Synthesis of 2-pyridone-fused 2,2′-bipyridine derivatives. An unexpectedly complex solid state structure of 3,6-dimethyl-9H-4,5,9- triazaphenanthren-10-one

Article abstract of DOI:10.1039/b417485b

2-Pyridone-fused 2,2′-bipyridine derivatives 1a and 1b were synthesised. X-Ray diffraction analysis of 1b revealed a highly complex solid state structure with a disordered molecule imbedded in a channel structure formed by a centrosymmetric lattice of hexagonally packed, hydrogen bonded columns. The columns are assembled from three symmetry independent molecules. Dimerisation of the self-complementary cis-amide hydrogen bond motif is overridden by the fulfilment of the proton coordination ability of the phenanthroline nitrogens in accordance with Etter's rules of hydrogen bond priorities. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b417485b

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
ꢀ
Synthesis of 2-pyridone-fused 2,2 -bipyridine derivatives.
An unexpectedly complex solid state structure of
3
,6-dimethyl-9H-4,5,9-triazaphenanthren-10-one
a
a
b
c
Stef a´ n J o´ nsson, Carlos Solano Arribas, Ola F. Wendt, Jay S. Siegel and
a
Kenneth W a¨ rnmark*
a
Organic Chemistry 1, Dept. of Chemistry, Lund University, P.O. Box 124, SE-22100 Lund,
Sweden. E-mail: kenneth.warnmark@orgk1.lu.se; Fax: +46 46 2224119; Tel: +46 46 2228217
b
Inorganic Chemistry, Dept. of Chemistry, Lund University, P.O. Box 124, SE-22100 Lund,
Sweden
c
Organisch-chemisches Institut, Universit a¨ t Z u¨ rich, Z u¨ rich, CH-8057, Switzerland
Received 18th November 2004, Accepted 2nd February 2005
First published as an Advance Article on the web 22nd February 2005
ꢀ
2
-Pyridone-fused 2,2 -bipyridine derivatives 1a and 1b were synthesised. X-Ray diffraction analysis of 1b revealed a
highly complex solid state structure with a disordered molecule imbedded in a channel structure formed by a
centrosymmetric lattice of hexagonally packed, hydrogen bonded columns. The columns are assembled from three
symmetry independent molecules. Dimerisation of the self-complementary cis-amide hydrogen bond motif is
overridden by the fulfilment of the proton coordination ability of the phenanthroline nitrogens in accordance with
Etter’s rules of hydrogen bond priorities.
Introduction
Janus molecule owing to its two different faces, one of hydrogen
12
bonding, and one of metal coordination. Upon complexation
with metals, it should form solid state networks of alternating
hydrogen bonding and metal coordination, the network dimen-
sions being governed by the coordination properties of the metal.
Small and rigid, highly functionalised heterocyclic molecules are
of interest as scaffolds for the construction of superstructures.
Heterocycles provide hydrogen bonding and metal coordinating
functionalities, offering strong and directional intermolecular
1
interactions. Appropriate positioning of those in a molecular
skeleton will ideally lead to the formation of an aggregate of the
2
desired structure. For solution based aggregates, careful design
is most often enough for success, given that entropy factors and
solvent interactions are not neglected. In the solid state however,
controlling or predicting the form of aggregation is much more
difficult, since nucleation kinetics and packing requirements can
severely perturb the thermodynamics of solution assembly. In
addition, close packing brings the molecules in close contact,
rendering weak intermolecular forces, not operating in solution,
3
much more important in the solid state. Arrays of bi- and
tridentate ligands have been synthesised and metal complexes
thereof form helices, rosettes, grids and ladders etc. in solution
4
and/or solid state. Numerous multicomponent assemblies of
5
hydrogen bonding subunits have also been reported. Molecular
aggregates based on both types of interactions include mainly
6
solid state network structures.
Fig. 1 (a) Janus molecules 1a (R = H) and 1b (R = Me). (b) The
anticipated cis-amide back-to-back hydrogen bonding.
We are interested in the application of the self-complementary
cis-amide hydrogen bonding motif for the assembly of functional
7
aggregates from simpler components. We wanted to integrate
Looking at the structure of the free ligand 1 (Fig. 1a), one
would anticipate it to aggregate to a back-to-back dimer due to
the presence of the self-complementary 2-pyridone hydrogen
bonding motif (Fig. 1b). However, to our surprise, the X-
ray structural analysis of 1b revealed a far more complicated
structure. We now present the synthesis of ligands 1a and 1b
as well as the surprisingly complex crystal structure of ligand
this motif into a metal coordinating ligand, and investigate
both the solution and solid state behaviour of the free ligand
as well as its metal complexes. We chose the 2-pyridone
moiety since it has been widely applied in the construction
8
of molecular assemblies in solution and solid state, in purely
9
10
ꢀ
organic structures as well as coordination complexes. 2,2 -
Bipyridine was chosen as the ligand framework, being a rigid,
bidentate ligand, so no competition should be expected from
the cis-amide moiety in terms of coordination by a metal.
Complexes with phenanthroline (phen) and terpyridine (tpy)
based ligands with integrated hydrogen bonding functionalities
have been utilised by other groups in constructing solid state
1
b, featuring self inclusion of a disordered guest molecule in a
centrosymmetric host lattice, and the complete absence of the
back-to-back dimer.
