Mechanochemical synthesis of phthalimides with crystal structures of intermediates and products

Article abstract of DOI:10.1039/c5ce00038f

A series of phthalimides have been successfully synthesized in the solid state by grinding (or kneading) of substituted phthalic anhydride and aniline derivatives. Selected products and intermediates were crystallized and characterized by crystallography leading to a potential rationale for the solid-state reactivity that suggests that co-crystals could serve as intermediates. This journal is

Full text of DOI:10.1039/c5ce00038f

CrystEngComm
View Article Online
View Journal | View Issue
PAPER
Mechanochemical synthesis of phthalimides with
crystal structures of intermediates and products†
Cite this: CrystEngComm, 2015, 17,
a
b
b
2523
*
Melwin Colaço,‡ Jean Dubois‡ and Johan Wouters
Received 7th January 2015,
Accepted 12th February 2015
A series of phthalimides have been successfully synthesized in the solid state by grinding (or kneading) of
substituted phthalic anhydride and aniline derivatives. Selected products and intermediates were crystallized
and characterized by crystallography leading to a potential rationale for the solid-state reactivity that sug-
gests that co-crystals could serve as intermediates.
DOI: 10.1039/c5ce00038f
www.rsc.org/crystengcomm
The reaction between various substituted phthalic anhy-
drides and aniline derivatives (Scheme 1) has been selected
Introduction
Mechanochemistry is now recognized as a clean and efficient
synthesis approach to prepare coordination polymers, metal–
organic frameworks or metallodrugs.1 It plays a major role in
the development of solvent-free chemical synthesis
approaches and has been successfully applied to a variety of
organic transformations including oxidations and reductions,
Wittig condensations,2 fullerene dimerization,3 oxime
for the present study.
The most important synthesis of phthalimides is the
dehydrative condensation of phthalic anhydride at high tem-
peratures with primary amines, when the amine is available.
When the amine is not readily accessible, the direct
N-alkylation of phthalimides with alcohols under Mitsunobu
conditions11 and of potassium phthalimide with alkyl halides
(Gabriel Synthesis)12 are popular alternative approaches.
Among other applications, phthalimides are suitable protec-
tive groups in peptide synthesis.
The present work shows that mechanochemistry is an
interesting alternative for the synthesis of these compounds.
Our study further supports the use of this approach for the
solid state reaction of amines and carbonyl compounds.13
functionalization,4
and
coupling
reactions.5
Mechanochemisty, either by neat grinding or liquid-
assisted grinding (LAG), also plays a major role in prepara-
tion of pharmaceutical co-crystals.6 Real-time and in situ
monitoring of mechanochemical reactions is now possible
and shows the highly dynamic process that leads to
reaction rates comparable or greater than those observed
in solution.7
Molecular-level
mechanisms
underlying
grinding
Experimental
Materials and methods
mechanosynthesis still need to be studied. Concerning LAG
(also known as solvent-drop grinding or kneading, as
coined by Braga et al.8) no detailed mechanism underlying
the process is known but molecular diffusion enhanced by
the added liquid has been proposed to explain the effi-
ciency of the reactions.9 Co-crystallization has also been
proposed to play a major role during the solid-state process
and led Zaworotko to develop the concept of co-crystal con-
trolled solid state synthesis C3S3.10 Formation of co-crystals
in the same lattice brings functional groups close enough
to react.
Reagents were purchased from Sigma-Aldrich, TCI and Santa
Cruz BioTechnologies. Solvents were purchased from Acros
a Department of Chemistry, St Joseph's College, P O Box 27094, Bangalore 560
027, India
b Unité de Chimie Physique, Théorique et Structurale, University of Namur,
Namur, Belgium. E-mail: johan.wouters@unamur.be; Fax: +32 081 725440;
Tel: +32 081 724550
† Electronic supplementary information (ESI) available: Powder X-ray diffraction
data, extra crystal structures, NMR data. CCDC 1021624–1021632. See DOI:
10.1039/c5ce00038f
Scheme
1 Formation of phthalimides (3) from corresponding
‡ Both authors contributed equally.
substituted phthalic anhydrides (1) and aniline derivatives (2).
