Hexagonal host framework of sym-aryloxytriazines stabilised by weak intermolecular interactions

Article abstract of DOI:10.1080/15421400590958601

2,4,6-Tris(4-halophenoxy)-1,3,5-triazine 1 is a convenient C₃ starting material for the self-assembly of hexagonal open frameworks mediated via the halogen...halogen trimer synthon and the π-stacked Piedfort Unit (PU). We examine in this paper crystal structures of 2,4,6-tris(2-iodo-3- pyridyloxy)-1,3,5-triazine 2, 2,4,6-tris (3-iodophenoxy)-1,3,5-triazine 3, 2,4,6-tris(6-methyl-3-pyridyloxy)-1,3,5-triazine 4, and 2,4,6-tris[4-(4′- bromophenyl) phenoxy)]-1,3,5-triazine 5. Triazine 2 forms isostructural 2:1 host·guest adducts (guest = mesitylene, collidine) in the rhombohedral space group R3 such that the host architecture is stabilised by the C _3i-PU and a helix of C-H...N interactions. The crystal structure of 3 is different from its chloro/bromo derivatives signifying the importance of the more polarisable I atom compared to Cl, Br. Pairs of C-H...O and C-H...N hydrogen bonds and C_3i-PU sustain the columnar structure of 3 (space group R3). The PU has pseudo trigonal symmetry in picolinoxy triazine 4 (space group P2₁/n). In contrast to the phenyl derivatives, the extended aryl arms in biphenyl 5 do not adopt a trigonal conformation: two biphenyl groups are oriented parallel that participate in Br...Br and Br...π interactions. We note that 1 and 2 readily form hexagonal host lattices for guest inclusion, while 3, 4, and 5 crystallise in solvent-free form. Thermal measurements (TGA, DSC) indicate that guest release occurs at a higher temperature in the cage type host·guest clathrates compared to the channel inclusion compounds for the same solvent. Statistics from the Cambridge Structural Database using CSD Symmetry show that the phenoxytriazine scaffold is unique among the trigonal molecules for the carry-over of symmetry relation from molecule to crystal. The ease of predicting crystal packing and space group in this family of compounds (1, 2) makes them good candidates for the crystal engineering of host frameworks.

Full text of DOI:10.1080/15421400590958601

This article was downloaded by: [Dalhousie University]
On: 03 October 2014, At: 07:12
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK
Molecular Crystals and Liquid
Crystals
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/gmcl20
Hexagonal Host Framework of
sym-Aryloxytriazines Stabilised
by Weak Intermolecular
Interactions
Binoy K. Saha a , Srinivasulu Aitipamula a , Rahul
Banerjee a , Ashwini Nangia a , Ram K. R. Jetti b
,
Roland Boese b , Chi-Keung Lam c & Thomas C. W.
Mak c
a School of Chemistry, University of Hyderabad ,
Hyderabad, India
b Institut für Anorganische Chemie der Universität
Essen , Essen, Germany
c Department of Chemistry , The Chinese University
of Hong Kong , Shatin, New Territories, Hong Kong
SAR, P. R. China
Published online: 31 Aug 2006.
To cite this article: Binoy K. Saha , Srinivasulu Aitipamula , Rahul Banerjee ,
Ashwini Nangia , Ram K. R. Jetti , Roland Boese , Chi-Keung Lam & Thomas C. W.
Mak (2005) Hexagonal Host Framework of sym-Aryloxytriazines Stabilised by Weak
Intermolecular Interactions, Molecular Crystals and Liquid Crystals, 440:1, 295-316,
DOI: 10.1080/15421400590958601
To link to this article: http://dx.doi.org/10.1080/15421400590958601
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the
information (the “Content”) contained in the publications on our platform.

However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness,
or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the
Content should not be relied upon and should be independently verified with
primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages,
and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the
Content.
This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan,
sub-licensing, systematic supply, or distribution in any form to anyone is
expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions

Mol. Cryst. Liq. Cryst., Vol. 440, pp. 295–316, 2005
Copyright # Taylor & Francis Inc.
ISSN: 1542-1406 print=1563-5287 online
DOI: 10.1080/15421400590958601
Hexagonal Host Framework of sym-Aryloxytriazines
Stabilised by Weak Intermolecular Interactions
Binoy K. Saha
Srinivasulu Aitipamula
Rahul Banerjee
Ashwini Nangia
School of Chemistry, University of Hyderabad, Hyderabad, India
Ram K. R. Jetti
Roland Boese
¨
Institut fu¨r Anorganische Chemie der Universitat Essen,
Essen, Germany
Chi-Keung Lam
Thomas C. W. Mak
Department of Chemistry, The Chinese University of Hong Kong,
Shatin, New Territories, Hong Kong SAR, P. R. China
2,4,6-Tris(4-halophenoxy)-1,3,5-triazine 1 is a convenient C3 starting material for
the self-assembly of hexagonal open frameworks mediated via the halogenꢀ ꢀ ꢀ halo-
gen trimer synthon and the p-stacked Piedfort Unit (PU). We examine in this paper
crystal structures of 2,4,6-tris(2-iodo-3-pyridyloxy)-1,3,5-triazine 2, 2,4,6-tris
(3-iodophenoxy)-1,3,5-triazine 3, 2,4,6-tris(6-methyl-3-pyridyloxy)-1,3,5-triazine
4, and 2,4,6-tris[4-(40-bromophenyl) phenoxy)]-1,3,5-triazine 5. Triazine 2 forms
ꢁ
isostructural 2:1 host guest adducts (guest ¼ mesitylene, collidine) in the rhombo-
ꢀ
hedral space group R3 such that the host architecture is stabilised by the C3i-PU
and a helix of CꢂHꢀꢀꢀN interactions. The crystal structure of 3 is different from
its chloro=bromo derivatives signifying the importance of the more polarisable I
atom compared to Cl, Br. Pairs of CꢂHꢀꢀꢀO and CꢂHꢀꢀꢀN hydrogen bonds and
ꢀ
C3i-PU sustain the columnar structure of 3 (space group R3). The PU has pseudo
trigonal symmetry in picolinoxy triazine 4 (space group P21=n). In contrast to the
phenyl derivatives, the extended aryl arms in biphenyl 5 do not adopt a trigonal
An thanks the Department of Science and Technology for research funding
(SR=S5=OC-02=2002). BKS and SA thank the CSIR, and RB thanks the UGC for fellow-
ship support. The UPE programme of UGC supports UoH. TCWM thanks the Hong
Kong Research Grants Council Earmarked Grant CUHK 402003.
Address correspondence to Ashwini Nangia, School of Chemistry, University of
Hyderabad, Hyderabad 500 046, India. E-mail: ashwini_nangia@rediffmail.com
295

