Identification of molecular crystals capable of undergoing an acyl-transfer reaction based on intermolecular interactions in the crystal lattice

Article abstract of DOI:10.1002/chem.201301058

Investigation of the intermolecular acyl-transfer reactivity in molecular crystals of myo-inositol orthoester derivatives and its correlation with crystal structures enabled us to identify the essential parameters to support efficient acyl-transfer reactions in crystals: 1)a the favorable geometry of the nucleophile (-OH) and the electrophile (C-O) and 2)a the molecular assembly, reinforced by C-H×××π interactions, which supports a domino-type reaction in crystals. These parameters were used to identify another reactive crystal through a data-mining study of the Cambridge Structural Database. A 2:1 co-crystal of 2,3-naphthalene diol and its di-p-methylbenzoate was selected as a potentially reactive crystal and its reactivity was tested by heating the co-crystals in the presence of solid sodium carbonate. A facile intermolecular p-toluoyl group transfer was observed as predicted. The successful identification of reactive crystals opens up a new method for the detection of molecular crystals capable of exhibiting acyl-transfer reactivity. Reactive crystal gazing? Characterization of functional groups in small organic molecules allows the prediction of their reactivity in solution, and often the structure of the resulting products with fairly good accuracy. This, however, is not true for reactions in the solid state. It was demonstrated that an understanding of the non-covalent interactions in molecular crystals can aid in identifying crystal structures that support chemical reactions (see figure). Copyright

Full text of DOI:10.1002/chem.201301058

FULL PAPER
DOI: 10.1002/chem.201301058
Identification of Molecular Crystals Capable of Undergoing an Acyl-Transfer
Reaction Based on InterACTHNUTRGNEUGmN olecular Interactions in the Crystal Lattice
Majid I. Tamboli,[a] Shobhana Krishnaswamy,[b] Rajesh G. Gonnade,[b] and
Mysore S. Shashidhar*[a]
Abstract: Investigation of the intermo-
lecular acyl-transfer reactivity in mo-
lecular crystals of myo-inositol ortho-
reaction in crystals. These parameters
were used to identify another reactive
crystal through a data-mining study of
the Cambridge Structural Database. A
2:1 co-crystal of 2,3-naphthalene diol
and its di-p-methylbenzoate was select-
ed as a potentially reactive crystal and
its reactivity was tested by heating the
co-crystals in the presence of solid
sodium carbonate. A facile intermolec-
ular p-toluoyl group transfer was ob-
served as predicted. The successful
identification of reactive crystals opens
up a new method for the detection of
molecular crystals capable of exhibiting
acyl-transfer reactivity.
ACHTUNGTRENNUNGester derivatives and its correlation
with crystal structures enabled us to
identify the essential parameters to
support efficient acyl-transfer reactions
in crystals: 1) the favorable geometry
Keywords: acylation · domino reac-
tions · intermolecular interactions ·
solid-state reactions · X-ray diffrac-
tion
ꢀ
of the nucleophile ( OH) and the elec-
trophile (C=O) and 2) the molecular
ꢀ
assembly, reinforced by C H···p inter-
actions, which supports a domino-type
Introduction
examples.[20–24] This is perhaps because most of these reac-
tive crystals were comprised of one chemical entity and
hence the reactions investigated were either intramolecular
reactions or homomolecular reactions (implying reactions
between one kind of molecules).
Carbon–carbon and carbon–heteroatom (O, N, S) bond-
forming reactions in molecular crystals have been investigat-
ed in the last few decades because of their facility and often,
unique product selectivity.[1–6] The crystal lattice provides a
structured environment for specific reactions, which results
in high product selectivity compared with the corresponding
solution-state reactions. Reactions in solids, apart from
being potential environmentally green or benign reaction
systems, also contribute to the understanding of differences
in the stability of small molecules in different phases, includ-
ing drug formulations.[7–10] The pioneering work of
Schmidt[11–13] and later others[14–19] helped formulate require-
ments in terms of inter-atomic distances, design and create
templates and new substrates for the addition reactions of
Transesterification reactions, involving nucleophilic addi-
tion to carbonyl groups (followed by elimination), are fre-
quently encountered organic reactions in the solution state
and in living systems. Acyl group-transfer reactions, albeit
few, have also been reported to occur in the organic solid-
state.[7–9,25,26] The earliest report was the thermally induced
intramolecular O!N acyl migration in O-acylsalicyl-
AHCTUNGTRENNUNG
amide.[25] Vyas and co-workers correlated the facile intramo-
lecular acyl group migration with the geometry of the reac-
tive centers in the crystalline state.[26] Transacylation was
also observed in tablet mixtures of aspirin and drugs with
easily acylated functionalities such as phenylephrine hydro-
chloride, codeine, and acetaminophen.[7–9] We had reported
(the first instance[27] of) a facile intermolecular benzoyl
group transfer in crystals of racemic 2,4-di-O-benzoyl-myo-
inositol 1,3,5-orthoformate (1), its orthoacetate analogue
2,[28] as well as in co-crystals 1·2[29] (Scheme 1), which yielded
the corresponding tribenzoates 3 and 4 and the diols 5 and
6.
A comparison of the crystal structures of all the com-
pounds in which the intermolecular acyl-transfer reaction
occurred and their polymorphs in which the corresponding
reaction was not facile, helped us determine the minimum
conditions necessary for the occurrence of intermolecular
acyl-transfer reaction in molecular crystals.[30,31] The present
article reports identification of a molecular co-crystal (in the
light of prior knowledge generated in our laboratory), by
ꢀ
olefins in crystals, resulting in C C bond formation. In con-
trast, most of the reports on reactions in molecular crystals
involving migration of relatively larger group of atoms be-
tween the reacting centers have been sporadic stand-alone
[a] M. I. Tamboli,+ Dr. M. S. Shashidhar
Division of Organic Chemistry, CSIR–National Chemical Laboratory
Dr. Homi Bhabha Road, Pashan, Pune-411008 (India)
Fax : (+91)20-25902629
E-mail: ms.shashidhar@ncl.res.in
[b] Dr. S. Krishnaswamy,+ Dr. R. G. Gonnade
Center for Materials Characterization
CSIR–National Chemical Laboratory, Dr. Homi Bhabha Road
Pashan, Pune-411008 (India)
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201301058.
