Comprehensive Study of the Organic-Solvent-Free CDI-Mediated Acylation of Various Nucleophiles by Mechanochemistry

Article abstract of DOI:10.1002/chem.201501325

Acylation reactions are ubiquitous in the synthesis of natural products and biologically active compounds. Unfortunately, these reactions often require the use of large quantities of volatile and/or toxic solvents, either for the reaction, purification or isolation of the products. Herein we describe and discuss the possibility of completely eliminating the use of organic solvents for the synthesis, purification and isolation of products resulting from the acylation of amines and other nucleophiles. Thus, utilisation of N,N′-carbonyldiimidazole (CDI) allows efficient coupling between carboxylic acids and various nucleophiles under solvent-free mechanical agitation, and water-assisted grinding enables both the purification and isolation of pure products. Critical parameters such as the physical state and water solubility of the products, milling material, type of agitation (vibratory or planetary) as well as contamination from wear are analysed and discussed. In addition, original organic-solvent-free conditions are proposed to overcome the limitations of this approach. The calculations of various green metrics are included, highlighting the particularly low environmental impact of this strategy.

Full text of DOI:10.1002/chem.201501325

DOI: 10.1002/chem.201501325
Full Paper
&
Solvent-Free Synthesis
Comprehensive Study of the Organic-Solvent-Free CDI-Mediated
Acylation of Various Nucleophiles by Mechanochemistry**
Thomas-Xavier MØtro,* Julien Bonnamour, Thomas Reidon, Anthony Duprez,
[
a]
Jordi Sarpoulet, Jean Martinez, and FrØdØric Lamaty*
In memory of Professor Bertrand Castro
Abstract: Acylation reactions are ubiquitous in the synthesis
of natural products and biologically active compounds. Un-
fortunately, these reactions often require the use of large
quantities of volatile and/or toxic solvents, either for the re-
action, purification or isolation of the products. Herein we
describe and discuss the possibility of completely eliminat-
ing the use of organic solvents for the synthesis, purification
and isolation of products resulting from the acylation of
amines and other nucleophiles. Thus, utilisation of N,N’-car-
bonyldiimidazole (CDI) allows efficient coupling between
carboxylic acids and various nucleophiles under solvent-free
mechanical agitation, and water-assisted grinding enables
both the purification and isolation of pure products. Critical
parameters such as the physical state and water solubility of
the products, milling material, type of agitation (vibratory or
planetary) as well as contamination from wear are analysed
and discussed. In addition, original organic-solvent-free con-
ditions are proposed to overcome the limitations of this ap-
proach. The calculations of various green metrics are includ-
ed, highlighting the particularly low environmental impact
of this strategy.
[
4]
Introduction
scale. Although enabling organic reactions to be performed
in the absence of solvents, the recovery of the reaction mixture
from the milling jar is often performed with the use of an or-
ganic solvent, raising questions concerning the “solvent-free”
denomination of such processes. For our part, we have devel-
oped a strategy allowing the complete elimination of organic
solvents in the synthesis, purification and isolation of products
resulting from the acylation of amines and other nucleophiles.
Other methods avoiding the use of organic solvent from reac-
tion to product recovery have been previously published but
they remain scarce. As an example, solvent-free recovery by
the sublimation of nitrosobenzenes resulting from a reaction
Acylation reactions are widely used in the synthesis of natural
[1]
products and biologically active molecules, but they usually
require large amounts of volatile and toxic solvents such as
N,N-dimethylformamide or dichloromethane to be performed
with efficiency. Indeed, a SciFinder survey of 680000 amidation
reactions has shown that 83% of the reactions involve the use
[2]
of either DMF (47%) or CH Cl (36%). The removal of organic
2
2
solvents from the reaction media to reduce environmental
impact would be highly attractive as they account for the
major part of the material used to manufacture active pharma-
[3]
[5]
ceutical ingredients. Unfortunately, solvent-free synthesis
under classical batch stirring often implies either solid or
highly viscous reaction mixtures along with mass-transfer limi-
tations. In a wide number of examples, these difficulties could
be efficiently overcome by using mechanically induced agita-
tion provided by ball mills from the laboratory to the pilot
performed in a ball mill was described recently. Although car-
ried out in the total absence of any organic solvent, this ap-
proach is limited to molecules presenting high vapour pressure
properties. In some studies, the reaction mixture has been re-
covered from a ball mill reactor by using a cyclone under inert
gas flow or by physical withdrawal directly from the milling jar
[
6]
followed by drying. When water-soluble inorganic salts were
produced during the course of the reaction, they were re-
moved by washing with water and/or trituration followed by
filtration and drying, thereby avoiding the use of organic sol-
[
a] Dr. T.-X. MØtro, Dr. J. Bonnamour, T. Reidon, A. Duprez, J. Sarpoulet,
Prof. Dr. J. Martinez, Dr. F. Lamaty
Institut des BiomolØcules Max Mousseron (IBMM)
UMR 5247, CNRS, UniversitØ Montpellier, ENSCM
UniversitØ de Montpellier, Campus Triolet
[
7]
vents. Unfortunately, these approaches were limited to addi-
tion or condensation reactions in which water or water-soluble
Place Eug›ne Bataillon, 34095 Montpellier Cedex 5 (France)
E-mail: txmetro@univ-montp2.fr
ˇ
inorganic salts were the only by-products. More recently, Stru-
kil et al. described the mechano-mediated synthesis of aromat-
ic amides and dipeptides by using N-ethyl-N’-(3-dimethylami-
nopropyl)carbodiimide hydrochloride (EDC·HCl) as coupling
agent, nitromethane and NaCl as grinding assistants, 4-dimeth-
frederic.lamaty@univ-montp2.fr
[
**] CDI=N,N’-carbonyldiimidazole
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201501325.
Chem. Eur. J. 2015, 21, 12787 – 12796
12787
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ylaminopyridine (DMAP) as base and water to recover and
The latter was not isolated but its formation was followed
[8]
[14]
purify the final products. Although water-based treatment
and filtration are critical steps for the success of such organic-
by ex situ IR analysis of reaction samples.
These samples
were taken directly from the milling jar, immediately after the
5 min ball-milling had ended. They were analysed as such,
without any solubilisation process. Thus, the disappearance of
[9]
solvent-free processes, to the best of our knowledge, limita-
tions concerning the water solubility and physical state (solid
or liquid) of the products have never been considered or dis-
cussed previously. In addition, contamination from wear of the
milling material is a phenomenon that has been widely de-
scribed and analysed for other applications, but not for mill-
À1
the hydrocinnamic acid IR band at 1692 cm was observed
along with the appearance of the acylimidazole IR band at
À1
1730 cm (Figure 1). The addition of 1.0 equivalent of benzyl-
[10]
ing-mediated organic synthesis.
Although critical for the
purity of organic molecules synthesised by ball-milling in the
complete absence of organic solvent from reaction to product
isolation, product contamination from wear of milling material
[11]
has never been analysed and discussed in depth. For our
part, we have previously described the mechanosynthesis of
amides from carboxylic acids and amine hydrochlorides by
treatment with N,N’-carbonyldiimidazole (CDI) without using
[12]
organic solvent from reaction to product recovery. Pursuing
our efforts to provide the organic chemistry community with
a set of solutions enabling to avoid the use of problematic sol-
[
13]
vents, we herein describe a comprehensive study of the or-
ganic-solvent-free CDI-mediated acylation of N-, O-, S- and C-
nucleophiles by mechanochemistry, including analysis, discus-
sion and solutions to limitations related to 1) the physical state
and water solubility of the products and 2) contamination
from wear.