This journal is © The Royal Society of Chemistry 2015
CrystEngComm, 2015, 17, 2523–2528 | 2523

View Article Online
Paper
CrystEngComm
Organic (Geel, Belgium) and used with no further
purification.
oxide jar. The total mass of all reactants did not exceed 1 g.
The mixture is ground using a vibration ball mill Retsch
MM400 for 90 minutes at 30 Hz. The melting point or phase
transition temperature of the ground mixture was initially
determined using a Koffler bench apparatus. The bulk was
then heated to 10 °C below this temperature in a heating
metal block thermostat (MBT 250, Kleinfeld). The tempera-
ture was maintained for 60 to 90 minutes to give imide 3.
General procedure
2.5 mmol of aniline 2, 1 equivalent of corresponding anhydride 1
and 2–10 balls are added to a 2 mL plastic vial or a zirconium
§ Crystal Data for PA-DMA (3ai). Prism colourless crystals, orthorhombic, Pbc21,
a = 9.837(2), b = 16.124(4), c = 8.643(3) Å, α = 90.00, β = 90.00, γ = 90.00°, V =
1370.9(6) Å3, Z = 4, ρcalc = 1.217 g cm−3, F000 = 528, λ Mo Kα = 0.71073 Å, θmax
25.0°, 4759 total measured reflections, 1994 independent reflections (Rint
=
=
Single crystals
Single crystal X-ray diffraction was performed on a Gemini
Ultra R system (4-circle kappa platform, Ruby CCD detector)
using Mo (λ = 0.71073 Å) or Cu Kλ radiation (λ = 1.54056 Å).
Selected crystals were mounted on the tip of a quartz pin
using cyanoacrylate (commercial glue). Cell parameters were
estimated from a pre-experiment run and full data sets col-
lected at room temperature. Structures were solved by direct
methods with sir92 (v.3.0)14 program and then refined on F2
using SHELXL-97 software.15 Non-hydrogen atoms were
anisotropically refined and the hydrogen atoms (not impli-
cated in H-bonds) in the riding mode with isotropic tempera-
ture factors fixed at 1.2 times UIJeq) of the parent atoms (1.5
times for methyl groups). Hydrogen atoms implicated in
H-bonds were localized by Fourier difference maps (ΔF).
CCDC 1021624–1021632 entries contain the supplementary
crystallographic data.§
0.031), 1322 observed reflections (I > 2σIJI)), μ = 0.081 mm−1, 136 parameters, R1
(observed data) = 0.0880, S = GooF = 1.09, Δ/s.u. = 0.00, residual ρmax = 0.27 e Å−3
ρmin = −0.26 e Å−3. Crystal Data for BTA-DMA 1st form (3bi). Plates colourless
,
¯
crystals, triclinic, P1, a = 8.3591(3), b = 10.4196IJ7), c = 19.0455(9) Å, α =
88.407(5), β = 88.481(3), γ = 76.340IJ4)°, V = 1610.95IJ15) Å3, Z = 1, ρcalc = 1.311 g cm−3
,
F000 = 665, λ Mo Kα = 0.71073 Å, θmax = 25.0°, 11980 total measured reflections,
5663 independent reflections (Rint = 0.024), 3983 observed reflections (I > 2σIJI)),
μ = 0.093 mm−1, 438 parameters, R1 (observed data) = 0.0468, S = GooF =