Chem. Eur. J. 2013, 19, 12867 – 12874
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12867

tures and theoretical calculations on a model system, by
Bꢂrgi and Dunitz.[32] Their results (from crystal structure
data) indicated that the N···C=O and O···C=O angles were
in the range (105ꢁ5)8 for all the distances smaller than
2.5 ꢁ between the electrophile and the nucleophile. Hy-
droxy acids with intramolecular O···C=O angles of ꢂ988
show the highest rate of intramolecular lactonization[33] and
Bender[34] also postulated that a perpendicular approach of
the nucleophile to the p-electron system should be preferred
over a coplanar approach in order to maximize the overlap
between the nucleophile and p-electrons of the carboxyl
group during intramolecular lactonization. Hence, it is rea-
sonable to expect larger deviation of O···C=O angle (from
the tetrahedral angle) as the El···Nu distance increases
(beyond 2.5 ꢁ), as observed in reactive crystals of myo-ino-
sitol orthoester derivatives. Subsequent investigations of
structure and benzoyl group-transfer reactions in crystals of
other myo-inositol 1,3,5-orthoester derivatives[29–31] helped
us realize the importance of helical molecular pre-organiza-
tion in reactive crystals,[35] which functions as a reaction
channel and contributes to the facility and almost quantita-
tive conversion of the reactants to products. Hence, we won-
dered whether 1) El···Nu geometry between potentially re-
active centers in molecules; 2) the assembly of these react-
ing pairs in the crystal, in the form of a channel for the reac-
tion to propagate; and 3) weak intermolecular interac-
Scheme 1. Solid-state transesterification reactions in co-crystal 1·2 of
myo-inositol orthoester derivatives.
search of the Cambridge Structural Database (CSD, Version
5.34, November 2012), which had the potential to support
intermolecular acyl-transfer between its constituent mole-
cules and the experimental verification of the facility of
acyl-transfer reaction as anticipated.
A large body of data on reactions in solutions, accumulat-
ed over more than a century, helped chemists to categorize
organic reactions in terms of “functional groups” and hence
increased the probability of fairly accurate prediction of the
facility of reactions of organic molecules and the structure
of the possible products, based on the structure of the reac-
tant molecules. In contrast, the complex nature of non-cova-
lent intermolecular interactions precludes the design and
synthesis of crystals that contain the reactive centers (of
constituent molecules) in the right relative orientation for
successful covalent bond formation. The prevalence of poly-
morphic modifications and their thermal phase transitions in
the solid state also contribute to the uncertainty in predic-
tion of the reactivity of molecules in their crystals. The re-
sults presented here show that the knowledge generated by
systematic analysis of crystal structures that facilitate a
chemical reaction, can be utilized in identifying other reac-
tive crystalline solids.
[38,39]
tions,[36,37] such as C H···p,
which help maintain the
ꢀ
topochemical control, could be used as parameters to pre-
dict the facility of acyl group-transfer reactivity in molecular
crystals.
A survey of the CSD for compounds containing a carbon-
yl group and an oxygen nucleophile revealed that when the
distance between the electrophile and the nucleophile was
lesser than the sum of their van der Waals radii (3.22 ꢁ),
the angle of approach of the nucleophile towards the elec-
trophile lay mostly in the range of 80–1008.[31] The CSD
searches, depending on the distance between the ester car-
bonyl carbon (El) and the hydroxyl oxygen (Nu) (HO···C=
O, 3.1–3.5 ꢁ, defined as a non-bonded contact) and the cor-
responding angle (85–958, see Figure 1), yielded ꢂ200 hits.
About 100 of these structures were scrutinized and ten of
these were selected as “potentially reactive” crystals.[40–49]
The co-crystal 7·8 (CSD reference code: IJAGIJ) was
Results and Discussion
Efficient intermolecular oxygen-to-oxygen benzoyl group-
transfer reactions were observed in myo-inositol orthoester
derivatives 1, 2, and co-crystals 1·2. Single-crystal X-ray dif-
fraction analysis of these reactive crystals[28,29] revealed that
the distance between the reaction centers (HO···C=O), and
ꢀ
the angle of approach of the OH (nucleophile=Nu) to-
wards the ester carbonyl group, C=O (electrophile=El) lay
in the range of 3.1–3.3 ꢁ and 84–908, respectively.[28] These
interatomic distances and angles are close to those arrived
for El···Nu interactions through the study of crystal struc-
Figure 1. Relative orientation of the neighboring molecules of 7 and 8 in
co-crystals 7·8 at ꢀ1738C. Thermal ellipsoids are drawn at 50% probabil-
ity level and H-atoms are depicted as spheres of arbitrary radii.
12868
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 12867 – 12874

Identification of Molecular Crystals
FULL PAPER
chosen for the actual experiment due to its ease of prepara-
tion in larger amounts and also due to contemporary inter-
est in the preparation and the study of properties of co-crys-
tals. The grounds for rejection of other “hits” in the CSD
were 1) crystal comprising of one kind of molecule (i.e., El
and Nu present in the same molecule) since we had earlier
reported[27.28] reactions in such crystals; 2) solvated crystal:
Potential loss of solvent and hence possibility of phase
change on heating, which could make the structure–reactivi-
ty correlation difficult; 3) multistep synthesis for obtaining
the desired molecule and hence its crystal; 4) compound ob-
tained by isolation in small quantities from natural sources;
5) hydrated crystal: Potential loss of water and/or possibility
of hydrolysis during the acyl-transfer reaction in crystals. In
the 2:1 co-crystal 7·8 of 2,3-naphthalene-diol (7) and its di-
p-methylbenzoate (8) the distance between the nearest Nu
ꢀ
( OH of 7) and the El (C=O of 8) was 3.166 ꢁ and the
angle of approach of Nu towards El was 85.68 (Figure 1; for
similar association at higher temperatures, see Figure S1–S4,
the Supporting Information).