Figure 1. IR spectra for the mechanosynthesis of N-benzylhydrocinnamide.
IR spectrum of hydrocinnamic acid (top); IR spectrum after mixing hydrocin-
namic acid with CDI for 5 min (middle); IR spectrum after addition of BnNH
and mixing for 10 min (bottom).
2
Results and Discussion
Optimisation of the reaction based on stoichiometry and
amine acidic/basic form
amine followed by mixing for 10 min led to the complete con-
sumption of the acylimidazole, as indicated by the disappear-
In most cases, coupling carboxylic acids with amine hydro-
chlorides in the presence of CDI has the advantage of produc-
ing easily separable compounds, namely solid hydrophobic
amides, gaseous carbon dioxide and water-soluble by-products
À1
ance of its IR band at 1730 cm and the formation of N-ben-
À1
zylhydrocinnamide (new IR band at 1636 cm ). The addition
of deionised water to the jar followed by mixing for 5 min pro-
vided a suspension that could easily be transferred onto
a glass frit, allowing separation of non-soluble N-benzylhydro-
cinnamide from the by-products (imidazole and possible traces
of hydrocinnamic acid and benzylamine). Drying under
(
carboxylic acids, imidazole and amines, the last two existing
either as free bases or as hydrochloride salts). Thus, the treat-
ment of hydrocinnamic acid with 1.0 equivalent of CDI in
a 12 mL stainless-steel jar containing fifty 5 mm diameter balls
in a planetary ball mill agitated at 500 rpm for 5 min furnished
the corresponding acylimidazole (Scheme 1).
vacuum over P O furnished pure N-benzylhydrocinnamide in
2 5
a yield of 76% (Table 1, entry 1). Interestingly, using benzyl-
Table 1. Optimisation of CDI-mediated mechanosynthesis of N-benzylhy-
[a]
drocinnamide and N-benzylbenzamide.
Entry
Carboxylic acid
Amine
Amine
equiv]
Isolated yield
[%]
[
1
2
3
4
5
6
7
hydrocinnamic acid
BnNH
BnNH
BnNH
BnNH
BnNH
BnNH
BnNH
2
2
2
2
2
2
2
1.0
1.0
0.9
0.9
0.9
0.9
0.9
76
82
76
94
88
73
·HCl
·HCl
·HCl
benzoic acid
[b]
·HCl
82
Scheme 1. Mechanosynthesis of N-benzylamides from carboxylic acids and
CDI.
[a] Performed on a 1.35 mmol scale. [b] Performed on a 2.7 mmol scale.
Chem. Eur. J. 2015, 21, 12787 – 12796
www.chemeurj.org
12788
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
amine hydrochloride salt (instead of benzylamine) provided N-
benzylhydrocinnamide with an improved yield of 82% (Table 1,
entry 2). Although the yield was not improved by using
We next evaluated the efficiency of this approach in a vibra-
tory ball mill. Thus, utilising a 10 mL screw-top stainless-steel
jar equipped with two 10 mm diameter stainless-steel balls agi-
tated at 30 Hz within a vibratory ball mill furnished N-benzyl-
benzamide in a yield of 85% (Table 2, entry 4). Unfortunately,
ICP-MS analysis of the product revealed higher amounts of
iron and chromium than when using the planetary ball mill
(1400 ppm Fe and 1000 ppm Cr; Table 2, entry 4). When using
a smaller container and balls (5 mL jar and two 5 mm diameter
balls), both the iron and chromium content in the product
were lower (642 ppm Fe and 176 ppm Cr; Table 2, entry 5),
which can be explained by a lower intensity of collisions be-
0
.9 equivalents of benzylamine (Table 1, entry 3), the treatment
of hydrocinnamic acid with 1.0 equivalent of CDI followed by
the addition of 0.9 equivalents of benzylamine hydrochloride
furnished N-benzylhydrocinnamide in a yield of 94% (Table 1,
entry 4). This 12% increase in reaction yield might be ex-
plained by a higher conversion of BnNH ·HCl when hydrocin-
2
namic acid and CDI were used in excess (Table 1, entry 4) com-
pared with when hydrocinnamic acid, CDI and BnNH ·HCl were
2
utilised in equimolar amounts (Table 1, entry 2). It was hypoth-
esised that the 10% excess of acylimidazole was hydrolysed to
hydrocinnamic acid and imidazole during ball-milling with
water and eliminated in the filtrate during the filtration step.
This hypothesis was confirmed by HPLC analysis, which re-
vealed the presence of both hydrocinnamic acid and imidazole
in the filtrate. When hydrocinnamic acid was replaced with
benzoic acid, N-benzylbenzamide was obtained in a yield of
[
10]
tween the balls and walls. The previous experiments were
realised in jars that had been used several times, and we ex-
pected to reduce contamination by using jars that had never
been used before. Unfortunately, higher metal contamination
was observed when N-benzylbenzamide was synthesised in
brand new jars (Table 2, entry 6 vs. entry 4 and entry 7 vs.
entry 5). This phenomenon could be explained by the presence
of fragilities on the metallic surface created during machining
of the milling material. These fragilities would disappear after
a few hours of milling. In these experiments, the lower yields
of N-benzylbenzamide can be explained by material losses
during jar manipulation due to a less hermetic closure system
compared with the screw-top jars utilised previously (59 vs.
85% in 10 mL jars and 67 vs. 75% in 5 mL jars). It was next en-
visaged to reduce contamination from wear by performing re-
actions with milling material lighter than stainless steel
8
8% (Table 1, entry 5). As for the hydrocinnamic acid, the use
of benzylamine as a free base furnished N-benzylbenzamide in
a lower yield of 73% (Table 1, entry 6).
Finally, increasing the filling ratio by performing the same re-
action on a larger scale (2.7 mmol instead of 1.35 mmol) in the
same combination of jar and balls furnished the amide in
a slightly lower yield of 82% (Table 1, entry 7). It appears that
after optimisation of these latter parameters, optimal yields
were obtained when carboxylic acid was treated with
À3
1
.0 equivalent of CDI in a 12 mL stainless-steel jar containing
(1stainless steel =7.8 gcm ). Thus, changing the jar material to
À3
fifty 5 mm diameter balls followed by the addition of 0.9 equiv-
alents of benzylamine hydrochloride on a 1.35 mmol scale
ZrO2 (1zirconium oxide =5.9 gcm ) furnished N-benzylbenzamide
[
16]
with a satisfying yield of 86%, albeit as a powder containing
12200 ppm zirconium (Table 2, entry 8). On the other hand,
the lowest contamination resulting from wear was obtained
when using a planetary ball mill with the jar and balls made of
(
Table 1, entries 4 and 5).
Influence of milling material on reaction efficiency and con-
tamination from wear
agate, a much lighter milling material than stainless steel
À3
(
1agate =2.65 gcm ). Indeed, less than 100 ppm silicon and
As these amides were isolated without being solubilised and
filtered, one could wonder about the presence of contami-
48 ppm iron were obtained in N-benzylbenzamide when the
reaction was performed with the jar and balls made from
[10]
[15]
nants arising from wear of the balls and container. To ad-
dress this point, the content of metal contaminant in N-benzyl-
benzamide was measured either by inductively coupled
plasma mass spectrometry (ICP-MS) or by gravimetric analysis.