0.99, Δ/s.u. = 0.02, residual ρmax = 0.16 e Å−3, ρmin = −0.17 e Å−3. Crystal Data for
¯
BTA-DMA 2nd form (3bi). Plates colourless crystals, triclinic, P1, a = 8.4040(5),
b = 8.4040(5), c = 15.240(1) Å, α = 97.881(6), β = 93.354(5), γ = 96.810IJ6)°,
V = 1468.73IJ18) Å3, Z = 4, ρcalc = 1.335 g cm−3, F000 = 616, λ Mo Kα = 0.71073 Å,
θmax = 23.3°, 10 179 total measured reflections, 4188 independent reflections
(Rint = 0.078), 2204 observed reflections (I > 2σIJI)), μ = 0.096 mm−1, 400 parameters,
R1 (observed data) = 0.0705, S = GooF = 1.01, Δ/s.u. = 0.03, residual ρmax = 0.19 e
Å−3, ρmin = −0.21 e Å−3. Crystal Data for 3NPA-DBA open form (3ck). Needle
¯
colourless crystals, triclinic, P1, a = 9.188(5), b = 13.284(7), c = 14.477(8) Å,
α = 68.21(5), β = 81.24(4), γ = 75.69IJ4)°, V = 1586.2(16) Å3, Z = 2, ρcalc = 1.857 g cm−3
,
F000 = 862, λ Mo Kα = 0.71073 Å, θmax = 25.0°, 11 937 total measured reflections,
5562 independent reflections (Rint = 0.052), 3222 observed reflections (I > 2 σIJI)),
μ = 5.136 mm−1, 415 parameters, R1 (observed data) = 0.0655, S = GooF = 1.04,
Δ/s.u. = 0.00, residual ρmax = 0.88 e Å−3, ρmin = −0.59 e Å−3. Crystal Data for 3NPA-
pAni (3cn). Needle yellow crystals, Monoclinic, P21, a = 3.8289(1), b = 23.3314IJ7),
c = 7.4554(2) Å, α = 90, β = 98.153(3), γ = 90°, V = 659.3(1) Å3, Z = 2, ρcalc = 1.502
g cm−3, F000 = 308, λ Cu Kα = 1.54184 Å, θmax = 66.6°, 5907 total measured reflec-
tions, 2304 independent reflections (Rint = 0.0182), 2285 observed reflections
(I > 2σIJI)), μ = 0.977 mm−1, 200 parameters, R1 (observed data) = 0.0270, S = GooF =
1.08, Δ/s.u. = 0.00, residual ρmax = 0.16 e Å−3, ρmin = −0.14 e Å−3. Crystal Data for
4NPA-DBA (3dk). Plates colourless crystals, monoclinic, I2/a, a = 14.8231IJ9),
b = 7.7177(6), c = 25.714(2) Å, α = 90, β = 97.692(7), γ = 90°, V = 2915.2(4) Å3,
Z = 8, ρcalc = 1.941 g cm−3, F000 = 1648, λ Mo Kα = 0.71073 Å, θmax = 25.0°, 6774
total measured reflections, 2563 independent reflections (Rint = 0.079), 1465
observed reflections (I > 2σIJI)), μ = 5.580 mm−1, 392 parameters, R1 (observed
X-Ray powder diffraction (XRPD)
Powder X-Ray Diffraction (PXRD) data were collected on a
PANalytical reflexion-geometry diffractometer, using Ni-
filtered Cu Kα radiation (λ = 1.54179 Å) at 40 kV and 40 mA
with an X'Celerator detector. Each sample was analyzed
between 4 and 50° 2θ with a step size of ca. 0.0167° 2θ and a
total scan time of 3 min 48 s.
Nuclear magnetic resonance (NMR)
All NMR (1H) were recorded on JEOL spectrometer (JNM EX-
400) at 25 °C. Chemical shifts are reported in parts per
million (ppm) using the solvent residual peak as reference
(CDCl3: 7.26 ppm; DMSO-d6: 2.50 ppm; CD3OD: 3.31 ppm).