296
B. K. Saha et al.
conformation: two biphenyl groups are oriented parallel that participate in
Brꢀ ꢀ ꢀBr and Brꢀ ꢀ ꢀp interactions. We note that 1 and 2 readily form hexagonal host
lattices for guest inclusion, while 3, 4, and 5 crystallise in solvent-free form.
Thermal measurements (TGA, DSC) indicate that guest release occurs at a higher
ꢁ
temperature in the cage type host guest clathrates compared to the channel
inclusion compounds for the same solvent. Statistics from the Cambridge Struc-
tural Database using CSD Symmetry show that the phenoxytriazine scaffold is
unique among the trigonal molecules for the carry-over of symmetry relation from
molecule to crystal. The ease of predicting crystal packing and space group in this
family of compounds (1, 2) makes them good candidates for the crystal engineering
of host frameworks.
Keywords: DSC; halogen. . .halogen trimer; hexagonal crystal strctures; Piedfort units;
sym-aryloxytriazunes; TGA
INTRODUCTION
The design and self-assembly of host–guest inclusion compounds [1–3]
and interpenetrated network structures [4] is a contemporary theme
in the crystal engineering of organic, inorganic and metal–organic
solids. These studies are of fundamental importance because they lead
to rational strategies for the construction of open frameworks and
porous architectures. The empty void space (Nature abhors a vacuum!)
is filled in one of two possible ways: guest inclusion or interpenetration.
These two modes of space filling in crystals are not mutually exclusive
because examples are known in which they occur in the same structure
[5,6]. Host–guest crystals and interpenetrated networks find applica-
tions in supramolecular materials for NLO and magnetism, gas storage
devices, chiral catalysts, topochemical reactions, and drug delivery
systems [1–4].
A number of strategies are being developed for the crystal engineer-
ing of one dimensional (1D) helices [7], chains and ladders [8], 2D
layers, hexagons, parquet grids and square patterns [9], and 3D archi-
tectures based on the adamantane network [10]. Our recent studies
focus on understanding the tecton [11] ! synthon [12] ! crystal self-
assembly model for the design of target honeycomb structures and
the synthesis of functional solids. Thus, starting from suitably functio-
nalised building blocks with appropriate symmetry, recognition is
mediated via robust, high-probability supramolecular synthons that
in turn lead to the desired crystal architecture. In this paradigm, C3
molecules lead to hexagonal or honeycomb networks and Td tectons
produce diamondoid solids. We discuss in this paper the transform-
ation of some C3 molecules to hexagonal open frameworks, exemplified
in Scheme 1 with the archetypal trimesic acid (TMA). In 2,4,6-tris

Hexagonal Host Framework
297
(4-halophenoxy)-1,3,5-triazine 1 (X-POT), self-assembly is mediated via
the recurring halogen trimer synthon to a hexagonal framework with
˚
an internal diameter of ca. 12 A [13]. These hexagons stack without
offset to give channel type host architectures that include a variety
of aromatic guest species [14]. Cage type lattice inclusion host [13,14]
and polar host–guest assembly [15] have been obtained depending on
the nature of weak halogenꢀ ꢀ ꢀhalogen and CꢂHꢀꢀꢀX (X ¼ halogen, oxy-
gen, nitrogen) interactions. Modification of the 4-halophenoxy group in
1 with different aromatic and heteroaromatic groups (2, 3, 4, 5) and its
influence on intermolecular interaction synthons and crystal packing
are discussed in this paper. Using the recently released software
CSD Symmetry by the Cambridge Crystallographic Data Centre [16],
we show that C3 triazine molecules reproducibly form high-symmetry
hexagonal host lattices, a unique feature of the X-POT series not

298
B. K. Saha et al.
SCHEME 1 (a) Archetypal honeycomb network in trimesic acid (TMA).
(b) Triazine and halogen trimer synthons alternate at the nodes, e.g. as in 1.
(c) The nodes are occupied by triazine molecule 2 or TMA.
shared by other trigonal molecules. Thermal analysis (TGA, DSC) of
some clathrates of halophenoxytriazine 1 reported earlier [13,14] and
iodopyridyloxytriazine 2 are also presented.
RESULTS AND DISCUSSION
Open frameworks may be constructed in inorganic systems (e.g.,
zeolites), with metal atom bonded to organic ligand (coordination
polymers, hybrid structures), and with purely organic molecules.