When the standard procedure was applied to the synthesis of
N-benzylbenzamide, ICP-MS analysis indicated the presence of
very low amounts of the two main components of the stain-
less-steel grinding jar (366 ppm Fe and 169 ppm Cr; Table 2,
agate (Table 2, entry 9). Although the yield was slightly lower
in this case (67%), completion of the reaction indicated that
heavy balls such as those made of stainless steel were not
mandatory for the success of this CDI-mediated mechanosyn-
thesis of N-benzylbenzamide. Indeed, performing the same re-
action in a 25 mL PTFE jar containing one 10 mm diameter
À3
PTFE ball with a steel core (1PTFE ball =3.3 gcm ) also produced
N-benzylbenzamide in a yield of 69% (Table 2, entry 10). Be-
cause under these conditions wear produced PTFE particles,
the contaminant content was quantified by gravimetric analy-
sis. This indicated that 2400 ppm contaminant was present in
N-benzylbenzamide when produced with PTFE milling material
(Table 2, entry 10). To our delight, it was discovered that even
lighter milling material such as rubber can induce the mecha-
[
15]
entry 1). As expected, solubilising the product in EtOAc fol-
lowed by filtration either through cotton or through a 0.2 mm
PTFE filter eliminated the contaminants to such an extent that
neither iron nor chromium were detectable by ICP-MS analysis
(
Table 2, entries 2 and 3). In this case, using an organic solvent
can be regarded as lowering the interest of the present meth-
odology. Nevertheless, it has to be noted that in this peculiar
situation, the only property expected from the organic solvent
is to be able to solubilise the product, which leaves the chem-
ist with a large number of possibilities in the choice of accept-
able solvent.
À3
nosynthesis of N-benzylbenzamide (1rubber ꢀ1.3 gcm ). Thus,
utilising an 18 mm diameter rubber ball in a PTFE jar furnished
N-benzylbenzamide in a yield of 80% (Table 2, entry 11). In this
case, gravimetric analysis of the product indicated a contami-
nant content of less than 1000 ppm. As the reliability of gravi-
Chem. Eur. J. 2015, 21, 12787 – 12796
www.chemeurj.org
12789
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. Influence of the milling material on reaction efficiency and contamination from wear.
Entry Material & size
of jar
Material, number & Type of
Jar
Post-treatment
Reaction
scale [mmol] [%] analysis
Yield Type of
Contaminant con-
tent in ppm (metal)
diameter of balls
ball mill status
1
2
3
4
5
6
7
8
stainless steel,
2 mL
stainless steel,
50”5 mm
planetary used suspension in H
2
O and filtration 1.35
88
ICP-MS
ICP-MS
ICP-MS
ICP-MS
ICP-MS
ICP-MS
ICP-MS
ICP-MS
ICP-MS
gravimetric 2400
gravimetric <1000
gravimetric 42000
gravimetric 3400
366 (Fe); 169 (Cr)
1
[
a]
solubilised in EtOAc and filtered
through cotton
solubilised in EtOAc and filtered
through 0.2 mm PTFE filter
<30 (Fe); <30 (Cr)
[a]
<30 (Fe); <30 (Cr)
stainless steel,
stainless steel,
vibratory used suspension in H
2
O and filtration 1.13
85
75
59
67
86
1400 (Fe); 1000 (Cr)
642 (Fe); 176 (Cr)
7400 (Fe); 1400 (Cr)
1200 (Fe); 314 (Cr)
1
0 mL screw-top 2”10 mm
stainless steel,
stainless steel,
used
new
new
used
0.56
1.13
0.56
1.13
5
mL screw-top 2”5 mm
stainless steel,
stainless steel,
2”10 mm
stainless steel,
2”5 mm
1
0 mL
stainless steel,
mL
ZrO , 10 mL
5
[
b]
c]
2
ZrO
2
, 2”12 mm
151 (Fe); 12200 (Zr);
740 (Y)
screw-top
[a]
9
0
agate, 50 mL
PTFE, 25 mL
screw-top
PTFE, 25 mL
screw-top
agate, 4”12 mm
PTFE with steel core, vibratory
1”10 mm
planetary
2.7
5.0
67
69
<100 (Si); 48 (Fe)
[
1
[a]
11
rubber, 1”18 mm
1.35
22.5
22.5
80
92
91
[
[
b]
b]
1
1
2
3
stainless steel,
stainless steel,
50”10 mm
rubber, 10”18 mm
planetary
2
50 mL
stainless steel,
50 mL
2
[
a] Contaminant content below the detection limit of the analytical method. [b] Corrected yield based on the contaminant content. [c] Agitation performed
at 20 Hz.
metric analysis performed on a 100 mg sample is much lower
than that of ICP-MS analysis, both the efficiency and wear re-
sistance of the rubber milling material in the production of N-
benzylbenzamide was tested on a larger scale. First, the reac-
tion was performed on a 22.5 mmol scale in a 250 mL jar con-
taining fifty 10 mm diameter stainless-steel balls, which pro-
from procedures avoiding the use of organic solvent from re-
action to product recovery.
Influence of milling time on contamination from wear
The influence of milling time on contamination content was
also studied. The reaction mixture of the N-benzylhydrocinna-
mide synthesis was sampled every 5 min and analysed by
atomic absorption spectroscopy (AAS) to determine the iron
content. Hydrocinnamic acid and CDI were first agitated for
5 min in the ball mill. After the addition of benzylamine fol-
lowed by mixing for 5 min, the iron content reached 190 ppm
(Figure 2). When the reaction mixture was sampled after 10
and 15 min of agitation, the iron content had increased to 216
and 319 ppm, respectively. These results confirm that contami-
nation from wear increased with time, highlighting the impor-
tance of the optimisation of reaction time in completely sol-
vent-free ball-mill-mediated organic syntheses to reduce prod-
uct contamination from wear. This analytical study of the
metal traces issuing from the wear of balls and container
shows that as long as the presence of metal traces is compati-
ble with the intended use of the products, synthesis and purifi-
cation can be realised in the complete absence of organic sol-
[16]
duced N-benzylbenzamide in a yield of 92%, albeit contain-
ing 42000 ppm metallic contaminants, as analysed by gravi-
metric analysis (Table 2, entry 12). Changing the stainless-steel
balls to ten 18 mm diameter rubber balls furnished N-benzyl-
benzamide in a yield of 91% with a much lower contaminant
content of 3400 ppm (Table 2, entry 13). The higher contami-
nant content obtained with rubber balls on a 22.5 mmol scale
(
Table 2, entry 13) compared with the contamination resulting
from the experiment realised with one rubber ball on
a 1.35 mmol scale (Table 2, entry 11) can be explained by the
higher intensity of collisions between the balls and walls. How-
ever, these results clearly show that contamination from wear
can be substantially reduced by a judicious choice of milling
material. As solvent-free organic syntheses often imply “soft”
reaction mixtures, selecting lighter grinding materials instead
of classical heavy materials such as stainless steel could greatly
help in reducing the contaminant content of products issuing
Chem. Eur. J. 2015, 21, 12787 – 12796
www.chemeurj.org
12790
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
a polar nitro group on the aromatic ring of phenylacetic acid
prevented the amides 4 and 5 from being isolated in good
yields (69 and 91%, respectively; Table 3). Similarly, the treat-
ment of a-methylphenylacetic acid with CDI and benzylamine
hydrochloride led to isolation of N-benzyl-2-phenylpropana-
mide (6) in a yield of 70% (Table 3, compound 6). The higher
yield of amide 3 (93%) compared with that of amide 6 could
be explained by a greater steric hindrance in 2-phenylpropion-
ic acid compared with in phenylacetic acid. However, HPLC
analysis revealed the presence of amide 6 in the aqueous fil-
trate. Thus, EtOAc extraction of the filtrate obtained from the
synthesis of 6 followed by purification on silica gel led to an
increased yield of 6 of 86%. This result indicates that a lower
yield of amide 6 (without extraction of the filtrate) could be ex-
plained by a higher water solubility of the product. Similarly,
CDI activation of highly hindered acids such as 1-adamantane-
carboxylic acid is not a limitation in this approach as the latter
could be transformed into amide 7 in a good yield of 77%
Figure 2. Iron content as a function of time during the synthesis of N-ben-
zylhydrocinnamide.
