenylmorpholine (4a)
and 3-phenyloxazolidin-2-one (5a) were also detected along with
other unidentified by-products (total amount 3–12%). The results
are reported in Table 3 which, for a more complete comparison,
includes also the data related to compound PIL1 (entries 5–6 of
Table 1).
Scheme 6 The reaction of aniline and ethylene carbonate with added
water or aq. HBr.
Traces of residual acidity (HBr) from the synthesis of ionic
liquids (Scheme 5), reasonably accounted for the presence of the
by-product 4a in Table 3. However, any attempt to purify PIL9
and 10 to improve their performance, was not successful.
The kinetic characterization of the reaction of aniline with EC
At 170 ◦C, after 2 h, the comparison of our model PILs showed
that: i) high aniline conversions (80–100%) were reached in all
cases; ii) bromide salts (PIL9–12) were by far, more efficient than
tosylate ones (3) for the transformation of the mono-N-alkyl
derivative 2a into the final desired bis-N-alkyl product 3a. The
ratio 3a/2a was 4–6.5 for bromides PIL9–12, (entries 9–12), while
it ranged between 0.3 and 1 for tosylates PIL1–4 (entries 1, 3, 5,
and 7). Overall, the bromide-catalysed processes were substantially
over after 2 h, while the tosylate-catalysed reactions were far from
complete even after 4 h (entries 2, 4, 6, and 8). In the case of PIL9
In the presence of small aliquots of PILs, mixtures of aniline and
EC were perfectly homogenous and retained their liquid state
even at room temperature. Hence, the alkylation of aniline with
ethylene carbonate (Scheme 7) could be adequately followed by 1H
decoupled 13C NMR spectroscopy, in a measuring tube equipped
with a sealed capillary of [D6]DMSO, for locking and homogeneity
purposes. In order to slow down the reaction and to make its
kinetic analysis easier, the experiments were performed at 140 ◦C,
by placing a mixture of 1a, EC, and a PIL (in a 1 : 2 : 0.05 molar
ratio, respectively), in a NMR tube. The tube was dipped into a
5190 | Org. Biomol. Chem., 2010, 8, 5187–5198
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
d[EC]/dt = -k1[EC][1a] -k2[EC][2a]
d[1a]/dt = -k1[EC][1a]
d[2a]/dt = k1[EC] [1a] - k2[EC][2a]
d[3a]/dt = k2[EC][2a]
No analytically integrated form of this system was available.
Therefore we resorted to the Runge–Kutta numerical integration
and to the simplex fitting thereof.20 For all the investigated
reactions, it emerged that: i) without the introduction of the
activity coefficients a(t), the routine was unable to reach a correct
fit, but ii) only one a(t) function was sufficient, holding for all
involved concentrations; iii) the optimized parameters k1, k2 and
a(t) were biased by high co-variance, and could not be utilized for
coherent comparisons within the examined panel of reactions.
Therefore we had to look for an independent method for
retrieving the correct values of k1 and k2. According to the initial
rate method:
Scheme 7 Consecutive reactions for the formation of compounds 2a and
3a.
thermostatted silicon oil bath. The highest possible concentrations
assured the swift acquisition of the signals, that were also well
separated. The modalities for the correct translation of the signal
intensities into molar concentrations were detailed elsewhere.20
The concentrations at t = 0 were corrected by a measured factor
of 1.10 due to the thermal expansion of the fluid. In addition, at
140 ◦C, the release of CO2 caused a significant volume shrinking
(Scheme 7). For example, a volume reduction by a factor of 0.71
was measured for the uncatalysed process after 8 h. Both chemical
and physical features evolved during the reaction time. As detailed
elsewhere,20 the chemical evolution was described by the kinetic
constants k1 and k2, the physical evolution by the time dependent
activity coefficients a(t).
d[EC]/dt = d[1a]/dt = -k1[EC]0[1a]0
However, the rates could not be derived from the first con-
centrations of EC and 1a, because of the action of the physical
evolution as represented by a(t), but rather from the first order
coefficient of the polynomial fittings (8th degree) of the whole sets
of these same concentrations. The first order coefficients of these
polynomials were the zero order coefficients of the derivatives, and
were the only surviving coefficients at zero time. They are reported
in Table 4. Their degree of similarity is a test for the procedure.
The k1 values are derived from their average. The k2 values are
measured at the time when the concentration of 2a reaches the
highest value. Here d[2a]/dt = 0 and, because of the uniqueness of
a(t), k2/k1 = [1a]/[2a].20
Fig. 2 exemplifies the behavior of the reaction of aniline and
EC catalysed by PIL11 [MeP(n-Hex)3Br]: the disappearance of
reactants, the transient formation of 2a and the final conversion
into 3a are quantitatively displayed.
The kinetic analysis qualitatively agreed with the general
behaviour of the PIL catalysts reported in Fig. 1 and Table 1
and 3. Moreover, the quantitative evaluation of k1 and k2 kinetic
constants helped to highlight some features of the process and
some differences among the PIL catalysts. In particular: i)
experiments confirmed that the reaction proceeded up to complete
conversion of aniline to the bis-N-alkyl derivative 3a, even in
the absence of PILs (entry 1). However, with respect to the non
catalysed process, the rate constant k1 was almost doubled by the
presence of tosylate salts (PIL1–4; entries 2–4); ii) the reaction
rate was considerably improved by bromide salts (PIL9–12) which
allowed an up to 8-fold increase of k1 (entries 4–8); iii) the k2/k1
ratio was between 0.25 and 0.37 (far right column), essentially
similar for all processes, either in the presence or in the absence
of the catalyst, meaning that the bis-N-alkylation of aniline was
always slower than the formation of the mono-derivative 2a, but
the relative rates of the two alkylation steps were not affected by the
nature of the catalysts (nor changed in their absence). In addition,
the k2/k1 ratio also suggested that the catalyst was operating on
the only specie present in both processes of Scheme 7, EC; iv)
within each set of tosylate and bromide salts, the cations had
poor, if any, effects: no appreciable variation of the reaction rate
was observed as the size of the alkyl P-substituents was increased
(compare entries 2–4 and 4–7); v) the iodide exchanged PIL13
(entry 8) offered an intermediate behaviour between PIL1 and
Fig. 2 The reaction at 140 ◦C of aniline 1a and ethylene carbonate
EC (3.70 and 7.35 mol L-1, respectively), in the presence of PIL11
([MeP(n-Hex)3Br], 0.19 mol L-1). Time dependence of the molarities of
EC, 1a, N-2-hydroxyethylaniline 2a, and bis-N-(2-hydroxyethyl)aniline 3a.
In analogy to Fig. 2, the same kinetic analysis was carried out for
reactions catalyzed by PIL1, PIL3–4, PIL9, PIL12–13 as well as
for the uncatalysed process. The kinetic constants k1 and k2 and the
time dependent activity coefficients a(t) were necessary parameters
for an acceptable fitting of the experimental concentrations into
the system of differential equations describing the reactions of
Scheme 7.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5187–5198 | 5191
©

View Article Online
Table 4 Kinetic analysis for the reaction of aniline 1a with EC, carried out both with and without PILs as catalyst.a
Initial concentration/mol L-1 Initial rate/mol L-1 h-1
-d[EC]/dt -d[1a]/dt
PIL
[EC]
[1a]
[PIL] (¥10)
k1 (¥102)/L mol-1 h-1
k2 (¥102)/L mol-1 h-1
k2/k1
None
i-Bu3PMe TosO
n-Hex3PMe TosO 7.19
n-Oct3PMe TosO 7.01
i-Bu3PMe Br
8.29
7.32
4.15
3.70
3.58
3.53
3.78
3.70
3.61
3.56
0.58
0.84
0.82
0.72
3.78
3.54
2.53
1.12
0.61 1.71
0.43
0.76
0.81
0.79
4.47
4.63
3.26
1.32
0.25
0.25
0.25
0.26
0.37
0.36
0.34
0.30
PIL1
PIL3
PIL4
PIL9
1.82
1.82
1.76
1.88
1.86
1.82
1.78
0.81 3.05
0.86 3.26
0.78 3.03
3.59 12.9
3.45 12.9
2.44 9.57
1.10 4.39
7.53
7.35
7.19
7.10
PIL11 n-Hex3PMe Br
PIL12 n-Oct3PMe Br
PIL13 n-Oct3PMe I
a All reactions were carried out at 140 ◦C, in a NMR tube charged with a mixture of 1a, EC, and a PIL (in a 1 : 2 : 0.05 molar ratio, respectively).