Results and discussion
11
networks.
Ligand 1a was the original target molecule. It was synthesised
ꢀ
13
We have thus fused the 2-pyridone and 2,2 -bipyridine moi-
in two steps, starting from chloropyridines 2a and 3a as shown
eties to obtain 9H-4,5,9-triazaphenanthren-10-one (1, Fig. 1), a
in Scheme 1. Despite obvious yield limitations, the statistical
9
9 6
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 9 9 6 – 1 0 0 1
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5

View Article Online
Scheme 1
ꢀ
Ullman coupling was chosen for the synthesis of 2,2 -bipyridine
derivative 4a, since initial efforts to prepare organometallic
intermediates required for cross coupling were not fruitful. Thus,
4
a was obtained in 39% yield using copper powder in DMF. In
the following reductive cyclisation, iron–acetic acid gave the best
results out of the methods investigated, affording 1a in 39% yield.
Ligand 1a proved extremely insoluble in even the most polar
aprotic solvents, making both purification and characterisation
difficult. Furthermore, in all attempts to grow crystals, it
crystallised as clusters of microscopic needles, and no single
crystal of appropriate dimensions could be obtained.
Fig. 2 DIAMOND drawing with atomic numbering of 1b. Molecule
D is shown in only one of its two possible orientations. It was refined
isotropically and, as seen, the disorder gave rise to high isotropic
displacement parameters. Hydrogens are omitted for clarity.
suggesting a polycyclic aromatic structure. Considering the lack
of centrosymmetry in 1b we concluded this to be a disordered,
half molecule of 1b. Attempts to resolve the disorder, either by
solving the structure in the less symmetrical P1 space group or
by low temperature data collection were unsuccessful.
Thus, in the hope for more amenable solubility and crys-
tallisation properties, neocuproine analogue 1b was synthesised
following the same route as for 1a (Scheme 1). Esterification of
6
-methyl-2-chloronicotinic acid to produce 2b resulted in low
The three well resolved molecules arrange into infinite
columns propagating along the [111] axis, held together by four
hydrogen bonds. Three of those are bifurcated, with the phen
type nitrogen pairs acting as bidentate acceptors. The column
structure is presented in Fig. 3 and the relevant hydrogen bond
distances and angles are listed in Table 1. Molecule B serves as
a bridge between p-stacked sandwiches of molecules A and C,
respectively, fully utilising its hydrogen bonding capabilities.
The carbonyl oxygen of molecule A (O1A) forms weak,
intercolumnar hydrogen bonds to aromatic C–H donors C2B
and C2C. These hydrogen bonds provide direct contact points
to four out of six neighbouring columns. As the columns pack
together, channel structures are formed along the a axis (Fig. 4).
Inside these channels, the disordered guest molecules line up
in a plane parallel to the a axis. This was initially thought to
yields using a range of methods, including the one used to make
2
a. However, treatment with POCl
3
followed by MeOH afforded
2b in an acceptable 69% yield. Chloropyridine 3b was synthesised
14
in two steps, by nitration of 2-amino-6-picoline, followed by
15
diazotisation in hydrochloric acid. Finally, 1b was obtained via
Ullman coupling in 44% yield followed by reductive cyclisation
in 31% yield, using the same procedures as for 1a.
The two methyl substituents greatly improved the solubility,
and rather large, thick, platelike single crystals were grown by
slow diffusion of diethyl ether into a solution of 1b in a 4 : 1
mixture of chloroform and methanol.
The crystal structure was solved in the triclinic, centrosym-
¯
metric space group P1, with seven molecules in the unit cell (Z).
The asymmetric unit contained three well resolved molecules
16
(
A, B and C, Fig. 2). After all the atoms in those three molecules
be a clathrate. However, molecular modelling of 1b inside the
rigid channel with the disordered part of the crystal structure
removed, revealed that random distribution of guest molecules
along the channel is highly unlikely. Two symmetry-equivalent
had been located, the final difference Fourier map contained a
number of peaks from which no reasonable molecular structure
could be discerned. All the peaks were in the same plane,
Fig. 3 Crystal structure of 1b. Ball and stick model showing the arrangement of molecules A, B and C into a column structure along the [111]
ꢀ
axis. Hydrogen bonds are shown as broken lines. Intercolumnar hydrogen bonds are orange. The prime symbols ( ) refer to different neighbouring
columns.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 9 9 6 – 1 0 0 1
9 9 7

View Article Online
Table 1 Hydrogen bond geometry in the infinite column structure
a
a
D · · · A/A˚
D–H · · · A/◦
Donor atom
Acceptor atom
N10A
N10A
C16B
C16B
N10B
N10B
N10C
N4B
N5B
N5A
N4A
N4C
N5C
O1B
O1A
O1A
3.200(35)
3.057(33)
3.551(64)
3.548(43)
3.298(38)
3.011(21)
2.845(11)
3.592(73)
3.383(19)
141.19(4)
144.85(5)
179.23(3)
119.83(3)
146.28(4)
141.28(3)
164.52(5)
155.13(3)
127.59(2)
b
C2Cb
C2B
a
b
For numbering see Fig. 2. Intercolumnar hydrogen bonds.
