Propensity of formation of zipper architecture vs. Lincoln log arrangement in solvated molecular complexes of melamine with hydroxybenzoic acids

Article abstract of DOI:10.1039/c1ce05179b

Molecular complexes of melamine with hydroxy and dihydroxybenzoic acids have been analyzed to assess the collective role of the hydroxyl (OH) and carboxyl (COOH) functionalities in the recognition process. In most cases, solvents of crystallization do pla

Full text of DOI:10.1039/c1ce05179b

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
CrystEngComm
Cite this: CrystEngComm, 2011, 13, 4886
www.rsc.org/crystengcomm
PAPER
Propensity of formation of zipper architecture vs. Lincoln log arrangement in
solvated molecular complexes of melamine with hydroxybenzoic acids†
*
Sunil SeethaLekshmi and T. N. Guru Row
Received 7th February 2011, Accepted 4th May 2011
DOI: 10.1039/c1ce05179b
Molecular complexes of melamine with hydroxy and dihydroxybenzoic acids have been analyzed to
assess the collective role of the hydroxyl (OH) and carboxyl (COOH) functionalities in the recognition
process. In most cases, solvents of crystallization do play a major role in self-assembly and structure
stabilization. Hydrated compounds generate linear chains of melamine molecules with acid molecules
pendant resulting in a zipper architecture. However, anhydrous and solvated compounds generate
tetrameric units consisting of melamine dimers together with acid molecules. These tetramers in turn
interweave to form a Lincoln log arrangement in the crystal. The salt/co-crystal formation in these
complexes cannot be predicted apriori on the basis of DpKa values as there exists a salt-to-co-crystal
continuum.
acid.5 In recent times, ML formed the central stage, when several
Introduction
companies in China were implicated in a scandal involving infant
feed formulation which was adulterated with melamine, resulting
in renal failure among children.6 Lu and co-workers developed
sensors based on the recognition patterns existing between
melamine and cyanuric acid to establish the presence of mela-
mine in milk products.7 This clearly emphasizes the importance
of an understanding of the recognition process involving mela-
mine. With carboxylic acids, ML generally forms hydrated salts
as revealed from a CSD analysis.8 It was noted that in the
molecular complexes of ML with gallic acid, tartaric acid and
citric acid, carboxyl and hydroxyl groups participate in extensive
hydrogen bonding.9 However, the collective influence of
hydroxyl and carboxyl groups in the recognition with ML have
not been systematically evaluated. In this context, we have
prepared and analyzed molecular complexes of ML with
hydroxybenzoic acids and isomers of dihydroxybenzoic acids
(Scheme 2). We report and discuss these molecular complexes in
terms of synthon formation, structural variations due to hydra-
tion or solvation and salt-to-co-crystal continuum.
Understanding molecular co-crystals and salts in terms of the
spatial arrangement of intermolecular interactions is of utmost
importance in supramolecular synthesis.1 This knowledge is
usually transformed to the design of robust synthons with much
generality and predictability. Such an approach is usually found
in the development of molecular complexes, for example novel
pharmaceutical formulations with better stability, solubility and
bioavailability.2 Some of the best studied interactions in crystal
engineering are the homomeric and heteromeric synthons of
carboxylic acids and carboxamides.3 Studies pertaining to the
acid-triazine heterosynthons are limited compared to the copious
acid-pyridine heterosynthons.3e This has prompted us to study
the interactions present in a series of complexes with acids and
a triazine.
The triazine compound melamine (2,4,6-triamino-1,3,5-
triazine, ML) is interesting in a crystal engineering perspective
due to its symmetry and the availability of several hydrogen bond
donor and acceptor functionalities. In the solid state, ML forms
complementary arrays of N–H/N hydrogen bonds resulting in
a 2D network (Scheme 1A). It has been demonstrated that
depending on the extent of protonation of the triazine ring, the
dimensionality of the hydrogen bonded network decreases from
2D to 1D and eventually to 0D (Scheme 1B and C).4
Experimental
All compounds were purchased from Sigma-Aldrich and were
used without further purification. Co-crystallization experiments
were carried out by dissolving equimolar compounds from a 1 : 1
CH3CN–H2O solution followed by slow evaporation at room
temperature. In 6, the co-crystallization experiments were carried
out from a 1 : 1 CH3CN–DMF (N,N-dimethylformamide)
solution. Single crystals suitable for X-ray diffraction studies
were obtained over a period of one week. Co-crystallization of
3-hydroxybenzoic acid with ML did not yield crystals suitable for
One of the most intriguing and well-known assemblies
involving ML is the rosette architecture formed with cyanuric
Solid State & Structural Chemistry unit, Indian Institute of Science,
Bangalore, 560 012, India. E-mail: ssctng@sscu.iisc.ernet.in; Fax: +91
80 2360 1310; Tel: +91 80 2293 2796
† Electronic supplementary information (ESI) available. CCDC
reference numbers 782608–782613. For ESI and crystallographic data
in CIF or other electronic format see DOI: 10.1039/c1ce05179b
4886 | CrystEngComm, 2011, 13, 4886–4894
This journal is ª The Royal Society of Chemistry 2011

View Online
Scheme 1 Hydrogen bond patterns exhibited by the ML molecule (A) 2D, (B) 1D and (C) 0D depending on the extent of protonation.