PIL9 which were less and more active catalysts than PIL13,
respectively. This further substantiated the role of the anion for
the reaction outcome.
N-[(p-substituted)phenyl]morpholines (4: 1–4%) and N-[(p-
substituted)phenyl]-oxazolidin-2-ones (5: 2–4%) along with other
unidentified by-products (total amount 2–8%) were also detected.
Compounds 3b–d were isolated and fully characterised by 1H and
13C NMR, while the structures of 2b–d, 4b–d and 5b–d, were
assigned from their MS spectra. The results are reported in Table 5
where, for a more complete comparison, also the data related to
aniline (Tables 1 and 2) are included.
Three main considerations emerged: i) the bromide exchanged
PIL was more active than the tosylate salt. At 150 ◦C for example,
both p-anisidine and p-toluidine were completely converted in
4–7 h and in 15–18 h, using PIL12 [MeP(n-Oct)3Br] and PIL1
[MeP(i-Bu)3TosO], respectively (entries 8–9 and 4–5). ii) The
reaction was of a general scope for different amines. On average,
p-anisidine (entries 3–4, and 8) was more reactive than p-toluidine
(entries 5 and 9), aniline (entries 1–2 and 7), and p-chloroaniline
(entry 6). iii) as for the case of aniline, the FCC-purification
of bis-N-alkyl derivatives 3b–d was rather difficult: these com-
pounds however, were isolated in reasonably good yields (62–74%;
entries 4–6).
The reaction of different anilines with ethylene carbonate
PIL1 and PIL12 [MeP(i-Bu)3TosO and MeP(n-Oct)3Br),
respectively] were chosen as catalysts for the reaction of primary
aromatic amines (p-anisidine, p-toluidine, and p-chloroaniline:
compounds 1b–d, respectively) with ethylene carbonate (EC). A
mixture of the primary amine (8.6 mmol), ethylene carbonate, and
a PIL in a 1 : 2.2 : 0.1 molar ratio, respectively, was set to react
at temperatures in the range of 150–170 ◦C. All reactions were
followed by GC-MS at different time intervals. Major products
were the mono- and the bis-N-(2-hydroxy)ethyl derivatives of the
reactant amines (Scheme 8: compounds 2 and 3, respectively).
The reaction of primary aromatic amines with propylene carbonate
Three different PILs, namely tri-i-butylmethylphosphonium to-
sylate and bromide [PIL1 and PIL9: MeP(i-Bu)3TosO and
MeP(i-Bu)3Br), respectively] and tri-n-octylphosphonium bro-
mide [PIL12: MeP(n-Oct)3Br)], were selected as catalysts for the
Scheme 8 The products observed in the reaction of different anilines with
ethylene carbonate catalysed by PIL1 and PIL12.
Table 5 The reaction of different anilines with ethylene carbonate catalysed by PIL1 and PIL12 a
Products (%) b
#
XC6H4NH2
Catalyst
T/◦C
t/h
Conversion (%)b
2
3
4
5
Others
3 (Yield, %) c
1
2
3
4
5
6
7
8
9
1a X = H
PIL1 [MeP(i-Bu)3] Tos
170
150
170
150
150
170
170
150
150
6
6
4
15
18
14
2
4
7
99
95
23
51
21
3
74
38
72
86
87
72
77
71
74
2
5
4
6
2
4
1a X = H
PIL1
1
2
4
2
3
1
2
1
1b X = p-MeO
1b X = p-MeO
1c X = p-CH3
1d X = p-Cl
1a X = H
1b X = p-MeO
1c X = p-CH3
PIL1
100
100
100
100
100
99
1
1
PIL1
74
70
62
PIL1
9
PIL1
17
18
16
13
4
4
2
4
PIL12 [MeP(n-Oct)3] Br
PIL12
PIL12
8
7
99
a All reactions were carried out using a mixture of the primary amine (8.6 mmol), ethylene carbonate, and a PIL in a 1 : 2.2 : 0.1 molar ratio, respectively.
b The reaction conversion was referred to the primary amine. Both the conversion and the amounts of products were determined by GC-MS analyses.
Others: total amounts of unidentified by-products. c Isolated yields of bis-N-alkyl derivatives 3b-d.
5192 | Org. Biomol. Chem., 2010, 8, 5187–5198
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Table 6 The reaction of different anilines with propylene carbonate in the presence of PIL catalystsa
Products (%)b
#
XC6H4NH2
Catalyst
None
T/◦C
t/h
Conversion (%)b
6
7
8
9
Others
7 (Yield %)c
1
2
3
4
5
X = H
170
190
190
170
170
16
10
54
35
40
1
3
98
96
99
1
3
11
13
9
PIL1 MeP(i-Bu)3 TosO
PIL9 MeP(i-Bu)3 Br
PIL12 MeP(n-Oct)3 Br
41
9
6
36
63
65
10
11
15
1
4
58
55
6
7
8
X = p-MeO
PIL1 MeP(i-Bu)3 TosO
PIL9 MeP(i-Bu)3 Br
PIL12 MeP(n-Oct)3 Br
170
150
150
21
29
24
89
100
100
48
6
6
18
60
58
4
3
6
1
6
7
18
25
23
a All reactions were carried out using a mixture of the primary amine (8.6 mmol), propylene carbonate, and a PIL in a 1 2.2 : 0.1 molar ratio, respectively.
b The reaction conversion was referred to the primary amine. Both the conversion and the amounts of products were determined by GC-MS analyses.
Others: unidentified by-products were constituted by two isomers (m/z = 193). c Isolated yields of bis-N-alkyl derivatives 7a–b
Scheme 9 The products observed in the reaction of different anilines with propylene carbonate catalysed by PIL1, PIL9, and PIL12.
reaction of primary aromatic amines (aniline and p-anisidine)
with propylene carbonate (PC). A mixture of the primary amine
(8.6 mmol), racemic propylene carbonate, and the chosen PIL in
a 1 : 2.2 : 0.1 molar ratio, respectively, were set to react at temper-
atures in the range of 150–190 ◦C. An additional experiment was
carried out also in the absence of PILs. All reactions were followed
by GC-MS. The asymmetry of PC implied the formation of a
number of products whose structures are outlined in Scheme 9.
Major products derived from the N-alkylation of the reactant
amines. The double nucleophilic attack of the amine at the C in
position 5 of PC gave rise to bis-N-(2-hydroxy)propyl derivatives
(compounds 7a–b) as a pair of diastereoisomers in a 1 : 1 ratio.