A view of the crystal packing along the ab-diagonal (Fig-
ure 2a) revealed a layered structure in which the naphtha-
lene rings of 7 and 8 are perpendicular to each other. Such a
molecular assembly brings the hydroxyl oxygen (Nu) of the
diol 7 in close proximity to the carbonyl carbon (El) of the
adjacent diester 8, and generates a short El···Nu contact
(O2···C16=O4), satisfying one of the pre-requisites for the
acyl group migration from 8 to 7 (Figure 1 and Table 1).
Table 1. El···Nu parameters for the reacting molecules in the co-crystal
7·8 at various temperatures.
El···Nu
T [8C]
100
ꢀ173
25
125
25[a]
C16···O2
aO4=C16···O2
3.162
85.8
82
3.272
84.6
84
3.341
84.4
85
3.269
84.7
82
3.277
84.5
84
ꢀ
aH2A O2···C16
Figure 2. Packing of molecules in co-crystal 7·8. a) Layered structure of
molecules in crystals 7·8 displaying a sandwiched arrangement of diol
molecules between the diesters; b) molecular packing viewed down the b
ꢀ
aC3 O2···C16
102.4
101.3
100.7
101.3
101.1
[a] Crystal was heated to 1368C on hot-stage microscope and then cooled
to 258C.
ꢀ
axis revealing bridging of neighboring reactive layers through strong O
ꢀ
H···O and C H···O interactions; c) view of the packing of molecules
down the reaction channel showing discrete alignment of the layers.
This molecular architecture is further supported by the for-
ꢀ
mation of off-centered C H···p interactions involving naph-
thalene ring protons of the diester 8 and p-cloud of the adja-
cent diol 7 molecules (Figure 2a, for color graphic see Fig-
ure S5, the Supporting Information). The neighboring layers
are connected to each other through strong intermolecular
ꢀ
ꢀ
ꢀ
O H···O hydrogen bonds (O2 H2A···O4), and linear C
ꢀ
H···O (C18 H18···O1) interactions (Table S2, the Support-
ing Information), thus providing well-directed discrete reac-
tion channels throughout the crystal lattice (Figure 2b and
2c; for a color graphic see Figure S5, the Supporting Infor-
mation). Hence, the co-crystal 7·8 had all the structural fea-
tures (see points (1)–(3) mentioned earlier), for an intermo-
lecular p-toluoyl group-transfer reaction between 7 and 8
(Scheme 2).
The experiment for verifying the expected reactivity in
co-crystals 7·8 was interesting because the structure of the
Scheme 2. Intermolecular toluoyl group transfer in co-crystals 7·8.
Chem. Eur. J. 2013, 19, 12867 – 12874
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12869

M. S. Shashidhar et al.
constituent molecules (inositol orthoester derivatives) of
crystals that were used to arrive at the necessary conditions
for acyl transfer are hugely different from the constituent
molecules (naphthalene derivatives) of co-crystals 7·8 in
which the reaction was expected to occur.
significant interactions between 7 and 8 revealed compara-
ꢀ
ble values for H O···C=O, that is, El···Nu contact
(ꢀ67.8 kJmolꢀ1 e) and O H···O hydrogen-bonding interac-
ꢀ
ꢀ1
ꢀ
tion (ꢀ70.7 kJmol ), whereas, interaction energies for C
ꢀ1
ꢀ1
ꢀ
H···O (ꢀ19.6 kJmol ) and C H···p contacts (ꢀ5.3 kJmol )
were noticeably lower. These values reveal the relative im-
portance of the El···Nu contacts (O2···C16=O4) in the lay-
ered arrangement of one diester 8 and two diol 7 molecules
(Figure 2) in co-crystals 7·8.
The acyl-transfer reactivity in 7·8 was tested as described
in the experimental section to isolate the p-toluate 9 in a
very good yield (91%). Since the diol 7 and the diester 8
are present in the molar ratio of 2:1 in the co-crystal, ꢂ1
equivalent of the diol 7 was also obtained at the end of the
reaction. No acyl-transfer reactivity was observed at temper-
atures below 1108C or in the absence of sodium carbonate.
The acyl-transfer reaction was less efficient above the melt-
ing point of co-crystals 7·8 (140–1458C, 5 h, 46% yield of 9)
as compared to the reaction below the melting point (122–
1258C, 6 h, 80% yield of 9). Isolation of p-toluic acid as one
of the products in the reaction above the melting point of
the co-crystals 7·8 indicated that the acyl-transfer reaction
was less specific in the molten state (due to loss of topo-
chemical control) and was accompanied by hydrolysis of the
ester group. The diester 8 predominantly underwent hydrol-
ysis in DMF solution, when allowed to react with 7 in the
presence of sodium carbonate. Use of p-xylene as the sol-
vent resulted in the formation of the monotoluate 9 (9%)
and about 40% of the unreacted 8 was recovered. A com-
parison of all these results clearly indicates the role played
by the crystal lattice during the reaction between 7 and 8 in
their co-crystals.
The El···Nu parameters as well as the molecular assem-
blies shown in Figures 1 and 2 are derived from the X-ray
diffraction data of co-crystals 7·8 collected at low tempera-
ture. However, the acyl-transfer reaction in 7·8 occurred at
much higher temperature. Hence, we wanted to observe the
effect of temperature on the crucial intermolecular interac-
tions shown in Figures 1 and 2. The X-ray diffraction inten-
sity data measurement at 1008C was carried out successfully,
however data collection at temperatures higher than 1008C
was not initially successful due to changes at the crystal sur-
face (Figure S7, the Supporting Information) leading to frag-
mentation of the exterior into small crystallites (the interior
of the crystal was intact). The diffraction spots confirmed
the single crystalline nature but high mosaicity and bad
least-squares of the orientation matrix suggested a phase
change (this could be responsible for a small endothermic
hump in the differential scanning calorimetry (DSC) of 7·8
prior to melting, Figure S12, the Supporting Information).