single crystal X-ray diffraction studies, even after repeated
attempts from several solvents.
and O atoms were positioned geometrically and refined using
a riding model. Hydrogen atoms with water of crystallization
were located from a difference Fourier synthesis and were refined
isotropically. In 6, the water molecules and the disordered DMF
molecules were refined isotropically. The ORTEP diagrams of
the complexes are provided in the ESI.† For 2–7, chꢀaracteristic
Crystallography
Single crystals of the complexes 2–7 were chosen using an
Olympus microscope supported by a rotatable polarizing stage.
Single crystal data for the complexes 2–7 were collected on an
Oxford single crystal X-ray diffractometer (Microsource: Mova;
Detector: Eos) with a four-circle k goniometer employing
ꢀ
data of hydrogen bonds (bond lengths in A, angles in ) are given
in Table 2.
X-ray powder diffraction (PXRD) data were collected on
a Philips X’pert Pro X-ray powder diffractometer (Cu–Ka
radiation) equipped with an X’cellerator detector. The scan
range, step size, and time per step were 2q ¼ 5.00–40ꢀ, 0.02ꢀ, and
25 s, respectively. The combined PXRD plot of the complexes is
given in the ESI.†
ꢀ
a graphite-monochromatized Mo–Ka (lMo-Ka ¼ 0.71073 A)
radiation. The measured intensities were corrected for Lorentz
and polarization effects. The data were reduced using CrysA-
lisRED (special programs available with the diffractometer) and
an analytical absorption correction (after Clark and Reid) was
applied.10 Structure solution and refinements were performed by
the SHELX97 using the WinGX suite.11 Table 1 lists all the
relevant crystallographic information. The non-hydrogen atoms
were refined anisotropically and hydrogen atoms bonded to C, N
Thermal analysis
Thermogravimetric analyses were carried out using a Mettler
Toledo TG/DSC-1 thermogravimetric analyzer and the
Scheme 2 Molecular structures of the hydroxyl acids studied with melamine.
This journal is ª The Royal Society of Chemistry 2011
CrystEngComm, 2011, 13, 4886–4894 | 4887

View Online
Table 1 Crystallographic information of 2–7
2
3
4
5
6
7
Formula
2(C3H7N6),2(C7H5O3), (C3H7N6),
3(C3H7N6),
(C3H7N6),
(C7H5O3), 2(H2O) 3(C7H5O4), 11(H2O) (C7H5O4), 2(H2O) 2(C3H7NO), 2(H2O)
2(C3H6N6), 2(C7H6O4), (C3H7N6),
(C2H3N)
782612
569.57
(C7H5O4)
782611
280.26
Monoclinic
P21/c
14.372(1)
8.549(1)
20.671(2)
90
114.23(1)
90
2316(4)
8
1.607
CCDC no.
Formula wt
782613
300.29
Triclinic
782608
1016.77
Triclinic
782609
316.29
Monoclinic
P21/c
782610
730.64
Monoclinic
C2/c
Crystal system Monoclinic
Space group
ꢁ
ꢁ
P21/c
P1
P1
ꢀ
a (A)
16.691(8)
8.577(4)
20.692(8)
90
117.83(3)
90
2620(2)
4
1.444
298(2)
0.110
50.00
7.571(1)
10.857(1)
16.477(1)
87.12(5)
85.51(5)
82.63(5)
1337.96(15)
4
1.491
298(2)
0.121
50.46
10.811(1)
12.298(1)
19.580(1)
82.23(4)
79.12(5)
64.53(5)
2303.7(3)
2
1.466
150(2)
0.127
50.48
15.788(1)
12.160(1)
7.161
90
96.56(5)
90
38.011(5)
9.212(5)
20.430(5)
90
99.44(5)
90
ꢀ
b (A)
ꢀ
c (A)
a (ꢀ)
b (ꢀ)
g (ꢀ)
V (A )
Z
3
ꢀ
1365.78(14)
4
1.538
7057(4)
8
1.375
Dcalc (g cmꢁ3
T (K)
)
298(2)
0.128
50.20
298(2)
0.111
50.48
150(2)
0.128
50.48
m (mmꢁ1
)
2q range
(deg)
total reflns
unique reflns
reflns used
no. of
5980
4203
1247
371
25272
4818
2815
379
43060
8323
4456
640
13382
2437
1363
223
34754
6376
3172
469
21344
4184
1676
361
parameters
GOF on F2
Final R1,
wR2
0.689
0.0712, 0.1620
0.944
0.0459, 0.1032
1.112
0.0797, 0.2058
0.900
0.0659, 0.1608
0.972
0.0784, 0.2389
0.771
0.0464, 0.0838
experiments were carried with a heating rate of 5 ꢀC minꢁ1 from
35–350 ꢀC, under a nitrogen atmosphere. The TG plots are given
in the ESI.†
role. Acetonitrile molecules interact with these tetramers through
N–H/N hydrogen bonds. Two adjacent tetramer units form N–
H/Oꢁ interactions and they interweave to form a Lincoln log
arrangement in three-dimensions.