Coupling constants (J) are reported in Hertz (Hz). The reso-
nance multiplicity is described as s for singlet, bs for broad
singlet, d for doublet, t for triplet, q for quadruplet and m for
multiplet. Data are provided as ESI.†
data) = 0.0491, S = GooF = 0.96, Δ/s.u. = 0.00, residual ρmax = 0.36 e Å−3, ρmin
=
−0.40 e Å−3. Crystal Data for 4NPA-DBA open form (3dk). Plates colourless crys-
tals, monoclinic, P21/n, a = 14.0897IJ3), b = 4.8183(1), c = 22.6757(3) Å, α = 90,
β = 90.555(9), γ = 90°, V = 1539.3(1) Å3, Z = 4, ρcalc = 1.916 g cm−3, F000 = 864,
λ Cu Kα = 1.54184 Å, θmax = 66.6°, 8457 total measured reflections, 2710 inde-
pendent reflections (Rint = 0.0344), 2466 observed reflections (I > 2σIJI)), μ =
6.946 mm−1, 216 parameters, R1 (observed data) = 0.0465, S = GooF = 1.07, Δ/s.
u. = 0.00, residual ρmax = 0.53 e Å−3, ρmin = −0.68 e Å−3. Crystal Data for 4NPA-
pAni (3dn). Triangular yellow crystals, monoclinic, Pn, a = 6.9750IJ12), b =
3.876(1), c = 23.998(3) Å, α = 90, β = 91.227IJ12), γ = 90°, V = 648.6(2) Å3, Z = 2,
ρcalc = 1.527 g cm−3, F000 = 308, λ Mo Kα = 0.71073 Å, θmax = 23.3°, 1900 total
Results and discussion
measured reflections, 1462 independent reflections (Rint
observed reflections (I > 2σIJI)), μ = 0.117 mm−1, 200 parameters, R1 (observed
data) = 0.0724, S = GooF = 1.14, Δ/s.u. = 0.00, residual ρmax = 0.32 e Å−3, ρmin
= 0.045), 1224
Solid mixtures of 1 and 2 (Scheme 1) were ground using a
Retsch MM400 ball mill operating at 30 Hz for 30 min. For
2,6-dimethyl-aniline (2i, Scheme 1), which is a liquid at room
temperature, the resulting wet paste was ground following
the general procedure. The ground products obtained were
further heated in a metal block thermostat (MBT 250,
Kleinfeld). In each case, the outcome was characterized by
calorimetry (melting point), NMR, powder X-ray diffraction
=
−0.34 e Å−3. Crystal Data for 4NPA-pTol co-crystal (3dm). Triangular yellow crys-
tals, Orthorhombic, Fdd2, a = 59.519(2), b = 12.942(5), c = 7.623(2) Å, α = 90,
β = 90, γ = 90°, V = 5872(3) Å3, Z = 16, ρcalc = 1.440 g cm−3, F000 = 2656, λ Mo
Kα = 0.71073 Å, θmax = 25.0°, 20 263 total measured reflections, 2560 indepen-
dent reflections (Rint
= 0.052), 2335 observed reflections (I > 2σIJI)), μ =
0.113 mm−1, 217 parameters, R1 (observed data) = 0.0433, S = GooF = 1.01, Δ/s.
u. = 0.04, residual ρmax = 0.26 e Å−3, ρmin = −0.15 e Å−3
.
2524 | CrystEngComm, 2015, 17, 2523–2528
This journal is © The Royal Society of Chemistry 2015

View Article Online
CrystEngComm
Paper
(PXRD) and, when possible, by single X-ray crystallography
(SCXRD).
Grinding of 4-carboxylic acid phthalic anhydride (1b) with
aniline derivatives usually resulted in changes in colour of the
solids. In most cases, grinding led to amorphous solids which,
upon heating, transformed to crystalline material (Table 1). This
is well illustrated by the synthesis of 4-carboxylic-N-phthaloyl-
2,6-dimethylaniline (3bi, m.p. 222 °C), obtained by grinding of
1b and 2i and heating to 200 °C (Fig. 2). The amorphous solid
obtained by grinding (kneading) of 1b and 2i transforms to a
crystalline white powder (Fig. S2a,† 2θ = 5, 12, 13, 14, and 17°).
