Cation sensors containing a (bpy)Re(CO)3 group linked to an azacrown ether via an alkenyl or alkynyl spacer: Synthesis, characterisation, and complexation with metal cations in solution

Article abstract of DOI:10.1039/b401092b

Two [(bpy)Re(CO)₃L]⁺ complexes (bpy = 2,2′-bipyridine), where L contains an aza-15-crown-5 ether which is linked to Re via an alkenyl- or alkynyl-pyridine spacer, have been synthesised along with model complexes. Solutions of the com

Full text of DOI:10.1039/b401092b

View Article Online / Journal Homepage / Table of Contents for this issue
Cation sensors containing a (bpy)Re(CO)3 group linked to an
azacrown ether via an alkenyl or alkynyl spacer: synthesis,
characterisation, and complexation with metal cations in solution
Jared D. Lewis and John N. Moore*
Department of Chemistry, The University of York, Heslington, York, UK YO10 5DD
Received 23rd January 2004, Accepted 26th February 2004
First published as an Advance Article on the web 26th March 2004
Two [(bpy)Re(CO)3L]ϩ complexes (bpy = 2,2Ј-bipyridine), where L contains an aza-15-crown-5 ether which is linked
to Re via an alkenyl- or alkynyl-pyridine spacer, have been synthesised along with model complexes. Solutions of
the complexes in acetonitrile have been studied by UV-Vis absorption spectroscopy, and by 1D and 2D 1H NMR
spectroscopy. Strong UV-Vis bands, assigned to intraligand charge-transfer transitions localised at the L ligands,
blue shift on protonation of the azacrown nitrogen atom or on complexation of alkali-metal (Liϩ, Naϩ and Kϩ) or
alkaline-earth metal (Mg2ϩ, Ca2ϩ and Ba2ϩ) cations to the azacrown; the magnitude of the blue shift is dependent
on the cation, with protonation giving the largest shift of ca. 100 nm. Cation binding constants in the range of log
K = 1–4 depend strongly on the identity of the metal cation. Protonation or cation complexation causes downfield
shifts in the 1H NMR resonances from most of the azacrown and L ligand protons, and their magnitudes correlate
with those of the blue shifts in the UV-Vis bands; shifts in the azacrown 1H NMR resonances report on how the
different metal cations interact with the macrocycle. UV-Vis and 1H NMR spectra of the free L ligands enable the
effect of the Re centre to be assessed. Together, the data indicate that the alkene spacer gives a more responsive
sensor than the alkyne spacer by providing stronger electronic communication across the L ligand.
Introduction
There is considerable interest in molecules which bind cations in
solution, and which incorporate functional groups that signal
cation binding via the change in a molecular property that is
easily detected.1,2 Such cation sensors have potential uses
ranging from components in photonic communication devices
to monitors of cation-dependent biological processes.3,4 Many
reported systems contain crown ethers or their derivatives as
receptors,5 with the binding of a cation inducing changes in the
geometric or electronic structure of a signalling unit attached
to the crown; such changes can usually be detected by con-
ventional spectroscopy or by other simple physical techniques.
Azacrown ethers are common motifs in cation sensors
because the azacrown nitrogen atom has a lone pair that can
form part of a delocalised electron system, and so can provide
very effective electronic communication between the receptor
and the rest of the molecule, including the signalling unit. In
many examples, the azacrown is attached to an organic dye and
cation binding elicits a colour change that is easily monitored
by UV-Vis spectroscopy.5–7 Our interest in such systems extends
from sensors to “light-controlled ion switches” which are
designed so that excitation of the chromophore results in cation
ejection from the azacrown through effective electronic com-
munication with the nitrogen atom which lowers the binding
constant significantly on excitation.8,9
In two recent communications, we presented two novel
systems in which an azacrown ether is linked to a (bpy)Re(CO)3
group via an alkenyl (1a)10 or alkynyl (2a)11 spacer (Scheme 1),
and we reported two notable photochemical properties:
protonation of the azacrown regulates trans–cis photoisomeris-
Scheme 1 Structures of complexes 1a, 1b, 2a and 2b.
ation in the alkenyl system,10 and Ba2ϩ bound to the azacrown is
ejected rapidly on excitation of the alkynyl system.11 The
rationale behind these molecular designs is that incorporating a
transition-metal redox centre may bring additional excited-state
decay routes and greater versatility in design than all-organic
chromophore-azacrown systems. In particular, the (bpy)-
Re(CO)3 group is being used increasingly in sensors because its
metal-to-ligand charge-transfer (MLCT) excited state emits
with a moderate quantum yield, and in photochemically active
systems because it can undergo excited-state electron- and
energy-transfer reactions.12–14 If the intention of incorporating
a transition-metal centre is to mediate the binding properties
of the azacrown, then clearly strong communication between
these groups is required; consequently, complexes 1a and
2a have been designed with two different conjugated spacers
in order to explore whether this may enhance their effective-
ness as ion switches over other designs with weaker com-
munication, such as those incorporating amide spacers.13
Moreover, designs with strong electronic communication across
the molecule may also be expected to act as effective sensors, on
the basis of our earlier work on azacrowns conjugated with
organic dyes.7,15
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
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 4
1376

View Article Online
Scheme 2 Syntheses of L1a and L2a.
In this paper we focus on the ground states of 1a and 2a, and
in place of 6 to prepare L2b. All of the Re complexes were
on their binding of metal cations. We report in full details of
the synthesis and characterisation of these two azacrown
complexes, their models 1b and 2b, and the respective L ligands
prepared by a procedure reported for different L ligands:13,21
(bpy)Re(CO)3Cl was prepared initially and then CF3SO3H was
used to produce (bpy)Re(CO)3(CF3SO3), which was reacted
with <1 equivalent of the appropriate L ligand to generate its
Re complex.
1
without the Re group attached. We report UV-Vis and H
NMR studies of metal cation binding to the azacrown com-
plexes 1a and 2a, including the measurement of binding con-
stants. Parallel studies of the ligand L1a enable the effect of the
Re centre within this generic design to be assessed, while a
comparison of the results from 1a and 2a enables the effect of
using alkene versus alkyne spacers to be considered. In addition
to providing details of these systems as cation sensors, this
work establishes a quantitative understanding of the ground-
state spectroscopy and cation binding: this is an essential
requirement prior to photochemical studies of ion switching.
Characterisation
The UV-Vis spectra of the 1a, 1b, 2a and 2b complexes, and the
L1a and L1b ligands, all in acetonitrile, are shown in Fig. 1.
Both of the 1a and 2a complexes give intense visible absorption
bands, at 428 and 404 nm, respectively, and a comparison of the
1a and L1a spectra shows that they arise from transitions
localised on the respective L ligand: they are assigned to intra-
ligand charge-transfer (ILCT) transitions, in which charge is
transferred from the azacrown nitrogen donor to the ethenyl-
pyridyl or ethynylpyridyl acceptor.13 The model complexes 1b
and 2b give similar bands, at ca. 330 and 310 nm, respectively,
which are also assigned to transitions localised on the respective
L ligand: the shift to higher energy and the decrease in absorp-
tion strength may be attributed to a loss of charge-transfer
character due to the absence of an azacrown electron donor in
these ligands. All of the complexes also show weaker bands at
Results and Discussion
Synthesis
The ligands L1a and L2a were prepared according to Scheme 2.