Results and discussion
4-Hydroxybenzoic acid–ML complex, 3
Benzoic acid–ML complex, 1
4-Hydroxybenzoic acid (4HBA) with ML crystallizes as
ꢁ
a hydrated salt (Triclinic, P1) and the asymmetric unit consists of
Benzoic acid (BA) forms salts with ML, incorporating two
molecules of water. Two polymorphic structures of the complex
are reported, one crystallizing in a monoclinic space group (C2/c,
Z ¼ 8)12 and the other in an orthorhombic space group (Pbca, Z
¼ 8).13 In both cases, ML forms one-dimensional tapes through
centrosymmetric N–H/N hydrogen bonds. The benzoate
molecules arrange as pendants to the ML tapes through N–H/
two molecules each of 4HBA and ML along with four water
molecules. The ML units form one-dimensional tapes and the
benzoate units pendant to this, as observed in 1. Such units are
further connected with each other through several hydrogen
bonds (O–H/Oꢁ, O–H/O and N–H/O), involving the OH
group and lattice water molecules, to generate a zipper archi-
tecture in two-dimensions (Fig. 3).
Oꢁ and N+–H/Oꢁ (R2 (8)) interactions, as shown in Fig. 1. The
2
lattice water molecules hold the adjacent tapes together through
N–H/O, O–H/O and O–H/Oꢁ hydrogen bonds and the two
polymorphic structures show significant differences in the crystal
packing (see ESI†). The packing features of the BA–ML
complexes will be employed as a reference in subsequent
discussions to evaluate the synergic effect of both hydroxyl and
carboxyl groups in hydrogen bond formation with melamine.
2,3-Diydroxybenzoic acid–ML complex, 4
Upon co-crystallizing from an acetonitrile–water (1 : 1) mixture,
2,3-dihydroxybenzoic acid (23DHB) and ML yield a 1 : 1 molec-
ular complex, 4. Three molecules each of 23DHB and ML together
with eleven water molecules constitute the asymmetric unit. The
acid units are deprotonated and form a salt with ML. It is inter-
esting to note that among the three symmetry independent 23DHB
molecules, two have their hydroxyl groups in a syn-anti confor-
mation and the third has a syn-syn orientation. The ortho-hydroxyl
groups of the acids exhibit a similar orientation as in 2 and have
intramolecular interactions with the carboxylate moiety.
2-Hydroxybenzoic acid–ML complex, 2
The asymmetric unit of 2 consists of two molecules each of 2HBA
and ML along with a molecule of acetonitrile (Space group P21/c;
Z ¼ 4). The ML is mono-protonated and forms dimers through
centrosymmetric N–H/N (R22(8)) hydrogen bonds. The 2HBA
units cap the ML dimers resulting in a tetramer (Fig. 2). In 2HBA
the hydroxyl group is locked in an intramolecular interaction with
the carboxylate and hence is devoid of any structure stabilization
In the crystal, ML molecules form one-dimensional tapes and
the 23DHB units pendent to the tapes. The molecular tapes
constitute symmetry independent benzoate units with 23DHB in
the syn-syn and syn-anti (green and yellow respectively in Fig. 4a)
4888 | CrystEngComm, 2011, 13, 4886–4894
This journal is ª The Royal Society of Chemistry 2011

View Online
a
ꢀ
ꢀ
Table 2 Characteristic data of hydrogen bonds [bond lengths in A, angles in
]
N–H/N N–H /O O–H /O
N–H /Oꢁ N+–H /Oꢁ
O–H /Oꢁ
O–H/N
C–H/O
C–H/N
2
3
1.93 2.93 170 2.02 2.88 141 1.82 2.82 172 1.72 2.72 168
2.04 3.01 161 2.00 2.99 168 1.89 2.79 146 1.84 2.79 154
1.67 2.53 143
1.56 2.44 146
2.52 3.50 151
2.85 3.76 142
2.16 3.17 175
2.28 3.02 130
1.94 2.90 156
2.14 3.00 142
1.97 2.98 173 2.00 2.99 166 1.88 2.85 161 1.62 2.63 173 1.70 2.66 166 1.80 2.77 168
2.04 3.04 175 2.08 3.08 165 1.89 2.87 162 1.63 2.63 170 1.75 2.67 155 1.82 2.79 166
2.92 3.82 140
2.06 3.06 176 2.15 2.95 135
2.17 3.10 154
1.90 2.83 157 1.87 2.85 170
1.92 2.85 158 1.91 2.83 155
1.92 2.84 155
4
1.93 2.93 169 1.91 2.89 162 1.89 2.86 161 1.61 2.61 168 1.80 2.78 174 1.63 2.49 144
