Alkali metal ion selectivities of podands that form pseudo-cyclic structures by intramolecular hydrogen bonding

Article abstract of DOI:10.1039/b201525k

The podands capable of intramolecular hydrogen bonding were synthesized. ¹³C NMR titration experiments and FAB mass spectra of the podands showed that they formed pseudo-cyclic structures under guest-free conditions. The X-ray crystal structure

Full text of DOI:10.1039/b201525k

View Article Online / Journal Homepage / Table of Contents for this issue
Alkali metal ion selectivities of podands that form pseudo-cyclic
structures by intramolecular hydrogen bonding
Yoichi Habata,*a Yoji Fukuda,a Sadatoshi Akaboria and Jerald S. Bradshawb
1
a Department of Chemistry, Faculty of Science, Toho University, Funabashi, Chiba 274-8510,
Japan
b Department of Chemistry and Biochemistry, Brigham Young University, Provo, UT 84602,
USA
Received (in Cambridge, UK) 11th February 2002, Accepted 22nd February 2002
First published as an Advance Article on the web 8th March 2002
New podands capable of intramolecular hydrogen bonding (1–3) have been synthesized. 13C NMR titration
experiments and FAB mass spectra of the podands show that 1 and 2 form pseudo-cyclic structures under guest-free
conditions. The X-ray crystal structure of the 2–NaSCN complex is also reported.
Introduction
Podands are open chain analogues of the crown compounds,
and many kinds of podands have been prepared.1 Podands are
also important as biomimetic compounds for naturally occur-
ring ionophores such as nigericin2 and lasolocid.3 Several nat-
ural ionophore-mimic podands and semi-synthetic podands
have been synthesized.4 They each have a carboxylic acid at one
end and a hydroxy group at the other end of the molecule. In
these systems, hydrogen bonding between the end-hydroxy and
end-carboxy groups is an important factor in forming a cyclic
structure when these podands bind cations.
Recently, we reported that podand 1 formed a pseudo-cyclic
structure in the solid state and in solution.5 Those results
prompted us to prepare new podands in order to investigate
whether intramolecular pre-organization of the podands by
weak hydrogen bonding influences complexing ability and
cation selectivity. Here, we report the molecular design,6 prep-
aration, crystal structure, and alkali metal cation selectivity of
new podands that form pseudo-cyclic structures by weak
hydrogen bonding in the gas phase.
Fig. 1 Schematic drawings of podands 2 and 3.
Results and discussion
Two new podands 2 and 3 were designed and prepared (Scheme
1). We expected that podand 2 having o-hydroxyphenyl end
groups would form a pseudo-cyclic structure because the two
end OH groups would hydrogen bond with each other as shown
in Fig. 1. Podand 3 with o-hydroxybenzoyl end groups would
Fig. 2 Optimized structures of podands 2 and 3 by ab initio 3-21G(*)
calculations.
form a non-cyclic structure because the C᎐O and OH groups in
᎐
each salicylic moiety would form hydrogen bonds as shown. To
confirm our prediction, a conformer search using molecular
mechanics (MMFF was used as a mechanic engine) and then
optimization using semi-empirical AM1 calculations followed
by ab initio 3-21G(*) calculations were carried out. As we
expected, the global minima of 2 and 3 exhibited cyclic (Fig. 2a)
and non-cyclic structures (Fig. 2b), respectively.
Podand 2 was prepared by the reaction of 2,6-bis(tosyloxy-
methyl)pyridine with catechol in the presence of Na2CO3 in
acetonitrile in 41% yield. Dimerization product 47 was also
isolated as a by-product in 3% yield. Podand 3 was prepared
from pyridinedimethanol and salicylic acid in the presence of
DCC–DMAP in CH2Cl2 in 41% yield. The structures of the
DOI: 10.1039/b201525k
J. Chem. Soc., Perkin Trans. 1, 2002, 865–869
This journal is © The Royal Society of Chemistry 2002
865

View Article Online
Scheme 1
new podands 2 and 3 were confirmed by 1H and 13C NMR, MS,
FT-IR, elemental analyses, and X-ray crystallography. In the
FT-IR spectra (Fig. 3), the O–H stretching vibrations of
Fig. 5 ORTEP diagram of podand 3 (30% probability level).