Removal of the small crystallites from the surface and rapid
collection of the data sets (at 1258C) gave the structure of
the co-crystal 7·8 with a good R-value. We also determined
the structure of the co-crystal 7·8 (at 25 (2)8C), which was
heated up to 1368C (close to melting temperature 1398C)
on hot stage polarizing microscope and then cooled to room
temperature. The structure overlay of the reacting molecules
at different temperatures matched very well (Figure S6, the
Supporting Information).
The co-crystal 7·8 was reported[42] as an intermediate in
the solvent-free acylation of the diol 7 with p-toluoyl chlo-
ACHTUNGTRENNUNGride at or below 1258C. The existence of this intermediate
in the reaction mixture was revealed by IR spectroscopy,
and the same intermediate was later crystallized using 7 and
8. Although the co-crystal 7·8 was described as an inter-
mediate during the O-acylation of 7, it was not clear wheth-
er co-crystals 7·8 functioned as reactive intermediates
during the solvent free acylation of 7 with p-toluoyl chlo-
These experiments revealed that there were no major
changes in the intermolecular El···Nu contacts, with varia-
tion in temperature (Figure 1 and Table 1). The variable
temperature powder X-ray diffraction patterns recorded at
temperatures of 25, 80, and 1258C matched reasonably well
(Figures S14–S17, the Supporting Information), thus suggest-
ing a modest effect of heating on the packing arrangements
of molecules in the crystal lattice of co-crystals 7·8. Interest-
ingly, the thermal anisotropies of molecules of 7 and 8 in co-
crystals 7·8 at 1008C showed a significant change; the lone
pair on the oxygen of the hydroxyl group was found to be
better oriented towards the carbonyl carbon, as compared
with that observed at other temperatures (Figure 1 and Fig-
ures S1–S4, the Supporting Information). The thermal
motion analysis of the reactants revealed a larger internal
motion of O2 towards C16 at 1008C. This motion diminish-
ed at 1258C suggesting stabilization of the movement of
these groups towards each other at higher temperatures.
This perhaps implies the onset of the acyl-transfer reaction
between 7 and 8 in co-crystals 7·8.
ACHTUNGTRENNUNGride. This is especially because the solvent free benzoylation
of 7 with p-toluoyl chloride at 1208C was complete in
15 min to yield the diester 8. The acyl-transfer reaction in
co-crystals 7·8 on the other hand, needed several hours
above 1208C for completion. Hence it is unlikely that the
acylation of 7 under solvent free conditions proceeded ex-
clusively through
a topochemically controlled benzoyl
group-transfer reaction in the intermediate co-crystals 7·8.
The facile formation of co-crystals 7·8 just by co-grinding
the diol 7 and its diester 8 revealed a strong affinity between
these two molecules (Figure S13, the Supporting Informa-
tion). The significant interactions observed between 7 and 8
ꢀ
in their co-crystals were, (i) O H···O hydrogen-bonding in-
teractions between OH of the diol 7 and the C=O group of
ꢀ
8; (ii) short El (C=O) and Nu (OH) contact; (iii) C H···O
ꢀ
contacts and (iv) C H···p interactions between the naphtha-
lene ring proton of 8 and aromatic ring of the diol 7 (Table
S2, the Supporting Information). An estimation of the lat-
tice energy[50] of the co-crystal 7·8 gave
a value of
ꢀ233 kJmolꢀ1. The computation of interaction energies for
12870
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 12867 – 12874

Identification of Molecular Crystals
FULL PAPER
The mechanism of the acyl-transfer reaction depicted in
Scheme 3 rationalizes the experimentally observed high
extent of conversion of reactants to product. Sodium car-
progress well into the crystal, by generation of phenoxide
ions due to a series of intramolecular toluoyl group transfer
(Step 2, Scheme 3) and proton transfer (Steps 3 and 4,
Scheme 3).
ACHTUNGTRENNUNGbonACHTUNGTRENNUNGate possibly initiates the reaction at the base of a molec-
ular layer (Step 1, Scheme 3) after which the reaction pro-
gresses by successive proton and p-toluoyl group transfers,
resulting in a clean reaction and high yield of the mono
ester 9. A high percentage of conversion of reactants to
products observed is consistent with this mechanism since
the acyl-transfer reaction initiated at the surface of the crys-
tals by the base (sodium carbonate, Step 1, Scheme 3) can
The intramolecular toluoyl group transfer proposed in
Step 2 of Scheme 3 is supported by the perpendicular ap-
proach of the phenolic oxygen (Nu) of 8 towards the ester
carbonyl group (C=O, El) of the same molecule (Figure 3,
for a color graphic see Figure S10, the Supporting Informa-
tion). The El···Nu distance observed in the crystal of 7·8 for
intramolecular toluoyl group transfer is 2.935(2) ꢁ and the
angle of approach is 80.98. The
intramolecular transfer of the
toluoyl group is necessary to
sustain a domino-type reaction
in co-crystals 7·8.