1.95 2.95 170 1.93 2.91 164 1.91 2.85 155 1.66 2.67 173 1.94 2.84 150 1.71 2.56 143
2.67 3.72 163
2.86 3.76 140
2.93 3.94 154
1.95 2.96 177 2.00 2.95 156 1.93 2.92 169 1.79 2.77 165
1.97 2.98 177 2.02 3.00 163
2.02 3.02 171 2.08 2.97 146
2.03 3.04 179 2.09 2.87 132
2.10 2.87 131
1.73 2.59 144
2.12 2.99 144
2.04 3.01 160 1.95 2.94 168 1.85 2.85 172 1.65 2.66 174 1.76 2.71 158 1.70 2.56 143
5
6
2.44 3.43 151
2.06 3.05 164 2.17 3.09 150
1.93 2.83 151 1.86 2.82 164
2.09 2.98 151 1.91 2.76 143
1.63 2.60 169
2.00 3.01 176 1.95 2.94 167 1.88 2.88 171
2.01 3.02 174 1.96 2.94 164 1.89 2.87 163
2.36 3.04 124 2.01 2.99 163 2.31 3.13 137
2.10 3.09 167
1.70 2.69 177 2.32 3.18 135 2.64 3.56 143
1.71 2.68 167 2.57 3.59 156 2.93 3.99 169
2.78 3.58 130
1.65 2.50 143
1.68 2.53 142
1.79 2.77 175
2.78 3.77 152
2.22 3.02 135
7
1.98 2.98 173 1.96 2.93 160 1.85 2.81 159
2.08 3.09 176 1.96 2.96 170 1.86 2.82 155
2.10 2.91 136 1.93 2.88 157
2.11 2.92 134 2.08 3.02 155
2.13 2.89 130 2.11 3.00 146
2.24 3.09 141
1.61 2.49 146
1.66 2.54 146
1.68 2.55 145
2.27 3.12 140
a
ꢀ
ꢀ
The three columns correspond to H/A, D/A (A) distances and D–H/A ( ) angle [A¼ acceptor, D ¼ donor].
Fig. 2 The hydrogen bonding pattern observed in 2.
intramolecular interaction with the carboxylate and is not
involved in any structure directing role. The 4–OH moiety
together with the lattice water (through O–H/O and N–H/O
hydrogen bonds) stabilizes the zipper structure (Fig. 5).
Fig. 1 A typical arrangement of molecules in 1.
conformations make distinct tapes which are arranged in an
XYY XYY sequence in two-dimensions. Water molecules
occupy the cavity generated by the inefficient crystal packing and
form a unique one-dimensional undecameric water cluster with
2,5-Diydroxybenzoic acid–ML complex, 6
Two molecules each of 25DHB, ML, water and DMF (which are
disordered) constitute the asymmetric unit of 6. Of the two
symmetry independent 25DHB molecules, the OH groups of one
is in the syn-syn and other is in the syn-anti conformation. The
ML units make dimers through N–H/N hydrogen bonds and
their further extension to a one-dimensional tape is restricted by
the interaction of 25DHB units (Fig. 6a). This in fact is similar to
14
ꢀ
an average O/O distance of 2.846 A (Fig. 4b).
2,4-Diydroxybenzoic acid–ML complex, 5
The crystal packing in 24DHB with ML is quite similar to 3. This
is not surprising as the ortho-hydroxyl group is involved in an
This journal is ª The Royal Society of Chemistry 2011
CrystEngComm, 2011, 13, 4886–4894 | 4889

View Online
Fig. 5 The zipper assembly obtained in 5.
crystal lattice, it has no structure directing role, as in the previous
hydrated structures 1, 3, 4 and 5, rather it stabilizes the assembly
together with the DMF molecules.
Fig. 3 Zipper architecture observed in 3. ML tapes and the acid
pendants are shown in two different colours.
2,6-Diydroxybenzoic acid–ML complex, 7
the case observed in 2. However, unlike in 2, one of the 25DHB
units retains its COOH functionality while the other undergoes
a proton transfer towards the ML unit (C–O distance is 1.22,
26 DHB yielded a salt with ML. Tetramer units consisting of ML
dimers and the 26DHB molecules as capping agents constitute
the basic recognition unit. It is interesting to note that even in the
absence of any solvent molecules, 7 gives a recognition pattern
similar to the one observed in 2 and 6 and forms a Lincoln log
network (Fig. 7). Several attempts to obtain the hydrated/
solvated complex with a variety of solvent combinations were
unsuccessful. This can be attributed to the absence of any
hydrogen bonding functionality at the rear of the tetramer units.
Both the hydroxyl groups are locked in an intramolecular
ꢀ
ꢀ
1.30 A; 1.25, 1.26 A, respectively, for the protonated and
deprotonated acid units).