The benzaldehyde 3 served as a common precursor to both of
these ligands, and was prepared from the phenylaza-15-crown-5
via a Vilsmeyer reaction.16 In the preparation of L1a, 4-methyl-
pyridine was selectively deprotonated at the methyl group using
LDA and reacted with 3 to produce the alcohol 4 which was
dehydrated to generate L1a.17 In the preparation of L2a, 3 was
converted first to the alkene 5 via a Wittig reaction with the
ylide reagent generated from mixing PPh3 with CBr4 under
anhydrous conditions (Corey–Fuchs method).18 This dibromo-
vinyl derivative 5 was reacted with BuLi to generate a lithium
acetylide19 which afforded 6 on aqueous work-up; 6 was
coupled to the 4-position of a pyridine ring through a
Sonogashira–Hagihara cross-coupling reaction20 to generate
L2a. The models were prepared in an identical manner to the
respective azacrown analogues except that benzaldehyde was
used in place of 3 to prepare L1b, and ethynylbenzene was used
<320 nm which can be assigned to π
on bipyridine and L ligand groups. The Re
π* transitions localised
bpy MLCT
absorption bands of the complexes, which occur at ca. 350 nm
in related complexes and typically have peak absorption
coefficients of εmax ≈ 4000 molϪ1 dm3 cmϪ1,14,22 underlie the
strong L-centred bands, which have εmax ≈ 40000–50000 molϪ1
dm3 cmϪ1.
The 500 MHz 1H NMR spectrum of 1a in CD3CN is shown
in Fig. 2, and assignments according to the numbering in
Scheme 3 are given in Table 1 along with those for the model
complex 1b and the ligand L1a. The spectra of 2a and 2b are
similar to those of 1a and 1b, and the same numbering scheme
is used; H12 and H13 are absent in the alkynes.
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
1377

View Article Online
The spectrum of 1a is assigned as follows. The group of
resonances at ca. δ 3.5, with a combined integration of 20H, is
readily assigned to protons on the azacrown ether ring, with
individual protons assigned according to the literature.15 The
Scheme 3 Numbering scheme for assignment of 1H NMR resonances
of 1a; an identical numbering scheme was used for 2a, with H12 and
H13 absent.
other resonances are assigned according to the expected
pattern, in which the bipyridyl group gives two doublets and
two triplets, in the range δ 7.5–9.3 reported for 2,2Ј-bpy,23 the
pyridyl and aryl groups of L each give a characteristic AAЈBBЈ
pattern of two doublets, and the alkene group gives two doub-
lets. 1H–1H COSY and NOESY spectra were used to assign the
resonances of 1a to each of these fragments. COSY cross-peaks
between the triplets at δ 8.26 and 7.79 and the doublets at δ 8.37
and 9.22, respectively, indicated that these four resonances
could be assigned to the bpy ligand. A NOESY cross-peak
between the bpy resonance at δ 9.21 and the doublet at δ 8.01
allowed these resonances to be assigned to H(24,29) on bpy
and to H(21,22) on the pyridyl ring of L, respectively. The
remainder of the bpy resonances and the other pyridyl reson-
ance were then assigned using the COSY cross-peaks. Two
NOESY cross-peaks were also present between the azacrown
resonances at δ 3.53–3.68 and a doublet at δ 6.70 which was
therefore assigned to H(16,18) on the aryl ring of L, enabling
Fig. 1 UV-Vis absorption spectra of complexes 1a, 1b, 2a and 2b, and
the ligands L1a and L1b, in acetonitrile.
Fig. 2 500 MHz 1H NMR spectra of 1a and of 1a with selected cation salts added in large excess, all in CD3CN. Dashed lines show shifts in the
resonances of H(15,19) H(20,23), H(12), and H(16,18).
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
1378

View Article Online
Table 1 Chemical shifts, δ (in ppm) from the 500 MHz 1H NMR spectra of 1a, 1b, L1a, 2a and 2b in CD3CN, along with the shifts in the resonances,
∆δ = δbound Ϫ δfree, on complexation of 1a and 2a with cations; numbering given in Scheme 3
Compound H(5,6) H(1,10) H(3,4,7,8) H(2,9) H(16,18) H(12) H(20,23) H(13) H(15,19) H(25,29) H(21,22) H(26,30) H(27,31) H(24,28)
δ
1b
––
3.56
3.53
––
3.57
3.55
––
3.60
3.57
–– 7.40
3.71 6.73
3.68 6.70
7.08
6.89
6.76
7.40
7.40
7.25
7.47 7.59
7.35 7.44
7.35 7.39
7.81
––
7.79
8.14
8.47
8.01
8.28
––
8.26
8.39
––
8.37
9.24
––
9.22
L1a
1a
∆δ
1a–Liϩ
1a–Naϩ
1a–Mg2ϩ
1a–Ca2ϩ
1a–Ba2ϩ
1a–Hϩ
0.08
0.06
0.04
0.08 0.14
Ϫ0.13 0.25
0.04 0.57
0.10 0.48
0.02 0.40
0.21 1.11
0.07
0.10
0.29
0.14
0.16
0.47
0.04
0.03
0.14
0.02
0.06
0.19
0.00 0.05
Ϫ0.02 0.04
Ϫ0.07 0.11
Ϫ0.06 0.10
Ϫ0.02 0.15
0.17 0.42
Ϫ0.01
Ϫ0.01
Ϫ0.12
Ϫ0.08
Ϫ0.03
0.06
0.03
0.03
0.00
Ϫ0.02
0.02
0.19
Ϫ0.01
Ϫ0.04
Ϫ0.14
Ϫ0.10
Ϫ0.05
0.04
0.01
Ϫ0.01
Ϫ0.10
Ϫ0.08
Ϫ0.04
0.07
0.02
Ϫ0.05
Ϫ0.13
Ϫ0.11
Ϫ0.09
0.02
0.02 Ϫ0.23 Ϫ0.02
0.22 Ϫ0.27
Ϫ0.24
0.17
0.07
0.18
0.21
0.13
0.13
0.23
0.15
0.05
δ
2b
2a
––
3.53
––
3.55
––
3.57
–– 7.45
3.68 6.69
––
––
7.35
7.25
–– 7.56
–– 7.33
7.81
7.79
8.22
8.12
8.28
8.27
8.39
8.38
9.22
9.21
∆δ
2a–Liϩ
0.12
0.10
0.05
0.10 0.13
Ϫ0.10 0.21
0.07 0.50
0.10 0.50
0.07 0.40
–– Ϫ0.02
–– Ϫ0.01
–– 0.04
–– 0.03
–– 0.10
–– 0.11
–– 0.13
Ϫ0.01
Ϫ0.04
Ϫ0.14
Ϫ0.08
Ϫ0.05
0.02
Ϫ0.01
Ϫ0.02
Ϫ0.02
Ϫ0.01
Ϫ0.02
Ϫ0.06
Ϫ0.16
Ϫ0.11
Ϫ0.07
0.01
Ϫ0.04
Ϫ0.11
Ϫ0.09
Ϫ0.06
Ϫ0.03
Ϫ0.07
Ϫ0.15
Ϫ0.12
Ϫ0.10
2a–Naϩ
2a–Mg2
2a–Ca2ϩ
2a–Ba2ϩ
0.03 Ϫ0.20 Ϫ0.01
0.23 Ϫ0.26
Ϫ0.24
0.22
0.18
0.21
0.18
––
–– Ϫ0.04
–– 0.01
0.03
0.23
0.20
the final doublet at δ 7.25 to be assigned to H(20,23) on the
pyridyl ring of L. The two doublets at δ 6.76 and 7.35, with
integrations of 1H, were assigned to the vinyl protons; the large
coupling constants of ca. 17 Hz for these doublets indicate that
1a was prepared exclusively as the trans-isomer. One NOESY
cross-peak between the pyridyl resonance at δ 7.25 and the vinyl
resonance at δ 7.35 enabled it to be assigned to H(13) due to its
close proximity to H(23); and another cross-peak between the
aryl resonance at δ 7.39 and the other vinyl resonance at δ 6.76
enabled it to be assigned to H(12) due to its similarly close
proximity to H(19).
The resonances in the spectrum of 2a were assigned in an
identical manner to those of 1a, and those of 1b, L1a, and 2b
were assigned readily using the assignments for 1a and 2a.