We also solved the crystal
structure of the product 9 (Fig-
ure S8, Table S1, the Supporting
Information) for comparison
with the structure of the co-
crystal 7·8 and to inspect the
El···Nu geometry for inter- as
well as intramolecular acyl-
transfer. The intramolecular ge-
ometry
(El···Nu
distance
O2···C11=2.937 ꢁ and angle
ꢀ
O2···C11 O3=80.98) for acyl
transfer in crystals of the mono-
ester 9 matches (Figure 4; for a
color graphic see Figure S11,
the Supporting Information)
with the parameters observed
for the diester 8 in its co-crystal
7·8, supporting the possibility of
intramolecular
during the process shown in
Scheme 3.
acyl-transfer
AHCTUNGTRENNUNG
However, a catemeric associ-
ation of the c-glide-related mol-
ꢀ
ecules through strong O H···O
ꢀ
(O2 H2A···O3)
hydrogen-
bonding interactions between
the carbonyl oxygen atom and
the hydroxyl group (potential
reaction centers) of adjacent
molecules precludes the possi-
bility of intermolecular acyl-
transfer reaction in crystals of 9
(Figure S9, Table S3, the Sup-
porting Information).
Conclusion
Co-crystals and molecular com-
plexes have gained prominence
Scheme 3. A mechanism for the acyl group transfer from 8 to 7 in 7·8.
Chem. Eur. J. 2013, 19, 12867 – 12874
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12871

M. S. Shashidhar et al.
diol 7 (0.160 g, 1.0 mmol), the diester 8 (0.198 g, 0.50 mmol), and two–
three drops of ethyl acetate, using a pestle and mortar.
Solid-state reactivity of co-crystals 7·8: A mixture of freshly grown crys-
tals of 7·8 (0.093 g, 0.13 mmol) and activated sodium carbonate (0.110 g,
1.03 mmol) was ground into a fine powder using a mortar and pestle. The
mixture was transferred to a test tube filled with argon, and heated in an
oil bath (122–1258C) for 60 h. The reaction mixture was cooled to ambi-
ent temperature, suspended in water (20–30 mL) and then extracted with
ethyl acetate (20–30 mL). The organic extract was washed with water,
followed by brine. The organic layer was dried over anhydrous sodium
sulphate and evaporated under reduced pressure. The residue obtained
was purified by column chromatography (silica gel 100–200 mesh; eluent,
ethyl acetate/light petroleum 10:90, v/v) to obtain 9 (0.066 g, 91%) as a
solid. M.p. 197–1988C (crystals from chloroform); TLC (35:65 ethyl ace-
tate/light petroleum, v/v): Rf: 0.6; 1H NMR (200 MHz, CDCl3): d=8.12–
8.20 (m, 2H), 7.69–7.80 (m, 3H), 7.30–7.50 (m, 5H), 5.70 (s, 1H),
2.48 ppm (s, 3H); 13C NMR (50 MHz, [D6]Acetone): d=165.7, 149.7,
145.6, 141.8, 134.2, 131.3, 130.5, 129.6, 128.4, 128.2, 127.2, 127.1, 124.9,
121.9, 112.3, 22.0 ppm; IR (Nujol): n˜ =3391, 1723 cmꢀ1; elemental analy-
sis calcd (%) for C17H24O3: C 77.68, H 5.07; found: C 77.30, H 4.81. The
diol 7 (0.022 g, 53%) was recovered on elution with ethyl acetate/light
petroleum (25:75 v/v) (Figures S18–S22, the Supporting Information).
Figure 3. ORTEP of the diester 8 in a crystal of 7·8 showing relative ori-
entation between the phenolic oxygen and ester carbonyl group at
1258C. Thermal ellipsoids are drawn at 50% probability level and
H-atoms are depicted as spheres of arbitrary radii.
Reaction of 7·8 in melt: The co-crystals 7·8 (0.179 g, 0.25 mmol) and
sodium carbonate (0.212 g, 2 mmol) were ground together. The mixture
obtained was placed in a test tube and immersed in an oil bath pre-
heated to 1458C (reaction time 5 h). The reaction mixture was worked
up as above. The residue obtained from the ethyl acetate layer was chro-
matographed over silica gel (100–200 mesh) to obtain the mono-p-toluate
9 (eluent, ethyl acetate/light petroleum, 10:90 v/v, 0.064 g, 46%), the diol
7 (eluent, ethyl acetate/light petroleum, 20:80 v/v, 0.080 g, 67%) and the
diester 8 (0.014 g, 14%). The aqueous layer was acidified with hydro-
chloric acid and extracted with ethyl acetate. The organic extract was
dried over anhydrous sodium sulfate and evaporated to obtain p-toluic
acid (0.013 g, 19%).
Figure 4. Molecular overlap of the diester 8 in a crystal of 7·8 and 9 in its
crystal.
Reaction between 7 and 8 in solution: Diol 7 (0.032 g, 0.2 mmol), diester
8 (0.040 g, 0.1 mmol), sodium carbonate (0.085 g, 0.8 mmol) and dry p-
xylene (2.5 mL) were heated (1258C) for 24 h. Excess of dry p-xylene
(2.5 mL) was added and the heating continued for further 36 h. The reac-
tion mixture was concentrated under reduced pressure and the residue
worked up as above. Column chromatographic separation yielded 8
(0.016 g, 40%) and the monotoluate 9 (0.005 g, 9%). Use of dry DMF
(instead of p-xylene) as the solvent for the same reaction resulted pre-
dominantly in the hydrolysis of 8 and formation of 9 was not observed.
in the last two decades with numerous studies devoted to
developing methods of synthesis and potential applica-
tions.[51–60] Solid-state reactions in such multi-component
crystals are rare and constitute a largely unexplored field
due to inherent difficulties in obtaining molecular crystals
wherein the reacting centers are aligned in the proper orien-
tation for the desired reaction. Our investigations illustrate
the importance of systematic studies of group transfer reac-
tions in crystals paving the way for identification of reactive
crystals and co-crystals from the crystal structure database.
This provides a new route to obtain crystals that are capable
of undergoing chemical reactions. Structure–reactivity corre-
lation studies in crystals could also provide methods for
evaluating the chemical stability of functional multicompo-
nent solids.
DSC analysis: DSC curves for the co-crystal 7·8, a 2:1 mixture of 7 and 8,
as well as 7, were recorded on a Mettler Differential Scanning Calorime-
ter. Crystals (ꢂ3 mg) were placed in a sealed aluminium pan (40 mL)
and were analyzed from 45–1758C using an empty pan as the reference.