Single crystals of the resulting product could be obtained,
either from a saturated solution in toluene/MeOH or from a
solution in EtOAc.
The crystal structure from EtOAc confirms formation of
the 3bi phthalimide product (Fig. 2). The second structure
corresponds to a solvate (toluene) (Fig. S2†). The X-ray dif-
fraction pattern simulated from the single crystal structure
obtained in EtOAc confirmed that the phthalimide product
was obtained in the ground solid powder after heating (2θ =
6, 16, 17, 19, and 25°, Fig. 2S†).
Highlights of the results are summarized in Table 1.
Grinding of phthalic anhydride (1a) with aniline deriva-
tives led to either an amorphous solid (2i) or a physical mix-
ture (2k and 2n) of the starting reactants. Heating of the
resulting solids allowed formation of the corresponding
phthalimides (3ai, 3ak and 3an) as white crystalline powders.
In the case of N-phthaloyl-2,6-dimethylaniline (3ai, m.p.
204 °C), X-ray diffraction data measured on both powders
and single crystals (Fig. 1 and S1†) confirm unambiguously
the formation of the product. NMR data indicate 100% con-
version of the reactants.
Interestingly, the powder diffractogram of the ground prod-
uct (2θ = 7, 10, and 13.5°) is not only different from the starting
phthalic anhydride (1a, 2θ = 13, 17, and 23°) ij2,6-dimethyl-
aniline (2i) is a liquid] (Fig. S1†) but also from the final
N-phthaloyl-2,6-dimethylaniline (3ai, 2θ = 11, 15, 19 and 22°)
product obtained after heating (Fig. 1 and S1†). This suggests
that the unheated solid corresponds to an intermediate state
that needs to be heated to form the final phthalimide product.
Similarly, crystals of 4-carboxylic-N-phthaloyl-p-anisole
(3bn, m.p. >260 °C) could be grown from the solid resulting
Table 1 Outcome of grinding (and heating) of phthalic anhydrides (1a–d) with different aniline derivatives (2i–p)
R1 = H (1a)
R2 =
R1 = 4-COOH (1b)
2,6-DiMe (2i)
White
220
Amorphous
200
White
222
μCrystalline
2,6-DiMe (2i)
White
226
μCrystalline
200
White
2,6-DiBr (2k)
White
160
μCrystalline
150
White
4-OMe (2n)
White
158
Mixture
150
Brown
160
2,6-DiBr (2k)
Yellow
164
μCrystalline
140
White
4-OMe (2n)
White
>260
Amorphous
200
Yellow
Colour
m.p.a (°C)
Solid
Heated to (°C)
Colour
m.p. (°C)
Solid
204
μCrystalline
160
μCrystalline
160
μCrystalline
>260
μCrystalline
μCrystalline
R1 = 4-COOH
R2 =
Colour
2,6-DiCl (2j)
Yellow
2,4-DiCl (2l)
White
4-Me (2m)
White
3-NO2 (2o)
Yellow
2-COOH (2l)
Yellow
m.p. (°C)
Solid
Heated to (°C)
Colour
202
μCrystalline
180
222
Amorphous
180
>260
Amorphous
200
>260
Amorphous
200
230
Amorphous
180
White
White
Yellow
Grey
White
m.p. (°C)
Solid
230
μCrystalline
>260
μCrystalline
>260
μCrystalline
>260
μCrystalline
230
μCrystalline
R1 = 3-NO2 (1c)
R2 =
Colour
m.p. (°C)
Solid
Heated to (°C)
Colour
2,6-DiMe (2i)
White
206
μCrystalline
180
White
2,6-DiCl (2j)
Yellow
178
μCrystalline
150
White
2,6-DiBr (2k)
White
168
Mixture
150
White
4-OMe (2n)
Yellow
200
Amorphous
180
Yellow
4-Me (2m)
Yellow
170
Amorphous
150
Yellow
160
μCrystalline
3-NO2 (2o)
White
224
Amorphous
120
White
m.p. (°C)
Solid
206
μCrystalline
176
μCrystalline
190
μCrystalline
202
μCrystalline
224
μCrystalline
R1 = 4-NO2 (1d)
R2 =
Colour
m.p. (°C)
Solid
Heated to (°C)
Colour
2,6-DiMe (2i)
Brown
168
Amorphous
120
Brown
2,6-DiCl (2j)
Yellow
168
Mixture
160
Brown
168
μCrystalline
2,6-DiBr (2k)
Orange
178
Mixture
150
White
178
μCrystalline
4-OMe (2n)
Yellow
212
Amorphous
180
Yellow
4-Me (2m)
Yellow
178
Amorphous
150
Brown
160
μCrystalline
3-NO2 (2o)
Yellow
254
Amorphous
120
Brown
m.p. (°C)
Solid
168
μCrystalline
212
μCrystalline
254
μCrystalline
a
Tg for amorphous solids.