Hexagonal Host Framework
299
There are several advantages in designing porous frameworks built
from an organic core [3]. (1) Structural diversity can be grafted on
the molecule by functional group addition and=or manipulation using
synthetic reactions and reagents. (2) Supramolecular synthons (recur-
ring hydrogen bond motifs) permit the design and engineering of
diverse host lattices and porous architectures from functionalised
molecules. (3) The nature, size and shape of pores and cavities may
be systematically altered with hydrogen bonding and hydrophobic
interactions. (4) The guest molecule may be used as a template to
induce crystallisation of the adduct.
2,4,6-Tris(4-halophenoxy)-1,3,5-triazine
We have recently shown that a combination of molecular symmetry
and intermolecular interaction geometry may be exploited for the
design of crystal structures in high-symmetry space groups [3,13–
15,17]. Crystallisation of 2,4,6-tris(4-halophenoxy)-1,3,5-triazine 1
(halo ¼ chloro, bromo; Cl-POT, Br-POT) from a variety of solvents
and guest molecules (e.g., benzene, mesitylene, collidine, trinitroben-
zene, hexamethylbenzene, hexachlorobenzene, hexamethylphosphora-
mide, etc.) afforded diffraction-quality crystals in the hexagonal space
group P63=m. A slice of the honeycomb layer consists of alternating
triazine molecule and halogen trimer synthon at the nodes of the hexa-
gon. Such layers may stack without offset to give a continuous channel
structure (P63=m, P63) or a finite cavity such that triazine and halogen
ꢀ
trimers surround the guest (R3). The diameter of the hexagonal cavity
2
˚
˚
in 1 is 12–13 A with a cross-sectional area of 100–110 A . A stereoview
ꢁ
of the hexagonal layer in ClPOT TNB (2:1, TNB ¼ 1,3,5-trinitroben-
zene) is shown in Figure 1. The halogen trimer is a cooperative array
of CꢂXꢀꢀꢀXꢂC interactions having polarisation-induced type II
geometry (ClꢀꢀꢀCl 3.42 A, h1 170.6ꢃ, h2 110.6ꢃ) [18], the difference
˚
between the two angles being ca. 60ꢃ because of crystallographic sym-
metry. In idealised type I approach h1 ¼ h2 ꢄ 150ꢃ (halogen atoms
related across an inversion centre) and in type II geometry h1 ꢄ 180ꢃ,
h2 ꢄ 90ꢃ (halogen atoms related by a screw axis or a glide plane). The
fully ordered structure is ascribed to the numerous heteroatom inter-
actions between the nitro O atom of the guest and the O=Cl atoms on
˚
˚
the host (OꢀꢀꢀO 3.16 A, OꢀꢀꢀCl 3.53 A). Figure 1 shows the stacking of
TNB guest molecules in the 1D channel of Cl-POT at a distance of
˚
3.5 A (¼ c=2), a unique motif of p-stacked electron-deficient aromatic
rings at van der Waals separation in the Cambridge Structural Data-
base [19], an archive of over quarter of a million crystal structures. In
ꢁ
ꢁ
contrast to the channel structure of Cl-POT TNB and Cl-POT HCB

300
B. K. Saha et al.
FIGURE 1 (a) Stereoview of the hexagonal network formed by alternating
ꢁ
triazine 1 and halogen trimer synthons at the nodes in ClPOT TNB (2:1). Note
the Oꢀ ꢀ ꢀCl and Oꢀ ꢀ ꢀO hostꢀ ꢀ ꢀguest interactions. This type of honeycomb net-
work is depicted in Scheme 1(b). (b) Planar conformation of 1,3,5-trinitroben-
zene (phenyl ring and nitro oxygen atoms occupy the same plane) in the 1D
channel of Cl-POT. The centroid-to-centroid distance of adjacent phenyl rings
˚
is 3.5 A (¼ c=2).
ꢁ
(HCB ¼ hexachlorobenzene), the crystal structure of Br-POT HCB
ꢀ
(space group R3) is layered because the host molecules are stac-
ked with offset in adjacent layers. Similar to the TNB guest, the