Fig. 3 FT-IR spectra of podands 2 (dotted line) and 3 (solid line) in
the region for O–H stretching vibrations.
structure contains two molecules of 2. The O1–O4, O5–O6
and O5–O8, distances are 2.69, 2.71 and 2.80 Å, respectively
(Table 1). These distances are typical of a cyclic structure in
which the hydrogen atoms on O4 and O5 are hydrogen bonded
to O1 and O8, respectively. There was a slight difference
between the optimized structure and the crystal structure.
Although the hydrogen bond between the ring N atom and
the H atom of the HO group was predicted by the ab initio
3-21G(*) calculations, the hydrogen bond between them was
not observed in the crystal structure, because the H1–N1 and
H34–N2 atom distances are 2.75 and 2.82 Å, respectively. The
overall Rw factor of the podand 2 would be very large (12.5%),
because it is very difficult to obtain good single crystals for
X-ray crystallography. On the other hand, in podand 3 (Fig. 5),
the two aromatic rings (ring A and ring B) approximately
planar and the dihedral angle between ring B and ring C
is about 38Њ. There are hydrogen bonds between O2 and O3
(2.63 Å) and O5 and O6 (2.63 Å) in the podand. Thus, we have
prepared two podands, one of which has a pseudo-cyclic struc-
ture and the other a non-cyclic structure under guest-free
conditions.
podands 2 (dotted line) and 3 (solid line) appear at 3249 and
3250 cmϪ1, respectively. These IR data confirm that both
podands form hydrogen bonds,8 as estimated by the ab initio
3–21G(*) calculations.
Figs. 4 and 5 show the ORTEP diagrams of podands 2 and 3,
Fig. 4 ORTEP diagram of podand 2 (30% probability level).
respectively. As the molecular calculations indicated, podand 2
forms a pseudo-cyclic structure by hydrogen bonding between
the OH groups. The hydrogens of the OH groups, H1, H17,
H18, and H34, in podand 2 were located by difference Fourier
maps, and were not refined. As shown in Fig. 4, the unit
Association constants in CD3CN between these podands and
Naϩ were determined by 13C NMR titration experiments9 in
order to study their complexing abilities in solution, since
molecular modeling also suggested that the cavity sizes of
podands 1 and 2 are similar to the size of the Naϩ cation. The
866
J. Chem. Soc., Perkin Trans. 1, 2002, 865–869

View Article Online
Table 1 Intermolecular hydrogen bond parameters
D
H
A
D ؒ ؒ ؒ A
D–H
H ؒ ؒ ؒ A
D–H ؒ ؒ ؒ A
Podand 2
O(4)
O(5)
O(5)
H(17)
H(18)
H(18)
O(1)
O(6)
O(8)
2.69
2.71
2.80
1.10
1.07
1.07
2.11
2.22
1.85
110
105
144
Podand 3
O(3)
O(6)
H(1)
H(2)
O(2)
O(5)
2.63
2.64
1.15
1.05
1.65
1.75
139
140
Table 2 Fragment ion peaks and relative intensities for [M ϩ Li]ϩ,
[M ϩ Na]ϩ, [M ϩ K]ϩ, [M ϩ Rb]ϩ and [M ϩ Cs]ϩ in 1–alkali metals,
2–alkali metals and 3–alkali metals. The intensity of each peak was
normalized on the basis of that of [M ϩ H]ϩ (in parentheses)
1–Alkali metals
2–Alkali metals
3–Alkali metals
[M ϩ H]ϩ
[M ϩ Li]ϩ
[M ϩ Na]ϩ
[M ϩ K]ϩ
[M ϩ Rb]ϩ
[M ϩ Cs]ϩ
256 (100)
262 (150)
278 (1800)
294 (720)
340 (210)
388 (170)
324 (100)
329 (24)
346 (130)
362 (22)
408 (10)
456 (12)
380 (100.0)
386 (0.9)
402 (5.9)
418 (1.7)
464 (0.7)
512 (1.0)
log K values for 1–Naϩ, 2–Naϩ, and 3–Naϩ are 2.5, 2.3, and 1.3,
respectively. These data clearly suggest that podands 1 and 2,
which form a guest-free pseudo-cyclic structure, show higher
complexing ability for Naϩ than 3, which forms a non-cyclic
structure in solution.