There is a significant downfield shift in the resonances of the L
ligands on going from 1a and 2a to 1b and 2b, respectively, with
shifts of ca. 0.7 ppm for H(16,18) and ca. 0.2 ppm for H(15,19)
on the aryl rings, and ca. 0.1 ppm for H(20,23) and H(21,22) on
the pyridyl rings; the resonances from the vinyl protons of 1
also show downfield shifts of 0.32 ppm for H(12) and 0.12 ppm
for H(13). These shifts indicate that the L ligand protons are
more deshielded in 1b and 2b due to the absence of the electron-
donating N-atom of the azacrown; the magnitude of the shift
correlates broadly with the distance of the proton from the
azacrown. By contrast, the resonances from the bpy protons
show very small downfield shifts of 0.01–0.02 ppm, indicating
that these protons are essentially unaffected by the absence of
the azacrown.
Fig. 3 UV-Vis absorption spectra of 1a (ca. 2.5 × 10Ϫ5 mol dmϪ3) and
2a (ca. 2.1 × 10Ϫ5 mol dmϪ3) in acetonitrile containing tetrabutyl-
ammonium perchlorate at 0.1 mol dmϪ3 to maintain constant ionic
strength, and at 23 ЊC, with NaClO4 at (a) 0, (b) 2.7 × 10Ϫ4, (c) 1.4 ×
10Ϫ3, (d) 2.7 × 10Ϫ3, (e) 5.4 × 10Ϫ3, (f ) 1.1 × 10Ϫ2, (g) 2.7 × 10Ϫ2 and (h)
0.14 mol dmϪ3. Insets: absorbance versus NaClO4 concentration for 1a
at 428 nm and for 2a at 404 nm; the solid lines are the best fits to
eqn. (2).
Similar changes to those observed on addition of NaClO4
were observed on addition of other alkali or alkaline-earth
metal salts or HCl to 1a and 2a. The spectra obtained under
limiting conditions of high salt concentrations are shown in
Fig. 4, along with those for L1a (discussed below).
The largest blue shift of the ILCT band is caused by proton-
ation (Table 2), and the spectra of 1a–Hϩ and 2a–Hϩ (Fig. 4)
are notably similar to those of 1b and 2b (Fig. 1). These spectra
indicate that Hϩ binds exclusively to the azacrown N-atom,25
and that the electron-donor properties of the azacrown
nitrogen atom are effectively removed in both cases. The mag-
nitude of the blue shift is smaller for all of the other cations,
and it varies with cation identity (Fig. 4; Table 2). The shift is
larger for the di-cations than the mono-cations, and Fig. 5
shows that it correlates broadly with the charge density of the
cation.26 Thus, the small and doubly charged Ca2ϩ elicits a
larger shift than the larger Ba2ϩ, and both cause larger shifts
Protonation and cation complexation
The UV-Vis spectra of 1a and 2a in acetonitrile show a blue
shift in the ILCT absorption band as NaClO4 is added (Fig. 3):
no further changes were observed above a salt concentration
of ca. 0.14 mol dmϪ3, and the presence of isosbestic points
indicates that an equilibrium between two species is observed.
Similar blue shifts in ILCT absorption bands have been
observed on cation binding to the azacrown in related com-
plexes,5,7 and therefore the spectra indicate the formation of
1a–Naϩ and 2a–Naϩ, respectively, according to the general case
in eqn. (1).
(1)
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
1379

View Article Online
Table 2 Absorption maxima, λmax (in nm), of 1a, 2a and L1a in
acetonitrile, and blue shifts, ∆λ (in nm) on cation complexation, along
with cation radii for six-coordination, r (in Å),24 and charge-to-radius
ratios, (z/r) (in ÅϪ1
)
1a
2a
L1a
Cation
r
z/r
λmax
∆λ
λmax
∆λ
λmax
∆λ
none
Liϩ
––
––
428
402
385
404
342
351
362
319
––
26
43
24
86
77
66
109
404
386
376
390
327
341
350
307
––
18
28
14
77
63
54
97
374
361
342
355
324
319
329
––
––
13
32
19
50
55
45
––
0.76
1.02
1.38
0.72
1.00
1.49
––
1.32
0.98
0.72
2.78
2.00
1.34
––
Naϩ
Kϩ
Mg2ϩ
Ca2ϩ
Ba2ϩ
Hϩ
Fig. 5 The hypsochromic shift on complexation (∆λ) versus the cation
charge-to-radius ratio (z/r) for 1a, 2a and L1a; solid lines illustrate the
general trend.
Shifts in resonances on cation binding (∆δ = δbound Ϫ δfree) are
given for 1a and 2a in Table 1, and identified by dashed lines
in Fig. 2 for the H(16,18), H(15,19), H(12), and H(20,23)
resonances of 1a. In each case, the largest shift on binding is a
downfield shift in the resonance of H(16,18), located on the aryl
ring of L and nearest to the azacrown. There are also relatively
large downfield shifts in the resonances of the other aryl ring
protons H(15,19), and both up- and downfield shifts in the
resonances of various azacrown protons H(1–10) depending on
the identity of the bound cation. For 1a, there are relatively
large downfield shifts in the resonance of the vinyl proton
H(12), and generally small upshifts in that of the other vinyl
proton H(13). Shifts in the resonances of protons on the pyridyl
ring of L, and on the bpy ligand, are generally small. These
observations are consistent with cation binding at the aza-
1
crown, and with the H NMR assignments made above for 1a
and 2a.
The downfield shifts in the H(16,18) and H(15,19) resonances
of 1a and 2a may be attributed to deshielding of the aryl
protons due to the electrostatic effect of the cation bound to the
1
adjacent azacrown; similarly large shifts in the aryl ring H
Fig. 4 UV-Vis absorption spectra of 1a, 2a, and L1a in acetonitrile,
and in the presence of selected cations at concentrations (ca. 0.14 mol
dmϪ3) approaching saturation of the spectral changes.
resonances on cation binding have been reported for organic
azacrown-containing molecules with similar structures.15,27 The
magnitude of the shift in the aryl ring and vinyl proton
resonances versus cation identity shows a trend which is broadly
consistent with that of the magnitude of the ILCT band shifts
in the UV-Vis spectra of 1a and 2a: Hϩ elicits the largest
downfield shift in H(16,18) and H(12), followed by Mg2ϩ, Ca2ϩ
and Ba2ϩ; Naϩ and Liϩ cause significantly smaller shifts, with
the small effect of Liϩ being consistent with it binding preferen-
tially to the azacrown oxygen atoms. This trend is consistent
with Hϩ and Mg2ϩ causing the greatest disruption to the
electronic structure of the L ligand via a strong interaction with
the azacrown nitrogen atom.
than singly charged Naϩ and Kϩ, which are of similar respect-
ive size. Liϩ gives a shift which is smaller than that predicted by
the general trend, consistent with this hard cation binding
prefentially to the oxygen atoms and interacting only weakly
with the nitrogen atom, as discussed in the literature.7,25,27,28
Mg2ϩ is similar in size to Liϩ but, by contrast, it gives the largest
shift apart from Hϩ, indicating that Mg2ϩ interacts strongly
with the azacrown N-atom. The trend in bandshift versus cation
is similar for 1a and 2a, consistent with the presence of the same
azacrown in both complexes.