Hexagonal Host Framework
301
host–guest structure is fully ordered because of numerous weak
˚
heteroatom interactions (OꢀꢀꢀCl 3.53 A). The hexagonal channels of
the Br-POT host lattice can accommodate fullerene guests (C60, C70),
albeit in a disordered arrangement [20].
O-Aromatic/Heteroaromatic Groups on Triazine
Upon observing the structural robustness of the hexagonal framework
in the supramolecular architecture of Cl-POT and Br-POT mediated
by the halogen trimer synthon, we replaced the phenyl ring with the
pyridyl group, as in 2, for two reasons: (1) to find out if these molecules
retain the hexagonal symmetry in their crystal structures; (2) to prof-
itably use the pyridyl group for making CꢂHꢀꢀꢀN interactions [21]
along the channel walls. In preliminary crystallisation experiments
we obtained single crystals with mesitylene and collidine solvents.
ꢁ
ꢁ
X-ray diffraction showed the structures, 2 C9H12 and 2 C8H11N, to
be isomorphous solvent inclusion adducts in the rhombohedral space
ꢀ
ꢁ
group R3 (2:1 H G, Table 1). Triazine 2 self-assembles via the C3i-
˚
˚
PU (3.30 A, 3.28 A) [22] in the ab-plꢃane together with a helix of
ꢃ
˚
˚
CꢂHꢀꢀꢀN hydrogen bonds (2.42 A, 139.5 ; 2.39 A, 137.8 , Table 2) along
the c-axis to generate a hexagonal host framework (Fig. 2). The dis-
ordered guest molecules are sandwiched in the hexagonal cavity with
˚
a diameter of 11 A (excluding van der Waals thickness of the phenyl
ring) between the C3i-PU of host molecules. In contrast to Cl=Br-
POT structures 1 with halogen trimer synthon, crystal structure 2 is
stabilised by van der Waals stacking in the PU. Attempts to model
guest disorder from the diffuse electron density map did not gave a
better R-factor.
The inclusion of solvent=guest molecule in channel=cavity is the
norm in this family of crystal structures studied thus far (Scheme 2).
The following three examples illustrate packing in guest-free crystal
structures.
The crystal structure of meta-chloro=bromo=methyl-phenoxytria-
zine 3 was determined previously [23]. These compounds are isostruc-
ꢀ
tural (space group P3 c1) and contain both C3i-PU and D3-PU motifs.
On the other hand, meta-iodophenoxytriazine 3I, upon crystallisation
ꢀ
from THF=mesitylene mixture, afforded crystals in R 3 space group.
ꢃ
˚
The polarisation-induced Iꢀ ꢀ ꢀI interaction (type II, 3.85 A, 169.9 ,
83.0ꢃ) aggregates the molecules in a helical motif parallel to the c-axis
(Fig. 3). A combination of C3i-PU and three pairs of CꢂHꢀꢀꢀO hydrogen
bonds [23] control the self-assembly of the columnar structure. The
more polarisable iodine atom compared to chlorine and bromine is

302

Hexagonal Host Framework
303
TABLE 2 Geometry of Hydrogen Bonds and Intermolecular Interactions
˚
˚
Complex
Interaction
d, A
D, A
h, deg
ꢁ
2 collidine
CꢂHꢀꢀꢀO
CꢂHꢀꢀꢀN
2.588(3)
2.394(6)
2.742(5)
3.301(1)
3.494(7)
3.278(9)
3.580(8)
4.171(6)
3.88(2)
3.28
3.516(9)
3.322(9)
3.628(10)
4.205(8)
3.95(1)
3.30
3.563(8)
3.497(6)
4.021(6)
3.852(1)
3.422(3)
3.484(4)
3.645(3)
3.577(4)
3.533(3)
3.505(3)
3.647(6)
3.720(6)
3.818(6)
3.923(6)
3.648(6)
3.792(5)
3.891(4)
3.72
140.7(4)
137.8(4)
134.0(4)
138.2(4)
167.5(4)
CꢂHꢀꢀꢀI
Iꢀꢀꢀp
pꢀꢀꢀp
ꢁ
2 mesitylene
CꢂHꢀꢀꢀO
CꢂHꢀꢀꢀN
2.598(4)
2.423(6)
2.782(5)
3.335(1)
142.1(4)
139.5(4)
134.9(4)
138.2(4)
168.2(3)
CꢂHꢀꢀꢀI
Iꢀꢀꢀp
pꢀꢀꢀp
3
4
CꢂHꢀꢀꢀN
CꢂꢀꢀꢀO
CꢂHꢀꢀꢀI
IꢀꢀꢀI
2.688(5)
2.576(5)
3.053(8)
137.6(4)
142.4(2)
149.0(3)
169.9, 82.9
125.9(2)
144.1(2)
148.3(2)
142.0(2)
133.5(2)
133.6(2)
144.9(2)
139.0(2)
148.0(2)
153.4(2)
132.1(3)
141.4(2)
132.2(2)
149.3
CꢂHꢀꢀꢀN
2.674(2)
2.549(2)
2.679(2)
2.661(2)
2.700(2)
2.669(2)
2.707(4)
2.833(3)
2.857(4)
2.924(3)
2.833(1)
2.884(1)
3.080(1)
2.75
CꢂHꢀꢀꢀO
5
CꢂHꢀꢀꢀN
CꢂHꢀꢀꢀO
CꢂHꢀꢀꢀBr
CꢂHꢀꢀꢀp
2.87
3.91
160.1
2.80
3.76
148.2
2.73
3.80
172.2
2.70
3.54
133.8
BrꢀꢀꢀBr
Brꢀꢀꢀp
3.513(1)
164.9(2),
108.8(1)
149.4(1)
163.5(1)
3.503(4)
3.543(4)
responsible for this halogen-directed crystal structure. Triazine rings
˚
are separated by 3.42 and 3.87 A in the PU and connected via CꢂHꢀꢀꢀO
and CꢂHꢀꢀꢀN interactions in a columnar array through the ortho-H
ꢃ
ꢃ
˚
˚
˚
donors of the phenyl ring (2.57 A, 142.4 ; 2.68 A, 137.7 ꢃ). The meta-H
atoms participate in CꢂHꢀꢀꢀI interactions (3.05 A, 149.0 ).

304
B. K. Saha et al.
FIGURE 2 Crystal structure of iodo-pyridinoxy triazine 2. (a) C3i-PU at pꢀ ꢀ ꢀp
˚
distance of 3.3 A. (b) Disordered mesitylene guest sandwiched between the C3i-
PUs. The ortho-iodine atoms project into the cavity. (c) The helix of C–Hꢀ ꢀ ꢀN
interactions stabilizes the structure along the c-axis. (d) A view of the ab-layer
to show the hexagonal cavity of 2 and the included guest (mesitylene) mole-
cules. This type of honeycomb network is depicted in Scheme 1(c).

