Synthesis and reactivity of the macrobicyclic complexes (1,5,8,12-tetraaza-17-oxabicyclo[10.5.2]-nonadecane)cobalt(III) perchlorate ([Co(L1)(CIO4)]-(CIO4)2), [(chloro(1,4,8,11-tetraaza-17-oxabicyclo-[9.5.3]nonadecane)cobalt(III) perchlorate ([Co(L2)-(CI)](CIO4)2)

Article abstract of DOI:10.1139/v05-089

The potentially penta-coordinating ligands L₂ and L₃ have been synthesized by reaction of the 10-membered macrocycle 1,5-diaza-8-oxacyclodecane with either 1,9-dichloro-3,7-diazanonane-2,8-dione, and subsequent reduction of the diamide (3), or with 1,9-dichloro-3,7-(dimethyl) diazanonane-2,8-dione and reduction (L₃)-. A similar procedure is outlined for the dimethylated macrobicycle (L₄), based on the corresponding nine-membered 1,4-diaza-7-oxacyclononane. The Co(III) complexes of these ligands and of 1,5,8,12-tetraaza-17-oxabicyclo[10.5.2]nonadecane (L ₁) have been prepared. Spectrophotometric determinations on the L₁ complex ion confirm the presence of two hydrogen ion related equilibria, one of which (pK = 2.2 ± 0.2) is associated with the proposed replacement of the apical ether oxygen by a water molecule that is bound to the metal centre and hydrogen bonded to the ether. The species derived from L ₁ and L₂ exhibit remarkable kinetic stability. Studies on the anation of the [Co(L₁)(H₂O)](CIO₄) ₃ with chloride ions in acidic media are consistent with the reaction of both the proton-related complex and the [Co(L₁)(H ₂O)]³⁺ ion. Whilst complex ions containing two secondary NH groups maintain the metal-ion coordination in strongly acidic media, corresponding species with ligands containing all four tertiary amine sites are subject to attack by protons leading to a relatively facile demetallation of the complexes. The latter finding is supported by kinetic studies and mass spectrometric fragmentation patterns of the ions.

Full text of DOI:10.1139/v05-089

8
94
Synthesis and reactivity of the macrobicyclic com-
plexes (1,5,8,12-tetraaza-17-oxabicyclo[10.5.2]-
nonadecane)cobalt(III) perchlorate ([Co(L )(ClO )]-
1
4
(
ClO ) ), [(chloro(1,4,8,11-tetraaza-17-oxabicyclo-
4
2
[
9.5.3]nonadecane)cobalt(III) perchlorate ([Co(L )-
2
(
Cl)](ClO ) ), (4,8-dimethyl-1,4,8,11-tetraaza-17-
4
2
oxabicyclo[9.5.3]nonadecane)cobalt(III) perchlorate
[Co(L )(ClO )](ClO ) ), and (5,8-dimethyl-1,5,8,12-
(
3
4
4 2
tetraaza-17-oxabicyclo[10.5.2]nonadecane)co-
balt(III) perchlorate ([Co(L )(ClO )](ClO ) ) — Crystal
4
4
4 2
1
structure of the L complex
2
Theo Rodopoulos, Koji Ishihara, Mary Rodopoulos, M.J. Zaworotko, M. Maeder,
and A. McAuley
Abstract: The potentially penta-coordinating ligands L and L have been synthesized by reaction of the 10-membered
2
3
macrocycle 1,5-diaza-8-oxacyclodecane with either 1,9-dichloro-3,7-diazanonane-2,8-dione, and subsequent reduction of
the diamide (3), or with 1,9-dichloro-3,7-(dimethyl)diazanonane-2,8-dione and reduction (L ). A similar procedure is
3
outlined for the dimethylated macrobicycle (L ), based on the corresponding nine-membered 1,4-diaza-7-oxacyclononane.
4
The Co(III) complexes of these ligands and of 1,5,8,12-tetraaza-17-oxabicyclo[10.5.2]nonadecane (L ) have been pre-
1
pared. Spectrophotometric determinations on the L complex ion confirm the presence of two hydrogen ion related
1
equilibria, one of which (pK = 2.2 ± 0.2) is associated with the proposed replacement of the apical ether oxygen by a
water molecule that is bound to the metal centre and hydrogen bonded to the ether. The species derived from L and
1
L exhibit remarkable kinetic stability. Studies on the anation of the [Co(L )(H O)](ClO ) with chloride ions in acidic
2
1
2
4 3
3
+
media are consistent with the reaction of both the proton-related complex and the [Co(L )(H O)] ion. Whilst complex
1
2
ions containing two secondary NH groups maintain the metal–ion coordination in strongly acidic media, corresponding
species with ligands containing all four tertiary amine sites are subject to attack by protons leading to a relatively fac-
ile demetallation of the complexes. The latter finding is supported by kinetic studies and mass spectrometric fragmenta-
tion patterns of the ions.
Key words: cobalt(III), macrobicycle, hydrolysis, anation, spectroscopic analysis. 902
Résumé : Les ligands L et L qui pourraient donner lieu à une pentacoordination ont été synthétisés en faisant appel
2
3
à la réaction du macrocycle à dix chaînons 1,5-diaza-8-oxacyclodécane avec la 1,9-dichloro-3,7-diazanonane-2,8-dione
Received 3 December 2004. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 23 July 2005. Reposted on
the Web site with correction on 16 August 2005.
It is a pleasure to recognize the many contributions to chemistry in Canada by colleague and friend Howard Alper.
T. Rodopoulos. CSIRO, Clayton, Victoria 3169, Australia.
K. Ishihara. Department of Chemistry, Waseda University, Shinjuku-ku, Tokyo 169-8555, Japan.
M. Rodopoulos. Department of Chemistry, University of South Florida, Tampa, FL 33620, USA.
M.J. Zaworotko. Institute of Drug Technology, Boronia, Victoria 3155, Australia.
M. Maeder. Chemistry Department, University of Newcastle, Callaghan 2308, Australia.
2
A. McAuley. Chemistry Department, University of Victoria, Victoria, BC V8W 3V6, Canada.
1
This article is part of a Special Issue dedicated to Professor Howard Alper.