This journal is © The Royal Society of Chemistry 2015
CrystEngComm, 2015, 17, 2523–2528 | 2525

View Article Online
Paper
CrystEngComm
Evidence for formation of 3- and 4-nitro-N-phthaloyl-p-
anisole (3cn and 3dn) starting from 4-methoxy aniline (2n)
was obtained by the determination of the corresponding crys-
tal structures (Fig. 3).
In the case of the reaction of 4-nitrophthalic anhydride
with 2,6-dibromoaniline (2k), the corresponding 4-nitro-N-
phthaloyl-2,6-dibromoaniline (3dk) product is formed as
evidenced by determination of its crystal structure (Fig. 4a).
Very interestingly, we were also able to trap an open interme-
diate of the reaction (Fig. 4b), i.e. the open amide product of
3dk. The same open product was observed after reaction of
3-nitrophthalic anhydrides (1c) and 2,6-dibromoaniline (2k)
(Fig. S4†). In all cases the single crystals are obtained from
the heated sample.
NMR confirmed that there was more of the open form of
the product along with the expected nitro-N-phthaloyl-2,6-
dibromoanilines (3ck or 3dk) product. This is also consistent
with the fact that corresponding single crystals were obtained
as a mixture with amorphous powder.
Whether the open products result from partial hydrolysis
of the nitro-N-phthaloyl-2,6-dibromoanilines or correspond to
the first step of the nucleophilic attack of the aniline nitro-
gen on the nitrophthalic anhydrides is still to be studied.
However, the second hypothesis is favoured in the present
case.
Fig.
1 X-ray diffraction analysis of the solid corresponding to
N-phthaloyl-2,6-dimethylaniline (3ai), obtained after grinding of 1a and
2i and heating to 200 °C. a) Powder diffractograms recorded before
and after heating of the ground sample and b) single crystal structure.
Fig.
2 Single crystal structure of 4-carboxylic-N-phthaloyl-2,6-
dimethylaniline (3bi) obtained by grinding of 1b and 2i and heating to
200 °C. Recrystallization from EtOAc.
from solid state reaction between 1b and 4-methoxyaniline
(2n). This crystal structure experimentally confirmed the for-
mation of the phthalimide product (Fig. 3S†).
Mechanochemical reactions of 3- and 4-nitrophthalic
anhydrides (1c and 1d) with aniline derivatives have also been
studied. Like in the case of phthalic anhydride (1a) and 4-car-
boxylic acid phthalic anhydride (1b), grinding followed by
heating led generally to the corresponding nitro-N-phthaloyl
products (Table 1).
Fig. 3 Solid-state reaction of 3- or 4-nitrophthalic anhydride (1c or 1d)
with p-anisole (2n) leads to 3- and 4-nitro-N-phthaloyl-p-anisole
(a, b).