A non-negligible amount of other bis-N-alkylated compounds
originated also from the reaction of the amine at the highly
hindered C in position 4 of PC (compounds 7¢a–b): the ratio 7a–b/
7¢a–b was ~ 4–5.21 Two isomeric morpholines, i.e. 2,6-dimethyl-
and 2,5-dimethyl-N-[(p-substituted)phenyl] morpholines (com-
pounds 8a–b and 8¢a–b, respectively; total of 3–11%), as well as
two isomeric oxazolidin-2-ones, i.e. 5-methyl- and 4-methyl-N-
[(p-substituted)phenyl]-oxazolidin-2-ones (compounds 9a–b and
9¢a–b, respectively; total of 1–7%) were also detected.
The results are reported in Table 6.
In the absence of PILs, the conversion of aniline was negligible
even at high temperature (190 ◦C, entry 2). This remarkable
difference with respect to ethylene carbonate, was likely to be
due to the weaker electrophilic character of PC. However, the
use of a catalytic amount of PILs allowed the reaction of
aniline with propylene carbonate to proceed up to completion.
Bromide exchanged salts (PIL9 and 12) were also able to
promote the transformation of mono-N-alkyl products 6 to bis-
N-alkyl derivatives 7 (entries 4–5 and 7–8). This process instead,
was rather sluggish in the presence of the tosylated PIL1: for
example, in the reaction of aniline, the ratio 6a/7a was in
favor of the mono-alkyl compound (6a) even after 54 h at
190 ◦C (entry 3). A similar situation held true for p-anisidine
(entry 6).
In any case, the poorer reactivity of PC implied long re-
action times (>20 h). This increased the formation of by-
products at the point that the overall N-alkyl selectivity (SN-alk
=
[(6+7)/Conversion]¥100) could not exceed 75%. A value, by
far, lower than that achieved with ethylene carbonate (up to
97%, Table 5). However, to the best of our knowledge, this
was the first ever reported reaction in which organocatalysts,
particularly PILs, were able to activate PC as an alkylat-
ing agent of primary aromatic amines. Bis-N-alkyl deriva-
tives 7 were purified by FCC and the corresponding isolated
yields were in agreement to their GC amounts (entries 4
and 7).
Finally, pairs of isomeric unidentified by-products were ob-
served (m/z = 193 in the reaction of aniline; m/z = 223 in the
reaction of p-anisidine) in 11–25% yields. The isomeric ensemble
1
7a–b/7¢a–b were isolated and fully characterised by H and 13C
NMR, while the structures of 6, 8 and 9 were assigned from their
MS spectra.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5187–5198 | 5193
©

View Article Online
and oxygen atoms.24 An initial coordination of the P center of the
catalyst to the carboxylic oxygen of EC, produces the acid–base
adduct (I); a similar structure, with pentavalent phosphorus, has
been described in the literature.25
Discussion
The reaction of aniline with alkylene carbonates: the effects of
PILs
Although the carboxylic carbon of the complex (I) undergoes
electrophilic activation, no reaction occurs at this position because
of the rigid structure of EC (see above), and of the steric crowding
imposed by the bulky phosphonium cation.26 Aniline directs its
nucleophilic attack to the b-alkyl carbons (in positions 3 and 4)
of EC: both these atoms are rather accessible, and they experience
an activating inductive effect caused by the strong polarisation of
the C O bond in the complex (I). An amino carbonic acid (II)
forms. This unstable compound27 quickly releases CO2 to give the
final product 2a.
A first significant feature of the reaction of aniline with ethylene
carbonate comes from the analysis of Fig. 1 and Table 4. At
140–150 ◦C, the process not only proceeds in the absence of
catalysts but, under such conditions, it yields selectively the bis-
N-(2-hydroxy)ethyl derivatives of the amine with no concurrent
formation of carbamates. This behaviour is rather unexpected,
especially if compared to the literature data discussed in the
introduction (Scheme 1).9,10,14 Aniline is perfectly capable, as
a nucleophile, of discriminating between the two electrophilic
centres of ethylene carbonate (the alkyl and the carbonyl carbons,
respectively), with a preference for the b carbon. This preference is
not altered by the action of the PIL catalysts. The cyclic structure
of EC offers two plausible reasons for this result: i) on one hand,
the N-alkylation selectivity would be favoured by the rigid 5-
membered ring of EC which hinders access to the carbonyl carbon;
ii) on the other hand, the relief of some ring strain consequent
to the alkylation of aniline with EC, would allow even the non-
catalytic reaction to proceed. It should be noted that this last
effect is often offered to explain the higher reactivity of ethylene
carbonate with respect to linear dialkyl carbonates.22 Fig. 1 and
Tables 3 and 4 however, leave few doubts about the catalytic
action of PILs. In general, the alkylation rate of aniline with EC
is doubled by the use of tosylate salts (PIL1–4), while the reaction
is up to eight-fold accelerated by the more efficient bromide
exchanged PILs (PIL9–12). This effect is even more pronounced
when propylene carbonate is the alkylating agent (Table 6): N-(2-
hydroxy)propyl derivatives of aniline (6a and 7a) form only in the
presence of catalytic amounts of PILs, otherwise no reaction takes
place at all. Also in this case, bromide salts (PIL9 and 12) offer
a better performance than the tosylated PIL1. Table 4 however,
indicates that the relative rates (the ratio k2/k1) of mono- and bis-
N-alkylation steps of aniline is not substantially affected by the
nature of the catalysts. This suggesting that the catalyst should be
active on the only species present in both processes: EC. Hence,
a convenient reaction mechanism should consider the action of
both the cationic and the anionic parts of the PILs. Scheme 10
illustrates two plausible hypotheses (for simplicity, only the mono-
N-(2-hydroxy)ethylation of aniline is shown).
Path a) hardly explains why halide salts, more specifically
bromides, are remarkably more active than tosylate PILs. The size,
the polarisability, and the stability of anions alter the strength
of ionic pairs in PILs and therefore, the Lewis acidity of the
corresponding cations is also modified. However, a crucial role
is plausibly played by the nucleophilic properties of bromides
and tosylates. Path b) accounts for this last aspect in analogy
to a mechanism previously proposed for the reaction of EC with
carboxylic acids in the presence of tetraalkylammonium halides
([NR4 X-]) as catalysts.3 The following steps are considered: i)
+
the aperture of the ethylene carbonate ring takes place through
the attack of an halide anion (X-) to one b alkyl carbon, and I¢
is formed; ii) then, aniline displaces the X-group and the amino
carbonic acid II¢ is obtained; iii) finally, the evolution of CO2
from II¢ brings to the final derivative 2a. The greater nucleophilic
strength of bromide or iodide anions with respect to the tosylate
anions, makes path b) especially valid for PILs 9–12 and 13, but
not for the poor nucleophilic tosylate catalysts (PIL 1–4).28 This
may account for the activity displayed by the two types of PILs:
paths (a) and (b) may simultaneously operate with halides, while
only path (a) is active for tosylates.
The two PIL partners (cation and anion) only act on EC, and
therefore mechanisms (a) and (b) are valid for the attack of both
primary and secondary amines, 1a and 2a, to EC. This explains
why the ratios k2/k1 are similar for all PILs. Within these ratios,
k2 < k1: the steric features of the secondary aniline 2a seem to be
preeminent with regard to the enhanced nucleophilicity.
It should be noticed that species I–II and I¢II¢are intrinsically
short-lived compounds, especially under the conditions (140–
170 ◦C) explored here. This is plausibly the reason why the struc-
tures of these intermediates were never detected during the NMR
investigation of the reaction of aniline with EC catalysed by PILs.