Fig. 4 Crystal structure of 1b. A space filling model of a hexagonally
packed array of seven columns viewed along the a axis. Each column
has a representative colour. Two channels formed by the packing of the
columns are shown, one is left empty for clarity and the other is filled
with a guest molecule (grey).
Fig. 5 Crystal structure of 1b. Modelling of the guest in the rigid, empty
channel structure. (a) Overlay of both possible orientations of molecule
D according to molecular modelling (grey, C; red, O; blue, N) with the
15 carbons representing the disordered part of the crystal structure in
the initial refinement (green and orange). The green atoms are created
by centrosymmetry from the orange ones, while the two orientations of
molecule D were modelled independently of each other. (b) The modelled
guest inside the channel structure. The modelled structure is shown as a
stick model while the crystal structure is a wire model. Possible hydrogen
bonds are shown as pink dotted lines with the D · · · A distance.
orientations of 1b (molecule D) are stabilised by interactions
with the interior surface of the channel, making all other
positions unfavourable. Overlaying 1b in these orientations with
the disordered part of the crystal structure reveals an almost
perfect match. An inversion point in the centre of the disordered
1
2
structure, at ( ,1,0) interrelates the two modelled orientations
(
Fig. 5a). It was thus concluded that the disordered structure
Table 2 Some hydrogen bond distances between one of the two possible
orientations of molecule D (guest) and the interior surface of the channel
represents two defined spatial orientations of 1b at a specific
position in the unit cell, each with an occupancy factor of 0.5.
Two hydrogen bond acceptors (A), O1B and O1C, and one
donor (D), C15C, line the interior surface of the channel. The
potential interactions between these and the modelled guest
structure are illustrated in Fig. 5b. To complete the structure
refinement, the atomic coordinates of the modelled molecule
D were applied to the crystal structure as a rigid group and
the position of the rigid molecule was refined. The prediction
of the molecular modelling was thereby confirmed in the crystal
structure refinement. Interestingly, the occupancy factor for each
of the two orientations of molecule D refined to 0.48, implying
that 4% of the guest positions are empty (or perhaps filled with
solvent). Some hydrogen bond interactions between molecule D
and the surrounding lattice molecules are listed in Table 2.
Thus, despite the lack of symmetry in 1b, molecule D complies
with the centrosymmetry of the unit cell by adopting two
centrosymmetrically related orientations. So the high number
a
a
D · · · A/A˚
Guest atom
Atom in channel structure
C16D
C15D
C1D
N10D
O1D
O1C (x, y, −1 + z)
O1C (−1 + x, y, −1 + z)
O1C (1 − x, 2 − y, 1 − z)
O1B (−1 + x, y, z)
C15C (x, y, z)
3.876
3.460
3.544
2.965
3.558
a
For numbering see Fig. 2.
four possible halves of molecule D, all of which are equally
represented by the disordered part of the crystal structure.
¯
It appears that this unexpected crystal packing mode (P1, Z =
ꢀ
17
7, Z = 3.5) is quite an unusual one. A search in the Cambridge
18
Structural Database (CSD) revealed that only two examples of
19
this type of packing have been reported earlier. It is interesting
to note that in both of these cases, like in the present one, the
lack of centrosymmetry in the molecular structure is made up
ꢀ
of molecules in the asymmetric unit (Z = 3.5) does not fully
describe the complexity of the crystal structure, since there are
9
9 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 9 9 6 – 1 0 0 1

View Article Online
for in the crystal packing by disordering one molecular unit out
of four over two orientations at an inversion centre.
donor to a phen type nitrogen pair in an adjacent molecule of
1b. The crystallisation process results from a very subtle balance
of interactions. One favourable interaction can imply another
unfavourable one. This fact is manifest in the presented crystal
structure, where the strongest single interaction is sacrificed to
allow the interaction potential of the whole structure to be
satisfied.
Bifurcated hydrogen bonds to the two phen nitrogens are
prominent in the crystal structure of 1b. This observation
18
prompted a CSD search on the hydrogen bonding behaviour
of phen derivatives in the solid state. Out of the 54 purely
organic, nonionic 3D structures obtained, 36 included hydrogen
20
bond interactions with the phen nitrogens. 26 of these were
solvates (H-bond donor being an OH or acidic CH of a solvent
molecule from the crystallisation medium). While all of the
In conclusion, we have presented the synthesis of 2-pyridone
ꢀ
fused 2,2 -bipyridine derivatives, two faced (Janus) molecules
with one potential hydrogen bonding face and one poten-
tial metal coordinating face. Our results highlight the 1,10-
phenanthroline moiety as an efficient proton chelator, owing
to its preorganised nitrogen pair. This property was the main
driving force for the formation of an unexpectedly complex solid
state structure of the free ligand 1b, overriding the possible
dimerisation of the self-complementary cis-amide hydrogen
bond motif. Possible reasons for this could include the relatively
high acidity of the 2-pyridone type NH proton which makes it
a preferred hydrogen bond donor compared to solvent hydroxyl
groups. Further, we have presented an example of an unusual
crystal packing mode, containing three and a half molecules in
the asymmetric unit and featuring self-inclusion in a channel
type structure.