The ortho hydroxyl group is locked in an intramolecular
interaction. However, the OH functionality in the 5-position,
exhibiting two distinct conformations, makes O–H/O hydrogen
bonds with DMF molecules as shown in Fig. 6a. The acid–ML
tetramer units stack in a Lincoln log arrangement in three-
dimensions (Fig. 6b). Though there exists water molecules in the
Fig. 4 The recognition patterns existing between the acid and ML molecules in 4. (a) The symmetrically distinct molecular tapes formed by the acid
molecules with syn-syn (green) and syn-anti (yellow) conformations. The water molecules present in the cavity is given in red colour. (b) The undecamer
water cluster. The pendent water molecules to the one-dimensional water chain are represented in green.
4890 | CrystEngComm, 2011, 13, 4886–4894
This journal is ª The Royal Society of Chemistry 2011

View Online
Fig. 6 (a) The basic hydrogen bonding pattern present in 6. The disorder associated with the DMF molecules is not shown for clarity (b) The Lincoln
log arrangement. Inset showing the solvents of crystallization DMF and water stabilizing Lincoln log arrangement.
interaction with the carboxylate and are not available to interact
with any incoming molecular species.
instances of hydration. This is in agreement with the fact that
compared to carboxylic acid, the hydration affinity of carboxy-
late is high. From a CSD analysis it is evident that 2671 out of
7722 hits for carboxylate (35%) are hydrated. However, it is only
Analysis of the hydrated/solvated structures
Hydrates are multiple-component systems that contain included
water molecules in the crystal lattice. Although there is no defi-
nite rule for the prediction of hydrate formation, it is widely
accepted that water molecules are incorporated into the crystal
lattice when there is an imbalance in the number of donor–
acceptor functionalities.15 Infantes et al. reported that the
probability of organic compounds forming hydrated structures
increases with an increasing number of polar chemical groups in
the molecule.16 They also noted that molecules possessing groups
like >C]O, COOH, CONH2, NH, NH2 and OH have a greater
tendency to form hydrated structures. In the analysis of the
structural features of the complexes of ML presented in the
previous sections, these postulates have been found to be vali-
dated, particularly in terms of the number of polar functionalities
in the crystal lattice. A CSD analysis further suggests that the
molecular complexes of ML with acids have a greater tendency
to form hydrated structures.8 In the majority of these complexes
the COOH group is deprotonated, resulting in increased
Fig. 7 The Lincoln log network present in 7. A representative tetramer
unit is highlighted.
This journal is ª The Royal Society of Chemistry 2011
CrystEngComm, 2011, 13, 4886–4894 | 4891

View Online
2332 out of 13128 for carboxylic acid (18%). The values are
comparable with those reported by Infantes et al.16
include careful DSC/TG analysis of the loss of water molecules
and variable temperature XRD to systematically evaluate the
impact of dehydration in these structures.
In the hydrated complexes (1, 3–6), TG analysis reveals that
ꢀ
the onset of the loss of lattice water molecules is below 100 C
(Table 3). This, in fact, suggests that the water molecules are
loosely bound and belong to the Class-1 type of hydrates
reported by Zaworotko and co-workers.17 Since water of
hydration plays a pivotal role in the structure stabilization, they
can be categorized in the ‘water as integral part’, according to the
classification of water of crystallization by Varughese et al.18 The
hydrated complexes 3, 4 and 5 show a similar recognition pattern
and form a zipper architecture (Scheme 3A). This is clearly
distinct from the acid–ML tetramer units and the Lincoln log
arrangement observed in the solvated/non-solvated complexes 2,
6 and 7 (Scheme 3B). Although the ML–benzoic acid complex 1
has similar hydrogen bonding pattern with the benzoic acid
forming pendants as those in 3, 4 and 5, formation of a zipper
architecture is not observed. This can be attributed to the
absence of a hydrogen bonding functionality on the benzoic acid
moiety to extend hydrogen bonding features which are essential
in the zipper architecture. In the complexes of hydroxy acids with
3- and 4-hydroxyl functionalities, an extended network with
water results in a zipper architecture. Thus, it is evident that
a hydroxyl functionality on the pendent acid together with water
is playing a co-operative role and is a requisite element in the
transformation of the one-dimensional melamine chains to
zipper architectures in the ML–HBA complexes. It is noteworthy
that the structure of 7 adopts a Lincoln log network as it is not
solvated and lacks any functionalities at the opposite end of the
carboxylate moiety.
Conformational analysis
It is noteworthy that in 4 and 6 the DHBA molecules exhibit two
different conformations. In 4, two out of the three symmetry
independent 23DHB molecules exhibit a syn–anti conformation
while the remaining one make a syn–syn orientation (Fig. 8a). It
is quite interesting to note that two energetically distinct
conformers co-exist in the crystal lattice. It is noteworthy that in
the co-crystal/salt literature of hydroxybenzoic acids, there exists
only one example, in the case of 26DHB, wherein both the syn–
syn and syn–anti conformers co-exist (Data from the CSD
analysis is given in Table S1, ESI†). Similarly in 6, the 25DHB
molecules exhibit different conformations (syn–syn and syn–anti)
as given in Fig. 8b. The occurrence of different orientations
shows the conformational flexibility of the OH groups and its
reliability towards various molecular recognition phenomena.