The reaction of different anilines with EC and PC
Tables 5 and 6 show that both activated and scantly nucle-
ophilic amines react with alkylene carbonates to yield mainly
the corresponding bis-N-alkyl derivatives. Three major observa-
tions emerge: i) on average, the reactivity of different aromatic
amines agrees with the electron-donating properties of their ring
substituents, i.e. p-OCH3 > p-CH3> H ꢀ Cl; ii) PC is by far,
less reactive than EC. As reported by different authors and by
us,7a,21,29 the steric hindrance of the methyl group in PC, likely
accounts for its poor electrophilicity with respect to EC; iii) the
higher activity of bromide exchanged salts (PIL9 and 12) over the
Scheme 10 Mechanistic hypothesis for the mono-N-(2-hydroxy)-
ethylation of aniline with EC.
The hypothesis (a) stands on the weak Lewis acidity of
phosponium cations,23 and the great affinity between phosphorous
5194 | Org. Biomol. Chem., 2010, 8, 5187–5198
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
tosylate (PIL1) in the alkylation of p-anisidine and p-toluidine
with both EC and PC, suggests that the mechanism of Scheme 10
discussed for aniline, holds for p-substituted anilines as well.
PIL : 1a was 0.1). The chosen catalysts were PILs 1–4 and 9–12.
The reactor was then immersed in an oil bath thermostated at
the desired temperature (170 ◦C) and the mixture was kept under
magnetic stirring throughout the reaction. At intervals, samples
of the reaction mixture were withdrawn and analysed by GC-
MS. The procedure was used to run an experiment also in the
absence of PILs. The above described procedure was adapted for
the following experiments [(i)–(v)].
Conclusions
The reaction of primary aromatic amines with alkylene carbon-
ates is a rather unexplored transformation. This investigation
describes an advantageous methodology to carry out the bis-
N-hydroxyalkylation of anilines with EC and PC. Although the
methodology is rather energy intensive, several advantages can be
recognized from both synthetic and environmental standpoints: i)
the reaction is of a general scope for different anilines and for PC
and EC; ii) non toxic alkylene carbonates replace very harmful
hydroxyalkylating agents such as alkylene oxides or halohydrins;
iii) the reactions are chemoselective, occurring almost exclusively
at the b carbons of EC and PC. The competitive formation of car-
bamate products is ruled out; iv) the reaction setup is very simple
and it allows truly catalytic processes using robust systems based
on phosphonium salts (PILs). In particular, bromide exchanged
PILs have proven to be the most efficient organocatalysts; v) a
catalyst is not always required for the reaction: the quantitative
formation of N,N-bis(2-hydroxy) ethylaniline is achieved at a
reasonable temperature by merely mixing stoichiometric amounts
of aniline and EC; vi) no additional solvents are required.
Two mechanistic hypothesis are proposed to account for the effi-
ciency of the PILs organocatalysts: the Lewis acidic phosphonium
cations may electrophilically activate the alkyl carbons of alkylene
carbonates, while nucleophilic halide anions (but not tosylates)
assist the aperture of the carbonate ring through a direct attack at
the alkyl positions.
(i) Reactions with different catalyst loadings (Fig. 1) being all
other conditions unaltered, the molar ratio Q = PIL1 : 1a was set
to 0.05, 0.1, 0.2, 0.5 and 0.75, and the reaction temperature was set
to 150 ◦C; (ii) reactions with different ethylene carbonate loadings
(Table 1): being all other conditions unaltered, the molar ratio
EC : 1a was set to 2.2 and 4, and the reaction temperature was set
to 150 or to 170 ◦C; (iii) reactions with added water or aq. HBr
(Scheme 6): being all other conditions unaltered, water (0.015 g,
0.86 mmol, molar ratio H2O:1a was 0.1) or an aqueous solution
of HBr [48%] (0.1 mL, HBr 0.88 mmol, molar ratio HBr : 1a was
0.1) were added. The reaction temperature was set to 170 ◦C; (iv)
reactions of different anilines with ethylene carbonate (Table 5): the
glass reactor was charged with the amine (1b–d, 8.6 mmol; 1b:
1.06 g, 1c: 0.92 g, 1d: 1.09 g), ethylene carbonate (EC, 1.66 g,
18.9 mmol, molar ratio EC:1 was 2.2), and a PIL (0.86 mmol,
molar ratio PIL : 1 was 0.1). The chosen catalysts were PIL1 and
PIL12. The reactor was then immersed in an oil bath thermostated
at the desired temperature (150–170 ◦C); (v) reactions of different
amines with propylene carbonate (Table 6): the glass reactor was
charged with the amine (1a–b, 8.6 mmol; 1a: 0.8 g, 1b: 1.06 g),
propylene carbonate (PC, 1.93 g, 18.9 mmol, molar ratio PC : 1
was 2.2), and a PIL (0.86 mmol, molar ratio PIL : 1 was 0.1).
The chosen catalysts were PIL1, PIL9, and PIL12. The reactor
was then immersed in an oil bath thermostated at the desired
temperature (150–190 ◦C).
Propylene carbonate is a weaker electrophile than ethylene
carbonate: primary aromatic amines react with PC only in the
presence of catalytic amounts of bromide or tosylate onium salts.
1
Kinetic profiles via 13C{ H} NMR spectra (Fig. 2 and Table 4)
A solution (~1 mL) of aniline, ethylene carbonate and a PIL
(molar ratio 1a : EC : PIL were 1 : 2 : 0.05) was placed in a screw-
cap NMR tube equipped with a coaxial calibrated glass capillary
filled with pure [D6]DMSO for locking and homogeneity purposes.
The chosen catalysts were tosylate and halide salts (PILs◦1, 3–4
and 9, 11–13). The tube was dipped in an oil bath at 140 C. At
intervals, the tube was taken out from the thermostat, cooled at
room temperature, and finally, inserted in the NMR spectrometer.
All spectra were recorded at 25 ◦C. For each measure, it was
assumed that the heating and the cooling times of the mixture
were comparable and that a negligible reaction advancement
Experimental
Anilines 1a–d (p-XC6H4NH2: 1a: X = H; 1b: X = OMe; 1c: X =
Me; 1d: X = Cl), ethylene carbonate (EC), propylene carbonate
(PC) and tri-isobutylmethylphosphonim tosylate (PIL1) were
ACS grade and were employed without further purification. PIL2–
13 were prepared accordingly to a method recently reported by
us.17 GC-MS (EI, 70 eV) analysis were run using a HP5/MS
1
capillary column (30 m). H NMR were recorded at 400 MHz,
13C NMR spectra at 100 MHz, 31P NMR spectra were recorded
at 200 MHz. Chemical shifts were reported in d values downfield
from TMS; CDCl3 or CD3OD were used as solvents. 31P chemical
shifts were reported with respect to 85% orthophosphoric acid (as
external standard). IR spectra were recorded at room temperature
on KBr disks.
1
occurred at room temperature. The 13C{ H}NMR signals were
used to obtain a plot of concentration vs. time for all the reactants
and the products involved in the reaction. Fig. 2 exemplifies the
process carried out in the presence of PIL11. Analogous plots
were obtained for each of the catalysts used. The numerical
integration of these profiles was performed through a method
recently developed by us:20 accordingly, the kinetic constants of
Table 4 were determined.
Reaction procedure
The bis-N-(2-hydroxy)ethylation of aniline in the presence of
different ionic liquids (Table 3). A glass reactor (7 mL) shaped
as a test tube and equipped with a side screw-capped neck for the
withdrawal of samples and a condenser, was charged with aniline
(1a, 0.80 g, 8.6 mmol), ethylene carbonate (EC, 1.51 g, 17.2 mmol,
molar ratio EC : 1a was 2), and a PIL (0.86 mmol, molar ratio
Synthesis of PILs (Table 2)
Synthesis of methyltrialkylphosphonium methylcarbonate salts.