1
8 structures without hydrogen bonds to the phen nitrogens
were devoid of donors in the molecular structure, this was true
for only 17 of the solvate structures. This means that in only
about 50 percent of the cases do donor free phen derivatives
incorporate a donor containing solvent molecule into the crystal
to fulfil their proton coordination ability, although this certainly
relies on crystallisation conditions and medium. Further, of the
3
6 hydrogen bonded structures, 31 (86%) contained bifurcated
hydrogen bonds to the phen nitrogen pair and 7 (19%) non-
bifurcated ones. Two structures contained examples of both.
This ratio was maintained in both solvate and non-solvate
structures. Some recently published structures of phen–urea
21
cocrystals support the notion that a high (>1) donor–acceptor
ratio increases the probability of non-bifurcated hydrogen bonds
to the phen nitrogens.
Experimental
Due to an inappropriate bite angle, the two nitrogens rarely
contribute equally to these bifurcated interactions, neither do
they in our structure presented here. While the literature on
hydrogen bonding behaviour of phen derivatives is scarce, one
General methods
1
3
2
6
-Chloronicotinic acid methyl ester 2a and 2-chloro-3-nitro-
14,15
-picoline 3b
were synthesised according to literature proce-
22
spectroscopic study shows that they form hydrogen bonds with
phenols, that are 3–5 times stronger than expected from the
dures. All other chemicals were used as received from commer-
cial sources unless otherwise stated. TLC analyses (Merck 60
correlation with pK in the pyridine series. Briefly, both the
a
F
254 sheets) were visualised under UV light (254 nm). Column
chromatography (CC) was performed with silica gel (Matrex
.063–0.200 mm): column diameter 5 cm, length 12–18 cm and
frequency shift in the phenol O–H stretching band in the IR
spectrum, and the −DS term of the free energy of hydrogen
bond formation, are smaller than for pyridines, suggesting either
two low-energy positions for the proton in the field of the two
nitrogen lone pairs, with rapid switching, or a broad, shallow
0
fraction sizes 30 mL. Melting points were determined in capillary
tubes, using a Sanyo GallenKamp melting point apparatus and
are uncorrected. Infrared spectra were recorded on a Shimadzu
FTIR-8300 spectrometer. NMR spectra were recorded on a
Bruker DRX400 NMR spectrometer. Chemical shifts are given
in ppm relative to TMS, using the residual solvent peaks at
22
energy minimum in the vicinity of the two acceptor atoms.
Given the self-complementarity of the cis-amide motif,
23
the crystal structure of 1b was anticipated to consist of
closely packed hydrogen bonded back-to-back dimers of the
kind shown in Fig. 1b. Phenanthridone derivatives have been
1
13
1
7
(
.27 ( H) and 77.23 ( C) for CDCl
3
and 2.50 ( H) and 39.51
1
3
C) for DMSO-d
6
as internal standards. Assignments were
24
shown to form such dimers in solution and phenanthridone
accomplished by coupling constants and integrals, as well as
D correlation experiments when necessary.
25
itself crystallises as the dimer. Still, given the rather large
2
2
6a
difference between the crystal structures of phenanthrene
26b
and 1,10-phenanthroline, resulting from the highly differ-
ent charge distribution in the two compounds, one could
argue that 1b should not afford the same crystal structure as
phenanthridone. Other compounds, structurally related to 1b,
such as neocuproine, phenanthrene-9,10-dione and 1,10-
phenanthroline-5,6-dione have quite simple and symmetric
solid state structures.
Synthesis
2
-Chloro-6-methylnicotinic acid methyl ester (2b). A suspen-
sion of 2-chloro-6-methylnicotinic acid (6.125 g, 35.7 mmol) in
26c
26d
POCl
3
(25 mL) was heated at reflux for 2 h. A solution was
at reduced
26e
attained upon heating. After distilling off the POCl
3
pressure, MeOH (72 mL) was added to the residue and the
resulting solution was heated at reflux for 1.5 h. Evaporation
of the solvent afforded a yellow solid. The pure product was
So why does the expected back-to-back dimerisation yield
to the unusual, apparently weaker (judging from the D · · · A
distances in Table 1) bifurcated hydrogen bonds? Back-to-back
dimerisation could be expected to result in higher symmetry and
more efficient p-stacking but, without solvent incorporation, it
would leave the nitrogen lone pairs of the bipyridine moiety
◦
obtained by bulb to bulb distillation at 85 C and 0.3 mmHg,
following a small fore fraction. Eluting the distillate through a
silica pad with heptane–EtOAc (2 : 1) afforded pure 2b (4.56 g,
◦
2
4.6 mmol, 69%) as a colorless oil: mp 27–28 C; R
f
(heptane–
27
uncompensated. According to Etter’s rules, the best acceptor
in the crystallisation mixture, the phen nitrogen pair of 1b,
should hydrogen bond with the most acidic hydrogen. The amide
hydrogen is considerably more acidic than the methanol crys-
EtOAc 2 : 1) 0.26; anal. calc. for C
8
H
8
ClNO
2
: C, 51.77; H, 4.34;
−1
N, 7.55; found C, 51.73; H, 4.31; N, 7.59%; IR (KBr) cm
(400 MHz, CDCl
3H, s), 7.16 (1H, d, J = 7.9 Hz), 8.08 (1H, d, J = 7.9 Hz); d
100 MHz, CDCl ) 24.5, 52.9, 121.9, 123.7, 140.9, 149.6, 162.7,
65.2; HRMS (FAB+) m/z calc. for C ClNO : 185.0244;
found 185.0225.