Although the ortho- hydroxyl groups in these acids are locked in
intramolecular hydrogen bonds, there exist no such restrictions
for the meta –OH groups. It may be interpreted that the co-
existence of the energetically different conformationally flexible
molecules observed in 4 and 6 occur preferentially to augment the
formation of hydrogen bonds, leading to robust frameworks in
the lattice.
Salt–to-co-crystal continuum
The extent of hydration in 4, wherein the cavity formed by the
inefficient packing of the molecules is occupied by a one-
dimensional undecameric water cluster with a unique topology,
is of interest. The formation of one-dimensional water clusters
with this particular topology is apparently imposed by the shape
of the host channels. Water clusters play a vital role in the
stabilization of supramolecular systems both in solution and in
the solid state and one-dimensional water chains represent
a unique form of water.19 Many of the biological processes
appear to depend on the properties of water clusters.20 Usually
these water clusters are stabilized by strong hydrogen bonds
formed between neighboring water molecules as well as the one
between water molecules and donor/acceptor groups associated
with channels. Several experiments are currently underway which
The pKa difference between the constituents is usually regarded
as a decisive factor in the formation of co-crystal or salt.21 A well-
accepted rule of thumb in predicting the salt formation is that
a proton transfer from acid to base can be anticipated if the DpKa
(DpKa ¼ pKa of base ꢁ pKa of acid) is greater than 3.22
A
negative DpKa would yield a co-crystal. However in the case of
melamine, such a prediction is void as a CSD analysis indicates
that the salt-to-co-crystal ratio is very high for the complexes of
ML with carboxylic acids wherein the components recognize
each other through N+–H/Oꢁ rather than the O–H/N
hydrogen bonds. This trend is observed in the present series of
complexes as well and hence the observed formation of salt/co-
crystal cannot be predicted in terms of DpKa values. Indeed, the
DpKa values in the aforesaid complexes are between 0.82 and
Table 3 General details of the molecular complexesa
Solvent of
crystallization
Conformations of
the OH groups
Salt/co-crystal preferences
of the co-formers (CSD)
Compound
pKa
DpKa
Solvent loss (ꢀC)
ML
1
2
3
4
5
6
7
5.39
4.20
3.01
4.57
2.96
3.32
3.01
1.30
—
—
Water
Acetonitrile
Water
Water
Water
Water/DMF
Anhydrous
—
—
RT
70–150
<65
80–100
40–100
—
—
—
—
—
1.19
2.38
0.82
2.43
2.07
2.38
4.09
salt < co-crystal
salt > co-crystal
salt < co-crystal
salt < co-crystal
salt > co-crystal
salt < co-crystal
salt > co-crystal
syn–syn & syn–anti
syn–syn
syn–syn & syn–anti
syn–syn
a
The pKa values of the compounds were obtained from SciFinder.
4892 | CrystEngComm, 2011, 13, 4886–4894
This journal is ª The Royal Society of Chemistry 2011

View Online
Scheme 3 Schematic representation of the two distinct topologies observed in the molecular complexes of hydroxyl acids with melamine.
Fig. 8 Conformational variations exhibited by the DHBA molecules (a) in 4 and (b) in 6.
4.09 and therefore a continuum exists between the two extremes
Acknowledgements
(Table 3). A more careful analysis (both theoretical and experi-
mental) of the propensity of salt formation, in terms of energies
associated with the hydrogen bonded regions, may provide
further insights into the charge transfer mechanism in these
complexes.23
TNG thanks DST for the award of a J. C. Bose fellowship and SS
thanks UGC for a Dr D. S. Kothari post-doctoral fellowship.
References
1 (a) G. R. Desiraju, Angew. Chem., Int. Ed. Engl., 1995, 34, 2311–2327;
€
Conclusions
(b) C. B. Aakeroy, Acta Crystallogr., Sect.B, 1997, B53, 569–586; (c)
O. Almarsson and M. J. Zaworotko, Chem. Commun., 2004, 1889–
1896; (d) R. Custelcean, M. Vlassa and J. E. Jackson, Chem.–Eur.
J., 2002, 8, 302–308; (e) M. W. Hosseini, Chem. Commun., 2005,
5825–5829; (f) P. Metrangolo and G. Resnati, Chem.–Eur. J., 2001,
7, 2511–2519; (g) A. Nangia and G. R. Desiraju, Top. Curr. Chem.,
1998, 198, 58–95; (h) K. Rissanen, CrystEngComm, 2008, 10, 1107–
1113.