PIL5–8: a sealed 200 mL steel autoclave fitted with a pressure
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5187–5198 | 5195
©

View Article Online
gauge and a thermocouple for temperature control was charged
with a trialkylphosphine (R3P, 56 mmol; R = i-butyl, 15 mL, 12.2 g;
R = n-butyl, 15 mL, 12.2 g; R = n-hexyl, 20 mL, 16.2 g; R = n-octyl,
25 mL, 20.8 g), dimethyl carbonate (30 mL, 32.1 g, 356 mmol) and
methanol (30 mL). Three freeze–pump–thaw cycles were carried
out to ensure complete degassing of the mixture and air removal.
The empty volume was then filled with nitrogen. The autoclave
was heated for 24 h at the desired temperature (140 ◦C: R = n-
octyl, n-hexyl, n-butyl; 170 ◦C: R = i-butyl) with magnetic stirring.
Then, the reactor was cooled to room temperature and vented.
Methanol and the residual DMC were removed from the mixture
by rotary evaporation. A small amount (<1 equiv.) of methanol
could remain incorporated in the sample even after a prolonged
high vacuum was applied. Isolated yields were: PIL6 (17.4 g) 96%;
PIL7 (21.5 g) 94%; PIL8 (27.5 g) 100%. These salts were fully
4.1 (d, J(P,H) = 52 Hz, 1C; P–CH3). 31P NMR (CDCl3, 25 ◦C,
200 MHz): d (ppm): 31.7.
PIL5 (viscous pale yellow liquid) [(i-Bu)3MeP][OCO2CH3]:
1H NMR (CDCl3, 25 ◦C, 400 MHz): d (ppm): 3.13 (s, 3H;
CH3OCOO), 1.97 (dd, J(P,H) = 13 Hz, J(H,H) = 7 Hz, 6H; P–
CH2), 1.78–1.68 (m, 3H; P–CH2–CH(CH3)2), 1.72 (d, J(P,H) =
13 Hz, 3H; P–CH3), 0.76 (d, J(H,H) = 7 Hz, 18H; P–CH2–
CH(CH3)2); 13C NMR (CDCl3, 25 ◦C, 100 MHz): d (ppm): 157.8
(1C; C O), 51.2 (1C; CH3O), 29.5 (d, J(P,C) = 46 Hz, 3C; P–
CH2–CH(CH3)2), 23.7 (d, J(P,C) = 9 Hz, 6C; P–CH2–CH(CH3)2),
22.9 (d, J(P,C) = 5 Hz, 3C; P–CH2–CH(CH3)2), 6.2 (d, J(P,C) =
50 Hz, 1C; P–CH3). 31P NMR (CDCl3, 25 ◦C, 200 MHz): d (ppm):
29.0.
PIL6 (viscous pale yellow liquid) [(n-Bu)3MeP][OCO2CH3]:
1H NMR (CDCl3, 25 ◦C, 400 MHz): d (ppm): 3.00 (s, 3H;
CH3OCOO), 1.92–1.80 (m, 6H; P–CH2), 1.52 (d, J(P,H) = 14 Hz,
3H; P–CH3), 1.06 (brs, 12H), 0.50 (brt, 9H); 13C NMR (CDCl3,
25 ◦C, 100 MHz): d (ppm): 156.8 (1C; C O)), 50.7 (1C; CH3O),
22.5 (d, J(P,H) = 15 Hz, 3C), 22.3(d, J(P,H) = 4 Hz, 3C), 18.6 (d,
J(P,H) = 49, 3C; P–CH2), 12.1◦(3C), 2.8 (d, J(P,H) = 53 Hz, 1C;
P–CH3). 31P NMR (CDCl3, 25 C, 200 MHz): d (ppm): 31.6.
PIL7 (viscous pale yellow liquid) [(n-Hex)3MeP][OCO2CH3]:
1H NMR (CDCl3, 25 ◦C, 400 MHz): d (ppm): 3.44 (s, 3H;
CH3OCOO), 2.26–2.18 (m, 6H; P–CH2), 1.92 (d, J(P,H) = 14,
3H, P–CH3), 1.46–1.38 (m, 12H), 1.26–1.21 (m, 12H), 0.82 (brt,
9H); 13C NMR (CDCl3, 25 ◦C, 100 MHz): d (ppm): 157.7 (1C;
1
characterized as such, by H and 13C NMR, and used without
further purifications. PIL5 (15.8 g, 87%) was washed with n-hexane
(3 ¥ 15 mL) to remove some unreacted phosphine; then, it was
characterized and used.
General anion exchange reaction procedure. a 50-mL round-
bottomed flask was charged with an equimolar mixture of
methyltrialkylphosphonium methylcarbonate (PIL5–8; 6 mmol,
1.75–2.76 g) and of a Brønsted acid (H–A: A = TsO, Br, or I), and
methanol (5 mL) as a co-solvent. TsOH·H2O (1.14 g) was used as
a solid, while both HBr (0.68 mL) and HI (0.79 mL) were used in
aqueous commercial solutions (48% and 57%, respectively). The
mixture was kept under magnetic stirring for 1 h at 40 ◦C. Then,
water and methanol were removed by rotary evaporation. The
desired PIL2–4 and PIL9–13 were obtained in quantitative yields
and fully characterized as such, by 1H and 13C NMR. They were
used without further purification.
C
O)), 51.5 (1C; CH3O), 30.4 (3C), 29.6 (d, J(P,C) = 15 Hz, 3C),
21.7 (3C), 20.9 (d, J(P,C) = 4 Hz, 3C), 19.3 (d, J(P,C) = 49 Hz, 3C,
P–CH2), 13.3 (3C), 3.5 (d, J(P,H) = 52 Hz, 1C; P–CH3). 31P NMR
(CDCl3, 25 ◦C, 200 MHz): d (ppm): 31.8.
PIL8: a complete NMR characterization is given in reference
17 31P NMR (CDCl3, 25 ◦C, 200 MHz): d (ppm): 31.7.
1
PIL2 (viscous clear colourless liquid) [(n-Bu)3MeP][TosO]: H
PIL9 (white hygroscopic solid) [(i-Bu)3MeP][Br]: 1H NMR
(CDCl3, 25 ◦C, 400 MHz): d (ppm): 2.17 (dd, J(P,H) = 13 Hz,
J(H,H) = 7 Hz, 6H; P–CH2), 1.88 (d, J(P,H) = 13 Hz, 3H; P–CH3),
1.92–1.82 (m, 3H; P–CH2–CH(CH3)2), 0.84 (d, J(H,H) = 7 Hz,
18H; P–CH2–CH(CH3)2); 13C NMR (CDCl3, 25 ◦C, 100 MHz): d
(ppm): 29.9 (d, J(P,C) = 46 Hz, 3C), 23.9 (d, J(P,C) = 9 Hz, 6C),
23.0 (d, J(P,C) = 5 Hz, 3C), 7.5 (d, J(P,C) = 50 Hz, 1C). 31P NMR
(CDCl3, 25 ◦C, 200 MHz): d (ppm): 29.2.
NMR (CDCl3, 25 ◦C, 400 MHz): d (ppm): 7.61 (d, J = 8 Hz, 2H),
7.00 (d, J = 8 Hz, 2H), 2.21 (s, 3H; –CH3), 2.10–2.02 (m, 6H; P–
CH2), 1.72 (d, J(P,H) = 14 Hz, 3H; P–CH3), 1.30–1.25 (m, 12H),
0.77 (brt, 9H); 13C NMR (CDCl3, 25 ◦C, 100 MHz): d (ppm):
144.1 (1C), 138.9 (1C), 128.4 (2C), 125.8 (2C), 23.7 (d, J(P,H) =
16 Hz, 3C), 23.4(d, J(P,H) = 5 Hz, 3C), 21.2 (1C), 19.6 (d, J(P,H) =
49, 3C), 13.4 (3C), 4.0 (d, J(P,H) = 52 Hz, 1C). 31P NMR (CDCl3,
25 ◦C, 200 MHz): d (ppm): 31.8.