In order to investigate the complexing properties of podands
1–3 with alkali metals in the gas phase, the FAB mass spectra of
the podand–alkali metal systems were measured. Table 2 shows
the relative intensities for [M ϩ Li]ϩ, [M ϩ Na]ϩ, [M ϩ K]ϩ, [M
ϩ Rb]ϩ and [M ϩ Cs]ϩ in mixtures of 1–alkali metals, 2–alkali
metals, and 3–alkali metals. In the mixture of 1–alkali metals,
all the peak intensities of [M ϩ Li]ϩ, [M ϩ Na]ϩ, [M ϩ K]ϩ, [M
ϩ Rb]ϩ and [M ϩ Cs]ϩ are greater than that of [M ϩ H]ϩ and
the order is as follows; [M ϩ Na]ϩ > [M ϩ K]ϩ > [M ϩ Rb]ϩ >
[M ϩ Cs]ϩ > [M ϩ Li]ϩ. It is important to note that the inten-
sities of [M ϩ Na]ϩ and [M ϩ K]ϩ are 18 and 7 times greater
than that of [M ϩ H]ϩ, respectively. These results suggest that
podand 1 is selective for Naϩ. The results of the 2–alkali metal
experiments also suggest Naϩ selectivity, but the relative inten-
sities of the ion peaks arising from the 2–cation complexes,
except that of 2–Naϩ, were weaker than that of [M ϩ H]ϩ. In
the case of the 3–alkali metal system, all the intensities of the
ion peaks of [3–cations]ϩ were very small, although 3 shows
clear Naϩ selectivity. It is well known that FAB MS can be used
to determine crown ether–alkali metal cation stability con-
stants.10 Therefore, the FAB MS data suggest that the complex-
ing abilities of podands 1–3 with alkali metal cations are as
follows; podand 1 > podand 2 ӷ podand 3. The data also sug-
gested that podand 2, which forms a guest-free pseudo-cyclic
structure, showed higher complexing ability than 3, which
forms a non-cyclic structure. These results are consistent with
the log K values determined in solution.
A single crystal of the Naϩ complex with podand 2 was
obtained and the X-ray crystal structure was determined
(Fig. 6a). The sodium ion is centered in the pseudo-cyclic struc-
ture of podand 2 and is five-coordinated by N1, O1, O2, O3
and O4. The sodium ion lies in the same plane as the other
hetero atoms. Since the sodium ions (Na1 and Na1Ј) of the unit
structure are covered by benzene rings (C14Ј–C15Ј–C16Ј–
C17Ј–C18Ј–C19Ј and C14–C15–C16–C17–C18–C19) (Fig. 6b),
other donor atoms could not bind above or below the plane.
The Na1–C14Ј, Na1–C15Ј, Na1–C16Ј, Na1–C17Ј, Na1–C18Ј,
and Na1–C19Ј distances are 3.61, 3.31, 3.23, 3.45, 3.74, and
3.82 Å, respectively. These distances are comparable to those of
the Naϩ complex with the double-armed diaza-18-crown-6
Fig. 6 (a) ORTEP diagram and (b) packing diagram of the 2–NaSCN
complex (30% probability level).
having indole side chains, which exhibit Naϩ–π interactions
between the incorporated cations and the aromatic moieties of
the side chain.11 Thus, the complex would be stabilized by Naϩ–
π interactions. The Na1–S1, Na1–C20, and Na1–N2 distances
are 5.13, 4.47, and 4.31 Å, respectively. Therefore, atoms of the
counter SCNϪ ion do not participate in complex formation.