Hexagonal Host Framework
305
FIGURE 2 Continued.

306
B. K. Saha et al.
SCHEME 2 sym-Aryloxytriazines and their guest inclusion adducts.
The picolinoxy triazine 4 crystallises from EtOAc in P21=n space
group. Inversion related molecules are stacked with slight offset
˚
(0.36 A) as PU motif with pseudo trigonal symmetry in the crystal
structure. With the loss of C3 symmetry, two pyridyl rings are oriented
above the triazine plane while the third pyridiyl group is twisted down
with respect to the N atom (Fig. 4). These PUs are connected bꢃy
ꢃ
˚
˚
CꢂHꢀꢀꢀO and CꢂHꢀꢀꢀN interactions (2.70 A, 133.5 ; 2.68 A, 148.3 )
involving the activated methyl group. The molecules form 2D
pseudo-hexagonal layer motifs, with layers being separated by
˚
˚
3.38 A within the PU and by 3.68 A between PUs.
The effect of extending the aryl group is examined next. In bromo-
biphenyloxy triazine 5, crystallised from CHCl3 with a few drops of
ꢀ
meta-xylene added, the crystal structure is layered (P1) without

Hexagonal Host Framework
307
FIGURE 3 Crystal structure of meta-iodo triazine 3. (a) C3i-PU and C–Hꢀ ꢀ ꢀO
interactions. (b) Helical iodo trimer motif. (c) Columnar structure along the
c-axis stabilised by C–Hꢀ ꢀ ꢀO=N=I interactions.
included solvent. In contrast to previous structures in which the trig-
onal molecule adopts a C3 (or pseudo C3) conformation in the crystal,
the two aryloxy groups in 5 are oriented parallel and the third group
points in a different direction [24], in effect making a tuning fork-like
molecular conformation (Fig. 5). Two bromiꢃne atoms are involved in
ꢃ
˚
type II Brꢀ ꢀ ꢀBr interaction (3.51 A, 164.9 , 108.7 ) and the third

308
B. K. Saha et al.
FIGURE 4 (a) pꢀ ꢀ ꢀp stacking and C–Hꢀ ꢀ ꢀO=N hydrogen bonds in the PU of
picolinoxy triazine 4. The molecule has pseudo C3 symmetry. Note that N atom
of two pyridyl rings is oriented above the triazine plane while the third points
below. (b) The PUs are connected through C–Hꢀ ꢀ ꢀN bonds. The PU in this
structure has pseudo C3 symmetry.

Hexagonal Host Framework
309
FIGURE 5 Crystal structure of bromo-biphenyloxy triazine 5. Note the
absence of C3 symmetry and the tuning fork shape molecular conformation.
Type II Brꢀ ꢀ ꢀBr and C–Hꢀ ꢀ ꢀO interactions are shown.
ꢃ
˚
bromine engages in Brꢀ ꢀ ꢀp interaction (3.50 A, 149.4 ). Miscellaneous
CꢂHꢀꢀꢀX (X ¼ O, N, p) interactions are present in the layered structure
˚
with inter-layer separation of 3.35 and 4.10 A.
CSD Symmetry
Crystallographers are used to seeing symmetrical molecules crystal-
lise in space groups with lower symmetry. The only molecular sym-
metry element that carries over routinely into the crystal structure
is the inversion centre [25]. We note the high occurrence of space
groups with hexagonal=rhombohedral symmetry in C3 symmetry tri-
azine structures. This is not so common. The well-known trigonal mol-
ecule trimesic acid produces a honeycomb chicken wire pattern but in
monoclinic crystal system C2=c. Although matching of molecular and
supramolecular symmetry has been examined in a limited number
of crystal structures [26], the triazine scaffold is the first example of
a family of trigonal molecules that crystallises in the hexagonal crystal
system with high probability. CSD Symmetry [16] was used to search
for the carry-over of molecular symmetry to the crystal for some
common trigonal scaffolds used in crystal engineering: trimesic acid,
trinitrobenzene, phloroglucinol, mesitylene, guanidine, triazine,
phenoxytriazine (Table 3). If a trigonal molecule occupies the Wyckoff
ꢀ
ꢀ
ꢀ
position in a high symmetry space group, e.g., P63=m, P3c1, R3, P3,

310
B. K. Saha et al.
TABLE 3 CSD Symmetry Statistics on Molecular and Crystal Relation in
Some Trigonal Molecules
Compound
type
Structures in high
symmetry space group
Structures in low
symmetry space group
1
6
19
62
2
9
11
58
6
38
35
7
0
94