1
(
(
735 (C=O, ester); d
H
3
) 2.58 (3H, s), 3.93
C
tallisation cosolvent OH (pK 17 (2-pyridone) vs. 28 (MeOH) in
a
3
DMSO), and thus the best hydrogen bond donor. However, the
self-complementarity and cooperativity of the 2-pyridone motif
was expected to compensate for that, retaining the cis-amide
dimer. Nevertheless, with a hydrogen bond donor–acceptor ratio
of 1 : 3 in 1b, the bifurcated interactions are an efficient way to
compensate for the inherent donor deficiency. In one case, one of
the relatively acidic methyl hydrogens of 1b also serves as a weak
1
8
H
8
2
ꢀ
ꢀ
3 -Nitro-2,2 -bipyridinyl-3-carboxylic acid methyl ester (4a).
A mixture of 2a (3.0 g, 17 mmol), 3a (5.5 g, 35 mmol) and copper
powder (6.0 g, 94 mmol) in distilled DMF (20 mL) was stirred
◦
at 100 C. After 20 h, all starting materials were consumed
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 9 9 6 – 1 0 0 1
9 9 9

View Article Online
according to TLC. At room temperature, Na
2
EDTA (32.0 g,
to the procedure for 1a, except that CH
100 mL) of the aqueous phase afforded the crude product
2
Cl
2
extraction (4 ×
8
6 mmol) in water (300 mL) was added to the mixture. After
stirring for 10 min, the copper was filtered off and washed with
dichloromethane (5 × 30 mL). The filtrate was extracted with
dichloromethane (5 × 30 mL). The combined organic phases
directly, as a red solid (0.26 g). Trituration with CH
2
Cl
2
(2 ×
10 mL) afforded 1b (0.12 g, 0.53 mmol, 31%) as a white solid:
◦
mp 290 C dec; R
f
(CH
2
2
Cl –MeOH 9 : 1) 0.33; anal. calc. for
were dried (Na
2
SO
4
) and the solvent was removed under reduced
C
13
H
11
N
3
O: C, 69.32; H, 4.92; N, 18.66; found C, 69.28; H,
−
1
pressure to give 8.9 g of a brown oil containing some DMF. The
crude product was chromatographed on a silica column using
5.08; N, 18.54%; IR (KBr) cm 1666, 1589, 1481, 1434, 1363;
(400 MHz, DMSO-d ) 2.59 (3H, s), 2.73 (3H, s), 7.44 (1H,
d, J = 8.4 Hz), 7.48 (1H, d, J = 8.4 Hz), 7.59 (1H, d, J =
8.0 Hz), 7.64 (1H, d, J = 8.0 Hz), 11.81 (1H, br s); d (100 MHz,
DMSO-d ) 23.9, 25.0, 121.1, 124.1, 124.2, 125.2, 132.6, 134.5,
d
H
6
petroleum ether–EtOAc (1 : 2) as the eluent, affording 4a (1.41 g,
◦
5
.44 mmol, 32%) as a pale orange solid: mp 82–84 C; R
f
C
(
petroleum ether–EtOAc 1 : 2) 0.2; anal. calc. for C12
H
9
N
3
O
4
: C,
6
5
5.60; H, 3.50; N, 16.21; found C, 55.54; H, 3.55; N, 16.04%; IR
135.8, 150.0, 152.6, 160.4, 163.7; HRMS (FAB+) m/z calc. for
−
1
+
(
KBr) cm 1724 (C=O, ester), 1527, 1348 (N=O); d
) 3.72 (1H, s), 7.49 (1H, dd, J = 4.8, J = 8.0 Hz, 1H), 7.56
1H, dd, J = 8.3, J = 4.7 Hz), 8.41 (1H, dd, J = 1.7, J = 8.0 Hz),
H
(400 MHz,
C
13
H
12
N
3
O [M + 1] : 226.0975; found 226.0974.
CDCl
(
3
X-Ray crystallographic studies
8
1
5
1
1
.50 (1H, dd, J = 8.3, J = 1.5 Hz), 8.78 (1H, dd, J = 4.8, J =
.7 Hz), 8.86 (1H, dd, J = 4.7, J = 1.5 Hz); d (100 MHz, CDCl
2.7, 123.6, 123.7, 125.2, 132.7, 138.8, 144.5, 152.4, 152.9, 154.0,
Platelike single crystals were grown by slow diffusion of diethyl
ether into a solution of 1b in a 4 : 1 mixture of chloroform
and methanol. Intensity data were collected at 293 K on a
Bruker SMART CCD system using x-scans and a rotating
C
3
)
57.2, 165.6; HRMS (FAB+) m/z calc. for C12
H
10
N
3
O
4
[M +
+
] : 260.0666; found 260.0691.
28
˚
anode with Mo Karadiation (k = 0.71073 A). The intensity was
2
9
ꢀ
ꢀ
ꢀ
corrected for Lorentz and polarisation effects using SADABS.