2 (a) A. D. Bond, CrystEngComm, 2007, 9, 833–834; (b) S. Byrn,
R. Pfeiffer, G. Stephenson, D. J. W. Grant and W. B. Gleason,
Chem. Mater., 1994, 6, 1148–1158; (c) P. Vishweshwar,
J. A. McMohan, J. A. Bis and M. J. Zaworotko, J. Pharm. Sci.,
2006, 95, 499–516; (d) N. Schultheiss and A. Newman, Cryst.
Growth Des., 2009, 9, 2950–2967; (e) A. V. Trask,
W. D. S. Motherwell and W. Jones, Cryst. Growth Des., 2005, 5,
1013–1021; (f) D. P. NcNamara, S. L. Childs, J. Giordano,
A. Iarriccio, J. Cassidy, M. S. Shet, R. Mannion, E. O’Donnell and
A series of molecular complexes of ML with various hydroxy-
benzoic acids have been prepared and analyzed, wherein an
intricate array of homomeric and heteromeric interactions exist.
In the complexes, the hydroxyl and carboxyl groups exhibit
a synergic influence in the molecular recognition. Further to the
functional groups, topology of the resulting assemblies is influ-
enced by the presence/absence of the solvent of crystallization.
Zipper architectures are formed in the hydrated complexes
whereas a Lincoln log arrangement is obtained in the presence/
absence of other solvents. It is clear that ML preferentially forms
salts with hydroxy acids and the prediction of proton transfer
based on DpKa is inappropriate in these complexes. Database
analysis corroborates our claim that in most cases, the recogni-
tion of ML with carboxylic acids leads to salt formation and the
reason for the proton transfer is still elusive. For a quantitative
analysis of the salt/co-crystal features, in terms of the charge
transfer character at the critical point and acid strengths,
a detailed charge density study of the complexes of ML with both
strong and weak acids is required.
ꢂꢂꢃ
A. Park, Pharm. Res., 2006, 23, 1888–1897; (g) S. Karki, T. Friscıc,
ꢃ
L. Fabian, P. R. Laity, G. M. Day and W. Jones, Adv. Mater.,
2009, 21, 3905–3909.
ꢃ
3 (a) S. Varughese and V. R. Pedireddi, Tetrahedron Lett., 2005, 46,
2411–2415; (b) K. K. Arora, J. PrakashaReddy and
V. R. Pedireddi, Tetrahedron, 2005, 61, 10793–10800; (c)
€
C. B. Aakeroy, N. Schultheiss and J. Desper, J. Mol. Struct., 2010,
972, 35–40; (d) C. B. Aakeroy, B. M. T. Scott, M. M. Smith,
J. F. Urbina and J. Desper, Inorg. Chem., 2009, 48, 4052–4061; (e)
This journal is ª The Royal Society of Chemistry 2011
CrystEngComm, 2011, 13, 4886–4894 | 4893

View Online
As against 722 hits available for the acid-pyridine heterosynthons
there exist only hits for the acid–triazine interaction. The
Refcodes corresponding to the acid–triazine interactions are
ODEHIO, ROGQUA, WINMIQ and NUSHIT (CSD version 5.32,
February 2011 update).
11 (a) G. M. Sheldrick, Acta Crystallogr., Sect. A, 2008, 64, 112–122; (b)
L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837–838.
12 G. J. Perpetuo and J. Janczak, Acta Crystallogr., Sect. E, 2005, 61,
o287–o289.
13 X. Zhang, Chem. Res. Chin. Univ., 2008, 24, 396–400.
14 R. Ludwig, Angew. Chem., Int. Ed., 2001, 40, 1808–1827.
15 G. R. Desiraju, J. Chem. Soc., Chem. Commun., 1991, 426–428.
16 (a) L. Infantes, J. Chisholm and W. D. S. Motherwell,
4
4 (a) E. W. Hughes, J. Am. Chem. Soc., 1941, 63, 1737–1752; (b)
J. N. Varghese, A. M. O’Connell and E. N. Maslen, Acta
ꢃ
Crystallogr., 1977, B33, 2102–2108; (c) G. J. Perpetuo and
ꢃ
ꢃ
J. Janczak, J. Mol. Struct., 2008, 891, 429–436; (d) A. Cousson,
B. Nicolai and F. Fillaux, Acta Crystallogr., Sect. E, 2005, E61,
o222–o224.
5 (a) A. Ranganathan, V. R. Pedireddi and C. N. R. Rao, J. Am. Chem.
Soc., 1999, 121, 1752–1753; (b) A. Ranganathan, V. R. Pedireddi,
G. Sanjayan, K. N. Ganesh and C. N. R. Rao, J. Mol. Struct.,
2000, 522, 87–94; (c) C. T. Seto and G. M. Whitesides, J. Am.
Chem. Soc., 1993, 115, 905–916.
CrystEngComm, 2003, 5, 480–486; (b) L. Infantes, L. Fabian and
W. D. S. Motherwell, CrystEngComm, 2007, 9, 65–71; (c)
L. Infantes and W. D. S. Motherwell, Z. Kristallogr., 2005, 220,
333–339.