Hexagonal Host Framework
311
etc. then Z0 will be 0.33, 0.17 or 0.08 depending on whether 1=3, 1=6 or
1=12 molecule is present in the asymmetric unit. An automated search
of Z0 values for the sub-set of trigonal molecules in the CSD was car-
ried out and the frequencies tabulated. Perusal of Table 3 shows that,
in general, trigonal molecules do not crystallise in higher crystallo-
graphic symmetry. For trimesic acid and phloroglucinol, the frequency
for adopting one of the hexagonal symmetry space groups is 1=20 and
2=13, respectively. In the 100 or so guanidinium–sulfonate host–guest
structures [27], there is no occurrence of high-symmetry space groups.
On the other hand, the phenoxytriazine core shows a dramatic
increase in the preference for trigonal molecular symmetry being
related to the hexagonal crystal system: 35=42 hits (¼ 83%). This kind
of success rate is the highest among molecular symmetry elements,
save the inversion centre. The global average for C3 molecules occupy-
ing the 3-fold rotation Wyckoff position in crystal structures is 66%
[28]. Two specific cases of symmetry transformation deserve mention.
Tetraarylmethanes XPh4 and molecules with XPh3 fragment crystal-
lise in unit cells with tetragonal and rhombohedral symmetry. Brock
[29] and Dance [30], in independent studies, ascribe this to the occur-
rence of a herringbone motif that fills the voids through pairs of
diphenyl groups, and due to the stabilisation gained from bimolecular
hexaphenyl embraces, respectively, as the reason for these molecules
adopting high symmetry space groups. We believe that the carry
over of molecular symmetry to the crystal in phenoxytriazine
is because of the stabilisation from the C3=D3-Piedfort Unit supra-
molecular aggregate and the polarisation-induced cooperative halogen
trimer synthon [31].
DSC/TGA Measurements
In addition to crystallographic characterisation of host–guest struc-
tures, thermochemical analysis provides valuable information about
the strength of guest to host interactions. The difference between
the onset temperature for guest release from the host lattice compared
to the normal boiling point of that solvent, Ton – Tbp, is a measure of
how strongly the guest molecule is enclosed in the host (cage) architec-
ture [32]. Examination of the host framework suggests that the guest
molecule can escape from the channel type structure (P63=m) with
relative ease because there are no strong hostꢀ ꢀ ꢀguest interactions.
On the other hand, when the guest molecule is sandwiched between
ꢀ
the aromatic groups of the cage architecture (R3), evolution of the
guest will require severe fragmentation of the framework and so guest
release should occur at a higher temperature or at the melting stage.

312
B. K. Saha et al.
TABLE 4 Thermal Analysis (TGA=DSC) on Inclusion Compounds of Triazines
Calculated
weight
loss (%)
B.p.
Inclusion
compounda
Observed weight
loss from TGA (%)
Guest release of Guest DH (J gꢂ1
)
T
on=ꢃC
Tb=ꢃC
calibrated
ꢁ
BrPOT MES
16.6
17.3
1st step 8.4
2nd step 10.3
7.6
20.2
20.4
23.9
139
145
127
187
200
197
162–164
171–172
240–243
99.37
84.08
14.16
51.56
70.81
75.69
ꢁ
BrPOT CLN
ꢁ
BrPOT MNP
ꢁ
IPyOT MES
8.1
8.2
162–164
171–172
ꢁ
IPyOT CLN
7.4
aSee Scheme 2 for abbreviated names of host and guest.
TGA=DSC confirms this to be the case. Although the hostꢀ ꢀ ꢀhost and
hostꢀ ꢀ ꢀguest intermolecular interactions are weak in both channel
and cage type hosts, guest release occurs at a higher temperature from
the latter framework in TGA measurements (Table 4, Fig. 6). Mesityl-
ene and collidine escape from the channel structure of Br-POT 1 at
about their boiling point and the observed weight loss is in good agree-
ment with that calculated from the host:guest stoichiometry determ-
ined from the X-ray crystal structure. While collidine and mesitylene
are lost in a single endothermic step over a narrow range of ca.
20ꢃC, 1-methylnaphthalene evolution occurs in two slow steps,
120–150ꢃC and 150–200ꢃC. Since the MNP guest molecules are
disordered in the host channel, the stepwise TGA behaviour is difficult
to explain. DSC of channel inclusion adducts shows two endotherms,
first the release of solvent=guest and then melting of the host com-
pound. In contrast to the channel structures, release of collidine and
mesitylene guest from the cage lattice of iodo-pyridyloxy triazine
2 occurs at 50–70ꢃC higher temperatures. This shows that the host
architecture is quite strong and thermally stable even though it
is assembled with weak CꢂHꢀꢀꢀO=N and van der Waals interactions.
Some of the highest recorded values of Ton – Tbp are 320ꢃC and
370ꢃC for the release of CH4 and CF4 from the interstitial void of
calix[4]arene host lattice [33]. In the case of solid guest (e.g., 1,3,5-
tribromobenzene), the compound decomposes in a single step at the
melting point of the adduct crystal.
CONCLUSIONS AND FUTURE OUTLOOK
We are currently working on two different C3 symmetry molecules for
the construction of hexagonal and honeycomb layer architectures.

Hexagonal Host Framework
313
FIGURE 6 TGA and DSC trace of inclusion adducts of Cl-POT=Br-POT ¼ 1
and I-PyOT ¼ 2. See scheme 2 for guest abbreviation. The weight loss of guest
in TGA matches with the endotherm in DSC (a-e). There is no release of solid
guest in thermogram (f) and decomposition of the host occurs at its melting
point.
Some recent advances on the triazine scaffold are discussed in this
paper. Results with cocrystals of 1,3-cis,5-cis-cyclohexane tricarboxylic
acid (CTA) are published elsewhere [34]. Our studies with these two
C3 molecules for building hexagonal network structures prompt sev-
eral related questions. (1) Why do triazine molecules adopt hexagonal
crystal symmetry whereas TMA=CTA crystallise in monoclinic=
triclinic systems? (2) Why do triazines prefer guest inclusion whilst
the tricarboxylic acid adducts fill the porous network through