The 1H NMR spectra of 1a and 2a in CD3CN show shifts in
the resonances on addition of alkali and alkaline-earth metal
salts, and on addition of HCl.29 Sample spectra are shown
in Fig. 2 for 1a samples at limiting salt concentrations which
UV-Vis spectra confirmed were of fully complexed forms.30,31
The various up- and downfield shifts in the resonances of the
azacrown protons H(1–10) may be attributed to both the
location of the bound cation and to the changes it induces in
the conformation of the azacrown ring, both of which may be
expected to depend strongly on the identity of the cation. The
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
1380

View Article Online
Table 3 Binding constants (log K) for complexation of metal cations
with 1a and 2a in acetonitrile, along with ranges reported for various
related aza-15-crown-5-ether dyes in acetonitrile6,7,27,33,34
log K
Cation
1a
2a
N-15-C-5 dyes
Liϩ
2.63 0.03
2.05 0.03
1.40 0.15
2.98 0.07
4.30 0.11
3.16 0.24
2.44 0.06
1.82 0.03
1.04 0.04
2.76 0.06
3.32 0.08
2.80 0.06
1.7–2.8
1.4–2.2
1.3–2.2
1.6–2.7
1.7–4.2
1.9–3.6
Naϩ
Kϩ
Mg2ϩ
Ca2ϩ
Ba2ϩ
Scheme 4 Resonance structures of L1a.
pattern of the azacrown ring resonances reports on the location
of the cation. In the absence of a bound cation, the resonances
divide into four sets, indicating that four magnetic environ-
ments exist: H(1,10) adjacent to the N-atom; H(2,9) next
removed; H(3,4,7,8) in the centre; and H(5,6) furthest from the
N-atom (Fig. 2; Table 1). On binding Hϩ, Liϩ, or Mg2ϩ, which
are small cations that locate in specific parts of the ring, a
shifted but similar pattern to that in the absence of the cation
is maintained, with three (Hϩ) or four (Liϩ, Mg2ϩ) distinct
resonances indicating that the ring retains three or four differ-
ent environments. On binding Ba2ϩ, which is a large cation
that is expected to locate above the centre of the ring, a quite
different pattern is obtained with all of the resonances
occurring at a similar chemical shift (Fig. 2), indicating that all
of the protons experience a similar environment. Naϩ and Ca2ϩ,
which are of similar and intermediate size, both give an inter-
mediate pattern of two distinct resonances.
Fig. 6 Structures calculated by DFT for ML1a, ML2a, and ML1a-
Me, which model L1a, L2a, and 1a, respectively (see text); selected
bond lengths given in Å.
Cation binding constants
been adopted successfully in reported semiempirical and DFT
calculations.36–38
For 1:1 complexation according to eqn. (1), binding constants
K can be calculated from spectrophotometric titrations using
eqn. (2), and where: A0, A, and A∞ are the absorbances at a fixed
wavelength in the absence of Mnϩ, in the presence of Mnϩ, and
in the limiting case of infinite cation concentration, respect-
ively; and the total ligand and total metal concentrations
are [L]tot = [L] ϩ [LMnϩ] and [Mnϩ]tot = [Mnϩ] ϩ [LMnϩ], respect-
ively.32
The insets within Fig. 3 illustrate non-linear regression
analyses using eqn. (2), with A0 fixed at the value in the absence
of added NaClO4 and the limiting value of A∞ fitted as one of
the parameters. The data fit well to eqn. (2), substantiating the
validity of applying eqn. (1) and providing binding constants of
K = 112 8 and 66 5 dm3 molϪ1 for Naϩ with 1a and 2a,
respectively. All of the cation binding constants obtained by
this method are given in Table 3.
The binding constants obtained for 1a and 2a are comparable
to those reported both for an analogous complex in which
aza-15-crown-5 is linked to (bpy)Re(CO)3 by an amide spacer,13
and for other systems in which aza-15-crown-5 is linked to a
dye (Table 3),6,7,27,33,34 but they are smaller than those for aza-
15-crown-5 itself (e.g., log K = 5.68 for Naϩ).35 This effect has
been attributed to a decrease in the electron density on the
azacrown nitrogen atom arising from its interaction with an
attached chromophore in dye systems, as illustrated by the
resonance structures shown in Scheme 4,7 and it may be
stronger in the Re complexes 1a and 2a, which carry a positive
charge. To assist in rationalising the magnitudes of the binding
constants, we have performed DFT calculations on ML1a
and ML2a (Fig. 6), which serve as models for L1a and
L2a, respectively, using diethylamino groups in place of
azacrowns to simplify the calculation. This approach has
The calculated structures of ML1a and ML2a show that the
ligands are remarkably planar between the two nitrogen atoms,
which is consistent with observed X-ray crystallographic
structures of similar systems15 and indicates that there is strong
conjugation throughout the pyridyl and aryl ring systems and
the alkenyl or alkynyl spacers.7 Selected bond lengths given in
Fig. 6 show that the aryl and pyridyl rings are both partially
quinoidal, which is consistent with a strong contribution to the
ground-state electronic structure from the charge-separated
resonance form shown in Scheme 4. In both ML1a and ML2a,
the short bond length between the aryl ring and the azacrown
N-atom (ca. 1.392 Å), in particular, indicates that there is strong
interaction between the nitrogen lone pair and the aryl ring
π-electrons. Fig. 6 also shows the results of a DFT calculation
on ML1a-Me, in which ML1a has been methylated at the
pyridyl nitrogen atom to create a positively charged species
to model the effect of coordination to a positively charged Re
atom, as in 1a. The bond lengths, and particularly the shorter
bond between the aryl ring and the azacrown N-atom (1.364
Å), are consistent with an increased contribution from the
charge-separated form on attaching an electron-acceptor at the
pyridyl group. Thus, the DFT calculations on these models
provide further support for the interpretation of the relatively
low binding constants of 1a and 2a represented by Scheme 4.
The variation in the magnitude of the cation binding con-
stants with cation identity (Table 3) do not correlate with those
of the cation-induced bandshifts (Table 2), as also reported for
other azacrown sensors.7,36,39 However, the variation in binding
constant with cation identity observed here for both 1a and 2a
is similar to that reported for other systems in which aza-15-
crown-5 is linked to a dye (Table 3): it has been rationalised
(2)
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
1381

View Article Online
through a combination of effects, including size-matching
between azacrown and cation, the thermodynamics of cation
desolvation required for azacrown complexation, and the
nature of the interaction between these hard cations and the
hard oxygen and soft nitrogen atoms of the macrocycle.40 Ca2ϩ,
in particular, binds strongly to the azacrown: its diameter of
2.00 Å is well-matched to the aza-15-crown-5 cavity diameter of
1.7–2.2 Å. Naϩ has a similar diameter of 2.04 Å but a much
smaller binding constant: the higher charge density of Ca2ϩ
may mean that the enthalpy change on binding with the aza-
crown and/or the gain in entropy associated with desolvation
are more favourable.
alkene versus alkyne spacer systems,43 which attributes the
difference to comparatively poor sp2–sp π-overlap in aryl–
alkyne systems, arising from a mismatch in the respective
p-orbital energies;44 the DFT calculations suggest that the
systems studied here are both approximately planar and, hence,
that twisting of the aryl rings due to steric effects45 is unlikely to
be significant in either system. The trend in the magnitude of
the ILCT band shift with cation identity is the same for 1a and
2a (Table 2, Fig. 5), and the larger shifts for 1a than 2a may be
attributed to stronger electronic communication in the alkene
system. The differences in the UV-Vis spectra of 1a and 2a
provide useful information on the relative strength of electronic
communication, but they are relatively small; Figs. 4 and 5 show
that both complexes can act effectively as sensors.
Effect of the Re centre on sensor properties
1
The H NMR spectra of 1a and 2a are closely similar. The
The changes in the UV-Vis and NMR spectra of 1a and 2a on
cation complexation arise principally from changes that are
localised at the L ligand; the free ligands L1a and L2a them-
selves can act as cation sensors in the absence of the Re centre.
largest differences on going from 1a to 2a are a downfield shift
of 0.11 ppm in the resonance of H(21,22) on the pyridyl ring
and an upfield shift of 0.06 ppm H(15,19) on the aryl ring.