(
Table 3). Primary amines other than benzylamine could also
be acylated with high efficiency. Thus, phenethylamine and
2,4-dimethoxybenzylamine hydrochlorides furnished the corre-
sponding amides in yields of 96 and 84%, respectively (Table 3,
vents. If the presence of metal traces is inconsistent with the
required use of the products, these impurities can be largely
removed by solubilisation of the
amide in a low-toxic solvent
[a]
Table 3. Scope and limitations of the CDI-mediated mechanosynthesis of amides.
such as EtOAc followed by
simple filtration. In addition, con-
tamination from wear can be
greatly reduced by an appropri-
ate choice of milling material
and by optimisation of milling
time.
Scope and limitations for acyla-
tion of N-nucleophiles
To give a global view of the sub-
strate scope and limitations of
this approach, preliminary results
from a study that we published
previously have been included in
[12]
the following discussion.
Ap-
plication of this approach to
other carboxylic acids and
amines was realised on the
1
.35 mmol scale in
stainless-steel jar containing fifty
mm diameter balls under plan-
etary ball-milling agitation.
a 12 mL
5
Linear carboxylic acids such as
hexanoic acid, 4-phenylbutyric
acid or phenylacetic acid reacted
efficiently with CDI and benzyl-
amine hydrochloride to furnish
the corresponding amides in
high yields (Table 3, compounds
[a] Realised on a 1.35 mmol scale unless otherwise noted. [b] After extraction of the aqueous phase with
EtOAc. [c] Completed within 20 min. [d] Completed within 1.5 h. [e] Determined by chiral HPLC, see the Sup-
porting Information for details. [f] Completed within 30 min. [g] After agitation for 50 min with (S)-
methylbenzylamine hydrochloride on a 1.35 mmol scale. [h] Completed within 30 min with (S)-methylbenzyl-
amine on a 0.45 mmol scale. [i] Determined by H and C NMR spectroscopy. No signal corresponding to the
diastereoisomer was observed. [j] Realised with 0.5 mmol of (S)-(+)-2-(6-methoxy-2-napthyl)propionic acid,
1
13
1.0 equiv of CDl and 1.9 equiv of (S)-methylbenzylamine. [k] Completed within 6 h. [l] Completed within 3 h.
1
–3). It is worth noting here that
[m] Completed within 4 h. [n] Completed within 3 h 10 min.
neither bromine atom nor
a
Chem. Eur. J. 2015, 21, 12787 – 12796
www.chemeurj.org
12791
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
compounds 8 and 9). Non-hindered nitrogen nucleophiles
such as O-benzyl hydroxylamine and phenylhydrazine were
also efficiently benzoylated, furnishing amide 10 and hydrazide
starting from 2-methyl-3-nitrophenylacetic acid furnished 3.3 g
of amide 20 in a moderate yield of 46%, slightly better than
that previously reported in the literature using a different strat-
[
17]
11 in yields of 64 and 83%, respectively. a-Substituted primary
egy. It is worth noting that in these cases distillation of the
reaction mixture would also efficiently eliminate any contami-
nant issuing from wear. Another requirement for the success
of this approach is the hydrophobicity of the products. When
coupling benzoic acid with morpholine, the reaction mixture
was completely solubilised after the addition of water to the
jar, excluding filtration as a way to recover the product. Yet,
EtOAc extractions of this aqueous homogeneous phase fur-
nished N-benzoylmorpholine (21) in a yield of 74% (Table 3).
This result shows that the solubility of amide 21 in water pre-
vented its organic-solvent-free recovery and purification, but
not its formation from benzoic acid, CDI and morpholine in
the absence of solvent under mechanically-mediated agitation.
We next explored the reactions with aniline derivatives,
which are known to react sluggishly with acyl-imidazole inter-
amines reacted more slowly than non-substituted primary
amines. Indeed, agitation of 20 min and 1.5 h, respectively,
were necessary to obtain the full conversion of benzoylimida-
zole intermediates when benzhydrylamine and cyclohexyl-
amine hydrochlorides were used. Yet, the corresponding
amides 12 and 13 were both isolated in good yields (82 and
7
7%, Table 3). When enantiopure a-substituted primary amines
such as (S)-a-methylbenzylamine and phenylalanine tert-butyl
ester hydrochlorides were used as nucleophiles, the corre-
sponding amides 14 and 15 were obtained as enantiopure
compounds in yields of 68 and 78%, respectively.
We next evaluated the stereoselectivity of the process with
enantiopure carboxylic acids as starting materials. When
1
1
.5 mmol of Boc-protected l-phenylalanine was treated with
.0 equivalent of CDI and then with 0.9 equivalents of (S)-a-
[
18]
mediates. When 2-phenylpropionylimidazole, prepared from
2-phenylpropionic acid and CDI, was treated with aniline hy-
drochloride in the absence of solvent in a ball mill, the corre-
sponding amide 22 was isolated in a yield of 91% after only
10 min of reaction (Table 4). It is worth noting that when per-
formed in boiling [0.39m] EtOAc, the same reaction required
methylbenzylamine during 50 min, the amide 16 was isolated
in a very low yield (19%), which can be explained by a sticky
reaction media preventing efficient agitation of the reaction
mixture (Table 3). This limitation could be solved by both per-
forming the reaction on a 0.45 mmol scale and by using (S)-a-
methylbenzylamine as a free base. Under these conditions,
amide 16 was isolated in a yield of 49% without any loss of
chirality as no signal corresponding to the diastereoisomer was
[
19]
3 h to reach 86% conversion.
Similarly, N-phenylbenzamide (23) was obtained in a yield of
93% after 10 min when the reaction was performed without
solvent in a ball mill, whereas 1 h was necessary to reach 90%
conversion when the reaction was realised in N-methyl-2-pyr-
1
observed by H NMR spectroscopy. Similarly, the reaction of
1
.5 mmol of both enantiopure Naproxen and CDI followed by
[
18]
agitation with 1.35 mmol of (S)-a-methylbenzylamine hydro-
chloride furnished only traces of the amide 17. As previously,
performing the CDI-mediated activation of Naproxen on
a smaller scale (0.5 mmol in this case) followed by agitation
with 1.9 equivalents of (S)-a-methylbenzylamine as a free base
furnished the amide 17 in high yield (81%) and diastereomeric
excess (Table 3).
rolidone (NMP) at 508C (Table 4). As in solution, the presence
of an electron-withdrawing group such as a trifluoromethyl or
cyano group on the aromatic ring slowed the reaction. Indeed,
Table 4. Comparison of the rate of acylation of aniline hydrochloride de-
rivatives under solvent and solvent-free conditions.