Corresponding author (e-mail: amcauley@uvic.ca).
2
Can. J. Chem. 83: 894–902 (2005)
doi: 10.1139/V05-089
© 2005 NRC Canada

Rodopoulos et al.
895
et réduction subséquente du diamide cyclique (3) ou avec la 1,9-dichloro-3,7-(diméthyl)diazanonane-2,8-dione suivie
d’une réduction (L ). On décrit une méthode semblable pour le macrobicycle diméthylé (L ) qui est basée sur le com-
3
4
posé à neuf chaînons 1,4-diaza-7-oxacyclononane. On a préparé les complexes du cobalt(III) avec ces ligands et avec
le 1,5,8,12-tétraaza-17-oxabicyclo[10.5.2]nonadécane (L ). Des déterminations spectrophotométriques sur le complexe
1
L1 confirment la présence de deux équilibres apparentés des ions hydrogènes, dont l’un (pK = 2,2 ± 0,2) est associé au
remplacement proposé de l’oxygène de l’éther apical par une molécule d’eau qui est liée au centre métallique et relié à
l’éther par une liaison hydrogène. Les espèces qui dérivent des ligands L et L présentent une stabilité cinétique re-
1
2
marquable. Des études d’anation du [Co(L )(H O)](ClO ) avec des ions chlorures en milieu acide sont en accord avec
1
2
4 3
3
+
une réaction du complexe lié par un proton et l’ion [Co(L )(H O)] . Alors que les ions complexes comportant deux
1
2
groupes NH secondaires maintiennent la coordination métal–ion en milieu fortement acide, les espèces correspondantes
comportant des ligands sur chacun des quatre sites amines tertiaires sont sujettes à des attaques par des protons ce qui
conduit à une démétallation relativement facile des complexes. Cette dernière observation est supportée par des études
cinétiques et par les patrons de fragmentation en spectrométrie de masse des ions.
Mots clés : cobalt(III), macrobicycle, hydrolyse, anation, analyse spectroscopique.
[
Traduit par la Rédaction] Rodopoulos et al.
adecane (L ) were synthesized as indicated in the following.
Introduction
4
Acetonitrile (Caledon) was fractionally distilled from CaH2
and stored under nitrogen. Water, with a specific resistance
of >15 MΩ cm was prepared using a MilliQ-Reagent sys-
tem. Co(ClO ) ·6H O was synthesized from CoCO ·xH O
The synthesis of novel macrocyclic and polymacrocyclic
ligands and the significant thermodynamic and kinetic stabil-
ity imparted by them upon metal ion complexation has been
a feature of interest for over two decades (1–5). Such species
are now found in a variety of chemistries from competitive
anion binding sites (6), multimacrocyclic luminescent arrays
4
2
2
3
2
(Aldrich) through reaction with the stoichiometric amount of
HClO . The perchlorate salt was recrystallized three times
4
from water, dried under high vacuum for 48 h at room tem-
(
7), and polynuclear complexes of pendant-arm ligands with
perature, and stored over CaSO in a desiccator. Ethanol
4
unusual spin states (8). Reports from previous investigations
from this laboratory have described macrobicyclic systems
involving five-coordinate Cu(II) complexes (9) and six-
coordinate Ni(II) (10, 11) species of L and L , where, in the
(Commercial Alcohols, 100%), activated charcoal (Sigma,
250–350 mesh), HClO (Allied Chemicals, 70% assay), HCl
4
(Anachemia, 36.5%–38.0% assay), NaClO ·H O (Fisher
4
2
1
2
Scientific), NaOH (BDH), PbO (BDH), D O (Cambridge
2
2
8
latter, the d ion, being octahedral, exhibits a single substitu-
tion site. Such species exhibit extraordinary kinetic stability,
with no detectable decomposition over days even in strongly
acidic (4 mol/L) media. In addition, spectroscopic observa-
tions are consistent with the formation of nickel(III) species
of sufficient lifetimes to permit kinetic measurements not
only of redox reactions (12), but also, in some instances, the
Ni(III) ions formed are sufficiently stable kinetically to per-
mit rate studies of substitution by chloride (13). Hydrogen-
ion dependences in these investigations are consistent with
an unusually strongly acidic proton-related equilibrium asso-
ciated with the ether linkage (13, 14). In this paper, we pres-
ent the results of comparable studies of analogous cobalt(III)
Isotope Laboratories), and CD CN (Cambridge Isotope Lab-
3
oratories) were all used as purchased. Mass spectra were
recorded on a Kratos Concept mass spectrometer in the LSI-
MS mode. The matrix employed in all cases was m-
nitrobenzyl alcohol. Bruker WM-250 and AC-300 NMR
1
13
spectrometers were used to record the H and C NMR
1
3
spectra. C NMR chemical shifts are referenced to the re-
sidual methyl carbon in incompletely deuterated CD CN
3
1
(δ (CH CN) = 1.30 ppm) and H NMR chemical shifts to the
3
residual protons in incompletely deuterated D O (δ (H O) =
2
2
4.70 ppm). Elemental analyses were performed by Canadian
Microanalytical Services, Vancouver, British Columbia, Canada.
6
complexes where the d ion imparts much greater kinetic
Synthesis
stability than the nickel(III) ion, providing an opportunity to
examine not only kinetic features, but also to spectroscopi-
cally identify species involved in the various proton-related
equilibria. Included are kinetic observations involving a
tetra-N-substituted macrobicycle where the stability in acidic
media is diminished owing to protonation at the N-methyl
tertiary amines competing with the complexation at the
metal centre.