314
B. K. Saha et al.
interpenetration? (3) Is it possible to induce a guest-inclusion system
towards interpenetration, and vice versa? (4) What covalent and
non-covalent spacers should be used to modularly tune the pore diam-
eter? We hope to address some of these issues in our ongoing studies
on the crystal engineering of host–guest adducts and interpenetrated
network structures.
EXPERIMENTAL
Synthesis
Equimolar quantities of ArOH and KOH (4 mmol) were dissolved in
THF and stirred for 30 min at room temperature. The reaction mixture
was cooled to 0ꢃC and then slowly 1 mmol of cyanuric chloride was
added and stirred for 1 h at 0ꢃC. After continued stirring at room tem-
perature for 48 hours, the reaction mixture was poured into crushed
ice. The heavy white precipitate was filtered by vacuum suction and
washed with methanol, dried and purified by column chromatography.
All compounds showed satisfactory NMR and IR spectra.
2,4,6-Tris(2-iodo-3-pyridyloxy)-1,3,5-triazine 2: 1H-NMR (CDCl3,
200 MHz) d 8.27 (m, 3H), 7.38 (m, 3H), 7.28 (m, 3H). 13C-NMR d 173,
149, 148, 130, 124, 115. 2-Iodo-3-pyridinol was prepared by the iodina-
tion of 3-hydroxypyridine with NaOCl=NaI [35]. Instead of the reported
6-iodo-3-pyridinol isomer, 2-iodo-3-pyridinol was obtained by oxidation
at the most activated C atom between the OH group and N atom.
2,4,6-Tris(3-iodophenoxy)-1,3,5-triazine 3: 1H-NMR (CDCl3, 200
MHz) d 7.59 (m, 3H), 7.48 (s, 3H), 7.12 (m, 6H). 13C-NMR d 173,
151, 135, 131, 130, 121, 93.
2,4,6-Tris(6-methyl-3-pyridyloxy)-1,3,5-triazine 4: 1H-NMR (CD3
OD, 200 MHz) d 8.33 (s, 3H), 7.57 (d, J ¼ 8 Hz, 3H), 7.30 (d, J ¼ 8 Hz,
3H), 2.50 (s, 9H). 13C-NMR d 173, 155, 146, 142, 129, 123, 23.
2,4,6-Tris[4-(40-bromophenyl)phenoxy)]-1,3,5-triazine 5: 1H-NMR
(CDCl3, 200 MHz) d 7.52 (d, J ¼ 8 Hz, 6H), 7.51 (d, J ¼ 8 Hz, 6H),
7.35 (d, J ¼ 8 Hz, 6H), 7.20 (d, J ¼ 8 Hz, 6H). 13C-NMR d 168, 146,
134, 133, 127, 123, 122, 116.
Crystallography
ꢁ
Reflections were collected for the single crystal of 2 collidine,
ꢁ
2 mesitylene, 3 and 5 on a Siemens SMART CCD area detector system
at 123(2) K, 183(2) K, 223(2) K and 183(2) K, respectively (Mo-Ka radi-
˚
ation, k ¼ 0.71073 A). Data for 4 were collected on a Bruker SMART
CCD 1 K area detector at 293(2) K (Mo-Ka radiation). Empirical

Hexagonal Host Framework
315
absorption corrections using SADABS [36–38] were applied for 2, 3
and 5. Structure solution and refinement were performed with
SHELXS-97 and SHELXL-97 packages [39,40]. Hydrogen atoms were
generated with idealised geometries and isotropically refined using
the riding model. The guest molecules were refined isotropically.
Refinement of coordinates and anisotropic thermal parameters of
non-hydrogen atoms was carried out by the full-matrix least-squares
method. The final R indices and other crystallographic parameters
are listed in Table 1. All CꢂH distances are neutron normalised to
˚
1.083 A for the intermolecular contacts in Table 2.
Thermal Analysis
Differential scanning calorimetry (DSC) was performed on a Mettler
Toledo DSC 822e instrument and thermogravimetry (TGA) on a
Mettler Toledo TGA=SDTA 851e, and data manipulated in the STAR
software module. Crystals taken from the mother liquor were blotted
dry on a filter paper and placed in open alumina pans for TGA
measurements and in crimped aluminium but vented sample pans
for DSC measurements. Sample size in each case was 5–7 mg. The
temperature range was 30–300ꢃC at a heating rate of 10ꢃC=min. The
samples were purged with a stream of dry nitrogen at a flow rate of
150 ml=min for DSC and 50 ml=min for TGA.
REFERENCES
[1] MacNicol, D. D., Toda, F., & Bishop, R. (Eds.) (1996). Comprehensive Supramolecu-
lar Chemistry, Solid-State Supramolecular Chemistry: Crystal Engineering,
Pergamon: Oxford, Vol. 6.
[2] Bishop, R. (1999). Synlett, 9, 1351–1358.
[3] Nangia, A. (2001). Current Opin. Solid State Mater. Sci., 5, 115–122.
[4] Batten, S. R. & Robson, R. (1998). Angew. Chem. Int. Ed., 37, 1461–1494.
[5] Herbstein, F. H., Kapon, M., & Reisner, G. M. (1985). Acta Crystallogr. Sect. B, 41,
348–354.
[6] Evans, O. R., Xiong, R. G., Wang, Z., Wong, G. K., & Lin, W. (1999). Angew. Chem.
Int. Ed., 38, 536–538.
[7] Martz, J., Graf, E., Cian, A. D., & Hosseini, M. W. (2003). In: Crystal Design: Struc-
ture and Function; Perspectives in Supramolecular Chemistry, Desiraju, G. R.
(Eds), Wiley: Chichester, Vol. 6, 177–209.
[8] Nguyen, V. T., Ahn, P. D., Bishop, R., Scudder, M. L., & Craig, D. C. (2001). Eur. J.
Org. Chem., 4489–4499.
[9] Moultan, B. & Zaworotko, M. J. (2001). Chem. Rev., 101, 1629–1658.
[10] Evans, O. R. & Lin, W. (2002). Acc. Chem. Res., 35, 511–522.
[11] Wuest, J. D. (1995). In: Mesomolecules: From Molecules to Materials, Mendenhall,
G. D., Greenberg, A., & Liebmen, J. F. (Eds.), Chapman & Hall: New York, Vol. 1,
107–131.