Coupling with secondary amines was next envisioned. The
treatment of hydrocinnamic acid with CDI followed by the ad-
dition of dibenzylamine hydrochloride required agitation for
6
h to furnish amide 18 in a yield of 44% (Table 3). LC-MS anal-
ysis of the aqueous filtrate revealed no trace of amide 18 but
the presence of both hydrocinnamic acid and dibenzylamine.
Thus, the relatively low yield of 18 can be explained either by
incomplete conversion of the starting materials or by partial
hydrolysis of the hydrocinnamic acyl-imidazole during the pro-
cess. The recovery of the product after agitation with water is
efficient as long as the product is a solid. As N,N-di-n-propyl-3-
phenylpropanamide (19) is an oil that solubilises imidazole at
ambient temperature, we envisaged carrying out the reaction
on a larger scale to realise an organic-solvent-free treatment of
the liquid reaction mixture. Indeed, performing the reaction
with 25 mmol of hydrocinnamic acid, 1.1 equivalents of CDI
and 1.1 equivalents of di-n-propylamine allowed an easy trans-
fer of the liquid reaction mixture directly into a Kugelrohr ap-
paratus. Subsequent distillation followed by aqueous washings
and drying produced 3.6 g of N,N-di-n-propyl-3-phenylpropa-
namide (19) in a yield of 62% (Table 3). The same approach
[
a] Reaction time to reach 86% conversion for 22 and >90% conversion
for compounds 23–25. [b] See ref. [19]. [c] See ref. [18]. [d] From 4-amino-
benzonitrile, yield corrected based on contaminant content. [e] Deter-
mined by AAS.
Chem. Eur. J. 2015, 21, 12787 – 12796
www.chemeurj.org
12792
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
3
0 min and 3.5 h were required to produce amides 24 and 25
Table 5. Examples of the organic-solvent-free CDI-mediated acylation of
O-, S- and C-nucleophiles.
from p-trifluoromethylaniline hydrochloride and p-aminoben-
zonitrile, respectively (Table 4). Yet, these acylation reactions
were still more rapid than in [0.41m] NMP at 508C; under
these conditions, 29 h and more than 143 h were necessary to
reach 90% conversion. It is worth noting that the contamina-
tion from wear in the case of the reaction of benzoylimidazole
with p-aminobenzonitrile was relatively low (0.9% w/w) when
considering that the ball-mill-mediated agitation was per-
formed for 3.5 h.
[a]
This approach was next applied to the production of the
[20]
active metabolite of Leflunomide, which was approved for
commercialisation under the name Teriflunomide (trade name
Aubagio) by the FDA in 2012 for the treatment of multiple
[
21]
sclerosis. Indeed, the treatment of 5-methyl-4-isoxazolecar-
boxylic acid with CDI followed by mixing with 4-(trifluorome-
thyl)aniline hydrochloride during 5 h and then with water
during 5 min produced Leflunomide as a fine suspension in
water (Scheme 2). This amide was not isolated but submitted
to hydrolysis under acidic conditions with magnetic stirring to
furnish Teriflunomide in a yield of 81% after filtration and
drying. AAS analysis of the final product revealed only low
traces of iron (678 ppm). As one could expect higher wear con-
tamination resulting from a relatively long grinding time, this
low iron content could be explained by the partial oxidation of
metallic particles during the acidic aqueous treatment, water
solubilisation of the corresponding cations and elimination
during the filtration step.
[a] Realised on a 1.35 mmol scale. [b] Performed in a planetary mill rotat-
ed at 500 rpm. [c] Performed in a vibratory mill at 25 Hz. [d] Recovered
with EtOAc.
used as nucleophiles as phenylmethanethiol and 4-methoxy-
benzenethiol furnished the corresponding benzothioates 29
and 30 in yields of 77 and 81%, respectively (Table 5). When 3-
phenylpropanoylimidazole was ball-milled with thiophenol, the
corresponding S-phenyl 3-phenylpropanethioate (31) was iso-
lated in a yield of 91%.
We next envisioned applying this approach to C-nucleo-
philes. The first attempt using the 1,3-dione dimedone (5,5-di-
methyl-1,3-cyclohexanedione) as the nucleophile did not fur-
nish any acylated product (neither C- nor O-acylated). The ad-
dition of N,N-diisopropylethylamine (DIEA) as base to the reac-
tion mixture did not lead to a successful reaction. In contrast,
the treatment of isobutyroylimidazole resulting from the CDI
activation of isobutyric acid with malononitrile furnished enol
Scheme 2. Mechanosynthesis of Teriflunomide.
32 in a yield of 85% (Table 5). Yet, the solubility of the product
in both neutral and acidic water forced us to recover it from
the aqueous phase by extraction with EtOAc. Finally, benzoic
acid could be transformed into enol 33 in a yield of 95% after
reaction with CDI followed by the addition of malononitrile
(Table 5). Despite being completely solubilised after ball-milling
with water, enol 33 could be recovered without using organic
solvent by acidification of the aqueous reaction mixture and
filtration of the resulting precipitate. These results show that,
as long as the final products are solids and not soluble in
water, O-, S- and C-nucleophiles can be acylated with high effi-
ciency without using organic solvents from reaction to product
recovery.
Application to O-, S- and C-nucleophiles
This organic-solvent-free CDI-mediated acylation strategy was
also applied to other nucleophiles. Indeed, when benzoylimi-
dazole resulting from CDI activation of benzoic acid was ball-
milled with phenol, phenyl benzoate (26) was isolated in
a yield of 84% (Table 5). Similarly, the treatment of benzoic
acid with CDI in a vibratory ball mill agitated at 25 Hz followed
by the addition of 3-nitrophenylmethanol furnished 3-nitro-
benzyl benzoate (27) in a yield of 85% (Table 5). Phenyl 2-phe-
nylpropanoate (28) was isolated in a yield of 80% by using the
same approach, yet it had to be recovered from the milling jar
with EtOAc due to its low melting point. Thiols could also be
Chem. Eur. J. 2015, 21, 12787 – 12796
www.chemeurj.org
12793
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Evaluation of the environmental impact
water or organic-solvent recycling steps. With a simple E-factor
value of 2.36, the organic-solvent-free CDI-mediated mechano-
synthesis of N-benzylbenzamide presents no clear advantage
over the other strategies (Table 6). On the other hand, the CDI-
mediated mechanosynthesis of N-benzylbenzamide appears to
be the most interesting approach when considering every ma-
terial utilised (including water and organic solvent) with the
lowest complete E-factor score of 20.8 (Table 6). Another very
simple way to evaluate the environmental impact of an experi-
mental procedure, especially for the chemist who is going to
perform the reaction at the bench, is to look at the cumulative
number of GHS pictograms of the materials he is going to ma-
nipulate. Considering all the strategies reported for the synthe-
sis of N-benzylbenzamide, our approach has a cumulative
number of GHS pictograms of 6, whereas for all the other strat-
egies this number is equal to or greater than 8 (Table 6).
We then evaluated the environmental impact of this approach
based on the mechanosynthesis of N-benzylbenzamide and
compared it with previously reported reaction conditions
using various reagents: N,N-diisopropylbenzylamine-2-boronic
[
22]
[23]
acid (IBA),
N,N’-dicyclohexylcarbodiimide (DCC),
thionyl
[
24]
[25]
chloride (SOCl ), N,N’-carbonyldiimidazole (CDI) and sulfat-
2
26]
[
ed tungstate. First of all, it has to be highlighted that CDI in
THF under classical stirring or without solvent in a ball mill was
the most efficient reagent as N-benzylbenzamide was isolated
in the highest yield (92%, Table 6, entries 4 and 6). Yet, the use
Table 6. Comparison of the environmental impact of the various strat-
egies for the synthesis of N-benzylbenzamide.