(1,5,8,12-Tetraaza-17-oxabicyclo[10.5.2]nonadecane)co-
balt(III) perchlorate ([Co(L )(ClO )](ClO ) )
1
4
4 2
Ligand L (0.299 g, 1.106 mmol) was dissolved in dry,
1
3
distilled acetonitrile (30 cm ) and an equivalent of cobalt(II)
perchlorate hexahydrate (0.405 g, 1.106 mmol) was added
with stirring. The solution was acidified to approximately
pH 3 with HClO and an excess of lead(II) oxide (0.400 g,
4
1
.672 mmol) added. Upon bubbling the mixture with air
overnight, a maroon solution was obtained. The mixture was
centrifuged and the excess solid lead(II) oxide filtered leav-
ing the maroon-coloured solution. Concentration of the fil-
Experimental
The literature method was used to prepare 17-oxa-1,5,8,
3
1
1
2-tetraazabicyclo[10.5.2]nonadecane (L ) (11). The ligands
trate to approximately 10 cm and cooling in an ice bath
1
7-oxa-1,4,8,11-tetraazabicyclo[9.5.3]nonadecane (L2), 4,8-
resulted in the precipitation of a red solid. Crystallization of
the complex was achieved by redissolution of the solid in the
minimum amount of acetonitrile and allowing the solution to
dimethyl-1,4,8,11-tetraaza-17-oxabicyclo[9.5.3]nonadecane
and 5,8-dimethyl-1,5,8,12-tetraaza-17-oxabicyclo-[10.5.2]non-
(L3),
©
2005 NRC Canada

8
96
Can. J. Chem. Vol. 83, 2005
evaporate slowly. Yield: 0.165 g (24%). 13C NMR (D O):
Scheme 1.
2
7
5.7 (-CH -O-), 61.4, 58.0, 55.3, 53.2, 46.9 (-CH -N-), 24.45
2
2
O
(
-C-CH -C-). Anal. calcd. for [Co(C H N O)(ClO )](ClO )
2 14 30 4 4 4 2
H
N
(%): C 26.79, H 4.82, N 8.93; found: C 26.77, H 4.87, N 9.54.
NH
NH
O
2
2
CHCl / H O
3
Cl
Cl
2
+
2
1
1
7-Oxa-1,4,8,11-tetraazabicyclo[9.5.3]nonadecane (L2)
The synthesis was carried out as shown in Scheme 1.
K2CO3
Cl
Cl
N
H
,9-Dichloro-3,7-diaza-nonane-2,8 dione (1)
O
1
Propylenediamine (4.88 g, 0.0567 mol) was added under
constant stirring to a mixture of CHCl (69 mL) and H O
3
2
(
34 mL) and cooled in an ice bath. A mixture of chloro-
HN
HN
CH
3
CN
acetyl chloride (13.9 mL, 0.172 mol) in CHCl (86 mL) and
K CO (22.36 g, 0.162 mol) in water (860 mL) were added
simultaneously under a nitrogen atmosphere over a 1 h pe-
riod. The reaction mixture was stirred at room temperature
3
O
/Na
2
CO
3
2
3
2
for a further 2 h. The CHCl layer was separated, washed
3
O
with water (2 × 100 mL), and dried over MgSO . Removal
4
H
of solvent yielded a colourless oil that was dried in vacuo
H
N
N
N
overnight. Upon addition of CCl and cooling overnight, a
N
N
4
3
BH /thf
O
white solid precipitated. Filtration and concentration of the
O
solution followed by a further addition of CCl afforded a
N
O
N
4
N
H
second crop of material. The diamide (1) was dried under
H
vacuum. Yield: 9.01 g, 70%. The material was used as such.
3
1
H NMR (D O): 7.2 (N-H), 4.1 (-CH -C(O)-), 3.4 (CH -
L
2
2
2
2
1
3
NH), 1.8 (-C-CH -C-). C NMR (CDCl ): 170.1 (-C=O),
4
2
3
1
3
2.7 (CH -NH), 37.4 (-CH -C(O)-), 27.9 (-C-CH -C-).
C
2
2
2
NMR (D O): 169.7 (-C=O), 42.2 (CH -NH), 37.0 (-CH -
2
2
2
[(Chloro)(1,4,8,11-tetraaza-17-oxabicyclo[9.5.3]nona-
decane)cobalt(III)] perchlorate ([Co(L )(Cl)](ClO ) )
C(O)-), 27.5 (-C-CH -C-).
2
2
4 2
Cyclization with the free macrocycle, 1,5-diaza-8-oxa-
cyclodecane (2), was carried out in basic acetonitrile.
Doubly distilled acetonitrile (2.5 L) was placed in a 5 L
four-necked flask fitted with a stirring attachment and a con-
An aqueous solution of cobalt(II) perchlorate hexahydrate
(0.134 g, 0.366 mmol) was added to a refluxing ethanol–
3
H O (4:1, 30 cm ) solution of the ligand L (0.099 g,
2
2
0.366 mmol). The resulting solution turned green almost in-
stantaneously. A catalytic amount of activated charcoal was
then added and the solution refluxed for 2 h. After cooling,
air was bubbled through the mixture for several hours yield-
ing a purplish-red solution. The charcoal was filtered off and
the solvent removed under reduced pressure. An aqueous so-
lution of the crude complex was purified by passage down a
Dowex 50W-x8 (50–100 mesh) exchange column. The com-
plex was eluted with 6 mol/L HCl and the eluate evaporated
denser and two septa for syringes. A stream of N was at-
2
tached and Na CO (2.402 g, 22.66 mmol, 3.16 equiv.) was
2
3
added to the bulk CH CN. The macrocycle 2 (1.036 g,
3
7
.183 mmol) was dissolved in 200 mL of dry CH CN.
3
Diamide 1 showed limited solubility in acetonitrile and only
.408 g (1.797 mmol) was dissolved in 50 mL of solvent.
The CH CN–Na CO solution was heated to reflux and
0
3
2
3
5
0 mL each of the reactant solutions 1 and 2 were added
slowly (four portions) over a period of 40 h. Reflux was
continued for another 24 h. After filtration of the Na CO ,
the acetonitrile was removed in vacuo leaving a yellow,
to dryness, leaving a green solid, which was recrystallized
3
2
3
from a hot aq. 0.2 mol/L NaClO (10 cm ) solution, col-
4
lected, washed with ethanol and diethyl ether, and air dried.
1
3
quasi-crystalline solid. Dissolution in CHCl removed the re-
Yield: 0.082 g, 40%. C NMR (D O): 82.4 (-CH -O-), 67.4,
3
2 2
sidual amounts of diamide impurity yielding a yellow crys-
58.7, 58.0, 52.7, 50.1 (-CH -N-), 27.8, 24.0 (-C-CH -C-).