The heating rate was 58Cminꢀ1 and nitrogen gas was used for purging.
The DSC curve for 7·8 showed a small and broad endothermic hump just
above 1208C suggesting minor structural phase transition before the
melting endotherm observed at 139.58C (Figure S12, the Supporting In-
formation).
X-ray crystallography: Single-crystal X-ray intensity measurements for
co-crystal 7·8 at different temperatures and crystals of 9 at 258C were re-
corded on a Bruker SMART APEX II and SMART APEX I single-crys-
tal X-ray CCD diffractometer, respectively, with graphite-monochromat-
ized (MoKa =0.71073 ꢁ) radiation. The X-ray generator was operated at
50 kV and 30 mA. Diffraction data were collected with a w scan width of
0.38 for 9 (f settings 0, 90, 180, and 2708; the detector position (2q) was
fixed at ꢀ288) and 0.58 for 7·8 (at different settings of f and 2q). The
sample-to-detector distance was fixed at 6.145 cm for 9 and 5.00 cm for
7·8. The X-ray data acquisition was monitored by SMART (for 9)[61] or
APEX II (for 7·8) programs.[62] All the data were corrected for Lorentz-
polarization and absorption effects using SAINT and SADABS programs
integrated in APEX II program package.[62] The structures were solved
by the direct method and refined by full matrix least squares, based on
F2, using SHELX-97[63] (Table S1, the Supporting Information). Molecu-
lar diagrams were generated by using ORTEP-32.[64] Geometrical calcula-
Experimental Section
Crystallization: Naphthalene-2,3-diol (7, 0.32 g, 2 mmol) and its di-p-
methyl benzoate[42] 8 (0.396 g, 1 mmol) were dissolved in ethyl acetate
(10 mL) by warming; light petroleum (90 mL) was added and the result-
ing solution was stored in an open container, at ambient temperature.
Crystallization was complete in a few hours to yield the 2:1 co-crystals
7·8 (0.716 g), M.p. 139–1408C. The structure of the co-crystals 7·8 re-
vealed by single-crystal X-ray diffraction analysis was identical to that re-
trieved from the CSD.[42] Crystallization of a mixture of 7 and 8 in the
molar ratio 1:1 or 1:2 or 1:3 also yielded 2:1 co-crystals 7·8 consistently.
The co-crystals 7·8 could also be obtained by grinding a mixture of the
12872
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 12867 – 12874

Identification of Molecular Crystals
FULL PAPER
tions were performed using SHELXTL and PLATON.[65] All the hydro-
gen atoms in the structure of 7·8 determined at ꢀ1738C were obtained
from difference Fourier and refined isotropically. In structure of 7·8 de-
termined at room temperature (258C), hydrogen atoms bound to naph-
thalene ring were located in the difference Fourier map and refined iso-
tropically. Hydrogen atoms bound to two hydroxyls that is, H1A, H2A
[7] A. E. Troup, H. Mitchner, J. Pharm. Sci. 1964, 53, 375–379.
[8] A. L. Jacobs, A. E. Dilatush, S. Weinstein, J. J. Windheuser, J.
Pharm. Sci. 1966, 55, 893–895.
[9] K. T. Koshy, A. E. Troup, R. N. Duvall, R. N. Conwell, L. L. Shan-
kle, J. Pharm. Sci. 1967, 56, 1117–1121.
[10] R. S. Sardessai, S. Krishnaswamy, M. S. Shashidhar, CrystEngComm
2012, 14, 8010–8016.
[11] M. D. Cohen, G. M. J. Schmidt, J. Chem. Soc. 1964, 1996–2000.
[12] M. D. Cohen, G. M. J. Schmidt, F. I. Sonntag, J. Chem. Soc. 1964,
2000–2013.
ꢀ
were constrained to an ideal geometry (O H=0.82 ꢁ and Uiso(H)=
1.5 Ueq(O)), whereas the H-atoms associated with benzene rings (H18,
H19, H21, H22) and methyl groups (H23A, H23B, H23C) of the diester 8
ꢀ
were constrained to an ideal geometry (C H=0.93 ꢁ and Uiso(H)=
1.2 Ueq(C) for the phenyl H atoms and C-H=0.96 ꢁ and Uiso(H)=
1.5 Ueq(C) for the methyl H-atoms). All the H-atoms in the structure de-
[13] G. M. J. Schmidt, J. Chem. Soc. 1964, 2014–2021.
ˇˇ ´
[14] L. R. MacGillivray, G. S. Papaefstathiou, T. Friscic, D. Varshney,
ꢀ
termined at 1258C were placed in geometrically idealized position (O H
T. D. Hamilton, Top. Curr. Chem. 2005, 248, 201–221.
[15] J. Yang, M. B. Dewal, S. Profeta Jr., M. D. Smith, Y. Li, L. S. Shimi-
zu, J. Am. Chem. Soc. 2008, 130, 612–621.
ꢀ
ꢀ
(OH)=0.82 ꢁ, C H (phenyl)=0.93 ꢁ and C H (methyl)=0.96 ꢁ) and
constrained to ride on their parent atoms (Uiso(H)=1.5 Ueq(O) for the
hydroxyl H atom, Uiso(H)=1.2 Ueq(C) for the phenyl H-atoms and Uiso
(H)=1.5 Ueq(C) for the methyl H atoms). In the structure of the co-crys-
tal 7·8, which was heated to 1368C and cooled to room temperature
(258C) for data measurement, the H atoms, such as H1A, H2A, H19,
H21, H23A, H23B, and H23C were placed in geometrically idealized po-
ˇˇ ´
[16] L. R. Macgillivray, G. S. Papaefstathiou, T. Friscic, T. D. Hamilton,
D.-K. Bucar, Q. Chu, D. B. Varshney, I. G. Georgiev, Acc. Chem.