6
,6 -Dimethyl-3 -nitro[2,2 ]bipyridinyl-3-carboxylic acid methyl
ester (4b). The product was synthesised from 2b (0.87 g,
.68 mmol) and 3b (1.62 g, 9.38 mmol) according to the proce-
30
All reflections were integrated using SAINT. The structure
of 1b was solved by direct methods and refined by full-matrix
4
2
31
least-squares calculations on F using SHELXTL 5.1. Non-
disordered non-H atoms were refined with anisotropic displace-
ment parameters. Aromatic hydrogen atoms were constrained
to parent sites using a riding model, methyl hydrogens using
a rotating model. From the difference Fourier map obtained
after the three well-resolved molecules in the asymmetric unit
had been located, a number of electron density peaks were
obtained, all of them seemingly coplanar. Thus, the disordered
molecule was represented by 15 isotropically refined carbon
atoms in a plane, with an occupancy factor of 0.5, each of them
having two symmetry-related positions. This refinement gave
dure for 4a, except that the reaction time was 5 h and extraction
was done with EtOAc. Purification by chromatography on a
silica column using heptane–EtOAc (6 : 5) as the eluent provided
4
b (0.61 g, 2.12 mmol, 45%) as a light yellow solid: mp 126–
◦
127 C; R
f
(heptane–EtOAc 3 : 7) 0.3; anal. calc. for C14
H
13
N
3
4
O :
C, 58.53; H, 4.56; N, 14.63; found C, 58.41; H, 4.48; N, 14.55%;
−
1
IR (KBr) cm 1720 (C=O, ester), 1587, 1514, 1348 (N=O); d
H
(
400 MHz, CDCl
3
) 2.69 (3H, s), 3.71 (3H, s), 7.32 (1H, d, J =
.1 Hz), 7.38 (1H, d, J = 8.5 Hz), 8.30 (1H, d, J = 8.1 Hz),
.48 (1H, d, J = 8.5 Hz); d (100 MHz, CDCl ) 24.9, 25.0,
2.4, 122.0, 123.1, 123.2, 132.9, 139.0, 142.3, 153.8, 157.3, 162.5,
8
8
5
1
1
C
3
2
wR(F ) = 0.2666.
63.6, 165.6; HRMS (FAB+) m/z calc. for C14
H
14
N
3
O
4
[M +
+
The optimal orientation of 1b inside the channel according
to molecular modelling (see below) was overlayed with the 15
carbon atoms corresponding to the strongest electron densities
in the disordered part of the initial crystal structure refinement
] : 288.0979; found 288.0986.
9
H-4,5,9-Triazaphenanthren-10-one (1a). A mixture of 4a
(
(
1.51 g, 5.82 mmol), acetone (40 mL), glacial acetic acid
3.7 mL), water (3.7 mL) and powdered iron (1.35 g, 24.2 mmol)
(
see Fig. 5a). The agreement of these two structures was consid-
was refluxed for 3.5 h. The initial solution turned from yellow
to dark yellow upon addition of iron, and when heated it turned
intensely dark red, indicating complex formation. After cooling
to room temperature, acetone was removed under reduced
ered good enough to justify incorporation of the model structure
into the crystal structure refinement. The 15 carbon atoms were
thus removed and the atomic coordinates of the model structure
were applied using a FRAG/FEND instruction. The position of
the whole rigid molecule was refined and its overall occupancy
factor of the two symmetry related positions was refined to 0.96.
Non-hydrogen atoms were refined isotropically and hydrogen
atoms were not included.
pressure. To the residue was added Na EDTA (8.1 g, 21.8 mmol)
2
in water (80 mL), and NaOH (10 M, 7 mL). The mixture
was stirred for 10 minutes. Unreacted iron was retained in the
reaction flask using a magnet and the suspension was transferred
and extracted with CH
organic phase and a red aqueous phase with precipitate. The
organic phases were separated and dried (Na SO ). The solvent
2
Cl
2
(6 × 150 mL) to obtain a yellow
¯
Crystal data. C11
H
13
N
3
O. M = 225.25, triclinic, P1, a =
˚
1
0.0572, b = 13.7223, c = 15.4233 A, a = 72.173, b = 80.393,
◦
3
3
2
4
c = 86.903 , V = 1997.7 A , Dcalcd = 1.301 g cm , Z = 7, l =
˚
−
1
was removed under reduced pressure affording 80 mg of a
brown powder, consisting of a mixture of products. Additional
0
0
0
2
.086 mm . 21513 reflections measured, 11415 unique (Rint
=
2
.0440) which were used in all calculations. The final wR(F ) was
.3289 and the S value 1.161 (all data). The R(F) was 0.0977 (I >
r(I)).†
Na
the aqueous phase. After stirring for 10 minutes, the red color
turned yellow. A new extraction with CHCl
–MeOH (9 : 1) (4 ×
50 mL) provided pure 1a (450 mg, 2.28 mmol, 39%) as a light
2
EDTA solution (12 g in 150 mL of water) was added to
3
1
Molecular modelling
◦
yellow solid: mp 370 C dec; R
7
C
f
(MeOH–2 M NH
: 2 : 1) 0.52 (displays a blue spot at 366 nm); anal. calc. for
O: C, 67.00; H, 3.58; N, 21.31; found C, 66.78; H, 3.67;
4
Cl–MeNO
2
In the program Diamond V2.0e, the crystal structure of 1b
was extended along the crystallographic a axis using symmetry
operations. The disordered part of the structure was removed to
create a hollow channel defined by 33 molecules of 1b, spanning
11
H
7
N
3
−
1
N, 21.10%; IR (KBr) cm 1669, 1605, 1479, 1450, 1419; d
H
(
400 MHz, DMSO-d
6
) 7.61 (1H, dd, J = 8.3, J = 4.4 Hz), 7.77
˚
approx. 25 A along the a axis. The atomic coordinates were
(
2H, m), 8.62 (1H, dd, J = 4.4, J = 1.5 Hz), 8.64 (1H, dd,
16
transferred to the program Macromodel where the correct
atom types and bond orders were defined. Keeping all 33
molecules from the crystal structure fixed at their coordinates,
J = 8.0, J = 1.8 Hz), 9.14 (1H, dd, J = 4.5, J = 1.8 Hz); d
C
(
100 MHz, DMSO-d ) 123.9, 124.2, 124.9, 126.1, 135.1, 136.1,
6
1
C
36.2, 145.1, 151.0, 155.0, 161.0; HRMS (FAB+) m/z calc. for
+
11
H
8
N
3
O [M + 1] : 198.0662; found 198.0677.