2 2
talline solid 3. Yield: 2.07 g, 96%. 13C NMR (CDCl ): 174.8
3
Anal. calcd. for [Co(C H N O)(Cl)](ClO ) ·H O (%): C
14 30 4 4 2 2
(
-C=O), 66.7 (CH -O-), 62.1, 57.4, 55.9 (CH -N), 37.4
28.91, H 5.54, N 9.63; found: C 28.94, H 5.24, N 9.64.
Neutralizing an aqueous solution of the monochloride
complex with 0.1 mol/L NaOH solution resulted in a colour
change from green to orange-red. Leaving the solution open
to the atmosphere to allow the solvent to evaporate slowly
yielded the purple perchlorate complex, [Co(L )(ClO )]-
2
2
+
(-CH -CH -NH-), 29.8, 27.7 (-C-CH -C-). MS: 299 (M + 1) .
2
2
2
The fully saturated macrocycle (L ) was prepared by addi-
2
tion of a solution of BH ·THF (100 mL) to a solution of 3
3
(
2
1.002 g, 3.358 mmol) in dry THF (30 mL) and refluxing for
4 h. After cooling to room temperature, excess borane was
2
4
destroyed with water. Upon removal of the solvent under
vacuum, the white solid formed was treated with 6 mol/L
HCl and refluxed for 3 h. Upon cooling, basification with
(ClO ) , which was filtered out, washed with ethanol, and air
dried.
4
2
4
,8-Dimethyl-1,4,8,11-tetraaza-17-oxabicyclo[9.5.3]nona-
KOH pellets to pH 14 and extraction with CHCl , the solu-
3
decane (L3)
tion was dried, filtered, and the solvent removed, yielding a
slightly yellow oil. Yield: 0.742 g, 82%. 1 C NMR (CDCl3):
3
The ligand, with no ionizable protons, was prepared in a
manner similar to that shown for L (Scheme 1, vide supra).
7
3.6 (-CH -O), 58.5, 58.1, 56.9 (CH -N), 47.8, 47.3 (-CH -
2
2
2
2
N,N′-Dimethylpropylenediamine (5 g, 0.0489 mol) was added
CH -NH-), 28.2, 28.1 (-C-CH -C-).
2
2
©
2005 NRC Canada

Rodopoulos et al.
897
3
work-up (1.944 g, 85%) was observed to be a mixture. Puri-
fication was achieved by preparation of the corresponding
copper(II) complex (0.642 g) followed by decomplexation
Me 4
N
5
8
7
2
9
3
N
6
1
O
of the metal ion by use of Na S (9). The final yield of L
2
3
1
3
was 0.426 g (21% from macrocycle 2). C NMR (CDCl3):
2.1 (CH -O), 56.6, 56.2, 56.0, 55.5, 54.7 (CH -N), 43.2 (N-
N
N
7
2
2
Me
+
CH ), 26.6, 25.8 (CH -CH -CH ). MS: 298 (M ), 299 (M +
1
3
2
2
2
+
+
) , 327 (M + 29) .
L
3
4
,8-Dimethyl-1,4,8,11-tetraaza-17-oxabicyclo[9.5.3]nona-
decane-cobalt(III) perchlorate ([Co(L )(ClO )](ClO ) )
3
4
4 2
under constant stirring to a mixture of CHCl (69 mL) and
3
The ligand L (0.132 g, 0.442 mmol) was dissolved in an
ethanol–water mixture (48 mL : 12 mL) and heated to re-
H O (34 mL) and cooled in an ice bath. A mixture of
3
2
chloroacetyl chloride (22.11 g, 0.1957 mol) in CHCl3
flux. An aqueous solution of Co(ClO ) ·6H O (0.162 g,
(
86 mL) and K CO (22.36 g, 0.162 mol) in water (860 mL)
4 2
2
2
3
0
.442 mmol) was added and the solution turned a yellow-
was added simultaneously under a nitrogen atmosphere over
a 1 h period. The reaction mixture was stirred at room tem-
green colour. Several grams of activated charcoal were intro-
duced and the solution was refluxed for 2 h. Air bubbling of
the cooled solution (5 h) followed by filtration yielded a
pink-red solution. After removal of the residual ethanol and
perature for a further 2 h. The CHCl layer was separated,
3
washed with water (2 × 100 mL), and dried over MgSO4.
Removal of the solvent yielded a colourless oil that was
addition of 1 drop of concd. HClO , the solution was left to
dried in vacuo overnight. Upon addition of CCl and cooling
4
4
crystallize at 4 °C overnight. The crystals were washed with
overnight, a white solid precipitated. Filtration and concen-
tration of the filtrate followed by a further addition of CCl4
afforded a second crop of material. The diamide 4 was
washed with ethanol and dried. Yield: 9.5 g, 76%. The NMR
ethanol, ether, and air dried. Yield: 0.106 g, 37%. 13C NMR
(
3
D O): 66.9 (CH -O), 60.9, 55.0, 54.6, 52.3, 51.7 (CH -N),
2 2 2
9.0 (N-CH ), 19.7 (CH -CH -CH ). It appears that there is
3 2 2 2
1
3
overlap of the methylene peaks, since only eight and not
spectrum showed the presence of several isomers. C NMR
nine signals were observed. MS (negative ion): 654.05
(
4
CDCl ): 166.2, 166.7 (-C=O), 45.7, 45.6 (CH -NMe), 41.0,
3
2
–
(
5
[Co(L )(ClO )](ClO ) ] . UV–vis (λ , nm): 232 (ε =
1.4, 41.6 (-CH -C(O)-), 24.3, 26.6 (-C-CH -C-).
3
4
4 2
max
2
2
–1
–1
–1
–1
380 (mol/L) cm ) and 516 (ε = 17 (mol/L) cm ).