The Na1–N1, Na1–O1, Na1–O2, Na1–O3, and Na1–O4 bond
lengths are 2.41, 2.40, 2.41, 2.43, and 2.33 Å, respectively. These
bond lengths are comparable to those of the Naϩ complex of
monoaza-15-crown-5 derivatives12 rather than those of the Naϩ
J. Chem. Soc., Perkin Trans. 1, 2002, 865–869
867

View Article Online
Table 3 Crystal and selected experimental data for podands 2 and 3, and the 2–NaSCN complex
2
3
2–NaSCN
Formula
M
C38H34N2O8
646.70
C21H17NO6
379.37
C20H17N2O4SNa
404.42
Crystal system
Space group
Monoclinic
P21/a
Triclinic
P1
Monoclinic
P21/n
¯
a/Å
b/Å
c/Å
8.117(6)
19.978(7)
19.981(8)
8.254(9)
14.99(1)
8.05(1)
9.744(5)
16.192(8)
12.099(6)
α/Њ
β/Њ
102.1(1)
111.99(9)
78.40(9)
894(2)
92.91(5)
95.72(4)
γ/Њ
U/Å3
3236(3)
1899(1)
Z
4
2
4
Dc/g cmϪ3
1.327
1.408
1.414
F(000)
1360.00
0.94
0.25 × 0.10 × 0.60
20 (20.5–24.7)
369.00
1.04
0.50 × 0.25 × 0.50
25 (22.6–24.9)
840.00
2.23
0.75 × 0.50 × 0.75
µ(Mo-Kα)/cmϪ1
Crystal dimensions/mm
No. of reflections for unit cell
determination (2θ range)/Њ
Scan width/Њ
1.47 ϩ 0.30 tan θ
Ϫ8 р h р 0, 0 р k р 19,
Ϫ19 р l р 19
1.68 ϩ 0.30 tan θ
0 р h р 9, Ϫ17 р k р 17,
Ϫ9 р l р 8
1.52 ϩ 30 tan θ
Ϫ12 р h р 12,
0 р k р 21,
Ϫ15 р l р 0
4733
Limiting indices
No. of reflections measured
3879
3384
Unique (Rint
Used [all data], No
)
3447 (0.219)
3438
3147 (0.030)
3147
4365 (0.037)
4361
R
Rw
0.321
0.329
0.125
1.41
0.117
0.152
0.054
1.27
0.088
0.139
0.048
1.15
R1 [I > 2.0σ(I )]
Goodness of fit
No. parameters, Np
Maximum shift/error in final cycle
1020
0.004
1642
0.001
0.45, Ϫ0.41
50
2088
0.001
0.24, Ϫ0.33
55
Maximum, minimum peaks in final difference map/e ÅϪ30.78, Ϫ0.42
2θmax/Њ
45
2
½
½
Details in common: ω-2θ scan; Rw = [Σw(|Fo| Ϫ |Fc|)2/ΣwFo ] , R1 = Σ||Fo| Ϫ |Fc||/ΣFo|, goodness of fit [Σw(|Fo| Ϫ |Fc|)2/(No Ϫ Np)] .
complex of pyridinodibenzo-18-crown-6.13 This is the first
instance of a planar five-coordinated complex of Naϩ with
podands and crown ether compounds.14
In conclusion, we have demonstrated that a podand that
forms a pseudo-cyclic structure by hydrogen bonding shows
higher affinity for metal ions than a podand that forms a
non-cyclic structure.
s-Dipyridino-18-crown-6 (4). Mp 178.5–179.5 ЊC (lit.,7 183–
185.5 ЊC). 1H NMR (400 MHz, CDCl3) 5.17 (8H, s), 6.99–7.12
(8H, m), 7.31 (2H, d, J 7.7 Hz), and 7.58 (1H, t, J 7.7). FAB-MS
(m-NBA as a matrix) m/z 427 ([M ϩ 1]ϩ, 100%).
Compound 2. Mp 100–101 ЊC (Found: C, 70.8; H, 5.4; N,
1
4.3. Calc. for C19H17NO4: C, 70.6; H, 5.3; N, 4.3%). H NMR
(400 MHz, CDCl3) 5.62 (4H, s), 7.44 (2H, d, J 7.7), 7.78 (1H,
t, J 7.9), and 9.66 (2H, s). The aromatic moiety of the
2-hydroxybenzene fragment showed an ABCD pattern: δHA
6.83 (2H, triple doublet), δHB 7.05 (2H, dd), δHC 7.05 (2H, dd),
Experimental
δHD 6.99 (2H, triple doublet), for HA–HD, JAB(ortho) = JBC
=
Melting points were obtained with a Mel-Temp capillary
apparatus and are not corrected. EI and FAB mass spectra were
JCD = 7.7 Hz, JAC(meta) = JBD = 1.5 Hz. 13C NMR (CD3CN)
157.50, 148.92, 148.07, 139.72, 124.53, 122.44, 120.97, 117.95,
117.40, and 73.69. EI-MS m/z 323 (Mϩ, 60%).
recorded using a JEOL 600 H spectrometer. 1H NMR and 13
C
NMR spectra were measured in CDCl3 and CD3CN, respect-
ively, on a JEOL ECP400 (400 MHz for 1H) spectrometer.