These shifts are consistent with the interpretation that the
ground state of 1a has greater charge-transfer character than
that of 2a due to stronger electronic communication. The shifts
1
A comparison of the UV-Vis and H NMR spectra of the free
ligand L1a with those of the the Re complex 1a enables the
effect of the Re centre to be assessed.
The UV-Vis spectra of L1a obtained on addition of salts at
limiting concentrations are shown in Fig. 4. The ILCT band of
L1a (λmax = 374 nm; εmax = 24000 dm3 mo1Ϫ1 cmϪ1) is at shorter
1
in the H NMR resonances with cation identity are broadly
similar for 1a and 2a.
The cation binding constants of 1a and 2a are within an
order of magnitude of each other, but the values for 1a are
consistently higher than those for 2a (Table 3: typically, log K
ca. ϩ0.3 higher; K ca. ×2 higher). This is the converse of what
would be expected if the relative magnitudes of the binding
constants were determined simply by the extent of charge-
transfer in the ground state (Scheme 4), which is greater for 1a
than 2a. However, the cation binding constants of a large range
of alkene-linked aza-15-crown-5-dyes have been shown not to
correlate with the charge on the azacrown N-atom but, instead,
to correlate with the charge on the complete NR2 functionality,
including the alkyl groups.36 This correlation indicates that the
binding constant is determined not only by the charge at the
azacrown N-atom but also by the charges on the alkyl groups
resulting from hyperconjugation, i.e. R N electron donation
via σ–π mixing that increases with increasing pyramidalization
at the nitrogen atom.
The structures of the models calculated by DFT (Fig. 6)
show that ML1a is partly pyramidal whereas ML2a is essen-
tially planar, as quantified by the sums of the bond angles
around the azacrown nitrogen atom: the calculated values are
356.1Њ and 359.9Њ, corresponding to differences from planarity
of 3.9Њ and 0.1Њ for ML1a and ML2a, respectively. Moreover,
although the calculated negative charge on the N-atom itself
(QN) is lower for ML1a than ML2a, consistent with Scheme 4,
the calculated positive charge on the whole diethylamino group
(QDEA) is lower for ML1a than ML2a, consistent with increased
σ–π mixing due to greater pyramidalisation:36 the calculated
values are QN = Ϫ0.493 and Ϫ0.503, and QDEA = ϩ0.364 and
ϩ0.384 for ML1a and ML2a, respectively. Thus, the calcu-
lations on ML1a and ML2a suggest that the higher binding
constants of 1a than 2a can be rationalised in terms of partial
pyramidalisation at the azacrown N-atom in the alkene system
resulting in an increased partial positive charge on the whole
azacrown.
wavelength and is weaker than that of 1a (λmax = 428 nm; εmax
=
51000 dm3 mo1Ϫ1 cmϪ1), and the magnitude of the blue shift on
cation binding is smaller for L1a than for 1a (Table 2). The
observation that 1a is a more responsive UV-Vis sensor than
L1a may be attributed to the increased charge polarisation that
results from coordinating the positively charged Re centre, as
indicated by the DFT calculations discussed above. The ILCT
absorption bands of similar donor-acceptor ligands have also
been observed to red shift on coordination to metal centres,
with the magnitude of the shift dependent on the charge of the
metal centre and on ancillary ligands.41,42 Therefore, the
increased responsiveness in UV-Vis cation sensing reported
here on coordination of L1a to Re is likely to be a general effect
which could be enhanced by coordination to metal centres
which are stronger acceptors than the (bpy)Re(CO)3 group,
e.g., to highly charged metal centres with strong π-acid
ligands. However, an increase in the UV-Vis spectral response
due to increased charge polarisation may be accompanied by
a decrease in binding constants, and hence in a decrease in
sensitivity to cation concentration, as discussed above.
1
The largest changes in the H NMR spectra on going from
L1a to 1a are upfield shifts of 0.46 and 0.15 ppm in the
resonances of H(21,22) and H(20,23), respectively, on the pyr-
idyl ring (Table 1). The resonance of the vinyl proton H(12) also
shifts upfield, by 0.13 ppm, whereas that of the other vinyl
proton H(13) does not shift. The aryl proton resonances show
small upfield shifts of 0.05 and 0.03 ppm for H(15,19) and
H(16,18), respectively. These shifts indicate that the L ligand
protons are more shielded in 1a than L1a, and that the mag-
nitude of the effect correlates broadly with the distance of the
proton from the Re centre. These observations are consistent
with the interpretation that coordination to the Re centre
increases the electron demand at the pyridyl ring of the L ligand
and enhances the charge-transfer character of the ground state
(Scheme 4).
Conclusions
Alkene versus alkyne spacers
These studies establish a quantitative understanding of the UV-
1
The ILCT bands of 1a and L1a are at longer wavelength and
are more intense than those of 2a and L2a, respectively (Fig. 1),
indicating that the charge-transfer state is at lower energy and
that there is a larger transition dipole moment for 1a and L1a
than for 2a and L2a. Hence, the electronic communication
between the azacrown electron-donor and the pyridyl electron-
acceptor in these systems is stronger via an alkene than an
alkyne spacer. This is consistent with the general literature on
Vis and H NMR spectra of the azacrown-containing [(bpy)-
Re(CO)3L]ϩ complexes 1a and 2a, and of their protonation and
complexation of alkali and alkaline-earth metal cations. Strong
ILCT bands enable these systems to act as responsive UV-Vis
sensors to cations. A stronger spectral response than the free L
ligand is obtained by attaching the Re group, which causes an
increase in the charge-transfer character of the system; and a
stronger spectral response by 1a than 2a is attributed to
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
1382

View Article Online
stronger electronic communication via an alkene rather than an
alkyne spacer. The 1H NMR spectra of 1a and 2a show shifts in
the resonances on cation binding which provide complementary
information on structure, with the pattern of azacrown reson-
ances reporting on the location of the cation within the ring.
The lower cation binding constants of 1a and 2a than aza-15-
crown-5 ether itself are attributed to the charge-transfer
character of these systems, while stronger binding constants for
1a than 2a can be rationalised by the electronic effects of partial
pyramidalisation at the azacrown N-atom in 1a. Our interest
in 1a and 2a extends to their possible application as light-
controlled cation switches, and our studies of their photo-
chemistry will be reported elsewhere.
Syntheses
L ligands
4-(4,7,10,13-Tetraoxa-1-azacyclopentadecyl)benzaldehyde
(3). This was prepared by following a reported procedure.16 Pale
1
green crystals were obtained. Yield 72%; H NMR (270 MHz,
CDCl3) δ 3.6–3.8 (m, aliphatic crown) 6.67 (d, 2H, J 9.2 Hz,
aromatic), 7.68 (d, 2H, J 9.2 Hz, aromatic), 9.70 (s, 1H, CHO).
1-[4-(4,7,10,13-Tetraoxa-1-azacyclopentadecan-1-yl)phenyl]-
2]pyridin-4-ylethanol (4). This was prepared by following a
reported procedure for an NEt2 derivative.17 Diisopropylamine
(651 mg, 6.4 mmol) was dissolved into 2 cm3 of dry THF under
N2. BuLi solution (4 cm3, 6.4 mmol) in 5 cm3 THF was added
under N2 at 0 ЊC and stirred for 0.5 h. 4-Methylpyridine
(600 mg, 6.4 mmol) was added at Ϫ78 ЊC and the solution
turned deep red. After stirring for 90 min, 3 (1.98 g, 6.4 mmol)
in 10 cm3 THF was added dropwise and the now light-red
solution was stirred for 5 h. H2O (30 cm3) was then added and
the solution turned yellow. The mixture was extracted ×3 into
Et2O and the organic layer retained and dried over MgSO4. The
solvent was removed and the oily residue dissolved into the
minimum amount of hot EtOH. H2O was added on cooling
and the mixture left to stand overnight. Pale yellow needles
were obtained. 1H NMR (270 MHz, CDCl3) δ 3.00 (d, 2H, J 8.0
Hz, CH2CHOH), 3.6–3.8 (m, 20H, aliphatic crown), 4.79
(t, 1H, J 8.0 Hz, CH2CHOH), 6.62 (d, 2H, J 8.1 Hz, Ar),
7.1–7.2 (m, 4H, Ar, pyridyl) 8.40 (d, 2H, J 5.4 Hz, pyridyl).