316
B. K. Saha et al.
[12] Desiraju, G. R. (1995). Angew. Chem. Int. Ed. Engl., 34, 2311–2327.
[13] Jetti, R. K. R., Xue, F., Mak, T. C. W., & Nangia, A. (1999). Cryst. Eng., 2, 215–224.
[14] Jetti, R. K. R., Thallapally, P. K., Xue, F., Mak, T. C. W., & Nangia, A. (2000). Tetra-
hedron, 56, 6707– 6719.
[15] Jetti, R. K. R., Nangia, A., Xue, F., & Mak, T. C. W. (2001). Chem. Commun.,
919–920.
[16] Yao, J. W., Cole, J. C., Pidcock, E., Allen, F. H., Howard, J. A. K., & Motherwell, W.
D. S. (2002). Acta Crystallogr. Sect. B, 58, 640–646.
[17] Jetti, R. K. R., Thallapally, P. K., Nangia, A., Lam, C-K., & Mak, T. C. W. (2002).
Chem. Commun., 952–953.
[18] Bosch, E. & Barnes, C. L. (2002). Cryst. Growth Des., 2, 299–302.
[19] Allen, F. H. (2002). Acta Crystallogr. Sect. B, 58, 380–388.
[20] Su¨ss, H. I., Lutz, M., & Hulliger, J. (2002). Cryst. Eng. Comm., 4, 610–612.
[21] Desiraju, G. R. & Steiner, T. (1999). The Weak Hydrogen Bond in Structural
Chemistry and Biology, OUP: Oxford.
[22] The Piedfort Unit (PU) concept for hexahost design was first introduced by
MacNicol. Jessiman, A. S., MacNicol, D. D., Malinson, P. R., & Vallance, I.
(1990). J. Chem. Soc. Chem. Commun., 1619–1621. See also Ref. 1, 421–464.
[23] Thalladi, V. R., Brasselet, S., Weiss, H.-C., Blaser, D., Katz, A. K., Carrell, H. L.,
Boese, R., Zyss, J., Nangia, A., & Desiraju, G. R. (1998). J. Am. Chem. Soc., 120,
2563–2577.
ꢁ
ꢁ
ꢁ
ꢁ
[24] Fabian, L., Bombicz, P., Czugler, M., Kalman, A., Weber, E., & Hecker, M. (1999).
Supramol. Chem., 11, 151–167.
[25] Kitaigorodskii, A. I. (1973). Molecular Crystals and Molecules, Academic Press:
New York.
[26] Thallapally, P. K., Chakraborty, K., Katz, A. K., Carrell, H. L., Kotha, S., &
Desiraju, G. R. (2001). Cryst. Eng. Comm., 31, 1–3.
[27] Holman, K. T., Pivovar, A. M., Swift, J. A., & Ward, M. D. (2001). Acc. Chem. Res.,
34, 107–118.
[28] Pidcock, E., Motherwell, W. D. S., & Cole, J. C. (2003). Acta Crystallogr. Sect. B, 59,
634–640.
[29] Lloyd, M. A. & Brock, C. P. (1997). Acta Crystallogr. Sect. B, 53, 780–786.
[30] Dance, I. & Scudder, M. (2001). New J. Chem., 25, 1510–1515.
[31] Saha, B. K., Jetti, R. K. R., Reddy, L. S., Aitipamula, S., & Nangia, A. (2005).
Cryst. Growth Des., 5, 887–899.
[32] Nassimbeni, L. R. (2003). Acc. Chem. Res., 36, 631–637.
[33] Atwood, J. L., Barbour, L., & Jerga, A. (2002). Science, 296, 2367–2369.
[34] Bhogala, B. R. & Nangia, A. (2003). Cryst. Growth Des., 3, 547–554.
[35] Edgar, K. J. & Falling, S. N. (1990). J. Org. Chem., 55, 5287–5291.
[36] Higashu, T. (1995). ABSCOR: An Empirical Absorption Correction Based on
Fourier Coefficient Fitting, Rigaku Corporation: Tokyo.
[37] Sheldrick, G. M. (1996). SADABS: Program for Empirical Absorption of Area
¨
Detector Data, University of Gottingen: Germany.
[38] Kopfmann, G. & Huber, R. (1968). Acta Crystallogr. Sect. A, 24, 348–351.
[39] Sheldrick, G. M. (1997). SHELXS-97: Program for the Solution of Crystal
¨
Structures, University of Gottingen: Germany.
[40] Sheldrick, G. M. (1997). SHELXL-97: Program for the Refinement of Crystal
structures, University of Go¨ttingen: Germany.