[
28]
[
a]
[b]
Finally, when considering the Eco-scale score,
a green
Entry Reagent
Yield Atom econ- sEF cEF GHS picto- Eco-
[
c]
[d]
[
%] omy [%]
grams
scale
metric ranging from 0 (completely failed reaction) to 100 (ideal
reaction) taking into account yield, cost, safety, technical setup,
temperature and reaction time, workup and purification, the
CDI-mediated mechanosynthesis of N-benzylbenzamide has
the highest score (79), which corresponds to an excellent syn-
thesis. These excellent green metric scores can partially be ex-
plained by the use of non-toxic water in the process. One
could argue that the utilisation of water, a precious resource
on earth, results in the contamination of water with organic
material, implying costly aqueous-phase post-treatments and
an overall negative influence on the environmental impact of
the process. Yet, to the best of our knowledge, none of the
previously described acylation strategies have avoided water
contamination with organic material, due to the systematic
use of water washings to obtain pure acylated products. Con-
sequently, environmental impact superiority of the present ap-
proach over classical strategies does not rely only on the use
of water as an innocuous entity, but, inter alia, on its original
and specific use (water-assisted ball-milling) enabling the com-
plete elimination of organic solvents during the purification
and recovery steps.
[
22]
1
2
3
4
IBA
50
82
72
92
92.1
48.5
60.6
54.0
2.38 396.2 8
44
62
54
67
[23][e]
DCC
SOCl
CDI in
THF
sulfated
tungstate
solvent-free 92
CDI
2.75 66.1 8
2.32 24.6 8
2.01 41.9 9
[
2
24][e]
[25][e]
5
6
81
92.1
49.3
1.55 23.5 9
2.36 20.8 6
49.5
79
[26]
[
a] Simple E-factor [b] Complete E-factor. [c] Cumulative number of GHS
pictograms. [d] See ref. [28]. [e] See ref. [26] for raw data.
of CDI is detrimental to the atom economy of the reaction like
other reagents that have to be used in stoichiometric quanti-
ties. Indeed, the atom economy is 54.0 and 49.3% when CDI is
used in THF and without solvent, respectively (Table 6, en-
tries 4 and 6). In contrast, the atom economy reaches 92.1%
when catalysts such as IBA or sulfated tungstate are employed
(
Table 6, entries 1 and 5). The yield and atom economy are
both simple and powerful parameters for evaluating the envi-
ronmental impact of a reaction, but they omit critical issues
such as the amounts of organic solvents and water required to
perform the reaction and isolate the final product. Indeed, sol-
vents and water account for 56 and 32%, respectively, of the
Conclusion
The acylation of amines and other nucleophiles can be effi-
ciently performed in the absence of any organic solvent from
reaction to product recovery when combining CDI activation
of carboxylic acids and ball-milling agitation. This strategy is
applicable to a wide range of carboxylic acids and nucleophiles
provided that the corresponding final products are immiscible
in water and that their contaminant contents are compatible
with their intended use. However, we have shown that the
amount of contaminant issuing from wear can be severely re-
duced by optimisation of the reaction time and appropriate
choice of milling material. In particular, a “soft” reaction mix-
ture, such as those frequently encountered in solvent-free or
solventless organic syntheses, does not require dense milling
materials. Indeed, utilising lighter milling material such as
agate, PTFE, or even rubber enables a sharp reduction of con-
tamination from wear without limiting the reaction efficiency.
[
3]
total mass of material used in the pharmaceutical industry.
Although green metrics such as the environmental factor (E-
[
27]
[3]
factor) and process mass intensity (PMI) have been devel-
oped to take into account the quantity of waste produced
during a reaction or the total mass of materials required to
[3]
obtain the final product, respectively, the lack of homogenei-
ty in the consideration of water and organic solvents in the
calculations by the chemistry community has led to confusion
and difficulties in the interpretation of these parameters.
This observation led Roschangar et al. to propose definitions
of the “simple E-factor” and the “complete E-factor” as well as
[29]
their mode of utilisation.
The simple E-factor (noted sEF)
omits water and solvents in the calculations, whereas the com-
plete E-factor (noted cEF) takes into account every material en-
tering into the process without taking into consideration any
Chem. Eur. J. 2015, 21, 12787 – 12796
www.chemeurj.org
12794
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
In the limited number of cases in which the products are solu-
ble in water, organic-solvent-based extraction appears to be
a convenient solution to the recovery of the final products
from the aqueous phase, although preventing the whole pro-
cess from being organic-solvent-free from reaction to product
recovery. Yet, the absence of organic solvent during the reac-
tion is still a decisive advantage as a tremendous reduction of
the reaction time has been observed when performing the
acylation of aniline derivatives under mechanical agitation. Fi-
nally, calculation and analysis of the various green metrics
have shown that this approach can be an efficient way to miti-
gate the environmental impact of nucleophile acylation reac-
tions.
Synthesis of N-benzylhydrocinnamide
Following the general procedure, 3-phenylpropionic acid (0.225 g,
1
.5 mmol) was used as the acid and benzylamine hydrochloride
(
0.194 g, 1.35 mmol) as the nucleophile. N-Benzylhydrocinnamide
[
30]
was recovered as a white solid (0.302 g, 94%). M.p. 79–808C (lit.:
808C); H NMR (200 MHz, CDCl
1
): d=7.38–7.12 (m, 10H), 5.60 (brs,
3
1
7
1
2
H), 4.42 (d, J=5.7 Hz, 2H), 3.03 (t, J=7.6 Hz, 2H), 2.54 ppm (t, J=
13
.6 Hz, 2H); C NMR (75 MHz, CDCl ): d=172.1, 140.9, 138.3, 128.7,
3
28.6, 128.5, 127.8, 127.5, 126.3, 43.6, 38.5, 31.8 ppm; MS (ESI): m/z:
+
40.1 [M+H] .
Synthesis of N-benzylbenzamide on a 22.5mmol scale with
rubber balls
Benzoic acid (3.053 g, 25 mmol) and CDI (4.062 g, 25 mmol) were
introduced into a 250 mL stainless-steel grinding bowl with 10
rubber balls (18 mm diameter). The bowl was closed and subjected
to grinding for 5 min in the planetary ball mill operated at
Experimental Section
5
00 rpm. Then benzylamine hydrochloride (3.231 g, 22.5 mmol)
was added. The bowl was closed and subjected to grinding for
0 min in the planetary ball mill operated at 500 rpm. Deionised
All reagents were purchased from Aldrich Chemical Co., Fluka and
Rhône-Poulenc and used without further purification. The milling
treatments were carried out either in a Retsch PM100 Planetary
Mill or in a vibrating Retsch Mixer Mill 200. FTIR spectra were re-
corded on a Perkin–Elmer Spectrum 100 FT-IR spectrophotometer.