Experimental
Spectroscopic and computational methods
UV-Vis spectra were recorded with a Hitachi U-3010 spectro-
meter, using chromophore concentrations in the range (1.6–2.5)
× 10Ϫ5 mol dmϪ3 and at 23 ЊC. Solutions of 2a with added
cation perchlorate salt were prepared in the dark in order to
avoid any photoisomerisation.10 Anhydrous acetonitrile was
used to determine binding constants, and tetrabutylammonium
perchlorate at 0.1 mol dmϪ3 was used to maintain constant
ionic strength. UV-Vis spectra were processed using Grams 386
software (Galactic Industries Corp.), and binding constants
were obtained by nonlinear regression analyses performed
using SPSS software (SPSS Inc.) with uncertaintities quoted as
95% confidence limits. 270 MHz 1H NMR spectra were
recorded to characterise the synthesis products using a JEOL
EX spectrometer, with CDCl3 as solvent and referenced against
the protiated solvent signal at 7.24 ppm. 500 MHz 1D and 2D
1H NMR spectra were recorded for more detailed studies using
a Bruker AMX 500 spectrometer, with CD3CN as solvent and
referenced against the protiated solvent signal at 1.96 ppm.
NMR data were recorded at a temperature of 300 K and
at chromophore concentrations of ca. 1 × 10Ϫ3 mol dmϪ3.
Electrospray ionisation mass spectra (ESI-MS) were recorded
on a Finnigan LCQ instrument: typical isotope patterns were
observed, and the values quoted here correspond to the more
abundant 187Re isotope. Elemental analyses were performed by
Elemental Microanalysis Ltd. Density functional theory (DFT)
calculations were performed using Gaussian 98 software:46
all calculations used the B3LYP functional and the 6-31G(d)
basis set.
1-{4-[(E )-2-pyridinyl-4-yl]phenyl}-4,7,10,13-tetraoxa-1-aza-
cyclopentadecane (L1a). This was synthesised from 4 by follow-
ing a reported procedure for an NEt2 derivative.17 The alcohol
was dissolved into 30 cm3 of 10% (v/v) aqueous H2SO4 in a
round-bottomed flask and heated to 90 ЊC for 2 h. The
red–orange solution was cooled and neutralised by adding 1
mol dmϪ3 aqueous NaOH, to yield a yellow precipitate. The
precipitate was filtered, washed with H2O, dissolved into
CH2Cl2 and dried over MgSO4. The product was obtained by
evaporation of the solvent and recrystallised from CH2Cl2 by
adding MeOH. The product was obtained as an orange solid.
Yield (2 steps from 3) 27%; 1H NMR (270 MHz, CDCl3)
δ 3.5–3.8 (m, 20H, aliphatic crown), 6.64 (d, 2H, J 8.1 Hz, Ar),
6.76 (d, 1H, J 16.2 Hz, CH᎐CHAr), 7.21 (d, 1H, J 16.2 Hz,
᎐
CH᎐CHAr), 7.29 (d, 2H, J 5.9 Hz, pyridyl), 7.39 (d, 2H, J 8.1
᎐
Hz, Ar), 8.48 (d, 2H, J 5.9 Hz, pyridyl).
4-[(E )-2-phenylvinyl]pyridine (L1b). This was prepared in an
identical manner to L1a except that benzaldehyde was used as
the starting material instead of the azacrown derivative. The
product was obtained as a white solid. Yield (2 steps) 25%. 1H
Materials
Lithium perchlorate, sodium perchlorate, potassium thio-
cyanate, hydrochloric acid, trifluoromethanesulfonic acid,
diethylamine, diisopropylamine, 4-methylpyridine, BuLi (1.6
mol dmϪ3 in hexanes), zinc dust, PPh3, NH4Cl, CBr4, 4-bromo-
pyridine hydrochloride, 4-ethynylbenzene, copper() iodide,
dichlorobis(triphenylphosphine)palladium(), 2,2Ј-bipyridyl,
POCl3 and NH4PF6 (all Aldrich) were used as received. Mag-
nesium perchlorate, barium perchlorate, calcium perchlorate
(all Aldrich) were dried under vacuum at 230 ЊC and stored
under nitrogen before use; tetrabutylammonium perchlorate
(Aldrich) was dried similarly but at 60 ЊC. MgSO4 and NaOH
were obtained from Fisher, 1-phenyl-(4,7,10,13-tetraoxa-1-
azacyclopentadecane) (phenylaza-15-crown-5) from Acros
Organics and VWR International, P2O5 from Avocado, and
Re(CO)5Cl from Strem: all were used as received. General
solvents for synthetic work were obtained from Fisher;
anhydrous acetonitrile, THF, and CH2Cl2 were obtained from
Aldrich; CDCl3 was obtained from Cambridge Isotope Labor-
atories and CD3CN from Aldrich. Silica gel for column
chromatography (silica gel 60, 35–70 mesh) was obtained from
Fluorochem and TLC plates (silica gel 60, F254) from Merck.
NMR (270 MHz, CDCl ) δ 7.00 (d, 1H, J 17.0 Hz, CH᎐CHAr),
᎐
3
7.2–7.4 (m, 6H, Ar, pyridyl, CH᎐CHAr), 7.53 (d, 2H, J 8.1 Hz,
᎐
Ar), 8.56 (d, 2H, J 5.4 Hz, pyridyl).
1-[4-(2,2-Dibromovinyl)phenyl]4,7,10,13-tetraoxa-1-azacyclo-
pentadecane (5). This was synthesised from 4 by following a
reported procedure for a different aldehyde.18 Zinc (2.84 g, 44
mmol) and CBr4 (14.6 g, 44 mmol) were mixed in anhydrous
CH2Cl2 and stirred at Ϫ15 ЊC. A solution of PPh3 (11.6 g, 44
mmol) in anhydrous CH2Cl2 was added dropwise. The reaction
mixture was stirred for ca. 0.5 h at Ϫ15 ЊC and then stirred at
room temperature for ca. 3 h. 3 (6.5 g, 20 mmol) was added and
the reaction mixture (now dark brown) was stirred at room
temperature for 2 h. H2O was added and the organic phase was
extracted with CH2Cl2, washed with H2O and dried over
anhydrous MgSO4. Removal of the solvent yielded a brown
solid which was washed with acetone to remove insouble
by-products, and purified by repeated column chromatography
(silica, 5% MeOH/CH2Cl2). The desired product was obtained
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
1383

View Article Online
1
as a very pale yellow oil. Yield 2.3 g (24 %). H NMR (270
C36H38N4O7Re, 100); IR (vCO, CH2Cl2; cmϪ1): 2035, 1931. Anal.
Calcd. for C36H38N4O7RePF6: C, 44.58; H, 3.95; N, 5.78.
Found: C, 43.05; H, 3.85; N, 5.62.
MHz, CDCl3) δ 3.5–3.7 (m, aliphatic crown) 6.60 (d, 2H, J 6.8
Hz, Ar), 7.30 (s, 1H, Br C᎐CHAr) 7.44 (d, 2H, J 6.8 Hz, Ar).