The isomeric macrobicyclic ligand, 5,8-dimethyl-1,5,8,12-
To a solution of the bis(chloro)diamide 4 (1.718 g,
6
.73 mmol), in dry distilled CH CN (200 mL), was added
3
tetraaza-17-oxabicyclo[10.5.2]nonadecane (L ), was pre-
sodium iodide (5.548 g, 37.01 mmol). The macrocycle 2
4
4
pared in a manner similar to that outlined previously using
(
(
0.971 g, 6.73 mmol) was dissolved in dry distilled CH CN
200 mL) and solutions of the reagents 4 and 2 were added
3
N,N′-dimethylethylenediamine and 3-chloropropionyl chlo-
ride as initial reactants. Further reaction of the resultant
bis(chloro)diamide with the macrocycle 1,4-diaza-7-
oxacyclononane yielded the bicyclic diamide that upon re-
very slowly (over a 24 h period) under N in 50 mL batches
2
to a refluxing solution of dry distilled acetonitrile (2.5 L)
containing Na CO (9.415 g, 88.83 mmol). Heating of the
2
3
duction, as described previously, yielded ligand L . For the
mixture, now orange coloured, was continued for a further
0 h, and the mixture was cooled overnight. Filtration under
nitrogen, removal of the solvent, and addition of CHCl3
250 mL) yielded an orange solution and an off-white solid.
4
Co(III) complex [Co(L )(Cl)](ClO ) calcd. (%): C 32.48, H
1
4
4 2
5
(
.79, N 9.47; found: C 32.87, H 5.84, N 9.57. UV–vis
–
1
–1
λmax, nm): 461 (ε = 16 (mol/L) cm ).
(
Caution! Transition-metal perchlorates are known to be
After washing with water and further addition of CHCl , the
3
hazardous and must be treated with care, especially in the
presence of organic solvents.
solution was dried (Na SO ), filtered, and the solvent re-
2
4
moved yielding an orange-brown oil that was dried under
high vacuum overnight. The cyclised diamide, 4,8-dimethyl-
1
,4,8,11-tetraaza-17-oxabicyclo[9.5.3]nonadecane-2,9-dione
Crystallography
(
5), solidified as a yellow, crispy honeycomb that was quite
+
hygroscopic. Yield: 2.0 g, 90%. MS: 327 (M + 1) .
The crude diamide 5 (2.5 g, 10.46 mmol) was transferred
to a 1 L flask using dry, freshly distilled THF (170 mL). Af-
[Co(L2)(Cl)](ClO4)2
X-ray quality crystals were grown by slow evaporation of
an aqueous solution containing the complex and NaClO4.
The experimental parameters for the complex are listed in
Table 1.
A crystal of dimensions 0.10 mm × 0.15 mm × 0.40 mm
was mounted in a glass Lindemann tube and optically cen-
tred in an Enraf-Nonius CAD-4 diffractometer employing
graphite-monochromated MoKα radiation. The unit cell was
refined by using 24 reflections in the 2θ range 25°–33°. The
diffraction data were collected using the ω/2θ scan mode. A
total of 3554 reflections were collected. The structure solu-
tion and refinement were completed using the PC version of
ter addition of the BH ·THF complex (340 mL, 1 mol/L),
3
the solution was refluxed for 24 h and cooled to room tem-
perature. The solution was cooled in ice water to quench the
remaining borane. Evaporation to dryness in vacuo gave a
cream-coloured solid to which was added 6 mol/L HCl
(
140 mL), and this solution was heated to reflux for 3 h.
Cooling in ice followed by the addition of KOH pellets to
pH ~ 14 and extraction with CHCl , and filtration, gave a
pale-yellow solution of the macrobicycle L . Upon drying
under high vacuum, a yellow oil formed. The product after
3
3
3
T. Rodopoulos. Unpublished observations.
M. Rodopoulos. Unpublished observations.
4
©
2005 NRC Canada

8
98
Can. J. Chem. Vol. 83, 2005
2
+
Table 1. Experimental crystallographic data for [Co(L )Cl](ClO ) .
Fig. 1. ORTEP diagram of the [Co(L )(Cl)] ion with 25% ther-
2
4 2
2
mal ellipsoids.
Formula
CoC H N O Cl
1
4
30
4
9
3
Formula mass
Crystal colour
Crystal system
Space group
563.7
Green
Monoclinic
P2 /n
1
Cell dimensions
a (Å)
9.8483(7)
b (Å)
13.1344(9)
c (Å)
α (°)
β (°)
γ (°)
Dcalcd. (Mg m )
16.6774(12)
90
90.6830(10)
90
–
3
1.707
V (A3)
Z
2194.0(7)
4
F(000)
λ (Å) –
1168
0.7093
1.2
0.10 × 0.15 × 0.40
0.839–0.999
25–33
1
µ (mm )
3
Cryst. dim. (mm )
Transmis. factors
Kinetic studies
2θ Range (°)
Anation reactions of [Co(L )(H O)](ClO ) with chloride
1
2
4 3
No. refls. meas.
3554
ions under pseudo-first-order conditions were followed at
No. unique refls.
No. refls. with Inet > 3.00σ(Inet
3432
1740
2
2
5–40 °C using a Shimadzu UV-160A spectrophotometer at
40 nm. The rates of anation were determined by reaction
with a series of LiCl solutions ranging in concentrations of
50–400 molar excess. The experiments were carried out at
eight different HClO4 concentrations between 0.11 and
)
R (F0)
0.062
0.061
Rw
0
.96 mol/L (I = 1.0 mol/L, LiClO ). Prior to mixing, the so-
4
the NRCVAX system (15–18). Of the 3432 unique reflec-
lutions were pre-equilibrated at the required temperature and
were controlled to within ±0.1 °C. A minimum of three ki-
netic runs was collected for each pair of solutions, prior to
signal averaging with kinetic analysis by nonlinear, least-
squares fitting methods.
tions observed, 1740 reflections with Inet > 3.00σ(I ) were
net
included in the final least-squares refinement for 61 atoms
and 280 parameters. Absorption corrections were made and
all non-hydrogen atoms were refined anisotropically. Min
and max transmission factors were estimated to be 0.839 and
Although metal–ion complexes of ligands L and L were
1
2
0
.999, respectively. All hydrogen atoms were placed in cal-
shown to be kinetically stable, with no detectable decompo-
sition in 1 mol/L media over a >24 h period, the complexes
incorporating L and L , if left overnight in 1 mol/L acidic
culated positions (DC-H = 1.08 Å), and were given isotropic
thermal parameters based upon the atom to which they were
bonded. The refinement converged with a max shift to esti-
mated standard deviation (esd) of 0.009 in the final cycle,
3
4
media, exhibited complete decomposition. Studies on
[Co(L )(ClO )](ClO ) were undertaken at 25 °C at 232 nm
3
4
4 2
and a maximum peak of 0.640 e Å–3 at an R(F ) value of
where the rates of absorbance decrease over a range of acidi-
ties confirmed the acid dependence.