Ligand 1 was prepared as reported.5 Other starting materials
were used as purchased.
Preparation of pyridin-2,6-diyldimethyl disalicylate (3)
A mixture of pyridine-2,6-dimethanol (0.21 g, 1.4 mmol), sali-
cylic acid (0.42 g, 3.0 mol), N,N-dimethylaminopyridine (0.35 g,
2.7 mmol), and dicyclohexylcarbodiimide (0.58 g, 2.8 mmol) in
CH2Cl2 (10 mL) was stirred for 24 h under N2. The reaction
mixture was filtered and the filtrate was concentrated under
reduced pressure. The residual oil was chromatographed on
silica gel (ethyl acetate–hexane = 3 : 2 as eluent) to give 3 as a
white powder in 46% yield.
Preparation of 2,6-bis[(2Ј-hydroxyphenyl)oxymethyl]pyridine (2)
A mixture of catechol (0.25 g, 2.3 mmol), 2,6-bis(tosyloxy-
methyl)pyridine (0.50 g, 1.1 mmol), and K2CO3 (1.5 g, 11
mmol) in acetonitrile (50 mL) was refluxed for 24 h under N2.
The reaction mixture was cooled and filtered. The precipitate
was washed with CHCl3 (30 mL × 4). The CHCl3 layers were
combined, dried over Na2SO4 and evaporated. The residual
solid was extracted with ether (10 mL × 3), and the ether layer
was concentrated to give pure dimer 4 (s-dipyridino-18-crown-
6) in 3% yield. The precipitate was suspended in water, and
neutralized by a 1 mol LϪ1 HCl solution. The aqueous phase
was extracted with CHCl3 (20 mL × 3). The CHCl3 solutions
were combined, dried over Na2SO4 and evaporated. Crystalliz-
ation was induced by addition of ether (5 mL). The crystals
were recrystallized from MeOH to give 2 as white needles in
41% yields.
Mp 99.5–100 ЊC (Found: C, 66.3; H, 4.6; N, 3.6. Calc. for
1
C21H17NO6: C, 66.5; H, 4.5; N, 3.7%). H NMR (400 MHz;
CD3Cl): 5.52 (4H, s), 7.48 (2H, d, J 7.7), 7.87 (1H, t, J 7.7), and
10.55 (2H, s). The aromatic moiety of the salicylate showed an
ABCD pattern: δHA 7.00 (2H, dd), δHB 6.90 (2H, triple doub-
let), δHC 7.48 (2H, triple doublet), δHD 7.92 (2H, dd), for HA–
HD, JAB(ortho) = JBC = JCD = 7.9 Hz, JAC(meta) = 1.8 Hz,
JBD(meta) = 1.1 Hz. 13C NMR (CD3CN) 170.56, 162.42, 156.28,
139.12, 137.16, 131.13, 122.16, 120.56, 118.37, 113.49, and
68.06. EI-MS m/z 379(Mϩ, 67%).
868
J. Chem. Soc., Perkin Trans. 1, 2002, 865–869

View Article Online
Na؉ complex with 2
References
Podand 2 (0.01 mmol) in acetonitrile (1 mL) was reacted with
NaSCN (0.01 mmol) in methanol (1 mL). After the solvent had
been evaporated, the crystals were recrystallized from aceto-
nitrile. The crystals were dried with an Abderhalden’s dryer
(50 ЊC, 0.5 Torr). 2–NaSCN complex: found: C, 58.5; H, 4.2; N,
7.0. Calcd. for C20H17N2O4SNa ϩ ½H2O: C, 58.1; H, 4.4; N,
6.8%.
1 For example: (a) G. W. Gokel and O. Murillo, in Comprehensive
Supramolecular Chemistry, ed. G. W. Gokel, Pergamon, Oxford,
1996, Volume 1, pp. 1–34 and literature cited therein; (b) G. W.