᎐
2
1-(4-Ethynylphenyl)-4,7,10,13-tetraoxa-1-azacyclopenta-
decane (6). This was prepared by following a reported
procedure for a different alkyne.19 2.3 g of 5 (4.9 mmol) was
dissolved into 19 cm3 of anhydrous THF under N2, 2 equiv. of
BuLi (6.2 cm3, 9.85 mmol) in 5 cm3 THF was added under N2 at
Ϫ78 ЊC, and the mixture was stirred for 1 h at Ϫ78 ЊC. The
mixture was brought to room temperature and stirred for 1 h,
and then H2O was added. The organic phase was extracted with
CH2Cl2, dried over anhydrous MgSO4 and the solvent removed.
A yellow oil was obtained. 1H NMR (270 MHz, CDCl3) δ 2.93
[(bpy)Re(CO)3(L1b)]PF6 (1b). The product was obtained as a
crystalline yellow solid. H NMR (270 MHz, CDCl3) δ 6.85
1
(d, 1H, J 16.7 Hz, CH᎐CHAr), 7.03 (d, 2H, J 8.1 Hz, Ar), 7.23
᎐
(m, 6H, Ar, pyridyl, CH᎐CHAr), 7.66 (t, 2H, J 5.4 Hz, bpy),
᎐
8.10 (m, 4H, bpy, pyridyl) 8.39 (d, 2H, J 8.1 Hz, bpy), 9.14
(d, 2H, J 5.4 Hz, bpy); ESI MS m/z (relative intensity %) 608
(Mϩ, C26H18N3O3Re, 100); IR (vCO, CH2Cl2; cmϪ1): 2035, 1931.
[(bpy)Re(CO)3(L2a)]PF6 (2a). The product was obtained as a
1
yellow solid. H NMR (270 MHz, CDCl3) δ 3.5–3.7 (m, 20H,
᎐
(s, 1H, C᎐CH ) 3.5–3.7 (m, aliphatic crown) 6.53 (d, 2H, J 10.8
aliphatic crown), 6.54 (d, 2H, J 8.1 Hz, Ar), 7.22 (d, 2H, J 6.5
Hz, pyridyl), 7.29 (d, 2H, J 8.1 Hz, Ar), 7.72 (t, 2H, J 5.4 Hz,
bpy) 7.95 (d, 2H, J 6.5 Hz, pyridyl), 8.26 (t, 2H, J 8.1 Hz, bpy),
8.54 (d, 2H, J 8.1 Hz, bpy), 9.05 (d, 2H, J 5.4 Hz, bpy); ESI MS
m/z (relative intensity %) 823 (Mϩ, C36H36N4O7Re, 100);
IR (vCO, CH2Cl2; cmϪ1): 2035, 1931. Anal. Calcd. for
C36H36N4O7RePF6: C, 44.67; H, 3.75; N, 5.79. Found: C, 44.21;
H, 3.77; N, 5.68.
᎐
Hz, Ar), 7.29 (d, 2H, J 10.8 Hz, Ar).
1-(4-(Pyridin-4-ylethynyl)phenyl)-4,7,10,13-tetraoxa-1-aza-
cyclopentadecane (L2a). This was prepared by following a
reported procedure for a different alkyne.20 6 (1.4 g, 4.4 mmol),
4-bromopyridine hydrochloride (1.02 g, 5.3 mmol), dichloro-
bis(triphenylphosphine)palladium() (16.8 mg, 0.02 mmol)
and copper() iodide (6.9 mg, 0.04 mmol) were stirred in 5 ml
diethylamine for 3 h at room temperature followed by stirring
for 2 h at reflux. After cooling to room temperature, the
diethylamine was removed under vacuum on a Schlenk line and
then a saturated aqueous NH4Cl solution was added to the
residue. The organic material was extracted with diethyl ether,
washed with a solution of NH4Cl, dried over anhydrous
MgSO4, and the solvent removed to yield a dark solid. This was
purified via column chromatography (5% MeOH/CH2Cl2 on
silica) to yield the desired product as a pale yellow solid. Yield
[(bpy)Re(CO)3(L2b)]PF6 (2b). The product was obtained as
1
a crystalline yellow solid. H NMR (270 MHz, CDCl3) δ 7.21
(m, 7H, Ar, pyridyl), 7.72 (t, 2H, J 5.4 Hz, bpy), 8.19 (t, 2H,
J 8.1 Hz, bpy), 8.27 (d, 2H, J 6.7 Hz, pyridyl), 8.45 (d, 2H, J 8.1
Hz, bpy), 9.19 (d, 2H, J 5.4 Hz, bpy); ESI MS m/z (relative
intensity %) 606 (Mϩ, C26H16N3O3Re, 100); IR (vCO, CH2Cl2;
cmϪ1): 2036, 1932. Anal. Calcd. for C26H17N3O3RePF6:
C, 41.60; H, 2.28; N, 5.60. Found: C, 41.54; H, 2.34; N, 5.63.
1
856 mg (50 %). H NMR (270 MHz, CDCl3) δ 3.5–3.7 (m,
Acknowledgements
aliphatic crown) 6.60 (d, 2H, J 9.0 Hz, Ar), 7.30 (d, 2H, J 5.4
Hz, pyridyl) 7.37 (d, 2H, J 9.0 Hz, Ar), 8.52 (d, 2H, J 5.4 Hz,
pyridyl).
We thank Professor Robin Perutz for his assistance and advice,
Dr Trevor Dransfield for recording mass spectra, and Ms
Heather Fish and Ms Kirsten Ampt for recording NMR
spectra. We acknowledge financial support from ESPRC.
4-(Phenylethynyl)pyridine (L2b). This was prepared in an
identical manner to L2a except that ethynylbenzene (1.87 g, 18
mmol) was used as the starting material instead of 6. The
1
References
product was obtained as white crystals. Yield 1.7 g (52 %). H
NMR (270 MHz, CDCl3) δ 7.2–7.4 (m, 5H, Ar, pyridyl), 7.53
(m, 2H, Ar), 8.58 (d, 2H, J 5.4 Hz, pyridyl).
1 A. P. de Silva, H. Q. N. Gunaratne, T. Gunnlaugsson, A. J. M.
Huxley, C. P. McCoy, J. T. Rademacher and T. E. Rice, Chem. Rev.,
1997, 97, 1515.
2 L. Fabbrizzi and A. Poggi, Chem. Soc. Rev., 1995, 197.
3 A. P. de Silva and C. P. McCoy, Chem. Ind., 1994, 992.
4 B. Valeur and E. Bardez, Chem. Br., 1995, 216.
5 H.-G. Lohr and F. Vögtle, Acc. Chem. Res., 1985, 18, 65.
6 J. Bourson and B. Valeur, J. Phys. Chem., 1989, 93, 3871.
7 I. K. Lednev, R. E. Hester and J. N. Moore, J. Chem. Soc., Faraday
Trans., 1997, 93, 1551.
8 I. K. Lednev, R. E. Hester and J. N. Moore, J. Phys. Chem., 1997,
101, 7371.
9 R. M. Mathevet, G. Jonusauskas, C. Rullière, J.-F. Létard and
R. Lapouyade, J. Phys. Chem., 1995, 99, 15709.
10 J. D. Lewis, R. N. Perutz and J. N. Moore, Chem. Commun., 2000,
1865.
11 J. D. Lewis and J. N. Moore, Chem. Commun., 2003, 2858.
12 K. S. Schanze, D. B. MacQueen, T. A. Perkins and L. A. Cabana,
Coord. Chem. Rev., 1993, 122, 63.
13 D. B. MacQueen and K. S. Schanze, J. Am. Chem. Soc., 1991, 113,
6108.
14 J. D. Lewis, R. N. Perutz and J. N. Moore, J. Phys. Chem. A, 2002,
106, 12202.