0
0
.062 (R = 0.061).
w
UV–vis (pK measurements)
UV–vis spectra were recorded on a Varian Cary 5
spectrophotometer. Small increments of NaOH solutions
a
Results
Molecular structure
An ORTEP of the structure of the [Co(L )Cl] ion is
2
+
(
varying concentrations) were added with a Gilson Pipetman
2
P200 micropipette to approximately 1 mol/L HClO solu-
shown in Fig. 1. Collection details are provided in Table 1
and selected bond lengths and bond angles are provided in
Table 2. Atomic coordinates are shown in Table S1. The ge-
4
tions of the cobalt(III) complexes in quartz UV–vis cuvettes.
The pH of the solutions was measured after each addition of
base with a Fisher glass electrode connected to a Fisher Sci-
entific Accumet pH meter 915 followed by the immediate
recording of their UV–vis spectra. Measurements were made
over the pH range 0.5–10.
5
ometry at the cobalt metal centre is only slightly distorted
6
from an ideal octahedron expected for a d ion, with the L2
ligand occupying five of the coordination sites and a chlo-
ride ion the sixth site. The macrobicyclic ligand L can be
2
5
Supplementary data for this article are available on the Web site or may be purchased from the Depository of Unpublished Data, Document
Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0S2, Canada. DUD 3689. For more information on obtaining mate-
rial refer to http://cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml. CCDC 256926 contains the crystallographic data for this manuscript. These
data can be obtained, free of charge, via www.ccdc.cam.ac.uk/conts/retrieving.html (Or from the Cambridge Crystallographic Data Centre,
1
2 Union Road, Cambridge CB2 1EZ, UK; fax +44 1223 336033; or deposit@ccdc.cam.ac.uk).
©
2005 NRC Canada

Rodopoulos et al.
Table 2. Selected interatomic distances (Å)
899
a
Fig. 2. Spectra of the individual cobalt(III) ions. A =
a
4+
3+
2+
and bond angles (°) for [Co(L )Cl](ClO ) .
[CoHL (OH )] ; B = [CoL (OH )] ; C = [CoL (OH)] .
2
4 2
1 2 1 2 1
Bond distances (Å)
Co(1)—Cl(1)
Co(1)—O(1)
Co(1)—N(1)
Co(1)—N(2)
Co(1)—N(3)
Co(1)—N(4)
2.219(3)
1.924(6)
1.979(9)
2.007(9)
1.949(9)
1.978(8)
Bond angles (°)
Cl(1)-Co(1)-O(1)
Cl(1)-Co(1)-N(1)
Cl(1)-Co(1)-N(2)
Cl(1)-Co(1)-N(3)
Cl(1)-Co(1)-N(4)
O(1)-Co(1)-N(1)
O(1)-Co(1)-N(2)
O(1)-Co(1)-N(3)
O(1)-Co(1)-N(4)
N(1)-Co(1)-N(2)
N(1)-Co(1)-N(3)
N(1)-Co(1)-N(4)
N(2)-Co(1)-N(3)
N(2)-Co(1)-N(4)
N(3)-Co(1)-N(4)
a
178.53(22)
92.5(3)
92.4(3)
88.4(3)
88.3(3)
86.3(3)
86.9(3)
92.8(3)
92.6(3)
96.5(5)
177.0(5)
87.5(4)
86.4(5)
176.0(5)
89.7(4)
mer being slightly greater as expected owing to increased
steric repulsion at the tertiary nitrogens. However, in con-
trast with the relatively more open species above, in the
present complex, the distances are much more similar, prob-
ably owing to the constraints imposed by the macrobicyclic
ligand.
Estimated standard deviations are given in
parentheses.
Equilibrium studies
2
+
In the case of the ion [Co(L )(ClO )] , spectra were mea-
1
4
sured over the pH range 0.872–9.92. The spectra showed the
presence of more than one ion equilibrium (Figs. S1 and
S2). Measurements made at 10 nm intervals over the range
envisaged as the fusion of a 10-membered cyclodecane ring
with a 14-membered cyclotetradecane ring. The cyclam ring
5
in L adopts a trans-III configuration according to the termi-
210–310 nm were analysed by a second-order global analy-
sis procedure (22) developed to examine the existence of dif-
fering, but spectroscopically similar, species in titrations of
this kind that result in an inability to reliably fit a reasonable
model to the data. The EQUISPEC program (22) permits a
number of titrations with differing initial concentrations of
the equilibrium partners to be analysed simultaneously. The
result is the determination of not only profiles for each re-
agent, but also the pK for each equilibrium observed. By use
2
nology of Bosnich et al. (19). Numerous studies have
illustrated that for complexes of this type, the trans-III con-
2
+
figuration is lowest in energy. In contrast to [Co(L )(Cl)] ,
2
the trans-I configuration is adopted by ligands in the macro-
2+
2+
bicyclic ions [Cu(L )] and [Ni(L )(ClO )] and in the com-
1
1
4
plexes of the analogous thioether ligand (1,5,8,12-tetraaza-
7-thiabicyclo[10.5.2]nonadecane) (L ) (10). Species such
1
5
3
+
as [Co(L )CH CN] also incorporate nine-membered cyclo-
5
3
5
nonane and 14-membered cyclotetradecane rings. In effect,
of the data shown in Fig. S1, the absorption spectra of three
the only structural difference between L and L , (where n =
species have been determined and are presented in Fig. 2.
The concentration profiles and extent of agreement between
observed and calculated spectra are presented in Figs. S3
2
n
1
or 5), other than the type of apical donor atom, is the
placement of the five-membered apical bridge. The trans-III
configuration results in a greater ability of the ether oxygen
5
and S4. It was observed that there is no change in the UV–
+
in the decane ring of L to reach over and better encapsulate
vis spectrum over 24 h under conditions of 1 mol/L [H ].