2526 | CrystEngComm, 2015, 17, 2523–2528
This journal is © The Royal Society of Chemistry 2015

View Article Online
CrystEngComm
Paper
Fig. 4 Solid-state reaction of 4-nitrophthalic anhydride (1d) and 2,6-
dibromoaniline (2k) leads to 4-nitro-N-phthaloyl-2,6-dibromoaniline
(3dk) (a) but also to the open intermediate (b).
Indeed, together with the open intermediates described in
Fig. 4b and S4,† a new crystal form was trapped in the case of
the solid state reaction between 4-nitrophthalic anhydride
(1d) and 4-methylaniline (2m).
Fig.
6
Possible initial steps in the mechanism of formation of
phthalimides leading to the open amide monosubstituted adduct
derived from the crystal structure of the 1 : 1 1d·2m co-crystal.
Nucleophilic attack by the deprotonated nitrogen atom on one of the
carboxylic acids (two possible attacks are highlighted Nu(a) and Nu(b))
leads to two possible open intermediates (pathway (a) or (b)). Atoms
involved in the mechanism presented as spheres, HC atoms omitted for
clarity.
In this case, the intermediate corresponded to the co-
crystal of 4-methylaniline (2m) and 4-nitrophthalic acid, the
hydrolysed 4-nitrophthalic anhydride (1d) (Fig. 5). Evidence
for a co-crystal instead of a salt comes from the geometry of
the two carboxylic acids showing clear alterning single C–OH
and double CO bonds for both COOH groups. In the
co-crystal, 4-methylaniline (2m) and 4-nitrophthalic acid form
a 1 : 1 1d·2m complex stabilized by π–π stacking. Stability of
the complex is reinforced by intermolecular H-bonds (Fig. S5,
Table S1†).
This structure suggests that a co-crystal is a viable inter-
mediate during the solid-state reaction. In addition, the crys-
tal packing supports a mechanism in which molecules are
positioned in such a way that the reactive nucleophilic nitro-
gen of the aniline derivative is optimally placed to react with
the electrophilic carbon atom of one of the carboxylic acids
(Nu(a) or Nu(b), Fig. 6). Based on the isomer observed in the
open intermediate (see Fig. 4b in the case of 3dk), it is rea-
sonable to favour pathway (b) in Fig. 6.
Such a nucleophilic attack of the aniline should also be
energetically favoured over the reaction with an acid involv-
ing a molecule parallely stacked to the aryl-amine, as this
would lead to less rearrangement of the solid.
The presence of this co-crystal as a general intermediate
in the different reactions that were carried out is suggested
by the PXRD data. In most cases, the ground powder
obtained before heating (see for example Fig. 1a) is character-
ized by small angle diffraction peaks, corresponding to enti-
ties (probably co-crystals) crystallizing in a larger cell.
Fig. 5 Two views (a, b) of the π–π stacking 1 : 1 co-crystal observed
during the reaction of 4-nitrophthalic anhydride (1d) and 4-methylaniline
(2m).
This journal is © The Royal Society of Chemistry 2015
CrystEngComm, 2015, 17, 2523–2528 | 2527

View Article Online
Paper
CrystEngComm
Acknowledgements
Authors acknowledge Ir B. Norberg for assistance in data col-
lection and Prof S. Lanners for fruitful discussions and assis-
tance for analysis of NMR data. Financial support of the FNRS
(grant no. 2.4511.07) is acknowledged.
Notes and references
1 (a) S. L. James, et al., Chem. Soc. Rev., 2012, 41, 413–447; (b)
S. L. James and T. Friscic, Chem. Soc. Rev., 2013, 42,
7487–7740; (c) D. Braga and F. Grepioni, Angew. Chem., Int.
Ed., 2004, 43, 4002–4011; (d) T. Friscic, I. Halasz, V. Strukil,
M. Eckert-Maksic and R. Dinnebier, Croat. Chem. Acta,
2012, 85, 367–378; (e) D. Braga and F. Grepioni, Angew.