1
PIL10 (viscous clear colourless liquid) [(n-Bu)3MeP][Br]: H
PIL3 (viscous clear colourless liquid) [(n-Hex)3MeP][TosO]: 1H
NMR (CDCl3, 25 ◦C, 400 MHz): d (ppm): 7.73 (d, J = 8 Hz, 2H),
7.09 (d, J = 8 Hz, 2H), 2.31 (s, 3H; –CH3), 2.26–2.18 (m, 6H;
P–CH2), 1.93 (d, J(P,H) = 14, 3H, P–CH3), 1.47–1.34 (m, 12H),
1.28–1.22 (m, 12H), 0.85 (brt, 9H); 13C NMR (CDCl3, 25 ◦C,
100 MHz): d (ppm): 144.5 (1C), 138.6 (1C), 128.3 (2C), 125.8
(2C), 31.0 (3C), 30.1(d, J(P,C) = 15 Hz, 3C), 22.2 (3C), 21.4 (d,
J(P,C) = 5 Hz, 3C), 21.1 (1C), 19.8 (d, J(P,C) = 49 Hz, 3C, P–CH2),
13.8 (3C), 3.9 (d, J(P,H) = 52 Hz, 1C). 31P NMR (CDCl3, 25 ◦C,
200 MHz): d (ppm): 31.6.
NMR (CDCl3, 25 ◦C, 400 MHz): d (ppm): 2.03–1.96 (m, 6H; P–
CH2), 1.62 (d, J(P,H) = 13, 3H, P–CH3), 1.14–1.03 (m, 12H), 0.50
(brt, 9H); 13C NMR (CDCl3, 25 ◦C, 100 MHz): d (ppm): 22.6 (d,
J(P,C) = 16 Hz, 3C), 22.5 (d, J(P,H) = 4 Hz, 3C), 19.3 (d, J(P,H) =
49, 3C), 12.3 (3C), 4.1 (d, J(P,H) = 52 Hz, 1C). 31P NMR (CDCl3,
25 ◦C, 200 MHz): d (ppm): 31.6.
PIL11 (viscous pale yellow liquid) [(n-Hex)3MeP][Br]: 1H NMR
◦
(CDCl3, 25 C, 400 MHz): d (ppm): 2.29–2.23 (m, 6H; P–CH2),
1.91 (d, J(P,H) = 13, 3H, P–CH3), 1.42–1.28 (m, 12H), 1.15–1.12
(m, 12H), 0.71 (brt, 9H); 13C NMR (CDCl3, 25 ◦C, 100 MHz): d =
30.6 (3C), 29.9 (d, J(P,C) = 15 Hz, 3C), 21.9 (3C), 21.3 (d, J(P,C) =
4 Hz, 3C), 20.3 (d, J(P,C) = 49 Hz,◦3C), 13.5 (3C), 4.9 (d, J(P,H) =
53 Hz, 1C). 31P NMR (CDCl3, 25 C, 200 MHz): d (ppm): 31.7.
PIL4 (white solid) [(n-Oct)3MeP][TosO]: 1H NMR (CDCl3,
◦
25 C, 400 MHz): d (ppm): 7.70 (d, J = 8 Hz, 2H), 7.06 (d, J =
8 Hz, 2H), 2.28 (s, 3H; –CH3), 2.21–2.15 (m, 6H; P–CH2), 1.88
(d, J(P,H) = 14, 3H, P–CH3), 1.41–1.33 (m, 12H), 1.20 (brs, 24H),
0.83 (t, 9H); 13C NMR (CDCl3, 25 ◦C, 100 MHz): d (ppm): 144.0
(1C), 138.7 (1C), 128.3 (2C), 125.8 (2C), 31.6 (3C), 30.5(d, J(P,C) =
15 Hz, 3C), 28.9 (3C), 28.8 (3C), 22.5 (3C), 21.6 (d, J(P,C) = 5 Hz,
3C), 21.1 (1C), 19.9 (d, J(P,C) = 48 Hz, 3C, P–CH2), 13.9 (3C),
1
PIL12 (w◦hite hygroscopic solid) [(n-Oct)3MeP][Br]: H NMR
(CDCl3, 25 C, 400 MHz): d (ppm): 2.42–2.36 (m, 6H; P–CH2),
2.06 (d, J(P,H) = 13, 3H, P–CH3), 1.52–1.38 (m, 12H), 1.26–1.20
(m, 24H), 0.82 (brt, 9H); 13C NMR (CDCl3, 25 ◦C, 100 MHz): d
(ppm): 31.4 (3C), 30.4 (d, J(P,C) = 15 Hz, 3C), 28.9 (3C), 28.8 (3C),
5196 | Org. Biomol. Chem., 2010, 8, 5187–5198
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
22.5 (3C), 21.6 (d, J(P,C) = 5 Hz, 3C), 21.1 (1C), 19.9 (d, J(P,C) =
48 Hz, 3C), 13.9 (3C), 4.1 (d, J(P,H) = 52 Hz, 1C; P–CH3). 31P
NMR (CDCl3, 25 ◦C, 200 MHz): d (ppm): 31.6.
1H NMR (CD3OD, 400 MHz) d (ppm): 7.05 (d, J = 9,2 Hz, 2H),
6.64 (d, J = 9.2 Hz, 2H), 3.64 (t, J = 6.0 Hz, 4H), 3.45 (t, J =
6.0 Hz, 4H). 13C NMR (CDCl3, 100 MHz) d (ppm): 146.5 (1C),
129.0 (2C), 121.8 (1C), 113.8 (2C), 60.5 (2C), 55.2 (2C). IR (KBr):
n = 3314, 2866, 1618, 1865, 1593, 1500, 1383, 1062, 1037, 1014.
PIL13 (viscous yellow liquid) [(n-Oct)3MeP][I]: 1H NMR
◦
(CDCl3, 25 C, 400 MHz): d (ppm): 2.45–2.38 (m, 6H; P–CH2),
2.09 (d, J(P,H) = 13, 3H, P–CH3), 1.56–1.41 (m, 12H), 1.30–1.1.23
(m, 24H), 0.84 (brt, 9H); 13C NMR (CDCl3, 25 ◦C, 100 MHz): d
(ppm): 31.5 (3C), 30.4 (d, J(P,C) = 15 Hz, 3C), 28.7 (3C), 28.8 (3C),
22.4 (3C), 21.6 (d, J(P,C) = 5 Hz, 3C), 20.7 (d, J(P,C) = 48 Hz, 3C,
P–CH2), 13.8 (3C), 5.6 (d, J(P,H) = 52 Hz, 1C). 31P NMR (CDCl3,
25 ◦C, 200 MHz): d (ppm): 31.7.