2
the axial site. The bond to the apical oxygen is quite strong
with Co(1)—O(1) = 1.924(6) Å and is significantly shorter
Although there is maintenance of the Co–ether O linkage in
the solid state, it is considered that in solution the bond dis-
sociates and the presence of the lone pair on this oxygen
provides a template for formation of a hydrogen bond as
shown in Scheme 2. In increasingly basic conditions, there is
loss of the protonated solvent and, at higher pH, the release
of the proton from the trans aqua group with a significant
change in the spectrum. The pK values associated with these
equilibria are 2.2 ± 0.2 and 4.2 ± 0.2, respectively.
3
+ 6
than the Co—S distance (2.206 Å) in [Co(L )(CH CN)] ,
5
3
which displays strain owing to the isomeric ligand structure.
Furthermore, the Co(1)—Cl(1) bond length is 2.219(3) Å,
which is similar to the Co—Cl distance (2.246 Å) observed
2
+
in the crystal structure of [Co(daptacn)(Cl)] (20), where
daptacn is the pentaaza ligand 1,4-di(3-aminopropyl)-1,4,7-
triazacyclononane. In the latter complex ion, and in the
corresponding system, 1,4,7-tri(3-aminopropyl)-1,4,7-triaza-
cyclononane (21), there is a slight distinction between the
Co—N(tert) and Co—N(sec) bond distances, with the for-
An alternative scheme could be envisaged in which there
is retention of the Co—O (ether) bond and the site of the
+
solvated H interaction would be the remaining lone pair on
6
K.R. Coulter. Unpublished observations.
©
2005 NRC Canada

9
00
Can. J. Chem. Vol. 83, 2005
Scheme 2.
H
H
2+
4+
+
H
H
3+
H
H
H
O
O
H
O
Co
O
O
N
O
N
O
N
+
-
H
N
N
N
N
Co
N
H
H
H
H
Co
O
N
+
N
H
N
N
H
H
O
H
H
H
H
the oxygen. It is not possible to exclude the latter, but the
bond formation with the O atom to the Co(III) may be weak
in this case.
Scheme 3.
k1
CoHL
+
Cl
CoHLCl
The value for the first pK is higher than that (0.65) for the
corresponding strongly oxidizing Ni(III) species (14). How-
ever, the second pK, not observed in the Ni(III) system ow-
ing to decomposition, is similar to that for other trans aquo-
tetramminecobalt(III) species (23, 24).
k-
1
Ka
Kb
Similar results on the cobalt(III) complex of the
k2
k-
dimethylated ligand L showed two spectroscopic changes,
3
CoL + H + Cl
CoLCl + H
the first over the range pH 0.58–0.67, and a second more
clearly defined change with pK ~ 1. For the corresponding
2
complex of L , only a single pK ~1.05 is defined. In both
4
cases, however, there is acid decomposition of the com-
plexes owing to the protonation at the tertiary amines (vide
infra).
+
+
Fig. 3. Plot of slope (K + [H ]) vs. [H ] (eq. [1]).
a
Kinetic measurements
Anation reactions
Substitution reaction kinetics were investigated at 40 °C
with chloride concentrations of approximately 50–400 times
3
+
in excess halide. The rate data for the [Co(L )(H O)] reac-
1
2
5
tion, obtained at 280 nm, are tabulated in Table S2. In gen-
eral, the scatter of individual values for kobs was in the order
of ±5%. The observed first-order rate constants (kobs) vary
linearly with chloride concentration. However, the intercepts
+
and slope of each curve vary roughly inversely with [H ].
No evidence was observed for a second substitution step in
the anation.
A mechanistic scheme that accounts for the observed acid
dependence is described in Scheme 3, where (neglecting all
charges) CoHL and CoL represent the complex ions
4
+
3+
[
CoL(OH )(OH )] and [CoL(OH ) ] , and CoHLCl and
3 2 2 2
CoLCl, the corresponding chloro complexes.
The observed rate constant derived from this equilibrium
set may be represented as:
(
1
k–1 not statistically discernible), with intercept k K =
–2 b
–5 –1
0.4 ± 1.3 × 10 s . Assuming K ~ K , then, at 40 °C, k =
a b –2
–2 –1
1
.6 × 10
s
and K = 230 ± 60 mol/L. Comparison with
2
–1
–1
the analogous Ni(III) systems at 25 °C k = 142 (mol/L)
s
1
+
–
+
[
1]
kobs = (k [H ] + k K )[Cl ]/(K + [H ])
–1 –1
1
2
a
a
and k = 1400 ± 300 (mol/L)
pected, a significantly slower process for the d ion.
s
(14) indicates, as ex-
2
+
+
6
+
(k [H ] + k K )/(K + [H ])
–
1
–2
b
b
As can be seen from eq. [1], approximate inverse proton
dependences are predicted for both the intercepts and slopes
Decomposition of [Co(L )(ClO )](ClO ) in acidic media
3
4
4 2
Although complexes of L and L appeared indefinitely
1
2
–
+
of plots of k vs. [Cl ] (eq. [1]). A plot of slope(K + [H ])
stable, those incorporating the dimethylated ligands decom-
posed much more readily. Kinetic studies were made at
25 °C (I = 1.0 mol/L) over a range of acid concentrations us-
ing absorbance changes at 223 nm. The results are presented
obs
a
+
vs. [H ] is shown in Fig. 3 and, using the K value (6.3 ×
a
–
3
1
0
mol/L) obtained previously, the rate constants derived
–
1
–1
are k1 = 0.053 ± 0.002 (mol/L)
s
and k2 = 3.8 ±
–
1
–1
5
1
[
.0 (mol/L) s . However, the plots of intercepts (K +
in Table S3 and are consistent with the reaction scheme
b
+
+
H ]) vs. [H ] show a slope of zero or are slightly negative
(charges omitted):
©
2005 NRC Canada

Rodopoulos et al.
901
+
Fig. 4. Rate constants for the decomposition of
H + CoL
CoHL
3
+
5
[
Co(III)(L ))OH )] in acidic media. Curve fit: 10 k = 1.1 ±
3
2
1
K
a
–
1
5
–1 –1
0
.11 s ; 10 k = 9.9 ± 0.2 M s .