Chem., Int. Ed., 2004, 43, 4002–4011.
2 V. P. Balema, J. W. Wiench, M. Pruski and V. K. Percharsky,
J. Am. Chem. Soc., 2002, 124, 6244–6245.
3 G.-W. Wang, K. Komatsu, Y. Murata and M. Shiro, Nature,
1997, 387, 583–586.
Scheme 2 Possible steps in the pathway leading to formation of
phthalimides (3) from corresponding substituted phthalic anhydrides (1)
and aniline derivatives (2), involving the intermediates trapped in this
study.
4 C. B. Aakeroy and P. D. Chopade, Org. Lett., 2011, 13, 1–3.
5 G. Kaupp, CrystEngComm, 2009, 11, 388–403.
6 D. Braga, L. Maini and F. Gepioni, Chem. Soc. Rev., 2013, 42,
7638–7648; J. Wouters and L. Quere, Pharmaceutical Salts
and Cocrystals, RSC Drug Discovery, 2011; T. Friscic,
S. L. Childs, S. A. A. Rizvi and W. Jones, CrystEngComm,
2009, 11, 418–426.
Conclusions
In the course of our study of the solid state reactions of a
series of anilines and phthalic anhydrides, we have shown
that grinding of the reactants followed by heating of the
resulting solids led to the corresponding phthalimides, in
agreement with previous data from the literature.10,13 Evi-
dence for the formation of these products is provided by the
physico-chemical characteristics (m.p., NMR, powder X-ray
patterns) of the final samples and by a series of crystal struc-
tures. Interestingly, intermediates were also trapped in the
solid state. They do pave the way to formation of the final,
fully cyclized, phthalimide and suggest a potential mecha-
nism that could involve a co-crystal intermediate and the
open amide adduct (Scheme 2). Involvement of co-crystals
during the solid state synthetic process is in agreement with
evidences provided by Zaworotko et al.10
Influence of the nature and position of substituents (in
particular electronic effects) on the reactivity further needs to
be systematically studied. Implication of the hydrolysed,
instead of the closed, phthalic anhydride in the mechanism
is another open question that needs to be investigated. Devel-
opment of recent in situ methods now opens new opportuni-
ties to directly challenge the mechanism proposed in the
present work.
7 T. Friscic, I. Halasz, P. J. Beldon, A. M. Belenguer, F. Adams,
S. A. J. Kimber, V. Honkima and R. E. Dinnebier, Nat. Chem.,
2013, 5, 66–73.
8 D. Braga and F. Grepioni, Angew. Chem., Int. Ed., 2004, 43,
4002–4011.
9 G. A. Bowmaker, J. V. Hanna, R. D. Hart, B. W. Skelton and
A. H. White, Dalton Trans., 2008, 5290–5292.
10 M. L. Cheney, G. J. McManus, J. A. Perman, Z. Wang and
M. J. Zaworotko, Cryst. Growth Des., 2007, 7, 616–617;
M. L. Cheney, M. J. Zaworotko, S. Beaton and R. D. Singer,
J. Chem. Educ., 2008, 85, 1649–1651.
11 K. C. K. Swamy, et al., Chem. Rev., 2009, 109, 2551–2651;
S. E. Sen and S. L. Roach, Synthesis, 1995, 1995, 756–758.
12 S. Gabriel, Ber. Dtsch. Chem. Ges., 1887, 20, 2224–2236;
M. S. Gibson and R. W. Bradshaw, Angew. Chem., Int. Ed.
Engl., 1968, 7, 919–930.
13 G. Kaupp, J. Schmeyers and J. Boy, Tetrahedron, 2000, 56,
6899–6911.
14 A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi,
M. C. Burla, G. Polidori and M. Camalli, J. Appl. Crystallogr.,
1994, 27, 435.
15 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.,
2008, 64, 112–122.
2528 | CrystEngComm, 2015, 17, 2523–2528
This journal is © The Royal Society of Chemistry 2015