1
water (40 mL) was added and the bowl was closed and subjected
to grinding for 5 min in the planetary ball mill operated at
5
00 rpm. After this treatment, the fine suspension was transferred
1
onto a glass frit, washed with deionised water and dried under
H NMR spectra were recorded on a Bruker Avance DPX 200 MHz
vacuum over P O to obtain N-benzylbenzamide as a white solid
2
5
spectrometer and are reported in ppm using the solvent as the in-
(
4.33 g, 91%, corrected yield based on the content of contami-
ternal standard (CDCl at 7.24 ppm). The NMR data are reported as
3
[31]
1
nant). M.p. 104–1058C (lit.:
1048C); H NMR (200 MHz, CDCl3):
follows: s=singlet, d=doublet, t=triplet, m=multiplet, br=
broad; coupling constant in Hz; integration. C NMR spectra were
recorded on a Bruker Avance AM 75 MHz spectrometer and are re-
1
3
d=7.82–7.73 (m, 2H), 7.53–7.22 (m, 8H), 6.38 (brs, 1H), 4.64 ppm
13
(
d, J=5.7 Hz, 2H); C NMR (75 MHz, CDCl ): d=167.8, 138.4, 134.6,
3
1
31.8, 129.0, 128.8, 128.1, 127.8, 127.2, 44.3 ppm; MS (ESI): m/
ported in ppm using the solvent as the internal standard (CDCl at
3
+
z:212.2 [M+H] .
7
7.2 ppm). Analytical samples for NMR analysis were filtered
ꢁ
through Rotilabo syringe filters (PTFE, unsterile, 0.20 mm pore
size). Mass spectra were obtained by LC-MS by means of ESI using
a Water Alliance 2695 LC instrument coupled to a Waters ZQ spec-
trometer with an electrospray source, a simple quadrupole analyser
and a UV Waters 2489 detector. HRMS analyses were performed
with a Q-Tof (Waters, ESI, 2001) spectrometer. Enantiomeric excess-
es were measured by using a Beckman Coulter System Gold 126
Solvent Module HPLC machine and System Gold 168 Detector with
a 4.6”250 mm Daicel Chiralpak OD columns using n-hexane and
[
32]
Synthesis of Teriflunomide
5
(
-Methylisoxazole-4-carboxylic acid (191 mg, 1.5 mmol) and CDI
243 mg, 1.5 mmol) were introduced into stainless-steel grinding
bowl with 50 stainless-steel balls (5 mm diameter). The bowl was
closed and subjected to grinding for 20 min in the planetary ball
mill operated at 500 rpm. Then 4-trifluoromethylaniline hydrochlo-
ride (267 mg, 1.35 mmol) was added. The bowl was closed and
subjected to grinding for 5 h in the planetary ball mill operated at
2
-propanol as solvents. ICP-MS analyses were realised by the Ser-
5
00 rpm, with 1 min break every 10 min and inversion of the rota-
vice Central d’Analyse of the CNRS. Atomic absorption analyses
were realised on a Philips Pye Unicam SP9 spectrometer. Samples
were placed under an air/acetylene atmosphere.
tion direction after each break. Deionised water (4 mL) was added
and the bowl was closed and subjected to grinding for 5 min in
the planetary ball mill operated at 500 rpm. After this treatment,
the fine suspension was transferred to a round-bottomed flask, the
pH was adjusted to 1 with concentrated hydrochloric acid and the
suspension was stirred with a magnetic stirrer for 24 h. The sus-
pension was then filtered and dried under vacuum over P O to
General procedure for the acylation of N-, O-, S- and C-nu-
cleophiles
2
5
obtain a white solid. 2-Cyano-3-hydroxy-N-(4-trifluoromethylphen-
yl)-2-butenamide (Teriflunomide) as the major tautomer was recov-
ered as a white solid (0.297 g, 81%). M.p. 230–2328C; H NMR
The carboxylic acid (1.5 mmol) and CDI (243 mg, 1.5 mmol) were
introduced into a 12 mL stainless-steel grinding bowl with 50 stain-
less-steel balls (5 mm diameter). The bowl was closed and subject-
ed to grinding for 5 min in the planetary ball mill operated at
1
(
7
200 MHz, [D ]DMSO): d=10.97 (m, 2H), 7.75 (d, J=8.5 Hz, 2H),
.64 (d, J=7.6 Hz, 2H), 2.23 ppm (s, 3H); C NMR (150 MHz,
6
13
[
1
D ]DMSO): d=187.3, 166.5, 142.1, 126.3, 126.0, 123.7, 123.3, 122.7,
6
20.7, 120.2, 119.4, 80.5, 23.9 ppm; MS (ESI): m/z:271 [M+H] .
5
00 rpm. Then the nucleophile (1.35 mmol) was added. The bowl
+
was closed and subjected to grinding for 10 min in the planetary
ball mill operated at 500 rpm. Deionised water (4 mL) was added
and the bowl was closed and subjected to grinding for 5 min in
the planetary ball mill operated at 500 rpm. After this treatment,
the resulting fine suspension was filtered, washed with deionised
water and then dried under vacuum over P O to obtain the final
Acknowledgements
The CNRS and UniversitØ Montpellier (post-doctoral grant to
2
5
product.
J.B.) are acknowledged for their financial support. FranÅois
Chem. Eur. J. 2015, 21, 12787 – 12796
www.chemeurj.org
12795
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
MØtro is warmly acknowledged for production of the graphical
abstract.
[13] a) J. Bonnamour, T.-X. MØtro, J. Martinez, F. Lamaty, Green Chem. 2013,
1
5, 1116–1120; b) T.-X. MØtro, X. J. Salom-Roig, M. Reverte, J. Martinez,
F. Lamaty, Green Chem. 2015, 17, 204–208; c) T.-X. MØtro, E. Colacino, J.
Martinez, F. Lamaty, in Ball Milling Towards Green Synthesis: Applications,
Projects, Challenges, The Royal Society of Chemistry, Cambridge, 2015,
pp. 114–150.
Keywords: acylation
·
amides
·
green chemistry
·
mechanochemistry · solvent-free synthesis
[14] For previous examples of ex situ analysis, see reference [7c] and : a) X.
Ma, W. Yuan, S. E. J. Bell, S. L. James, Chem. Commun. 2014, 50, 1585–
[
1] a) N. I. Vasilevich, R. V. Kombarov, D. V. Genis, M. A. Kirpichenok, J. Med.
Chem. 2012, 55, 7003–7009; b) J. S. Carey, D. Laffan, C. Thomson, M. T.
Williams, Org. Biomol. Chem. 2006, 4, 2337–2347.
2] D. S. MacMillan, J. Murray, H. F. Sneddon, C. Jamieson, A. J. B. Watson,
Green Chem. 2013, 15, 596–600.
3] C. Jimenez-Gonzalez, C. S. Ponder, Q. B. Broxterman, J. B. Manley, Org.
Process Res. Dev. 2011, 15, 912–917.
4] a) S. L. James, C. J. Adams, C. Bolm, D. Braga, P. Collier, T. Fri sˇ cˇ i c´ , F. Gre-
pioni, K. D. M. Harris, G. Hyett, W. Jones, A. Krebs, J. Mack, L. Maini, A. G.
Orpen, I. P. Parkin, W. C. Shearouse, J. W. Steed, D. C. Waddell, Chem.
Soc. Rev. 2012, 41, 413–447; b) E. Boldyreva, Chem. Soc. Rev. 2013, 42,
1587; b) I. A. Tumanov, A. F. Achkasov, E. V. Boldyreva, V. V. Boldyrev,
CrystEngComm 2011, 13, 2213–2216; for previous examples of in situ
analysis, see: c) A. D. Katsenis, A. Pu sˇ kari c´ , V. Strukil, C. Mottillo, P. A.