Bis-N-(2-hydroxy)propyl aniline, 7a. The product was isolated
from the reaction carried out under the conditions of entry
4 of Table 6. The final mixture was purified by flash column
chromatography (FCC) on silica gel, using a gradient elution with
EA/PE solutions (initial EA : PE = 7 : 3 v/v, final EA : PE = 3 : 7
v/v). Compound 7a (sum of diastereoisomers) was isolated as a
pale yellow oil (58%). 1H NMR (CDCl3, 400 MHz) d (ppm): 7.26–
7.20 (m, 4H), 6.82–6.78 (m, 2H), 6.78 6.71 (m, 2H), 6.59–6.57 (m,
2H), 4.24–4.16 (m, 2H), 4.16–4.08 (m, 2H), 3.65 (dd, J1 = 15 Hz,
J2 = 2 Hz, 2H), 3.40 (dd, J1 = 15 Hz, J2 = 3 Hz, 2H), 3.18 (dd,
J1 = 15 Hz, J2 = 9 Hz, 2H), 3.02 (dd, J1 = 15 Hz, J2 = 10 Hz,
2H), 1.22 (d, J = 6 Hz, 6H), 1.20 (d, J = 6 Hz, 6H). 13C NMR
(CDCl3, 25 ◦C, 100 MHz): d (ppm): 149.2 (1C), 148.0 (1C), 129.2
(2C), 129.0 (2C), 117.4 (1C), 116.6 (1C), 114.0 (2C), 112.2 (2C),
65.9 (2C), 64.9 (2C), 62.4 (2C), 60.0 (2C), 20.2 (4C). IR (KBr): n =
3339, 2988, 1598, 1503, 1500, 1384, 1129, 1059, 1024.
Isolation and characterization of the products
All the reaction products 2a–d, 3a–d, 4a–d, 5a–d, 6a–b, 7a–b, 8a–b
and 9a–b were characterized by GC-MS (see ESI†). Compounds
3a–d and 7a–b were isolated and fully characterized by 1H and 13
C
NMR (see ESI†).
Bis-N-(2-hydroxy)ethyl aniline, 3a1d,30
. The product was iso-
lated from the reactions carried out under the conditions of entries
9 and 12 of Table 1. The final mixtures were purified by flash
column chromatography (FCC) on silica gel, using a gradient elu-
tion with methanol–diethyl ether/petroleum ether (PE) solutions
(initial MeOH–Et2O : PE = 0 : 1 : 4 v/v, final MeOH–Et2O:PE =
1 : 7.5 : 1.5 v/v). Compound 3a was a dark yellow oil (74 and 61%
respectively, Table 1). 1H NMR (CDCl3, 400 MHz) d (ppm): 7.27–
7.20 (m, 2H), 6.81–6.60 (m, 3H), 3.84 (t, J = 4.9 Hz, 4H), 3.57
(t, J = 4.9 Hz, 4H). 13C NMR (CDCl3, 100 MHz) d (ppm): 147.7
(1C), 129.1 (2C), 116.7 (1C), 112.4 (2C), 60.5 (2C), 55.1 (2C). IR
(KBr): n = 3368, 2881, 1597, 1504, 1384, 1033.
Bis-N-(2-hydroxy)propyl p-anisidine, 7b32. The product was
isolated from the reaction carried out under the conditions of
entry 6 of Table 6. The final mixture was purified by flash column
chromatography (FCC) on silica gel, using a gradient elution with
EA/PE solutions (initial EA : PE = 7 : 3 v/v, final EA : PE = 3 : 7
1
v/v). Compound 7b was a brown oil (55%). H NMR (CDCl3,
400 MHz) d (ppm): 6.86–6.78 (m, 6H), 6.61–6.58 (d, J = 9 Hz,
2H), 4.12–4.03 (m, 2H), 4.03–3.94 (m, 2H), 3.75 (s, 6H), 3.51 (dd,
J1 = 15 Hz, J2 = 2 Hz, 2H), 3.26 (dd, J1 = 14 Hz, J2 = 2 Hz, 2H),
3.02–2.92 (m, 4H), 1.17 (d, J = 6 Hz, 6H), 1.16 (d, J = 6 Hz, 6H).
13C NMR (CDCl3, 25 ◦C, 100 MHz): d (ppm): 152.8 (1C), 151.8
(1C), 143.8 (1C), 142.7 (1C), 117.4 (2C), 114.8 (2C), 114.5 (2C),
114.4 (2C), 65.8 (2C), 64.7 (2C), 63.0 (2C), 61.2 (2C), 55.7 (1C),
55.5 (1C), 20.1 (4C). IR (KBr): n = 3392, 2941, 1511, 1384, 1241,
1047.
Bis-N-(2-hydroxy)ethyl p-anisidine, 3b30b,31
. The product was
isolated from the reaction carried out under the conditions of
entry 4 of Table 5. The final mixture was purified by flash
column chromatography (FCC) on silica gel, using a ethyl
acetate (EA)/petroleum ether (PE) solution (EA:PE = 3 : 1 v/v).
Compound 3b was isolated as brown solid (74%). 1H NMR
(CDCl3, 400 MHz) d (ppm): 6.83 (d, J = 9.2 Hz, 2H), 6.76 (d,
J = 9.2 Hz, 2H), 3.79 (t, J = 5.0 Hz, 4H), 3.76 (s, 3H), 3.46 (t, J =
5.0 Hz, 4H). 13C NMR (CDCl3, 100 MHz) d (ppm): 152.4 (1C),
142.6 (1C), 115.7 (2C), 114.8 (2C), 60.5 (2C), 55.9 (2C), 55.7 (1C).
IR (KBr): n = 3305, 2950, 1651, 1511, 1250, 1036.
Acknowledgements
MIUR (Italian Ministry for Education, University and Research)
is gratefully acknowledged for the financial support of this work.
Bis-N-(2-hydroxy)ethyl p-toluidine, 3c31. The product was iso-
lated from the reaction carried out under the conditions of entry
5 of Table 5. The final mixture was purified by flash column
chromatography (FCC) on silica gel, using a gradient elution with
EA/PE solutions (initial EA:PE = 1 : 1 v/v, final EA:PE = 3 : 1
References
1 (a) P. Chin-Min Mao, J. F. Mouscadet, H. Leh, C. Auclair and L. F.
Hsu, Heterocycles, 2001, 55, 1263–1270; (b) S. Heilporn, F. Broeders,
D. Daloze and Breakman, Bull. Soc. Chim. Belg., 1994, 103, 309–319;
(c) B. D. Palmer, W. R. Wilson, S. M. Pullen and W. A. Denny, J. Med.
Chem., 1990, 33, 112–121; (d) M. H. Benn, L. N. Owen and A. M.
Creighton, J. Chem. Soc., 1958, 2800–2810.
2 www.sigmaaldrich.com: the safety sheets (MSDS) of ethylene and
propylene oxide report T labels.
3 T. Yoshino, S. Inaba, H. Komura and Y. Ishido, J. Chem. Soc., Perkin
Trans. 1, 1977, 1266–1272.
4 (a) Z. Y. Sun, E. Botros, A. D. Su, Y. Kim, E. Wang, N. Z. Baturay
and C. H. Kwon, J. Med. Chem., 2000, 43, 4160–4168; (b) F. Jourdain
and J. C. Pommelet, Tetrahedron Lett., 1994, 35(10), 1545–1548; (c) H.
Adkins and R. M. Simington, J. Am. Chem. Soc., 1925, 47, 1687–1689.
5 The safety sheets (MSDS) of ethylene bromo- and chloro-hydrin report
a T and T+ labels, respectively (www.sigmaaldrich.com).
1
v/v). Compound 3c was a dark brown solid (70%). H NMR
(CDCl3, 400 MHz) d (ppm): 7.04 (d, J = 8,7 Hz, 2H), 6.63 (d, J =
8,7 Hz, 2H), 3.81 (t, J = 4.9 Hz, 4H), 3.52 (t, J = 4.9 Hz, 4H), 2.25
(s, 3H). 13C NMR (CDCl3, 100 MHz) d (ppm): 145.7 (1C), 129.7
(2C), 126.2 (1C), 113.0 (2C), 60.5 (2C), 55.3 (2C), 20.0 (1C). IR
(KBr): n = 3391, 2934, 1618, 1520, 1382, 1082.