17 H. D. Clarke, K. K. Arora, H. Bass, P. Kavuru, T. T. Ong, T. Pujari,
L. Wojtas and M. J. Zaworotko, Cryst. Growth Des., 2010, 10, 2152–
2167.
18 S. Varughese and G. R. Desiraju, Cryst. Growth Des., 2010, 10, 4184–
4196.
6 (a) http://en.wikipedia.org/wiki/2008_baby_milk_scandal; (b) L. Zhu,
G. Gamez, H. Chen, K. Chingin and R. Zenobi, Chem. Commun.,
2009, 559–561.
19 (a) S.-P. Chen, M. Li, Y. Xuao, Y.-X. Yuan, L.-L. Pan and
L.-J. Yuan, CrystEngComm, 2008, 10, 1227; (b) L. E. Cheruzel,
M. S. Pometun, M. R. Cecil, M. S. Mashuta, R. J. Wittebort and
R. M. Buchanan, Angew. Chem., Int. Ed., 2003, 42, 5452; (c) Z. Fei,
D. Zhao, T. J. Geldbach, R. Scopelliti, P. J. Dyson, S. Antonijevic
and G. Bodenhausen, Angew. Chem., Int. Ed., 2005, 44, 5720; (d)
U. S. Raghavender, A. Kantharaju, S. Aravinda, N. Shamala and
P. Balaram, J. Am. Chem. Soc., 2010, 132, 1075; (e) A. Wakahara
and T. Ishida, Chem. Lett., 2004, 33, 354; (f) A. Wu, Y. Li,
F. Zheng, G. Guo and J. Huang, Cryst. Growth Des., 2006, 6, 444.
20 K. Murata, K. Mitsuoka, T. Hirai, T. Walz, P. Agre, J. B. Heymann,
A. Engel and Y. Fujiyoshi, Nature, 2000, 407, 599–605.
21 (a) S. L. Childs, G. P. Stahly and A. Park, Mol. Pharmaceutics, 2007,
4, 323–338; (b) C. B. Aakeroy, M. E. Fasulo and J. Desper, Mol.
Pharmaceutics, 2007, 4, 317–322; (c) S. Mohamed, D. A. Tocher,
M. Vickers, P. G. Karamertzanis and S. L. Price, Cryst. Growth
Des., 2009, 9, 2881–2889; (d) K.-S. Huang, D. Britton, M. C. Etter
and S. R. Byrn, J. Mater. Chem., 1997, 7, 713–720; (e)
B. R. Bhogala, S. Basavoju and A. Nangia, CrystEngComm, 2005,
7, 551–562.
7 K. Ai, Y. Liu and L. Lu, J. Am. Chem. Soc., 2009, 131, 9496–9497.
8 F. H. Allen and O. Kennard, Chem. Des. Automat. News, 1993, 8, 31–
37, Refcodes corresponding to the hydrated ML-carboxylic acid
complexes; CSD version 5.32, February 2011 update, AJOFUA,
ARUDAS, CELHOR, CELHUX, EFAZOA, ETIGUJ, FETTOO,
ISUTUL, JIWQIQ, MAQBEM, MAQBIQ, OMIJEZ, QETPIP,
ROJGIH, ROJGON, TECFUD, VANDEU, VORSUR, BOPTEG,
DUCSOK, TUPKAR.
9 (a) M. K. Marchewka, J. Barana, A. Pietraszko, A. Haznar, S. Debrus
and H. Ratajczak, Solid State Sci., 2003, 5, 509–518; (b) R. Thomas,
S. Pal, A. Datta, M. K. Marchewka, H. Ratajczak, S. K. Pati and
G. U. Kulkarni, J. Chem. Sci., 2008, 120, 613–620; (c)
K. U. Lakshmi, S. Thamotharan, K. Ramamurthi and B. Varghese,
Acta Crysttallogr., Sect. E., 2006, E62, o455–o457; (d)
M. K. Marchewka and A. Pietraszko, J. Phys. Chem. Solids, 2003,
64, 2169–2181; (e) H. Su, Y.-K. Lv and Y.-L. Feng, Acta
Crysttallogr., Sect. E., 2009, E65, o933; (f) G. J. Perpetuo and
J. Janczak, J. Mol. Struct., 2008, 891, 429–436.
10 Oxford Diffraction (2009). CrystAlis CCD and CrystAlis RED,
Version 1.171.33.31. Oxford Difraction Ltd. Abingdon,
Oxfordshire, England.
22 S. L. Johnson and K. A. Rumon, J. Phys. Chem., 1965, 69, 74.
23 V. R. Hathwar, R. Pal and T. N. G. Row, Cryst. Growth Des., 2010,
10, 3306–3310.
4894 | CrystEngComm, 2011, 13, 4886–4894
This journal is ª The Royal Society of Chemistry 2011