Re complexes
The Re complexes were prepared by following a procedure
reported for other L ligands;13,21 the synthesis of model 1b has
been reported previously.47 (bpy)Re(CO)3Cl was prepared by
refluxing Re(CO)5Cl with 3 equiv. of 2,2-bipyridyl in toluene
for ca. 3 h. (bpy)Re(CO)3Cl was then stirred with an excess of
CF3SO3H at room temperature for 1 h to yield (bpy)Re-
(CO)3(CF3SO3), which was refluxed overnight in THF under N2
with <1 equiv. of the appropriate L ligand. The product
was metathesized with NH4PF6 to yield [(bpy)Re(CO)3L]PF6,
which was purified by repeated recrystallization and column
chromatography (5% MeOH/CH2Cl2 eluting on silica) and
characterized by IR, ES-MS, 1H-NMR, and elemental analysis.
1H NMR showed that the trans-isomers of the styryl complexes
(1a and 1b) were formed exclusively. Yields were typically
40–50% following chromatography.
15 M. V. Alfimov, A. V. Churakov, Y. V. Fedorov, O. A. Fedorova,
S. P. Gromov, R. E. Hester, J. A. K. Howard, L. G. Kuz’mina,
I. K. Lednev and J. N. Moore, J. Chem. Soc., Perkin Trans. 2, 1997,
2249.
16 J. P. Dix and F. Vögtle, Chem. Ber., 1980, 113, 457.
17 A. Juris, S. Campagna, I. Bidd, J.-M. Lehn and R. Ziessel, Inorg.
Chem., 1988, 27, 4007.
18 A. de Meijere, S. Kozhushkov, T. Haumann, R. Boese, C. Puls,
M. J. Cooney and L. T. Scott, Chem. Eur. J., 1995, 1, 124.
19 E. J. Corey and P. L. Fuchs, Tetrahedron Lett., 1972, 3769.
[(bpy)Re(CO)3(L1a)]PF6 (1a). The product was obtained as
an orange–red solid. H NMR (270 MHz, CDCl3) δ 3.5–3.7
1
(m, 20H, aliphatic crown), 6.56 (d, 1H, J 16.7 Hz, CH᎐CHAr),
᎐
6.59 (d, 2H, J 8.1 Hz, Ar), 7.19 (d, 1H, J 16.7 Hz, CH᎐CHAr),
᎐
7.21 (d, 2H, J 7.0 Hz, pyridyl), 7.31 (d, 2H, J 8.1 Hz, Ar), 7.70
(t, 2H, J 5.4 Hz, bpy), 7.85 (d, 2H, J 7.0 Hz, pyridyl), 8.25
(t, 2H, J 8.1 Hz, bpy), 8.53 (d, 2H, J 8.1 Hz, bpy), 9.05 (d, 2H,
J 5.4 Hz, bpy); ESI MS m/z (relative intensity %) 825 (Mϩ,
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
1384

View Article Online
20 J. Okubo, H. Shinozaki, T. Koitabashi and R. Yomura, Bull. Chem.
Soc. Jpn., 1998, 71, 329.
35 R. M. Izatt, K. Pawlak, J. S. Bradshaw and R. L. Bruening, Chem.
Rev., 1991, 91, 1721.
36 K. Rurack, M. Sczepan, M. Spieles, U. Resch-Genger and W. Rettig,
Chem. Phys. Lett., 2000, 320, 87.
21 B. B. Sullivan and T. J. Meyer, Chem. Commun., 1984, 1244.
22 K. Kalyandasundaram, J. Chem. Soc., Faraday Trans., 1986, 82,
2401.
37 J.-F. Létard, S. Delmond, R. Lapouyade, D. Braun, W. Rettig and
M. Kriessler, Recl. Trav. Chim. Pays-Bas., 1995, 114, 517.
38 L. Antonov, M. Vladimirova, E. Stanoeva, W. M. F. Fabian,
L. Ballester and M. Mitewa, J. Inclusion Phenom. Macrocycl. Chem.,
2001, 40, 23.
23 C. J. Pouchert and J. Behnke, Aldrich Library of 13C and 1H FT
NMR spectra, Aldrich Chemical Company Inc., Milwaukee, WI, 1st
edn., 1993.
24 R. D. Shannon, Acta. Crystallogr., 1976, A32, 751.
25 B. Brzezinski, G. Schroeder, A. Rabold and G. Zundel, J. Phys.
Chem., 1995, 99, 8519.
39 Plots of log K versus the ionic index z2/r (ÅϪ1) do not give a linear
plot,27 although there is a broad correlation.
26 Alternative plots of the shift in peak absorbance in wavenumbers
(cmϪ1) versus the ionic index z2/r (ÅϪ1) do not give a linear plot,27
although there is a broad correlation which is more linear than that
shown in Fig. 5.
27 K. Rurack, J. L. Bricks, G. Reck, R. Radeglia and U. Resch-Genger,
J. Phys. Chem. A, 2000, 104, 3087.
28 Our time-resolved studies of the photochemical release and thermal
binding of Liϩ to Re-azacrown complexes also support this
interpretation; they will be reported elsewhere.
29 Addition of excess HCl or HClO4 to an NMR sample of 2a
(ca. 1 × 10Ϫ3 mol dmϪ3) resulted in a complicated 1H NMR spectrum
which was quite different from the others reported here and may
indicate that a reaction occurred at these high concentrations; this
effect was not studied further.
30 The purpose of these measurements was to determine the positions
of the resonances from 1a–Mnϩ and 2a–Mnϩ, and so samples of fully
complexed species were studied; samples comprising both metal-free
and metal-complexed forms may show single resonances with
positions dependent on the sample composition because the
literature indicates that exchange would be fast on the NMR time
scale.31
31 (a) F. Hallwass, M. Engelsberg, A. M. Simas and W. Barros,
Chem. Phys. Lett., 2001, 335, 43; (b) K. M. Brière and C. Detellier,
Can. J. Chem., 1992, 70, 2536; (c) P. Szczygiel, M. Shamsipur,
K. Hallenga and A. I. Popov, J. Phys. Chem., 1987, 91, 1252.
32 J. Bourson, J. Pouget and B. Valeur, J. Phys. Chem., 1993, 97, 4552.
33 S. Fery-Forgues, M.-T. LeBris, J.-P. Guetté and B. Valeur, J. Phys.
Chem., 1988, 92, 6233.
40 R. M. Izatt, K. Pawlak, J. S. Bradshaw and R. L. Bruening, Chem.
Rev., 1995, 95, 2529.
41 M. J. G. Lesley, A. Woodward, N. J. Taylor, T. B. Marder,
I. Cazenobe, I. Ledoux, J. Zyss, A. Thornton, D. W. Bruce and
A. K. Kakkar, Chem. Mater., 1998, 10, 1355.
42 D. Roberto, R. Ugo, S. Bruni, E. Cariati, F. Cariati, P. Fantucci,
I. Invernizzi, S. Quici, I. Ledoux and J. Zyss, Organometallics, 2000,
19, 1775.
43 L. De Cola and P. Belser, Coord. Chem. Rev., 1998, 177, 301.
44 L.-T. Cheng, W. Tam, S. R. Marder, A E. Stiegman, G. Rikken and
C. W. Spangler, J. Phys. Chem., 1991, 95, 10643.
45 T. E. O. Screen, I. M. Blake, L. H. Rees, W. Clegg, S. J. Borwick and
H. L. Anderson, J. Chem. Soc., Perkin Trans. 1, 2002, 320.
46 Gaussian 98, Revision A.7, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery, Jr.,R. E. Stratmann, J. C. Burant,
S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C.
Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi,
B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski,
G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski,
J. V. Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox,
T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon,
E. S. Replogle and J. A. Pople, Gaussian, Inc., Pittsburgh PA,
1998.
34 J. F. Létard, R. Lapouyade and W. Rettig, Pure Appl. Chem., 1993,
65, 1705.
47 V. W.-W. Yam, V. C.-Y. Lau and L.-X. Wu, J. Chem. Soc., Dalton
Trans., 1998, 1461.
D a l t o n T r a n s . , 2 0 0 4 , 1 3 7 6 – 1 3 8 5
1385