Gokel and S. H. Korzeniowski, Macrocyclic Polyether Syntheses,
Springer-Verlag, Berlin, 1982, Chapter 7 and literature cited therein;
(c) D. S. Lawrence, T. Jiang and M. Levett, Chem. Rev., 1995, 95,
2229.
2 R. Harned, P. Harter Hidy, C. J. Corum and K. L. Jones, Antibiot.
Chemother., 1951, 1, 594.
3 J. Berger, A. I. Rachlin, W. E. Scott, L. H. Sternbach and M. W.
Goldberg, J. Am. Chem. Soc., 1951, 73, 5295.
Computer-modeling calculations
4 (a) J. O. Gardner and C. C. Beard, J. Med. Chem., 1978, 21, 357;
(b) N. Yamazaki, S. Nakahama, A. Hirao and S. Negi, Tetrahedron
Lett., 1978, 2429; (c) H. Sieger and F. Vogtle, Liebigs Ann. Chem.,
1980, 425; (d ) K. Taguchi, K. Hiratani and H. Sugihara, Chem.
Lett., 1984, 1457; (e) K. Hiratani and K. Taguchi, Bull. Chem. Soc.
Jpn., 1987, 60, 3827; ( f ) W. Wierenga, B. R. Evans and J. A.
Woltersom, J. Am. Chem. Soc., 1979, 101, 1334; (g) W. C. Still,
P. Hauck and D. Kempf, Tetrahedron Lett., 1987, 2817; (h) P. W.
Smith and W. C. Still, J. Am. Chem. Soc., 1988, 110, 7917.
5 Y. Habata, M. Takeshita, Y. Fukuda, S. Akabori and J. S. Bradshaw,
J. Heterocycl. Chem., 2001, 38, 1323.
6 We carried out a document search using Sci Finder Scholer® ver.
2001 (American Chemical Society, 2001) for podands 1–3. There
were no references to these compounds.
7 M. Newcomb, J. M. Timko, D. M. Walba and D. J. Cram, J. Am.
Chem. Soc., 1977, 99, 6392.
The ab initio 3-21G(*) calculations using the MMFF geometry
from the conformer search were performed on a Power Book
G4 (500 MHz processor and 500 MB memory). Mac Spartan
Pro was used as the computer molecular modeling software.
13C NMR titration experiments
Titration experiments were carried out by addition of 0.5, 1.0,
1.5, 2.0 and 5.0 equiv. of NaSCN in CD3CN (0.002 mmol µLϪ1
)
to podands 1–3 in CD3CN (0.02 mmol in 0.65 mL) at 25 ЊC.
Crystallography†
Crystal and selected experimental data for podands 2 and 3,
and the 2–NaSCN complex are summarized in Table 3.
8 R. M. Silverstein, G. C. Bassler and T. C. Morrill, Spectrometric
identification of organic compounds, 4th edn., John Wiley & Sons,
1981, pp. 112–114.
9 R. M. Izatt, K. Pawlak and J. S. Bradshaw, Chem. Rev., 1991, 91,
2085 and literature cited therein.
Acknowledgements
10 N. Malhotra, P. Roepstorff, T. K. Hansen and J. Becher, J. Am.
Chem. Soc., 1990, 112, 3709.
11 S. L. De Wall, E. S. Meadows, L. J. Barbour and G. W. Gokel, J. Am.
Chem. Soc., 1999, 121, 5613.
12 (a) R. D. Gandour, F. R. Fronczek, V. J. Gatto, C. Miganti, R. A.
Schultz and G. W. Gokel, J. Am. Chem. Soc., 1986, 108, 4078;
(b) Y. Habata and S. Akabori, J. Chem. Soc., Dalton Trans., 1996,
3871.
This research was supported by Grants-in Aid for Scientific
Research (No. 09640698 and 12640566) from the Ministry of
Education, Culture, Sports, Science and Technology (Japan)
and The Nishida Research Fund For Fundamental Organic
Chemistry.
† CCDC reference numbers 175380–175382. See http://www.rsc.org/
suppdata/p1/b2/b201525k/ for crystallographic files in .cif or other
electronic format.
13 G. Weber, Acta Crystallogr., Sect. C, 1984, 40, 592.
14 We carried out a document search using ConQuest ver. 1.2
(Cambridge Crystallographic Data Centre, 2001).
J. Chem. Soc., Perkin Trans. 1, 2002, 865–869
869