†
CCDC reference numbers 256289. See http://www.rsc.org/suppdata/
3
,6-Dimethyl-9H-4,5,9-triazaphenanthren-10-one (1b). The
ob/b4/b417485b/ for crystallographic data in .cif or other electronic
format.
product was synthesised from 4b (0.48 g, 1.67 mmol) according
1
0 0 0
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 9 9 6 – 1 0 0 1

View Article Online
the inside of the channel structure was probed by one molecule
Rissanen, Eur. J. Inorg. Chem., 2001, 1515; U. Ziener, E. Breuning,
J.-M. Lehn, E. Wegelius, K. Rissanen, G. Baum, D. Fenske and G.
Vaughan, Chem.–Eur. J., 2000, 6, 4132.
of 1b by energy minimising it from variable starting orientations
32
and positions inside the channel, using the MMFF force field.
1
1
2 Janus was the Roman god of change and transition. He had two faces
looking in opposite directions, hence the analogy with 1.
CSD search
3 F. Leroy, P. Despr e´ s, M. Bigan and D. Blondeau, Synth. Commun.,
1
996, 26, 2257.
The search was conducted using CSD version 5.25 (November
1
1
1
4 J. Ohmori and H. Kubota, J. Med. Chem., 1996, 39, 1331.
5 R. Trosch u¨ tz and A. Karger, J. Heterocycl. Chem., 1996, 33, 1815.
6 Macromodel V6.0. F. Mohamadi, N. G. J. Richards, W. C. Guide, R.
Liskamp, M. Lipton, C. Caulfield, G. Chang, T. Hendrickson and
W. C. Still, J. Comput. Chem., 1990, 11, 440.
18
2
003). A query consisting of the parent phenanthroline struc-
ture resulted in 86 crystal structures of purely organic, nonionic
composition. Entries containing N-substitution or chelation by
alkali metals (24) were eliminated by hand, as well as those
missing a 3D structure (8), leaving 54 structures to be analysed.
ꢀ
1
7 For a discussion on Z and crystal packing frequencies see: T. Steiner,
Acta Crystallogr., Sect. B, 2000, 56, 673; C. P. Brock, Chem. Mater.,
1
994, 6, 1118.
Acknowledgements
18 CCDC, Cambridge Structural Database CSD 1.6. University of
Cambridge, UK, 2003.
We thank the Swedish Research Council as well as the Crafoord
Foundation, the Lars-Johan Hjerta Foundation, the Magn.
Bergvall Foundation and the Royal Physiographic Society for
generous grants. The faculty exchange program between Lund
University and the University of California is acknowledged for
a scholarship to K. W.
19 T. Steiner, J. van der Maas and B. Lutz, J. Chem. Soc., Perkin Trans. 2,
1997, 1287; L. Kaczmarek, R. Balicki, J. Lipkowski and P. Borowicz,
J. Chem. Soc., Perkin Trans. 2, 1994, 1603.
2
0 For selected hydrogen bonded structures see: D. Paul, F. Melin, C.
Hirtz, J. Wytko, P. Ochsenbein, M. Bonin, K. Schenk, P. Maltese
and J. Weiss, Inorg. Chem., 2003, 42, 3779; E. S. Lavender, R. M.
Gregson, G. Ferguson and C. Glidewell, Acta Crystallogr., Sect. C,
1
999, 55, 751; E. Garc ´ı a-Mart ´ı nez, E. M. V a´ zquez-L o´ pez and D. G.
Tuck, Acta Crystallogr., Sect. C, 1998, 54, 840; Y.-P. Tian, C.-Y.
Duan, X.-X. Xu and X.-Z. You, Acta Crystallogr., Sect. C, 1995, 51,
2309; W. H. Watson, J. Galloy, F. V o¨ gtle and W. M. M u¨ ller, Acta
Crystallogr., Sect. C, 1984, 40, 200.
References
1
S. C. Zimmerman and P. S. Corbin, Struct. Bonding (Berlin), 2000,
9
1
6, 63; J. A. R. Navarro and B. Lippert, Coord. Chem. Rev., 1999,
85–186, 653.
21 P. S. Donnelly, B. W. Skelton and A. H. White, Aust. J. Chem., 2003,
56, 1249.