Res. 2008, 41, 280–291.
[17] M. W. Ghosn, C. Wolf, J. Org. Chem. 2010, 75, 6653–6659.
[18] M. H. Mir, L. L. Koh, G. K. Tan, J. J. Vittal, Angew. Chem. 2010,
122, 400–403; Angew. Chem. Int. Ed. 2010, 49, 390–393.
[19] A. Gavezzotti, M. Simonetta, Nouv. J. Chim. 1978, 2, 69–72.
[20] P. Venugopalan, K. Venkatesan, J. Klausen, E. Novotny-Bregger, C.
Leumann, A. Eschenmoser, J. D. Dunitz, Helv. Chim. Acta 1991, 74,
662–669.
[21] R. Sekiya, K. Kiyo-oka, T. Imakubo, K. Kobayashi, J. Am. Chem.
Soc. 2000, 122, 10282–10288.
[22] K. Tanaka, A. Tomomori, J. L. Scott, Eur. J. Org. Chem. 2003, 2035–
2038.
[23] K. M. Sureshan, T. Murakami, T. Miyasou, Y. Watanabe, J. Am.
Chem. Soc. 2004, 126, 9174–9175.
[24] M. L. Cheney, G. J. McManus, J. A. Perman, Z. Wang, M. J. Zawor-
otko, Cryst. Growth Des. 2007, 7, 616–617.
[25] A. J. Gordon, Tetrahedron 1967, 23, 863–870.
[26] K. Vyas, H. Manohar, K. Venkatesan, J. Phys. Chem. 1990, 94,
6069–6073.
[27] T. Praveen, U. Samanta, T. Das, M. S. Shashidhar, P. Chakrabarti, J.
Am. Chem. Soc. 1998, 120, 3842–3845.
ꢀ
ꢀ
ꢀ
sitions (O H (OH)=0.82 ꢁ, C HACTHNUTRGENNG(U phenyl)=0.93 ꢁ and C HACHUTNGTREN(NNGU methyl)=
0.96 ꢁ) and constrained to ride on their parent atoms (Uiso(H)=
1.5 Ueq(O) for the hydroxyl H atom, Uiso (H)=1.2 Ueq(C) for the phenyl
H-atoms and Uiso (H)=1.5 Ueq(C) for the methyl H atoms), whereas the
remaining H-atoms were located in the difference Fourier map and re-
fined isotropically. The hydroxyl hydrogen atom (H2A) in 9 was located
in the difference Fourier map and refined isotropically; all the other hy-
drogen atoms were placed in geometrically idealized positions (C-H-
ꢀ
ACHTUNGTRENNUNG(phenyl)=0.93 ꢁ and C HACHTUNGTERN(NUGN methyl)=0.96 ꢁ) and constrained to ride on
their parent atoms (Uiso(H)=1.2 Ueq(C) for the phenyl H-atoms and
Uiso(H)=1.5 Ueq(C) for the methyl H atoms). The thermal motion analy-
sis of reactants in co-crystal 7·8 at different temperatures were carried
out using anisotropic displacement parameters (ADPꢃs) through the
THMA11 program integrated in WinGX package.[66] The program calcu-
2
2
lates the quantity DA,B =ZAB ꢀZBA along any inter atomic vector AB
2
(ZAB is the mean square displacement amplitude (MSDA)) of atom A in
the direction of atom B.
Powder X-ray diffraction (PXRD) analysis: PXRD patterns were record-
ed on PANalytical X’PERT PRO instrument at a continuous scanning
rate of 28 2qminꢀ1 using CuKa radiation (40 kV, 30 mA) with the intensity
of the diffracted X-ray being collected at intervals of 0.0178 2q. A nickel
filter was used to remove CuKb radiation. The PXRD pattern of the mix-
ture of co-crystals 7·8 and solid sodium carbonate obtained by grinding
them together revealed presence of diffraction peaks of individual com-
ponents, that is, of 7·8 and sodium carbonate, thus eliminating the possi-
bility of generation of any new crystalline phase during grinding (Figur-
es S14 and S15, the Supporting Information).
[28] C. Murali, M. S. Shashidhar, R. G. Gonnade, M. M. Bhadbhade,
Chem. Eur. J. 2009, 15, 261–269.
[29] M. P. Sarmah, R. G. Gonnade, M. S. Shashidhar, M. M. Bhadbhade,
Chem. Eur. J. 2005, 11, 2103–2110.
[30] S. Krishnaswamy, R. G. Gonnade, M. S. Shashidhar, M. M. Bhadb-
hade, CrystEngComm 2010, 12, 4184–4197.
[31] S. Krishnaswamy, M. S. Shashidhar, M. M. Bhadbhade, CrystEng-
Comm 2011, 13, 3258–3264.
[32] H. B. Bꢂrgi, J. D. Dunitz, Acc. Chem. Res. 1983, 16, 153–161.
[33] D. R. Storm, D. E. Koshland, J. Am. Chem. Soc. 1972, 94, 5815–
5825.
[34] M. L. Bender, Chem. Rev. 1960, 60, 53–113.
[35] I. Weissbuch, L. Leiserowitz, M. Lahav, Top. Curr. Chem. 2005, 259,
123–165.
Acknowledgements
M.I.T. and S.K. thank the CSIR, New Delhi for fellowships. Technical
help from Niharika Pedireddi and helpful discussion with Dr. Mohan M.
Bhadbhade are much appreciated. We thank Mangesh Mahajan for re-
cording the PXRD patterns. This work was supported by the Department
of Science & Technology (DST), New Delhi.
[36] G. R. Desiraju, T. Steiner, The Weak Hydrogen Bond in Structural
Chemistry and Biology, Oxford University Press, Oxford, 1999.
[37] J. Bernstein, M. C. Etter, L. Leiserowitz in Structure Correlation:
The Role of Hydrogen Bonding in Molecular Assemblies, Vol. 1
(Eds.: H. B. Bꢂrgi, J. D. Dunitz), Wiley-VCH, Weinheim, 2008,
pp. 431–507.