Julien, K. U zˇ arevi c´ , M.-H. Pham, T.-O. Do, S. A. J. Kimber, P. Lazi c´ , O. Mag-
dysyuk, R. E. Dinnebier, I. Halasz, T. Fri sˇ cˇ i c´ , Nat. Commun. 2015, 6, 6662;
d) D. Gracin, V. Strukil, T. Fri sˇ cˇ i c´ , I. Halasz, K. U zˇ arevi c´ , Angew. Chem. Int.
Ed. 2014, 53, 6193–6197; Angew. Chem. 2014, 126, 6307–6311.
ˇ
[
[
[
ˇ
[
15] Supplier material analysis indicates the following content of the stain-
less-steel jars (in %): Fe (82.925), Cr (14.5), Mn (1), Si (1), C (0.5), P
(0.045), S (0.03). For the jar made of agate: SiO
2 2 3
(99.91), Al O (0.02),
Na O (0.02), Fe (0.01), K O (0.01), MnO (0.01), MgO (0.01), CaO (0.01)
2
2
O
3
2
7
719–7738; c) T. Fri sˇ cˇ i c´ , S. L. James, E. V. Boldyreva, C. Bolm, W. Jones, J.
[
[
16] Corrected yield based on the content of contaminant.
17] R. K. Mylavarapu, G. C. M. Kondaiah, N. Kolla, R. Veeramalla, P. Koilkonda,
A. Bhattacharya, R. Bandichhor, Org. Process Res. Dev. 2007, 11, 1065–
Mack, J. W. Steed, K. S. Suslick, Chem. Commun. 2015, 51, 6248–6256.
5] I. Huski c´ , I. Halasz, T. Fri sˇ cˇ i c´ , H. Van cˇ ik, Green Chem. 2012, 14, 1597–
[
[
1
600.
1068.
6] a) G. Kaupp, J. Schmeyers, M. R. Naimi-Jamal, H. Zoz, H. Ren, Chem. Eng.
Sci. 2002, 57, 763–765; b) G. Kaupp, M. R. Naimi-Jamal, V. Stepanenko,
Chem. Eur. J. 2003, 9, 4156–4161; c) G. Kaupp, M. Reza Naimi-Jamal, J.
Schmeyers, Tetrahedron 2003, 59, 3753–3760; d) G. Kaupp, CrystEng-
Comm 2006, 8, 794–804; e) P. Nun, C. Martin, J. Martinez, F. Lamaty, Tet-
rahedron 2011, 67, 8187–8194; f) M. Ferguson, N. Giri, X. Huang, D. Ap-
perley, S. L. James, Green Chem. 2014, 16, 1374–1382.
[
[
[
[
18] E. K. Woodman, J. G. K. Chaffey, P. A. Hopes, D. R. J. Hose, J. P. Gilday,
Org. Process Res. Dev. 2009, 13, 106–113.
19] P. J. Dunn, W. Hoffmann, Y. Kang, J. C. Mitchell, M. J. Snowden, Org. Pro-
cess Res. Dev. 2005, 9, 956–961.
20] D. Magne, F. MØzin, G. Palmer, P. A. Guerne, Inflammation Res. 2006, 55,
469–475.
21] P. O’Connor, J. S. Wolinsky, C. Confavreux, G. Comi, L. Kappos, T. P.
Olsson, H. Benzerdjeb, P. Truffinet, L. Wang, A. Miller, M. S. Freedman, N.
Engl. J. Med. 2011, 365, 1293–1303.
[
7] a) Z. Zhang, Y.-W. Dong, G.-W. Wang, K. Komatsu, Synlett 2004, 61–64;
b) Z. Zhang, Y.-W. Dong, G.-W. Wang, K. Komatsu, Chem. Lett. 2004, 33,
1
68–169; c) V. Declerck, P. Nun, J. Martinez, F. Lamaty, Angew. Chem. Int.
[
22] K. Arnold, A. S. Batsanov, B. Davies, A. Whiting, Green Chem. 2008, 10,
Ed. 2009, 48, 9318–9321; Angew. Chem. 2009, 121, 9482–9485; d) M. R.
Naimi-Jamal, J. Mokhtari, M. G. Dekamin, G. Kaupp, Eur. J. Org. Chem.
124–134.
[
[
23] J. C. Sheehan, G. P. Hess, J. Am. Chem. Soc. 1955, 77, 1067–1068.
24] G. Chattopadhyay, S. Chakraborty, C. Saha, Synth. Commun. 2008, 38,
2
009, 3567–3572; e) L. Konnert, A. Gauliard, F. Lamaty, J. Martinez, E.
Colacino, ACS Sustainable Chem. Eng. 2013, 1, 1186–1191; f) D. Tan, C.
ˇˇ ´
4068–4075.
ˇ
Mottillo, A. D. Katsenis, V. Strukil, T. Friscic, Angew. Chem. Int. Ed. 2014,
[
[
25] R. Paul, G. W. Anderson, J. Am. Chem. Soc. 1960, 82, 4596–4600.
26] P. S. Chaudhari, S. D. Salim, R. V. Sawant, K. G. Akamanchi, Green Chem.
ˇ
5
3, 9321–9324; Angew. Chem. 2014, 126, 9475–9478; g) D. Tan, V. Stru-
kil, C. Mottillo, T. Fri sˇ cˇ i c´ , Chem. Commun. 2014, 50, 5248–5250.
2010, 12, 1707–1710.
ˇ
´
´
[
[
8] V. Strukil, B. Bartolec, T. Portada, I. “ilovic, I. Halasz, D. Margetic, Chem.
[
[
[
27] R. A. Sheldon, Green Chem. 2007, 9, 1273–1283.
28] K. Van Aken, L. Strekowski, L. Patiny, Beilstein J. Org. Chem. 2006, 2, 3.
29] F. Roschangar, R. A. Sheldon, C. H. Senanayake, Green Chem. 2015, 17,
Commun. 2012, 48, 12100–12102.
9] With the exception of ref. [8] in which nitromethane was used during
the process.
752–768.
[
[
10] C. F. Burmeister, A. Kwade, Chem. Soc. Rev. 2013, 42, 7660–7667.
11] a) R. Thorwirth, F. Bernhardt, A. Stolle, B. Ondruschka, J. Asghari, Chem.
Eur. J. 2010, 16, 13236–13242; b) T. L. Cook, J. A. Walker, J. Mack, Green
Chem. 2013, 15, 617–619; c) D. W. Peters, R. G. Blair, Faraday Discuss.
[30] I. Shiina, Y.-I. Kawakita, Tetrahedron 2004, 60, 4729–4733.
[31] T. Maki, K. Ishihara, H. Yamamoto, Org. Lett. 2006, 8, 1431–1434.
[32] K. Deo, S. Patel, S. Dhol, S. Sanghani, V. Ray, 2011, US20110105795.
2
014, 170, 83–91; d) D. A. Fulmer, W. C. Shearouse, S. T. Medonza, J.
Mack, Green Chem. 2009, 11, 1821–1825.
[
12] T.-X. MØtro, J. Bonnamour, T. Reidon, J. Sarpoulet, J. Martinez, F. Lamaty,
Chem. Commun. 2012, 48, 11781–11783.
Received: April 3, 2015
Published online on July 14, 2015
Chem. Eur. J. 2015, 21, 12787 – 12796
www.chemeurj.org
12796
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim