Synthesis, structural studies and photochemistry of cobalt(III) complexes of anthracenylcyclam macrocycles

Article abstract of DOI:10.1039/b305773a

This work reports the syntheses, structures and some photochemistry in DMF of the cobalt complexes trans-[Co^III(2)Cl₂]Cl·0.5CH₃OH and trans-[Co^III(3)Cl₂]Cl·4H₂O, where 2 is 6-(anthracen-9-ylmethyl)-1,4,8,11-tetraazacyclotetradecane-5,7-dione and 3 is 6-(anthracen-9-ylmethyl)-1,4,8,11-tetraazacyclotetradecane. In the preparation of the macrocyclic ligand, 3, the formation of a polycyclic bis(aminal) intermediate and its subsequent acid hydrolysis to 3 is a cleaner route than the traditional procedure in which the analogous dioxocyclam 2 is reduced with borane reagents. The crystal structure of trans-[Co^III(3)Cl₂]Cl·4H₂O shows that the macrocycle adopts the trans-III conformation, in which the anthracene moiety is extended away from the cobalt ion and the anthracene to Co separation is 7.22 A. For the related complex trans-[Co^III(2)Cl₂]Cl·0.5CH₃OH, however, the anthracene is bent over the highly conjugated tetracycle and significant interactions between the anthracene and the complex occur. A novel new complex, trans-[Co(12)Cl₂] (where 12 is 5,7-hydroxy-6-oxo-1,4,8,11-tetraazacyclotetradecane-4,7-diene) which is a degradation product of the complex trans-[Co^III(2)Cl₂]Cl is also reported.

Full text of DOI:10.1039/b305773a

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis, structural studies and photochemistry of cobalt(III)
complexes of anthracenylcyclam macrocycles†
Alison M. Funston, Kenneth P. Ghiggino, Martin J. Grannas, W. David McFadyen and
Peter A. Tregloan
School of Chemistry, University of Melbourne, Victoria, 3010, Australia
Received 22nd May 2003, Accepted 15th August 2003
First published as an Advance Article on the web 29th August 2003
This work reports the syntheses, structures and some photochemistry in DMF of the cobalt complexes
trans-[CoIII(2)Cl2]Clؒ0.5CH3OH and trans-[CoIII(3)Cl2]Clؒ4H2O, where 2 is 6-(anthracen-9-ylmethyl)-1,4,8,11-
tetraazacyclotetradecane-5,7-dione and 3 is 6-(anthracen-9-ylmethyl)-1,4,8,11-tetraazacyclotetradecane. In the
preparation of the macrocyclic ligand, 3, the formation of a polycyclic bis(aminal) intermediate and its subsequent
acid hydrolysis to 3 is a cleaner route than the traditional procedure in which the analogous dioxocyclam 2 is reduced
with borane reagents. The crystal structure of trans-[CoIII(3)Cl2]Clؒ4H2O shows that the macrocycle adopts the
trans-III conformation, in which the anthracene moiety is extended away from the cobalt ion and the anthracene to
Co separation is 7.22 Å. For the related complex trans-[CoIII(2)Cl2]Clؒ0.5CH3OH, however, the anthracene is bent
over the highly conjugated tetracycle and significant interactions between the anthracene and the complex occur.
A novel new complex, trans-[Co(12)Cl2] (where 12 is 5,7-hydroxy-6-oxo-1,4,8,11-tetraazacyclotetradecane-4,7-diene)
which is a degradation product of the complex trans-[CoIII(2)Cl2]Cl is also reported.
ylmethyl)-1,4,8,11-tetraazacyclotetra-decane-5,7-dione 2. In
Introduction
this report we describe the isolation and structural characteris-
Donor–acceptor systems have been reported as models for
ation of the complexes trans-[CoIII(3)Cl2]Cl and trans-
molecular devices such as photomolecular switches,1–3
[CoIII(2)Cl2]Cl as well as describing the results of some
sensors4–9 and machines.4 The operation of these relies on vari-
photochemical investigations of these systems.
ation of the fluorescence quenching properties of the metal and
ligands as a function of pH,4,5 metal concentration6–9 or metal
oxidation state.1–3 The use of a metal centre as the electron
acceptor unit in such systems, the stability of the metal ion in
different oxidation states and its distinct properties in those
states gives substantial flexibility in the molecular design and
diversity of any applications.
Experimental
Electrochemical experiments
Cyclic voltammetry was carried out using a three-electrode cell
configuration consisting of a 1 mm glassy carbon disk working
electrode, a platinum wire auxiliary electrode, and a Ag/AgClO4
in [Et4N]ClO4 reference electrode separated from the working
solution by a liquid junction incorporating a Vycor glass frit.
DMF (Merck) was dried over molecular sieves (4Å).
[Et4N]ClO4 (Fluka) was used as supporting electrolyte and
AgClO4 (BDH) as the source of Agϩ in the reference electrode.
The ionic strength of the reference electrode was identical to
that of the working solution. Formal potentials were deter-
mined from the mean of the oxidation and reduction cyclic
voltammetric peaks.
A number of the systems developed have made use of
quadridentate ligands such as 1,4,8,11-tetraazacyclotetra-
decane (cyclam)4,10 and 1,4,8,11-tetraazacyclotetradecane-5,7-
dione,7,11 extended and appended to a variety of chromophores
including anthracene and naphthalene. Work with these ligands
has centred on Ni() and Cu() complexes and excited state
quenching in the metal complexes ascribed to energy or electron
transfer processes. The nature of the quenching process has
been related to the structure of the spacer between the donor
and the metal acceptor groups, although for systems of this
type it is still not possible to predict whether an electron trans-
fer or energy transfer mechanism will operate.4 Solvent inter-
actions, which may well influence solution conformation, are
also likely to play a role in the quenching behaviour of these
donor–acceptor systems.12
Another potential application of the properties of donor–
acceptor systems is their use as tools for triggered ligand release
as a model for photoinduced drug delivery. Transient reduction
of a stable cobalt() acceptor centre to its labile cobalt()
analogue may provide such a vehicle for controlled ligand
release. For this concept to be realised it is important to better
understand the nature of the quenching process in donor–
acceptor complexes containing cobalt() as the acceptor
moiety. To this end, and as part of our investigations of photo-
activated donor–acceptor complexes we have examined some
cobalt() complexes of the macrocyclic ligands 6-(anthracen-
9-yl)-1,4,8,11-tetraazacyclotetradecane 3 and 6-(anthracen-9-
Instrumentation
Ultraviolet/visible spectra in DMF were recorded on either a
Shimadzu UV-2401PC UV-VIS recording spectrophotometer
or Varian Cary 50 Bio ultraviolet-visible spectrophotometer
unless otherwise specified. 1H and 13C NMR spectra were
measured in d4-methanol using Varian Unity 300 or Unity Plus
400 spectrometers. Chemical shifts (δ, positive downfield) are
given in ppm (methanol reference) and J values in Hz.
Photochemistry experiments
Sample preparation. Solutions were prepared in DMF
(Merck, spectroscopic grade). In the determination of extinc-
tion coefficients, samples were weighed on a Perkin Elmer
Autobalance AM-2 to 5 µg and the optical density (OD) of
solutions of known concentration measured using a Cary 50
UV-VIS specrophotometer. For steady state fluorescence
experiments, the OD at 370 nm was adjusted to ∼0.1 (corre-
sponding to a concentration of ∼0.8–1.0 × 10Ϫ5 M); for flash
photolysis experiments, the OD was ∼1. Solutions were
degassed using multiple freeze–pump–thaw cycles. Typically,
† Electronic supplementary information (ESI) available: Selected bond
lengths and angles and ORTEP diagrams for compound 8 and com-
plexes [Co(3)Cl2]Clؒ4H2O, [Co(12)Cl2] and [Co(2)Cl2]Clؒ0.5CH3OH.
See http://www.rsc.org/suppdata/dt/b3/b305773a/
D a l t o n T r a n s . , 2 0 0 3 , 3 7 0 4 – 3 7 1 2
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 3
3704

View Article Online
four cycles were sufficient to remove dissolved gases. Solutions
were prepared in the dark and protected from light when not in
use.
2-(9-Anthracenylmethyl)propane-1,3-ditosylate 5. 4-Toluene-
sulfonyl chloride (18.5 g, 97.2 mmol) was added in small por-
tions to a rapidly stirred solution of diol 4 (10.00 g, 44.2 mmol)
in dry pyridine (50 cm3). After reaction at ambient temperature
overnight, the mixture was poured into a slurry of crushed ice
in water (0.8 l) and the ditosylate that separated was collected,
washed with water, and recrystallised from MeCN. Yield 63%,
Steady-state fluorescence. Steady-state fluorescence spectra
were recorded using a Varian Eclipse Fluorimeter and corrected
for instrumental response. Low-temperature steady-state fluor-
escence measurements were performed in an ethanol (Merck,
spectroscopic grade) glass in cylindrical quartz cells, using an
Oxford Instruments Optistat DN with a liquid nitrogen cryo-
stat. Fluorescence quantum yields were determined by using
9-methylanthracene in ethanol (Φ = 0.33) as the reference.13
1
mp 141–144 ЊC. H NMR (300 MHz, CDCl3): δ 8.36 (s, 1H,
Ar-10); 8.07–7.97 (m, 4H, Ar-1,4,5,8); 7.64 (d, 4H, J 8.2, tosyl);
7.50–7.43 (m, 4H, Ar-2,3,6,7); 7.24 (d, 4H, J 8.2, tosyl); 4.00 (d,
4H, J 5.3, Ts–CH2); 3.63 (d, 2H, J 8.6, Ar–CH2); 2.50 (m, 1H,
CH); 2.42 (s, 6H, Ar–CH3).
2-(9-Anthracenylmethyl)-1,3-diiodopropane 6. The crude
ditosylate 5 (25.9 g, 45 mmol) was heated at reflux in a solution
of sodium iodide (27.0 g, 182 mmol) in acetone (500 ml) for
72 h. After removal of the solvent under reduced pressure, the
residue was partitioned between CH2Cl2 (300 mL) and water
(100 mL). The aqueous phase was extracted with further
CH2Cl2 (3 × 100 mL) and the combined organic phases were
dried (MgSO4), evaporated to dryness, then purified by chroma-
tography on SiO2 (2 : 1 light petroleum (bp 40–60 ЊC)–CH2Cl2).
The most mobile band was collected and the solvent evaporated
until a microcrystalline pale yellow solid separated. The solid
was collected, washed with light petroleum (bp 40–60 ЊC) and
Time-correlated single photon counting. Fluorescence decay
profiles were determined from time-correlated single-photon
counting data. The laser excitation source was a jet stream dye
laser (Spectra Physics model 3500), synchronously pumped by a
mode locked argon ion laser (Spectra Physics model 2030). The
output from the DCM dye was set at 700 nm using a 3-plate
birefringent filter. The output pulses were pulse-picked and fre-
quency-doubled in a KDP crystal to provide the 350 nm excit-
ation pulses of ∼5 ps (FWHM) at a repetition rate of 4 MHz.
Sample emission was collected through a polarizer set at the
magic angle (54.7Њ) relative to the vertically polarized excitation
source. Fluorescence decays were measured at 430 nm using
a microchannel plate photomultiplier (Hamamatsu model
R2809U-01) interfaced to a 512 channel multichannel analyser
and reconvoluted using non-linear least-squares iterative
procedures based on the Marquardt algorithm.14
1
air dried. Yield 95%, mp 134–136 ЊC. H NMR (300 MHz,
CDCl3): δ 8.42 (s, Ar-10); 8.28 (d, J 8.8, Ar-1,8); 8.04 (d, J 8.1
Ar-4,5); 7.59–7.47 (m, 4H, Ar-2,3,6,7); 3.76 (d, J 7.1 Ar–CH2);
3.41 (d, J 5.4, CH2I); 2.04 (m, –CH–).
2-(Anthracen-9-ylmethyl)decahydro-cis-10b,10c-dimethyl-
1H,6H-3a,5a,8a,10a-tetraazapyrene. 8. A solution of diiodo-
propane 6 (6.395 g, 13.15 mmol), anhydrous K2CO3 (9.09 g,
65.8 mmol) and tricyclic bis(aminal) 7 in dry MeCN (250 mL)
was heated at reflux for 26 h, in a flame-dried round-bottomed
flask fitted with a CaCl2 guard tube. The dark reaction mixture
was filtered, the filtrate was evaporated under reduced pressure
and the residue purified by squat SiO2 column chromatography.
After any unreacted 6 had been eluted with CH2Cl2, the band
that eluted with Me2CO and which appeared light blue under
365 nm irradiation was collected. Concentration of this frac-
tion yielded a crude mixture of the syn and anti isomers of 8
that was further purified by vacuum chromatography through a
squat SiO2 column followed by recrystallisation from acetone.
Yield = 3.06 g, 51%. Anal. Calc. for C29H34N4: C, 79.05; H, 8.23;
N, 12.72. Found: C, 78.63; H, 8.03; N, 12.81%. The anti isomer
is significantly less soluble in common solvents such as CHCl3
and Me2CO and was separated by crystallization from the
latter.
Flash photolysis. The excitation source in the flash photolysis
experiments was a Nd:YAG pulsed laser (Continuum NY-61).
The output was frequency tripled to 355 nm to give laser pulses
of ∼7–8 ns (FWHM) with a pulse intensity limited to ≥ 5 mJ
cmϪ2 pulseϪ1. The analysing light source was based on a 150 W
xenon arc lamp set perpendicular to the excitation beam. To
minimise the effects of excess heating and any UV photo-
reaction, the light was passed through UV-cutoff filters
(360 or 400 nm as appropriate) and a water bath, prior to the
sample solution. The monitoring wavelength was selected
using a dual port triple grating monochromator/spectrograph
(Acton Research Corporation SpectraPro model 300i). Two
optical observation systems were utilised for time resolved
optical detection. Kinetic measurements at a single wavelength
were detected using a fast response photomultiplier tube
(Hamamatsu R928) coupled to a digital recording oscilloscope
(Tektronix TDS-520). Transient spectra at specific delay times
were collected using a CCD camera (Princeton Instruments
I-MAX-512-T ICCD with ST-133 controller).
1
Anti isomer: mp 217–221 ЊC. H NMR (400 MHz, CDCl3):
δ 8.30 (s, 1H, Ar-10); 7.95 (d, 2H, J 8.4, Ar-1,8); 7.4–7.6
(m, 6H, Ar-2,3,4,5,6,7); 4.82 (t, 1H, J 11.8); 4.53 (t, 1H, J 13);
4.23 (t, 1H, J 10.3); 3.88 (m, 1H); 2.3–3.7 (m, 14H); 1.96 (m,
2H); 1.80 (s, 3H, C–CH3); 1.48 (d, 1H, J 13.9); 1.31 (s, 3H,
C–CH3).
Syn isomer: mp 176–179 ЊC. Poor solubility in common sol-
vents prevented the detailed analysis of the 1H NMR spectrum
of this isomer. 1H NMR (400 MHz, CDCl3): δ 8.30 (m); 7.96 (d,
J 8.5); 7.4–7.5 (m); 4.45 (m); 4.02 (m); 1.8–3.7 (m); 1.39 (s); 1.25
(m).
Steady-state irradiation. 3 mL of a solution containing either
[Co(3)Cl2]Cl (c = 4.70 × 10Ϫ3 M) or [Co(2)Cl2]Cl (c = 6.42 × 10Ϫ3
M) in DMF was placed in a glass cuvette and deoxygenated by
purging the solution with N2 for 10 min. The cuvette was
capped and sealed using Parafilm then irradiated for 5 h with
light from a 100 W high-pressure mercury arc source fitted with
a 372 nm narrow bandpass filter. A control solution from the
same stock solution was kept in the dark for the duration of the
irradiation.
Preparations
6-(9-Anthracenylmethyl)-1,4,8,11-tetraazacyclotetradecane
tetrahydrochloride hemihydrate 3ؒ4HClؒ0.5H2O. A mixture of
isomers of bis(aminal) 8 (2.00 g, 4.42 mmol) was hydrolysed in
aq. HCl (60 ml, 2.0 M) in EtOH (60 ml) heated overnight on a
steam-bath. The volatile components were removed under
reduced pressure, then the residue was triturated with EtOH
and evaporated to dryness again. The residue was washed with
cold ethanol, then with Et2O, to produce a colourless micro-
crystalline solid. Yield 2.33 g, 97%. Anal. Calc. for C25H38Cl4-
Reactions were performed in air using as-received solvents and
A.R. reagents unless otherwise specified. Pyridine was dried
over 4Å molecular sieves. 2-(Anthracen-9-ylmethyl)propane-
1,3-diol 4, the bis(aminal) 7 and the ligand 6-(anthracen-
9-ylmethyl)-1,4,8,11-tetraazacyclotetradecane-5,7-dione 2 were
prepared by literature methods.11,15 Elemental analyses were
performed by the Microanalytical Laboratory, Department of
Chemistry, University of Otago, Dunedin, New Zealand.
D a l t o n T r a n s . , 2 0 0 3 , 3 7 0 4 – 3 7 1 2
3705

View Article Online
Table 1 Crystallography data
Compound
8
[Co(3)Cl2]Clؒ4H2O
[Co(2)Cl2]Clؒ0.5CH3OH
[Co(12)Cl2]
Mr
440.62
Colourless
645.92
Green
599.84
Green
380.11
Green
Colour
Size/mm
Lattice
Space group
a/Å
b/Å
c/Å
0.02 × 0.07 × 0.11 × 0.12
Triclinic
0.05 × 0.09 × 0.37
Monoclinic
P21 (no. 4)
10.8717(15)
6.3706(8)
23.310(5)
90
91.034(12)
90
1614.2(4)
1.329
0.09 × 0.13 × 0.25
Triclinic
0.06 × 0.06 × 0.09
Orthorhombic
Pnnm (no. 58)
10.0111(11)
14.891(3)
9.692(2)
90
90
90
1444.8(4)
1.748
4
12.887
780
1.54180
150(1)
ω–2θ
5–70
7200
¯
¯
P1 (no. 2)
P1 (no. 2)
9.9553(6)
11.6873(5)
12.1534(5)
115.186(4)
96.967(4)
107.303(5)
1170.35(10)
1.250
10.408(4)
10.8133(8)
12.018(7)
93.22(3)
107.21(5)
94.824(15)
1282.8(9)
1.548
α/Њ
β/Њ
γ/Њ
V/Å3
Dc/g cmϪ3
Z
2
2
2
µ/mmϪ1
F(000)
λ/Å
0.586
476
1.54180
291(2)
ω–2θ
4–75
7200
0
5675
0.818
680
0.71073
293(2)
ω–2θ
2–25
7200
0
3639
1.016
618
0.71073
150(1)
ω–2θ
2–25
7200
0
5212
T/K
Scan method
θ Range/Њ
Standards
Decay (%)
Data collected
–
1274
Unique (Rint
)
4812 (0.0122)
3859
0.0443
0.1263
1.052
3112 (0.0271)
2066
0.0667
0.1849
1.087
4505 (0.0592)
3623
0.0489
0.1390
1.023
1274
848
0.0854
0.2470
1.040
Observeda
b
R1
c
wR2
S
2
a I ≥ 2σ(I ). b R1 (I ≥ 2σ(I )) = Σ|Fo| Ϫ |Fc||/Σ|Fo|. c wR2 (all data) = (Σw(Fo2 Ϫ Fc2)2)/Σw(Fo )2)1/2
.
N4ؒ0.5H2O: C, 55.05; H, 7.21; N, 10.27; Cl, 26.00. Found: C,
54.88; H, 7.14; N, 10.43; Cl, 25.57%.
nol (400 mL). Air was bubbled through the solution for 5 h and
the light green solution was then filtered. Upon the slow addi-
tion of concentrated hydrochloric acid (2 mL) the solution
became dark green. The solvent was evaporated under reduced
pressure to a volume of 5 mL during which time a dark green
crystalline solid formed. After cooling at Ϫ20 ЊC for 12 h the
product was collected and washed with ice-cold ethanol (5 ×
1 mL). The product was recrystallised from the minimum
amount of 0.67% v/v concentrated HCl–methanol, washed
with ice-cold methanol (5 × 1 mL) and dried in vacuo at 110 ЊC
for 8 h to yield [Co(6-(anthracen-9-ylmethyl)-1,4,8,11-tetraaza-
cyclotetradecane-5,7-dione)Cl2]Clؒ0.5CH3OH. Yield = 90 mg,
24%. Anal. Calc. for C25H30Cl3CoN4O2ؒ0.5CH3OH: C, 51.06;
H, 5.38; N, 9.34; Cl, 17.73. Found: C, 51.18; H, 5.58; N, 9.20;
Cl, 17.51%. ESI-MS data: parent ion m/z 549.5, 547.8. Visible
The hydrochloride was converted to the free base by treat-
ment with aqueous base (NaOH or NH3) and extraction into
CH2Cl2. 1H NMR (400 MHz, CDCl3): δ 8.32 (s, Ar-10); 8.26 (d,
J 8.6, Ar-1,8); 7.96 (d, J 8.1, Ar-4,5); 7.48–7.40 (m, 4H,
Ar-2,3,5,6); 3.56 (d, J 7.4, Ar–CH2); 2.87–2.50 (m, 16H,
CH2N); 2.34 (m, HC(CH2)3); 2.13 (br, 4H, NH, exchanges with
D2O addition); 1.68 (m, –CH2–).
(6-(9-Anthracenylmethyl)-1,4,8,11-tetraazacyclotetradecane)-
dichlorocobalt(III) chloride monohydrate. Aqueous NaOH (2.0
mL, 1.0 M, 2.0 mmol) was added to a hot solution of CoCl2ؒ
6H2O (119 mg, 0.500 mmol) and 3ؒ4HClؒ0.5H2O (273 mg,
0.500 mmol) in MeOH (40 mL), The pale red–brown solution
was heated at reflux overnight, then treated with 2.0 M aqueous
HCl (2.0 mL, 4.0 mmol). After a few min cooling, air was
bubbled into the mixture for 0.5 h whereupon the mixture rap-
idly became deep green. The solvent was then boiled off cau-
tiously on a steam-bath, with the occasional addition of small
portions (ca. 0.5 mL each) of water. When the mixture had been
concentrated to ca. 10 mL, it was placed in a screw-capped glass
jar containing water, which diffused slowly into the mixture
with concomitant MeOH diffusion into the water from the
reaction mixture. The green microcrystalline solid began separ-
ating within 30 min and was collected after 3 days, washed with
a minimum of cold water, then dried in vacuo at 80 ЊC for 2 h.
Yield = 153 mg, 60%. Anal. Calc. for C25H34Cl3CoN4ؒH2O: C,
52.32; H, 6.32; N, 9.76; Cl, 18.53. Found: C, 52.44; H, 6.56; N,
λmax: 678 nm, εmax: 69 L molϪ1 cmϪ1 1H NMR (400 MHz,
.
d4-MeOH): δ 8.63 (d, J 8.6, 2H, Ar CH); 8.49 (s, 1H, Ar CH);
8.07 (d, J 9.0, 2H, Ar CH); 7.51 (m, 4H, Ar CH); 6.34 (br, 2H,
NH); 4.57 (d, J 7.7, 2H, Ar–CH2); 4.17 (t, J 7.5, 1H, cyclam
CH); 3.80 (dd, J 13.4, 4.4, 2H, cyclam CH); 3.00–3.45 (m, 4H,
cyclam CH); 2.63 (d, J 11.9, 2H, cyclam CH); 2.20 (d, J 15.4,
1H, cyclam CH); 2.00 (q, J 15.6, 1H, cyclam CH).
Crystallography
Crystal data and details of data collection and refinement are
given in Table 1.
Crystals of [Co(3)Cl2]Clؒ4H2O suitable for X-ray structure
determination were grown by H2O diffusion into a MeOH
solution of the complex. Data were collected on the crystal
mounted in a sealed capillary tube in the presence of
mother-liquor. Crystals of the (bis-aminal) anti-isomer suitable
for X-ray structure determination were grown by evaporation
from Me2CO solution and data were collected from the crystal
mounted on a glass fibre at room temperature. Single X-ray
quality crystals of [Co(2)Cl2]Clؒ0.5CH3OH were grown by the
slow diffusion of acetone into a solution of the complex in
methanol containing two drops of trimethylorthoformate.
Single X-ray quality crystals of [Co(12)Cl2] were grown by the
9.67; Cl, 18.36%. Visible λmax: 626 nm, εmax: 33.2 L molϪ1 cmϪ1
.
1H NMR (300 MHz, d4-MeOH): δ 8.46 (s, Ar-9); 8.32 (d, J 8.7,
Ar-1,8); 8.06 (d, J 8.1, Ar-4,5); 7.59–7.46 (m, 4H, Ar-2,3,6,7);
6.22 (br, 2H, NH); 6.11 (br, 2H, NH); 3.70 (d, J 7.5, Ar–CH2).
(6-(Anthracen-9-ylmethyl)-1,4,8,11-tetraazacyclotetradecane-
5,7-dione)dichlorocobalt(III) chloride–0.5CH3OH. A solution of
CoCl2ؒ6H2O (167 mg, 0.704 mmol) in ethanol (10 mL) was
added to a solution of 6-(anthracen-9-ylmethyl)-1,4,8,11-tetra-
azacyclotetradecane-5,7-dione 2 (262 mg, 0.627 mmol) in etha-
D a l t o n T r a n s . , 2 0 0 3 , 3 7 0 4 – 3 7 1 2
3706

View Article Online
slow evaporation of a solution of (6-(anthracen-9-ylmethyl)-
1,4,8,11-tetraazacyclotetradecane-5,7-dione)dichlorocobalt()
chloride–0.5CH3OH in methanol over a number of months.
The single crystals of [Co(2)Cl2]Clؒ0.5CH3OH and [Co(12)Cl2]
were mounted on glass fibres, covered in low temperature oil,
and kept at 150 K during data collection.
Accurate unit cell parameters were calculated using a least-
squares procedure from the setting angles of 25 reflections.
Intensity data were collected on an Enraf-Nonius CAD4-
MachS single crystal X-ray diffractometer. For [Co(12)Cl2]
during the data collection the cryostat failed, resulting in a time
during which the crystal was at room temperature and some
solvent lost; as a result of this the intensities of the remaining
peaks collected were lower and the peaks broader. The data
collection could not be repeated due to the lack of another
suitable crystal. Corrections were applied for Lorentz and
polarization effects as well as for absorption, the corrections
being evaluated by Gaussian integration.
form the dimethyldecahydrotetraazapyrene 10, which was sub-
sequently hydrolysed to form cyclam 11 in high overall yield
as presented in Scheme 2.15 This prompted us to investigate
cognate alkylations using suitably 2-substituted 1,3-diactivated
propane derivatives. The desired anthrylmethylpropane was
prepared from the known diol 4, which was converted to the
diiodopropane 6 via the ditosylate 5. Significantly lower cycliz-
ation yields were obtained from the reaction of 5 with 7. The
anthracenylmethyl bis-aminal 8 was purified by chromato-
graphic workup. It eluted as a mixture of what we believe to be
a mixture of its syn and anti isomers, which differ in the orien-
tation of the anthracenylmethyl moiety relative to the butane-
dione-derived methyl groups of the decahydrotetraazapyrene
moiety. The single crystal X-ray structure of the anti isomer was
determined, and confirms the macrocyclic nature of the prod-
uct. Upon hydrolysis in HCl–H2O–EtOH solution both pro-
posed isomers yielded 6-anthracenylmethylcyclam 3 as its stable
tetrahydrochloride salt. The free base of 3 could be extracted
into CH2Cl2 from aqueous solutions of 3ؒ4HCl that had been
treated with base (e.g. NaOH or NH3).
Cobalt() complexes of 3 were prepared by neutralizing a
hot methanol solution of CoCl2ؒ6H2O and the tetrahydro-
chloride of 3 with NaOH solution. A long reaction time
allowed the complex to isomerize to its thermodynamically
stable state under the neutral to basic conditions where macro-
cyclic Co–N bonds, and thus conformational inversions, are
more labile.20 Subsequent acidification and oxidation allowed
the isolation of the cobalt() complex. The elemental analysis
is consistent with the formation of a clearly defined, pure prod-
uct which exists as the monohydrate in the crystalline form. Of
the many isomers possible, the equatorial trans-III isomer (see
crystal structure description below) crystallizes from the reac-
tion mixture in better than 90% isomeric purity, determined
from the 1H NMR spectrum of a CD3OD solution of the
product, together with some of the axial trans-III isomer. The
equatorial trans-III isomer obtained by the method described
exhibits a characteristic chemical shift (δ 3.70) for the methylene
group that links the anthracene to the cyclam, the axial isomer
has this group at δ 3.97. Other peaks of low peak area inte-
gration “shadow” the aromatic signals of the more abundant
isomer. A shorter reaction time always resulted in the final
product consisting of an even more complex mixture of other
conformations of the Co() complex, manifested by lower
yields of the characterized isomer, the appearance of other Ar–
CH2 signals, and concomitant additional complexity in both
the aromatic and aliphatic regions of the 1H NMR spectrum.
The trans-dichlorocobalt() complex of 6-(anthracen-9-
The structures were solved using a combination of direct
methods and difference synthesis.16,17 The site of the solvent
molecule in [Co(2)Cl2]Clؒ0.5CH3OH was found to be partially
occupied and was thus given a refineable occupancy factor. Due
to the disorder of the molecule no attempt was made to model
attached hydrogen atoms. For [Co(12)Cl2] the lattice solvent
molecule was also disordered and was included in the refine-
ment as an oxygen atom; no attempt was made to model
attached atoms. Protons bonded to the oxygen atoms were able
to be located using difference synthesis. These were modelled
and refined using a 50% occupancy for each of the protonation
sites. The structures were refined using a full-matrix least-
squares procedure based on F 2. Hydrogen atoms were located
from the difference map and were constrained to geometrical
estimates. Final refinement was carried out with anisotropic
displacement parameters applied to all non-hydrogen atoms
for [Co(2)Cl2]Clؒ0.5CH3OH. For [Co(12)Cl2] anisotropic dis-
placement parameters were applied to all non-hydrogen atoms
except the solvent molecule which, along with the hydrogen
atoms, were refined using isotropic displacement parameters. A
2
weighting scheme of type [σ2(Fo ) ϩ (aP)2 ϩ bP]Ϫ1 was used.
The atomic scattering factors were those incorporated in the
SHELXL-97 program system.17 Calculations were carried out
on a Vaxstation 4000VLC computer system. The Figures of the
cations and unit cells were prepared from the output of
ORTEPIII for Windows.18
CCDC reference numbers 211119–211122.
See http://www.rsc.org/suppdata/dt/b3/b305773a/ for crystal-
lographic data in CIF or other electronic format.
ylmethyl)-1,4,8,11-tetraazacyclotetradecane-5,7-dione
2 was
prepared in methanol by formation of the Co() complex in
situ, followed by the aerial oxidation of the Co() complex to
form the Co() complex. The product was isolated from an
acidic solution and recrystallised from acidified methanol to
yield the complex [Co(2)Cl2]Clؒ0.5CH3OH. Despite drying for
prolonged periods at 100 ЊC in vacuo the elemental analysis
revealed the presence of a half mole equivalent of solvate,
which was masked in the 1H NMR spectrum by the solvent (d4-
methanol) resonance. The same behaviour was observed when
the complex was recrystallised from ethanol, indicating the
presence of inaccessible voids in the solid state structure. A
single resonance, split into a doublet, is present in the 1H NMR
spectrum at δ 4.57 ppm due to the protons of the methylene
bridge. This is in accordance with the symmetry of the complex
and indicates the formation of only one isomer of the complex.
Puckering of the six-membered chelate ring incorporating
C(13) leads to an inequivalence of the C(13) protons. The
multiplets at δ 2.20 and 2.00 ppm, each integrating to one pro-
ton, are assigned to these protons. The IR spectrum of this
complex is consistent with the ligand in the complex existing as
the doubly protonated iminol tautomeric structure. The pos-
ition of the carbonyl stretch of the ligand when complexed to
Results and discussion
Syntheses of anthracenyl cyclam ligands and complexes
The anthracene-appended macrocyclic ligands were prepared
by the routes outlined in Scheme 1. The diamido-diamino
macrocycles (2) that can be prepared from aminolysis of suit-
able malonate diesters (1) initially seemed to be suitable
precursor compounds. This is the “standard” procedure for
preparing such C-alkylated macrocycles.19 Reduction of the
amide groups of dioxocyclam 2 using an excess of borane:
methyl sulfide complex in THF did not furnish the desired
1
anthrylmethylcyclam 3 cleanly. H NMR analysis of the prod-
ucts of several reduction attempts, performed under a variety
of conditions, indicated that the 1,4,8,11-tetraazacyclotetra-
decane (hereafter cyclam) salt (and its free base) obtained was
at best about 80% pure. We were unable to identify or remove
the contaminating byproducts. An alternative route to the syn-
thesis of these cyclams was thus investigated.
A recent report described the alkylation of the bis-aminal 7
(formed by condensation of 1,9-diamino-3,7-diazacyclononane
(2,3,2-tet) with 2,3-butanedione15) with 1,3-dibromopropane to
D a l t o n T r a n s . , 2 0 0 3 , 3 7 0 4 – 3 7 1 2
3707

View Article Online
Scheme 1 Reagents and conditions: (i) 2,3,2-tet, EtOH; (ii) BH3ؒSMe2, THF; (iii) LiAlH4, Et2O; (iv) TsCl, pyridine; (v) NaI, acetone; (vi) reagent 7,
K2CO3, MeCN; (vii) aq. HCl, EtOH; (viii) CoCl2ؒ6H2O, EtOH, air, then HCl; (ix) CoCl2ؒ6H2O, NaOH, air, then HCl.
Scheme 2 Reagents and conditions: (i) MeCOCOMe, CH3CN, Ϫ7 ЊC, 2 h; (ii) BrCH2CH2Br, K2CO3, CH3CN, 60 ЊC, 6 h; (iii) HCl, EtOH, 60 ЊC,
48 h.
the metal (compared to that of the free ligand) has been used as
an indicator of the structure and protonation state adopted by
ligands containing the 5,7-dioxo functionalisation.21 The free
ligand, 6-(anthracen-9-ylmethyl)-1,4,8,11-tetraazacyclotetra-
complex showed the parent ion at m/z 549.5 and 547 and the
isotopic distribution expected for an ion containing 35Cl and
37Cl. Peaks are also observed at m/z 511.5 and 475.6 consistent
with the loss of one and two axially coordinated chloride
ligands, respectively, in addition to one or two protons to
balance the charge.
decane-5,7-dione 2, displays a carbonyl stretch at 1665 cmϪ1
.
The carbonyl stretch of the corresponding cobalt() complex,
[Co(2)Cl2]Clؒ0.5CH3OH, appears in the IR spectrum at 1701
cmϪ1. The large frequency increase to 1700 cmϪ1 of this stretch
upon complexation to cobalt() is characteristic of the form-
ation of a doubly protonated iminol tautomeric structure.21 The
formation of the iminol tautomer is also consistent with the
Crystal structures
The tetracyclic bis(aminal) 8 crystallizes without lattice solvent
from acetone: the solid-state structure determined from a single
crystal X-ray study is depicted in Fig. 1 and shows the anti
disposition of the anthrylmethyl moiety with respect to the
butanedione-derived methyl groups. In the solid state the
1
multiplet at δ 6.34 ppm observed in the H NMR spectrum
which integrates to two protons and is assigned to the amine
protons of N(1) and N(11). The ESI-MS spectrum of the
D a l t o n T r a n s . , 2 0 0 3 , 3 7 0 4 – 3 7 1 2
3708

View Article Online
disordered as a result. Hydrogen bonding contacts between the
chloride ligands and the nitrogen atoms of the macrocyclic
skeleton link lying to the chloride and lattice water. Significant
aromatic coplanar interactions do not occur in the three-
dimensional lattice.
1H NMR experiments indicated that the complex [Co(2)Cl2]-
Clؒ0.5CH3OH was stable in methanol for a period of at least 2
days. However, in one attempt to grow crystals suitable for sin-
gle crystal X-ray diffraction studies, a solution of this complex
in d4-methanol was allowed to evaporate in the dark over a
period of some months during which time emerald green crys-
tals were deposited. An X-ray diffraction study of a represent-
ative crystal indicated the formation of a novel, and previously
unreported, Co() complex containing the highly conjugated
macrocyclic ligand 5,7-hydroxy-6-oxo-1,4,8,11-tetraazacyclo-
tetradecane-4,7-diene 12 as shown in Fig. 3. The growth of
crystals of [Co(2)Cl2]Clؒ0.5CH3OH suitable for diffraction
studies was accomplished by the addition of trimethyl-
orthoformate to a solution of [Co(2)Cl2]Clؒ0.5CH3OH and
using diffusion techniques instead of evaporative. Selected
bond lengths and angles for the compound 8 and complexes
[Co(3)Cl2]Clؒ4H2O, [Co(12)Cl2] and [Co(2)Cl2]Clؒ0.5CH3OH
are provided as ESI.† Diagrams of [Co(12)Cl2] and the cation of
[Co(2)Cl2]Clؒ0.5CH3OH are shown in Figs. 4 and 5.
Fig. 1 Diagram of the solid-state structure of 8. Ellipsoids are scaled
to encompass electron density at the 50% probability level. Hydrogen
atoms have been omitted.
aromatic anthracene moiety is bent away from the dimethyl-
decahydrotetraazapyrene framework although a slight twist
across the linking methylene bridge is present and C(17) and
C(12) approach to within 3.5 Å of one another. The methyl
groups of the dimethyldecahydrotetraazapyrene skeleton exist
in the cis-configuration, this configuration along with the fold-
ing of the decahydrotetraazapyrene framework (due to each of
the six-membered rings adopting a chair conformation) has
been reported for the non-methyl substituted derivatives.15,22
The twisted chiral molecule comprises the asymmetric unit of
the structure, but the centrosymmetric structure contains a
racemic mixture of both diastereoisomers. The central cross-
linking ethylene bridge has a torsion angle of Ϫ47.75(13) (for
C(28), C(26), C(27), C(29)). Three-dimensional stacking is
driven by aromatic coplanar interactions which occur with
planeؒ ؒ ؒplane distances of ca. 4.2 Å.
Fig. 3 Structure of trans-[CoIII(12)Cl2]
The cation in the solid-state structure of trans-[Co(3)Cl2]Clؒ
4H2O is depicted in Fig. 2. The Co–N and Co–Cl bond lengths
and angles of the slightly distorted octahedra are essentially
identical to those of the unelaborated [Co(cyclam)Cl2]Cl com-
plex.23 The chloride ligands are coordinated to the cobalt in the
trans-diaxial positions and the macrocycle adopts the trans-III
configuration, with the anthracenylmethyl group occupying an
equatorial orientation in the six-membered chair-conformed
chelate ring. The anthracene moiety is extended well away from
the cobalt cyclam system and does not come in close contact
with the atoms of the cyclam macrocycle or cobalt() ion. The
distance from the centre of the middle anthracene ring to the
cobalt atom is significantly greater than that found for trans-
[Co(2)Cl2]Cl, being 7.22 Å. The complexes within the unit cell
are stacked so that the anthracene chromophores lie directly
above one another and the cobalt ions also lie directly above
one another. The chloride ligands of one complex are within
hydrogen bonding distance of the nitrogen atoms of complexes
lying directly above and below the cobalt ion. There are
few hydrogen bonding contacts to lattice water, which is highly
Fig.
4
Diagram of solid-state structure of trans-[CoIII(12)Cl2].
Ellipsoids are scaled to encompass electron density at the 50%
probability level. Hydrogen atoms have been omitted.
Fig. 2 Diagram of the cation in the crystal structure of trans-
[CoIII(3)Cl2]Clؒ4H2O. Ellipsoids are scaled to encompass electron
density at the 35% probability level. Hydrogen atoms have been
omitted.
Fig. 5 Diagram of the cation in the crystal structure of trans-
[CoIII(2)Cl2]Clؒ0.5CH3OH. Ellipsoids are scaled to encompass electron
density at the 50% probability level. Hydrogen atoms have been
omitted.
D a l t o n T r a n s . , 2 0 0 3 , 3 7 0 4 – 3 7 1 2
3709

View Article Online
The crystal structure of [Co(12)Cl2] is interesting in a number
of respects. The complex exists in the iminol tautomeric form,
however only one of the oxygen atoms is protonated so that the
ligand has a single negative charge. As the cobalt() complex
also contains two trans-diaxial chloride ligands, the overall
charge of the complex is neutral. The complex crystallises in the
N-meso isomeric form with both of the amine protons on the
same side of the plane defined by the macrocycle. The crystal
packing of [Co(12)Cl2] is driven by close contact hydrogen
bonding between the protonated iminol oxygen atom of one
cation with the deprotonated iminol oxygen atom of a neigh-
bouring cation. This hydrogen bond has an O–H ؒ ؒ ؒ O angle
of 155(17)Њ. The ketone oxygen atom (O(6)) is bent towards
N(1) of an adjacent complex, forming a weak hydrogen bond
with an O–H ؒ ؒ ؒ O angle of 121(16)Њ. In addition to these,
close hydrogen bonding contacts between N(1) and O(5) of a
symmetry-related complex are observed. The chloride ligands
form no hydrogen bonding contacts and the solvent molecules
align in channels throughout the cation network.
of the anthracene to the cobalt atom occurs at C(18) and the
distance between these two atoms is only 4.65 Å, whilst
the distance from the cobalt ion to the centre of the middle
anthracene ring is 6.50 Å. The anthracene moiety is essentially
planar with the maximum deviation 0.0558(32) Å for C(25).
The least squares plane of the anthracene is at an angle of
29.14(8)Њ to that of the macrocycle.
Photochemistry
Photochemical investigations on the tethered donor–acceptor
complexes [Co(3)Cl2]Cl and [Co(2)Cl2]Cl have been carried out
and interesting photochemical behaviour is observed. The
absorption and emission data for the free ligands 2 and 3 and
their corresponding cobalt complexes along with appropriate
photophysical data are reported in Table 2. Steady-state fluor-
escence excitation and emission spectra of 2 and 3 are presented
in Fig. 6. The absorption and fluorescence spectra of the metal
free ligands display bands characteristic of the anthracene
fluorophore.
Within the complex, the cobalt atom lies within a slightly
distorted octahedral coordination environment. The Co–Cl
bond lengths and bite angles of the ethylene (85.6(3)Њ) and pro-
pylene bridges (94.9(4) and 93.8(4)Њ) are typical of those found
for related complexes. The Co–Nimine bond lengths (1.907(8) Å)
are similar to those found in cobalt() complexes contain-
ing derivatives of 5,7-dioxo-1,4,8,11-tetraazacyclotetradecane
lacking the oxo-group at C621 as well as [Co(5,7-dimethyl-6-
oxo-1,4,8,11-tetraazacyclotetradeca-4,7-diene)Cl2]ClO4.24 It is
noteworthy that none of these systems display the same degree
of conjugation as found in [Co(12)Cl2]. The highly conjugated
ring containing Co, N(4), C(5), C(6), C(7) and N(8) is distorted
from planarity and adopts a boat-like conformation. O(5) and
O(7) are bent 0.2198 Å below the plane defined by this ring
whilst O(6) is bent significantly above the plane, 0.7240 Å.
These trends are similar to that found for the related complex
[Co(5,7-dimethyl-6-oxo-1,4,8,11-tetraazacyclotetradeca-4,7-
diene)Cl2]ClO4,24 which also contains a ketone at C(6). It has
been postulated that oxidation at the C(6) position may lead to
a reduction in the strain of the macrocycle24 which may result
in a driving force for the decomposition of the cobalt()
tethered complex to this novel species.
The cyclam ligand backbone of [Co(12)Cl2] may also be
compared to that determined for the anthracene tethered
cobalt() complex, [Co(2)Cl2]Cl. The crystal structure of
[Co(2)Cl2]Cl reveals that this complex is isolated in the iminol
form, although, in contrast to [Co(12)Cl2], both of the oxygen
atoms are protonated and the ligand is neutral. The sp3 hybrid-
ised nitrogen atoms, N(1) and N(11), of [Co(2)Cl2]Cl are also in
the N-meso isomeric form.
Fig. 6 Steady-state fluorescence excitation and emission spectra of 3
[panel (a)] in DMF and 2 [panel (b)] in DMF.
Incorporation of the cobalt() ion into the ligand 2 leads to
a significant (approximately 3 nm) red shift of the anthracene
absorption bands indicating a ground state interaction between
the anthracene moiety and the cobalt ion. An analogous shift
has been observed in related nickel() complexes.11 In contrast
to the square-planer nickel() complex, however, for [Co(2)Cl2]-
Cl the octahedral geometry of Co() and the presence of the
axial ClϪ ligands hinders the approach of the anthracene to the
cobalt() ion. Despite this, the red shift indicates some inter-
action of the aromatic anthracene rings with a highly conju-
gated section of the macrocycle does occur, as shown by the
single-crystal X-ray diffraction studies. It is likely to be this
interaction between the anthracene and the orbitals of the
acceptor unit that causes the observed red shift in the anthra-
cene bands of [Co(2)Cl2]Cl.
The absorption spectrum of the donor–acceptor system
[Co(3)Cl2]Cl does not display a similar shift in the absorbance
of the anthracene bands. This is consistent with the fully
extended structure existing as the main conformer in solution,
as is observed in the solid state, and therefore the lack of inter-
action between the cobalt ion and anthracene chromophore in
the ground state.
The cobalt() atom in [Co(2)Cl2]Cl lies slightly out of the
least squares plane (0.1484(11) Å) of the macrocyclic ligand.
The coordination sphere is slightly distorted from
ideal octahedral geometry due to the restrictions imposed by
the five- and six-membered chelate rings of the macrocycle. The
observed bond lengths and bite angles of the macrocyclic
ligand are essentially identical to those found for related
complexes containing the 5,7-dioxo functionality21,24 as well as
[Co(12)Cl2]Cl. In contrast to the solid-state structure observed
in the corresponding four-coordinate nickel() complex,11 for
[Co(2)Cl2]Cl the appended anthracene moiety is not folded over
the top of the metal atom but is extended away from the macro-
cycle. Despite this, however, a small rotation about the C(6)–
C(15) bond is present leading to a torsion angle of 54.4(4)Њ for
C(7)–C(6)–C(15)–C(16). This results in the anthracene ring
atoms C(16), C(17), C(18) and C(19) lying directly over atoms
of the macrocycle (C(7), O(7), N(8) and C(9)). The closest
approach distance between the anthracene ring and the macro-
cycle occurs between C(16) and C(7) (3.02 Å) whilst the O(7)–
C(16) distance is 3.06 Å. Other approach distances for the
atoms listed above are 3.2–4.5 Å. The point of closest approach
Upon complexation of the ligand to cobalt(), the fluor-
escence of the anthracene fluorophore is quenched dram-
atically. The decay of fluorescence of the free ligand 3 did not
D a l t o n T r a n s . , 2 0 0 3 , 3 7 0 4 – 3 7 1 2
3710

View Article Online
Table 2 Photophysical properties of the donor–acceptor systems and the free ligands in DMF
Compound
λmax/nm
εmax/MϪ1 cmϪ1
Emission λmax/nm
Quantum yield
0.15
Lifetime(s)/ns
3
369
9250
417
1.68 (21%),
10.31 (79%)
0.06
a
[Co(3)Cl2]Cl
2
[Co(2)Cl2]Cl
370
370
373
9820
10000
11600
<0.01
0.68
<0.01
417
10.57
a
a
a Not measurable.
follow simple first-order kinetics but was comprised of two
first-order decay pathways, the major pathway (79%) having a
lifetime of 10.31 ns and a minor component (21%) with a life-
time of 1.68 ns. The presence of two lifetimes suggests that
there may be two conformational forms of 3 present in the
DMF solution, each of which have different energy relaxation
pathways. For the ligand 2, however, the fluorescence decay fol-
lows good first-order kinetics to yield a fluorescence lifetime of
10.57 ns.
electrode electrochemistry, the reduction potential of the CoIII/II
couple of [Co(dioxocyclam)Cl2]ϩ would be expected to be
approximately 100 mV more negative than that of [Co(cyclam)-
Cl2]ϩ, resulting in a more negative free energy change for the
electron transfer from the donor to acceptor in [Co(2)Cl2]Cl
than is calculated for [Co(3)Cl2]Cl. Hence for both the tethered
donor–acceptor systems the electron transfer process is
thermodynamically favourable.
Steady-state fluorescence experiments in ethanol glass (at
120 K) for both donor–acceptor complexes studied resulted
in no revival of the anthracene fluorescence. Whilst intra-
molecular electron transfer would be expected to be much more
dependent upon solvent interactions than intramolecular
energy transfer, the extent of destabilisation of the charge
separated state upon electron transfer in a glass and hence
recovery of fluorescence, is debatable. Additionally the
potential proximity of the donor and acceptor, particularly in
the case of [Co(2)Cl2]Cl, adds a further complication to the
interpretation of steady-state glass experiments.
The discussion above indicates that the quenching of the
anthracene fluorescence by both energy transfer or electron
transfer mechanisms is possible. Flash photolysis experiments
were unable to further elucidate the mechanism of quenching as
the formation of a charge separated state exhibiting the spectral
characteristics of the anthracene cation28 or changes in the
Co() centred ligand field bands at 678 nm due to the form-
ation of Co() were not observed. The lack of any significant
transient absorption in the spectra of these complexes indicates
that the lifetime of the species produced following either energy
or electron transfer from the donor to the acceptor is too short
for detection using the flash photolysis equipment employed.
The laser pulse used was of the order of 7–8 ns and this sets the
upper limit for the detection of short lived transients. Lifetimes
of ∼4 ns have been observed for the anthryl radical cation in
donor–acceptor systems containing polyimides as the acceptor
unit.28
The photochemistry of cobalt() complexes has been the
subject of considerable study in recent years.29 In preliminary
experiments with the present cobalt() complexes we have
shown that irradiation of the compounds at 372 nm in DMF
(which corresponds to the anthracene absorption bands) results
in a change in the visible absorption spectrum of both com-
pounds [Co(3)Cl2]Cl and [Co(2)Cl2]Cl. During the irradiation a
noticeable color change in the solutions occurred, consistent
with a change in the coordination sphere of the metal ion. For
both complexes a shift in the λmax and an increase in the absorb-
ance of the visible d–d band is observed along with the grow in
of an additional band. No change is observed over this time in
the spectra of the control solutions left in the dark. These
observations suggest that a change in the coordination sphere
of the cobalt() metal ion occurs as a direct result of irradi-
ation. It is known that the aquation of the cobalt() acceptor
unit of [Co(3)Cl2]Cl, [Co(cyclam)Cl2]3ϩ results in an increase in
the absorption coefficient around 570 nm as observed here.30 In
addition to this, prolonged heating of the acceptor unit of
[Co(2)Cl2]Cl, [Co(O2cyclam)Cl2]3ϩ in DMF led to an identical
change in the spectrum. It is likely that the change in the
absorption spectrum of both complexes is the result of the
replacement of the trans-diaxial chloride ligands with either
Complexation of 3 to cobalt() results in dramatic quench-
ing of the anthracene fluorescence by the coordinated cobalt
ion. However, for [Co(3)Cl2]Cl fluorescence decay analysis
could extract a very rapid major decay component of 0.06 ns
corresponding to the prescence of an additional non-radiative
process with a rate constant of 1.6 × 1010 Ϫ1. For [Co(2)Cl2]Cl
s
the rate of quenching is too fast to be detected . Metal com-
plexation thus in this case provides an extremely efficient fluor-
escence quenching pathway within the time resolution of our
time-correlated single-photon counting apparatus (<50 ps).
The mechanism of quenching of the anthracene excited state
by the cobalt ion may be via energy transfer (EnT) or electron
transfer (ET) or a combination of both mechanisms. Förster
resonance energy transfer (FRET) calculations were carried out
for the system [Co(3)Cl2]Cl using equations (2.16), (2.17) and
(2.18) of reference 25 with a J value of 6.369 × 10Ϫ17 MϪ1 cm3
and a κ2 value of 2/3. These calculations indicate that energy
transfer between the donor and acceptor is predicted to be effi-
cient (∼90%) and rapid (kEnT = 9.4 × 108 sϪ1). This is due to the
close proximity (7.22 Å) of the donor to the acceptor, which is
well within the calculated Förster distance of 10.6 Å. Although
these calculations show that energy transfer in [Co(3)Cl2]Cl is
efficient, the observed quenching is much higher than that pre-
dicted to occur by the Förster energy transfer mechanism. Thus
the energy transfer mechanism cannot explain all the quenching
observed. This difference could be explained by an electron
transfer mechanism occurring either solely or in competition
with energy transfer. Analogous calculations cannot be carried
out for the donor–acceptor system [Co(2)Cl2]Cl due to the close
proximity of the donor to the acceptor and resulting possibility
of energy transfer via the Dexter mechanism.
The driving force for electron transfer from the donor to the
acceptor can be estimated using the Rehm–Weller equation
which requires a knowledge of the reduction potentials of the
donor and acceptor and the excitation energy of the chromo-
phore.26 In DMF the cyclic voltammogram of the acceptor
unit, [Co(cyclam)Cl2]ϩ, shows quasi-reversible oxidation and
reduction processes for the CoIII/II couple allowing the
determination of the half-wave potential as E1/2 = Ϫ0.794 V
(with respect to Ag/0.01 Agϩ). Thus, the free energy change for
the electron transfer can be calculated as ∆GeT = Ϫ1.19 eV
(based on the experimentally obtained Eoo value of 2.973 eV
from the λmax emission peak of the free ligand at 417 nm and
E1/2 (Anϩ/An) = 0.99 V27), indicating that there is a strong driving
force for electron transfer from the donor to the acceptor in this
system. For [Co(2)Cl2]Cl the reduction potential of the acceptor
unit, [Co(dioxocyclam)Cl2]ϩ (dioxocyclam is 1,4,8,11-tetraaza-
cyclotetradecane-5,7-dione), could not be determined in DMF
with any certainty. However, based upon data obtained in aque-
ous solution using both cyclic voltammetry and rotating disk
D a l t o n T r a n s . , 2 0 0 3 , 3 7 0 4 – 3 7 1 2
3711

View Article Online
4 V. Amendola, L. Fabbrizzi, M. Licchelli, C. Mangano, P. Pallavicini,
L. Parodi and A. Poggi, Coord. Chem. Rev., 1999, 190, 649.
5 L. Fabbrizzi, M. Licchelli, P. Pallavicini and L. Parodi, Angew.
Chem., Int. Ed., 1998, 37, 800.
6 L. Fabbrizzi, M. Licchelli, P. Pallavicini, A. Perotti and D. Sacchi,
Angew. Chem., Int. Ed. Engl., 1994, 33, 1975.
7 L. Fabbrizzi, M. Licchelli, P. Pallavicini, A. Perotti, A. Taglietti and
D. Sacchi, Chem. Eur. J., 1996, 2, 75.
8 L. Fabbrizzi, M. Licchelli, P. Pallavicini, D. Sacchi and A. Taglietti,
Analyst, 1996, 121, 1763.
trace amounts of water or DMF and that the irradiation of the
complex results in ligand release from the donor–acceptor sys-
tems. It is noted that under the same conditions irradiation of
the acceptor unit for [Co(3)Cl2]Cl, [Co(cyclam)Cl2]3ϩ led to no
observable change in the UV-visible spectrum, whilst irradi-
ation of the acceptor unit for [Co(2)Cl2]Cl, [Co(O2cyclam)-
Cl2]3ϩ under the same conditions led to a change in the spec-
trum, with the grow in of a band at λ = 610 nm and the band at
λmax = 626 nm shifting to λmax = 676 nm along with an increase
in its intensity.
9 B. Juskowiak and W. Szczepaniak, Anal. Chim. Acta., 1994, 290,
121.
10 P. V. Bernhardt, E. G. Moore and M. J. Riley, Inorg. Chem., 2001, 40,
5799.
Conclusions
11 S. Boyd, N. M. Cabral, K. P. Ghiggino, M. J. Grannas,
W. D. McFadyen and P. A. Tregloan, Aust. J. Chem., 2000, 53, 651.
12 M. R. Wasielewski, Chem. Rev., 1992, 92, 435.
13 C. A. Parker and T. A. Joyce, Chem. Commun., 1967, 744.
14 D. V. O’Connor and D. Phillips, Time Correlated Single Photon
Counting; Academic Press, New York, 1983.
Although photo-activated ligand release occurs for the donor–
acceptor systems, the mechanism operating is not clear. Elec-
tron and energy transfer are possible in the anthracene tethered
donor–acceptor systems [Co(3)Cl2]Cl and [Co(2)Cl2]Cl and
both of these processes could lead to the observed behaviour.
To better understand the factors responsible we are presently
examining related complexes in which a range of axial ligands
are coordinated to a cobalt() cyclam moiety tethered to either
anthracene and naphthalene donors.
15 G. Herve, H. Bernard, N. Le Bris, J.-J. Yaouanc, H. Handel and
L. Toupet, Tetrahedron Lett., 1998, 39, 6861.
16 G. M. Sheldrick, Acta. Crystallogr., Sect. A., 1990, 46, 467.
17 G. M. Sheldrick, in SHELXL-97, Program for the Refinement of
Crystal Structures, University of Göttingen, Germany, 1997.
18 L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
19 I. Tabushi, Y. Taniguchi and H. Kato, Tetrahedron Lett., 1977, 12,
1049.
20 I. I. Creaser, T. Komorita, A. M. Sargeson, A. C. Willis and
K. Yamanari, Aust. J. Chem., 1994, 47, 529.
Acknowledgements
21 A. Jyo, Y. Terazono and H. Egawa, Anal. Sci., 1995, 11, 51.
22 P. Gluzinski, J. W. Krajewski, Z. Urbanczyk-Lipkowska, J. Bleidelis
and A. Kemme, Acta Crystallogr., Sect. B, 1982, 38, 3038.
23 M. E. Sosa-Torres and R. A. Toscano, Acta. Crystallogr., Sect. C,
1997, 53, 1585.
24 B. Durham, T. J. Anderson, J. A. Switzer, J. F. Endicott and
M. D. Glick, Inorg. Chem., 1977, 16, 271.
We thank the Australian Research Council for financial
support. A. M. F. acknowledges the support of an AINSE
scholarship. Assistance with the photochemical measurements
by Mr Nuno Cabral is gratefully acknowledged. The authors
would also like to acknowledge most helpful discussions with
Dr Brendan Abrahams concerning X-ray crystallography.
25 B. Wieb Van Der Meer, G. Coker III and S.-Y. Simon Chen,
Resonance Energy Transfer Theory and Data, VCH Publishers, Inc.,
1994.
26 A. Z. Weller, Phys. Chem. Neue Folge, 1982, 130, 93.
27 B. Case, N. S. Hush, R. Parsons and M. E. Peover, J. Electroanal.
Chem., 1965, 10, 360.
References
1 R. Bergonzi, L. Fabbrizzi, M. Licchelli and C. Mangano, Coord.
Chem. Rev., 1998, 170, 31.
2 G. D. Santis, L. Fabbrizzi, M. Licchelli, N. Sardone and A. H.
Velders, Chem. Eur. J., 1996, 2, 1243.
3 G. D. Santis, L. Fabbrizzi, M. Licchelli, C. Mangano and D. Sacchi,
Inorg. Chem., 1995, 34, 3581.
28 W. Li and M. A. Fox, J. Phys. Chem. B., 1997, 101, 11068.
29 D. M. Roundhill, Photochemistry and Photophysics of Metal
Complexes, Plenum Press, New York, 1994, pp. 42–49.
30 C. K. Poon and M. L. Tobe, J. Chem. Soc. A, 1967, 2069.
D a l t o n T r a n s . , 2 0 0 3 , 3 7 0 4 – 3 7 1 2
3712