[38] M. Nishio, Y. Umezawa, K. Honda, S. Tsuboyama, H. Suezawa,
CrystEngComm 2009, 11, 1757–1788.
[39] R. Boese, T. Clark, A. Gavezzotti, Helv. Chim. Acta 2003, 86, 1085–
1100.
[1] V. Ramamurthy, K. Venkatesan, Chem. Rev. 1987, 87, 433–481.
[2] Y. Ohashi, Reactivity in Molecular Crystals, Wiley-VCH, Weinheim
1993.
[40] M. Chino, K. Nishikawa, T. Tsuchida, R. Sawa, H. Nakamura, K. T.
Nakamura, Y. Muraoka, D. Ikeda, H. Naganawa, T. Sawa, T. Takeu-
chi, J. Antibiot. 1997, 50, 143–146.
[3] H. Kanazawa, Y. Ohashi, Mol. Cryst. Liq. Cryst. 1996, 277, 45–54.
[4] I. Rubinstein, G. Clodic, G. Bolbach, I. Weissbuch, M. Lahav, Chem.
Eur. J. 2008, 14, 10999–11009.
[41] K. Tanaka, D. Fujimoto, A. Altreuther, T. Oeser, H. Irngartinger, F.
Toda, J. Chem. Soc. Perkin Trans. 2 2000, 2115–2120.
[42] S. Nakamatsu, K. Yoshizawa, S. Toyota, F. Toda, I. Matijasic, Org.
Biomol. Chem. 2003, 1, 2231–2234.
[5] A. E. Keating, M. A. Garcia-Garibay, Mol. Supramol. Photochem.
1998, 2, 195–248.
[6] D. Braga, F. Grepioni, Angew. Chem. 2004, 116, 4092–4102; Angew.
Chem. Int. Ed. 2004, 43, 4002–4011.
Chem. Eur. J. 2013, 19, 12867 – 12874
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12873

M. S. Shashidhar et al.
[43] S. A. Talipov, B. T. Ibragimov, K. M. Beketov, K. D. Praliev, T. F.
Aripov, Kristallografiya (Crystallogr. Rep.) 2004, 49, 841–843.
[44] K. Ghosh, M. Datta, R. Frohlich, N. C. Ganguly, J. Mol. Struct. 2005,
737, 201–206.
[55] S. Basavoju, D. Bostrom, S. P. Velaga, Pharm. Res. 2008, 25, 530–
541.
[56] D. Daurio, C. Medina, R. Saw, K. Nagapudi, F. Alvarez-NꢄÇez,
Pharmaceutics 2011, 3, 582–600.
ˇˇ ´
[45] H.-Y. Peng, C. K. Lam, T. C. W. Mak, Z. Cai, W.-T. Ma, Y.-X. Li,
H. N. C. Wong, J. Am. Chem. Soc. 2005, 127, 9603–9611.
[46] J.-B. Yu, S.-W. Chen, G.-R. Zheng, L.-Y. Dai, Acta Crystallogr. Sect.
E: Struct. Rep. 2008, 64, o1653.
[57] T. Friscic, W. Jones, J. Pharm. Pharmacol. 2010, 62, 1547–1559.
[58] T. Itoh, T. Suzuki, T. Uno, M. Kubo, N. Tohnai, M. Miyata, Angew.
Chem. 2011, 123, 2301–2304; Angew. Chem. Int. Ed. 2011, 50, 2253–
2256.
[47] G. A. Wallace, T. D. Gordon, M. E. Hayes, D. B. Konopacki, S. R.
Fix-Stenzel, X. Zhang, P. Grongsaard, K. P. Cusack, L. M. Schaffter,
R. F. Henry, R. H. Stoffel, J. Org. Chem. 2009, 74, 4886–4889.
[48] E.-Y. Xia, J. Sun, R. Yao, C.-G. Yan, Tetrahedron 2010, 66, 3569–
3574.
[59] D. H. Oswald, C. R. Pulham, CrystEngComm. 2008, 10, 1114–1116.
[60] J. Bernstein, J. J. Novoa, R. Boese, S. Cirkel, Chem. Eur. J. 2010, 16,
9047–9055.
[61] Bruker (2003), SADABS Version 2.05, SMART Version 5.631, and
SAINT Version 6.45. Bruker AXS Inc., Madison, Wisconsin, USA.
[62] Bruker (2006). APEX2, SAINT, and SADABS. Bruker AXS Inc.,
Madison, Wisconsin, USA.
[49] S. Varughese, G. R. Desiraju, Cryst. Growth Des. 2010, 10, 4184–
4196.
[50] A. Gavezzotti, Acc. Chem. Res. 1994, 27, 309–314.
[63] G. M. Sheldrick, Acta Cryst. 2008, A64, 112–122.
[64] L. J. Farrugia, J. Appl. Cryst. 1997, 30, 565.
ˇˇ ´
[51] T. Friscic, S. L. Childs, S. A. A. Rizvi, W. Jones, CrystEngComm.
2009, 11, 418–426.
[65] A. L. Spek, J. Appl. Cryst. 2003, 36, 7–13.
[52] R. Kuroda, J. Yoshida, A. Nakamura, S. Nishikiori, CrystEngComm
2009, 11, 427–432.
[66] L. J. Farrugia, J. Appl. Cryst. 1999, 32, 837–838.
ˇˇ ´
[53] T. Friscic, W. Jones, Cryst. Growth Des. 2009, 9, 1621–1637.
[54] D. P. McNamara, S. L. Childs, J. Giordano, A. Iarriccio, J. Cassidy,
M. S. Shet, R. Mannion, E. OꢃDonnell, A. Park, Pharm. Res. 2006,
23, 1888–1897.
Received: March 19, 2013
Revised: May 29, 2013
Published online: August 9, 2013
12874
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 12867 – 12874