22 G. G. Siegel and Th. Zeegers-Huyskens, Spectrochim. Acta, Part A,
1989, 45A, 1297.
23 The literature does contain examples where the cis-amide motif fails
to form the hydrogen bonded dimer in the solid state, see: E. F.
Maverick, P. L. Wash and D. A. Lightner, Acta Crystallogr., Sect. C,
2001, 57, 284; E´ . Boucher, M. Simard and J. D. Wuest, J. Org. Chem.,
2
3
4
5
J.-M. Lehn, Supramolecular Chemistry-Concepts and Perspectives,
Wiley-VCH, Weinheim, 1995.
A. Gavezotti, Faraday Discuss., 1997, 106, 63; A. Gavezotti, Acc.
Chem. Res., 1994, 27, 309.
G. F. Swiegers and T. J. Malefetse, Chem. Rev., 2000, 100, 3483 and
references therein.
L. J. Prins, D. N. Reinhoudt and P. Timmerman, Angew. Chem., Int.
Ed., 2001, 40, 2382 and references therein.
1995, 60, 1408; M.-J. Brienne, J. Gabard, M. Leclercq, J.-M. Lehn, M.
Cesario, C. Pascard, M. Chev e´ and G. Dutruc-Rosset, Tetrahedron
Lett., 1994, 35, 8157.
6
7
A. M. Beatty, CrystEngComm, 2001, 3, 243 and references therein.
S. J o´ nsson, F. G. J. Odille, P.-O. Norrby and K. W a¨ rnmark, Chem.
Commun., 2005, 549; S. Ston cˇ ius, E. Butkus, A. Zilinskas, K. Larsson,
24 R. Kleppinger, C. P. Lillya and C. Yang, J. Am. Chem. Soc., 1997,
119, 4097.
¨
L. Ohrstr o¨ m, U. Berg and K. W a¨ rnmark, J. Org. Chem., 2004, 69,
5
196.
25 D. K. Sen, Acta Crystallogr., Sect. B, 1970, 26, 1629.
26 (a) V. Pet rˇ ´ı cˇ ek, I. C ´ı sa rˇ ov a´ , L. Hummel, J. Kroupa and B. B rˇ ezina,
Acta Crystallogr., Sect. B, 1990, 46, 830; (b) S. Nishigaki, H. Yoshioka
and K. Nakatsu, Acta Crystallogr., Sect. B, 1978, 34, 875; (c) D. K.
Sen, Acta Crystallogr., Sect. B, 1969, 25, 988; (d) M. Alleaume, R.
Darrouy and J. Housty, Acta Crystallogr., 1961, 14, 1202; (e) F.
Calderazzo, F. Marchetti, G. Pampaloni and V. Passarelli, J. Chem.
Soc., Dalton Trans., 1999, 4389.
8
9
S. C. Zimmerman and B. F. Duerr, J. Org. Chem., 1992, 57, 2215; M.
Gallant, M. T. P. Viet and J. D. Wuest, J. Org. Chem., 1991, 56, 2284;
Y. Ducharme and J. D. Wuest, J. Org. Chem., 1988, 53, 5789.
T. Olszewska, M. Gdaniec and T. Polonski, J. Org. Chem., 2004,
6
9, 1248; Y. Du, C. J. Creighton, B. A. Tounge and A. B. Reitz,
Org. Lett., 2004, 6, 309; M. R. Edwards, W. Jones and W. D. S.
Motherwell, Cryst. Eng., 2002, 5, 25; C. B. Aaker o¨ y, A. M. Beatty,
M. Nieuwenhuyzen and M. Zou, Tetrahedron, 2000, 56, 6693; E.
Murguly, R. McDonald and N. R. Branda, Org. Lett., 2000, 2, 3169;
M. Akazome, S. Suzuki, Y. Shimizu, K. Henmi and K. Ogura, J. Org.
Chem., 2000, 65, 6917; M. Simard, D. Su and J. D. Wuest, J. Am.
Chem. Soc., 1991, 113, 4696.
27 M. C. Etter, Acc. Chem. Res., 1990, 23, 120.
28 BrukerAXS, SMART, Area Detector Control Software, Bruker Ana-
lytical X-ray System, Madison, Wisconsin, USA, 1995.
29 G. M. Sheldrick, SADABS, Program for Absorption Correction,
University of G o¨ ttingen, Germany, 1996.
1
1
0 M. Munakata, L. P. Wu, M. Yamamoto, T. Kuroda-Sowa and M.
Maekawa, J. Am. Chem. Soc., 1996, 118, 3117; S. R. Breeze and S.
Wang, Inorg. Chem., 1993, 32, 5981.
30 BrukerASX, Saint, Integration Software, Bruker Analytical X-ray
System, Madison, Wisconsin, USA, 1995.
31 G. M. Sheldrick, SHELXTL 5.1, Program for Structure Solution and
Least Square Refinement, University of G o¨ ttingen, Germany, 1998.
32 T. A. Halgren, J. Comput. Chem., 1996, 17, 490.
1 D. G. Kurth, K. M. Fromm and J.-M. Lehn, Eur. J. Inorg. Chem.,
2
001, 1523; E. Breuning, U. Ziener, J.-M. Lehn, E. Wegelius and K.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 9 9 6 – 1 0 0 1
1 0 0 1