Bis-N-(2-hydroxy)ethyl p-chloroaniline, 3d31. The product was
isolated from the reaction carried out under the conditions of
entry 6 of Table 5. The final mixture was purified by flash column
chromatography (FCC) on silica gel, using a EA/PE solution
(EA:PE = 1 : 1 v/v). Compound 3d was a pale brown solid (62%).
6 (a) M. Selva, E. Militello and M. Fabris, Green Chem., 2008, 10, 73–
79; (b) F. Bonino, A. Damin, S. Bordiga, M. Selva, P. Tundo and A.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5187–5198 | 5197
©

View Article Online
Zecchina, Angew. Chem., Int. Ed., 2005, 44, 4774–4777; (c) M. Selva,
P. Tundo, A. Perosa and F. Dall’Acqua, J. Org. Chem., 2005, 70, 2771–
2777; (d) M. Selva and P. Tundo, Acc. Chem. Res., 2002, 35(9), 706–716;
(e) M. Selva, P. Tundo, A. Perosa and S. Memoli, J. Org. Chem., 2002,
67, 1071–1077; (f) P. Tundo, M. Selva and A. Bomben, Org. Synth.,
1999, 76, 169–177.
7 (a) B. M. Bhanage, S.-I. Fujita, Y. Ikushina and M. Arai, Appl. Catal.,
A, 2001, 219, 259–266; (b) B. Denise and R. P. A. Sneeden, Chemtech,
1982, 12(2), 108–112; (c) M. Lichtenwalter and J. Cooper, U.S. Patent
2 773 070, 1956. It should be noted that also the addition of CO2 to
glycol has been examined for the synthesis of EC; J. I. Kadokawa, H.
Habu, S. Fukamachi, M. Karasu, H. Tagaya and K. Chiba, J. Chem.
Soc., Perkin Trans. 1, 1999, 2205–2208. At the moment however, the
method has only a speculative interest.
16 FCC however, was not optimised.
17 M. Fabris, V. Lucchini, M. Noe`, A. Perosa and M. Selva, Chem.–Eur. J.,
2009, 15, 12273–12282.
18 G. Gattow and W. Behrendt, Angew. Chem., Int. Ed., 1972, 11(6), 534–
535.
19 Plausibly, HBr assisted the dehydration of 3a via an initial O-
protonation followed by an internal nucleophilic displacement releasing
water.
20 M. Fabris, V. Lucchini, M. Noe`, A. Perosa and M. Selva, paper
submitted.
21 In principle, an additional couple of diastereoisomers was expected,
coming from the double attack of the amine at the C4 position of
PC. However, plausibly for steric reasons, these compounds were never
observed.
8 (a) J. H. Clements, Ind. Eng. Chem. Res., 2003, 42, 663–674; (b) A. A. G.
22 (a) S. E. Sloop, J. K. Pugh, S. Wang, J. B. Kerr and K. Kinoshita,
Electrochem. Solid-State Lett., 2001, 4, A42–A44; (b) M. Selva, A.
Perosa and M. Fabris, Green Chem., 2008, 10, 1068–1077.
23 (a) O. Sereda, S. Tabassum and R. Wilhelm In Asymmetric Organocatal-
ysis, Topics in Current Chemistry, Springer, Berlin-Heidelberg, 2009;
(b) M. Terada and M. Kouchi, Tetrahedron, 2006, 62, 401–409; (c) X.
Wang and S.-K. Tian, Tetrahedron Lett., 2007, 48, 6010–6013.
24 (a) A. J. Kirby and S. G. Warren In The Organic Chemistry of
Phosphorus, Elsevier, Amsterdam, London, New York, 1967; (b) J.
Chen, Y. Chen, Y.-L. Niu, H. Fu and Y.-F. Zhao, J. Mass Spectrom.,
2002, 37, 934–939.
Shaikh and S. Sivaram, Chem. Rev., 1996, 96, 951–976.
9 (a) A. B. Shivakar, S. P. Gupte and R. V. Chaudari, Synlett, 2006,
9, 1374–1378; (b) E. Gulbins and K. Hamann, Chem. Ber., 1966, 99,
55–61.
10 (a) M. Selva, P. Tundo and T. Foccardi, J. Org. Chem., 2005, 70, 2476–
2485; (b) M. Selva, P. Tundo and A. Perosa, J. Org. Chem., 2003, 68,
7374–7378; (c) M. Selva, Synthesis, 2003, 18, 2872–2876; (d) M. Selva,
P. Tundo and A. Perosa, J. Org. Chem., 2001, 66, 677–680; (e) M.
Selva, A. Bomben and P. Tundo, J. Chem. Soc., Perkin Trans. 1, 1997,
1041–1045.
11 (a) L. Puppe, Angew. Chem., Int. Ed. Engl., 1975, 14, 620; (b) G. C. Bond
In Heterogeneous Catalysis Principles and Applications, Clarendon
Press, Oxford, 2nd edn, 1987; (c) D. Barthomeuf, J. Phys. Chem., 1984,
88, 42–45.
25 H. Schmidbaur, W. Buchner and F. H. Kohler, J. Am. Chem. Soc., 1974,
96, 6208–6210.
26 The size of the catalyst appears rather important to control the
chemoselectivity: for instance, if a small Lewis acid such as LiCl is
used as a catalyst (see ref. 9b), anilines are able to attack both alkyl and
carboxyl carbons of EC.
27 D. Belli Dell’Amico, F. Calderazzo, L. Labella, F. Marchetti and G.
Pampaloni, Chem. Rev., 2003, 103, 3857–3897.
28 E. V. Anslyn and D. A. Dougherty In Modern Physical Organic
Chemistry, University Science Books, 2006, Capt. 8.
29 D. Aurbach, Y. Gofer, M. Ben-Zion and P. Aped, J. Electroanal. Chem.,
1992, 339(1–2), 451–471.
12 M. Selva and M. Fabris, Green Chem., 2009, 11, 1161–1172.
13 (a) A. Loris, A. Perosa, M. Selva and P. Tundo, J. Org. Chem., 2004, 69,
3953–3956; (b) M. Selva, A. Perosa, P. Tundo and D. Brunelli, J. Org.
Chem., 2006, 71, 5770–5773.
14 In the presence of tryglime as solvent, at 160 ◦C, the same uncatalysed
process was reported to proceed to a much lower extent (the yield
of 2a was 4% after 2 h; ref. 9a). While, at 176 ◦C and under
solventless conditions, the reaction of aniline and ethylene carbonate
catalysed by LiCl, proceeded through competitive N-alkylation and N-
carboxyalkylation processes, to produce mixture of compounds 2a and
5a (ref. 9b).
30 (a) W. C. J. Ross, J. Chem. Soc., 1949, 183–191; (b) A. Clerici, A.
Ghilardi, N. Pastori, C. Punta and O. Porta, Org. Lett., 2008, 10(21),
5063–5066.
15 Although the increase of the amount of PIL1 (Fig. 1), favoured the
reaction rate, the Q ratio was set 0.1 for practical reasons: the separation
of the IL from the reaction products was easier and interferences in the
GC-MS analyses were greatly limited.
31 M. Freifelder and G. R. Stone, J. Org. Chem., 1961, 26(5), 1477–
1480.
32 A. H. Sommers, U. Biermacher, S. B. Brehm and J. H. Short, J. Med.
Chem., 1970, 13(3), 581–583.
5198 | Org. Biomol. Chem., 2010, 8, 5187–5198
This journal is
The Royal Society of Chemistry 2010
©