2
H+
k
k
2
1
products
products
for which the observed rate constant may be written in the
+
+ 2
form: kobs([H ] + K ) = k K + k [H ] . The data are pre-
a
1
a
2
sented in Fig. 4 and the individual rate constants (derived us-
ing the value of K = 0.7 derived spectroscopically) are
a
consistent with the reaction not only of the neutral species,
but also the protonated ion. The increased rate is attributed
to protonation at the tertiary nitrogen atoms, leading to a re-
duction in the macrocyclic coordination and eventual release
of the metal ion. A similar considerable increase in lability
over the unsubstituted complex has been observed previ-
ously for the Ni(II) tetramethyl-cyclam ion (25).
Confirmation of the locus of reaction is provided by a
comparison of the NMR spectra in neutral and acidic media.
1
3
The C NMR spectrum of [Co(L )(ClO )](ClO ) in D O–
3
4
4 2
2
CD CN is highly symmetrical with only eight distinct lines
3
5
exhibited (Table S4). Thus, the complex in solution pos-
sesses a time-averaged C axis of symmetry. Addition of a
2
sion of the metal centre was observed in the positive-ion
few drops of perchloric acid to the NMR tube containing
such a solution (pH ~ 0) resulted in the initial precipitation
of the purple complex. However, upon allowing the solution
to stand overnight, dissolution occurred with loss of colour.
The resultant spectrum reflects a high-symmetry species
where significant chemical shifts are seen at atoms C5, C6,
and C3 when compared to the Co(III) spectrum (at pH 7),
and additional shifts compared to the free ligand noted at C9
and C1 were consistent with demetallation and protonation
of the ligand. Bencini et al. (26) observed a similar
downfield shift by the resonances of the carbon atoms adja-
cent to methyl centres upon protonation of the ligand 4,10-
dimethyl-1,4,7,10-tetraazabicyclo[5.5.2]tetradecane.
mode spectra of the cobalt(III) complexes with ligands L or
1
L .
2
5
In the negative-ion spectra (Table S5), there is strong evi-
dence for the molecular ion [M] . This is the result of the re-
–
duction to Co(II) and maintenance of the outer sphere
perchlorates. Again, the geometric isomers [Co(L )(ClO )]-
1
4
(
ClO4)2 and [Co(L )(ClO )](ClO ) gave almost identical
2 4 4 2
–
mass spectra. Other than [M] , the only other significant
fragment observed was at 526 due to the loss of neutral
HClO . As seen in the positive-ion spectrum for [Co(L )-
4
3
(
ClO )](ClO ) , the negative-ion spectrum also exhibited
4 4 2
fragments with removal of the metal ion. Ions at 597 and
–
–
4
97 correspond to [(H L )(ClO ) ] and [(HL )(ClO ) ] , re-
2 3 4 3 3 4 2
spectively. However, there is evidence for cobalt(I) as the
–
Mass spectroscopy
Characterization by LSI-MS indentified both positively
and negatively charged ions, providing informative similarities
fragment at 355 corresponds to [Co(H L )] . The differ-
–
2 3
ences in mass spectra reflect the kinetic stabilities and reac-
tivities of the methylated and non-methylated complexes.
5
and differences. In the positive-ion spectra of [Co(L )-
1
From Table S5, it can be seen that [Co(L )(ClO )](ClO )
2
4
4 2
(
(
ClO )](ClO ) , [Co(L )(ClO )](ClO ) , and [Co(L )(ClO )]-
4 4 2 2 4 4 2 3 4
and the chloro-related complex[Co(L )(Cl)](ClO ) gave
2 4 2
5
ClO4) (see Table S5) , there was no evidence for the
2
very similar negative-ion mass spectra when the loss of ei-
+
+
–
pseudo-molecular ion [M + H ] where M = [Co(L )-
n
ther HClO or HCl from [M] is taken into account.
4
(
(
ClO )](ClO ) . The geometric isomers [Co(L )(ClO )]-
4 4 2 n 4
In summary, the Co(III) complexes with ligands L and L2
1
ClO ) (n = 1 or 2) gave almost identical positive-ion mass
4
2
are kinetically more stable than their Ni(III) analogues, but
exhibit similar proton-related equilibria that are exhibited in
the kinetic analysis. Whilst the species with secondary NH
groups maintain macrocyclic coordination in acidic media,
those incorporating all four tertiary nitrogens are subject
to attack by protons leading to an unexpectedly ready
demetallation of the complexes, a feature also observed in
the mass spectroscopic fragmentation of the complexes.
spectra with the highest ion observed at 428 ([Co(L )-
n
+
(
ClO )] ) owing to the inital loss of two outer-sphere per-
4
chlorate ions with the next highest ion at 328 being due to
the subsequent loss of neutral HClO ([Co (H L )] ). There
is no evidence for any Co(III) fragments in either decay. In
contrast, for the molecule [Co(L )(ClO )](ClO ) , while a
II
+
4
–1 n
3
4
4 2
similar initial loss of two perchlorate ions to give a
II
+
[
Co (L )(ClO )] fragment at 456 is observed, there is also a
3 4
+
fragment at 355, [Co(H L )] , in which the cobalt centre re-
mains in the +3 oxidation state. Thus, in the same spectrum,
there is evidence for cobalt in both the +2 and +3 oxidation
states. The only other significant fragment observed at 299
is assignable to the protonated free ligand HL . No extru-
–
2 3
Acknowledgements
We thank the Natural Sciences and Engineering Research
Council of Canada (NSERC) and the University of Victoria
for support. TR acknowledges the tenure of an NSERC In-
+
3
©
2005 NRC Canada

9
02
Can. J. Chem. Vol. 83, 2005
ternational Post-doctoral Fellowship, and KI, sabbatical as-
sistance from Waseda University.
13. M. Rodopoulos, T. Rodopoulos, J.N. Bridson, L.I. Elving, S.J.
Rettig, and A. McAuley. Inorg. Chem. 40, 2737 (2001).
1
1
4. A.M. Ingham, C. Xu, T.W. Whitcombe, C. Xu, J.N. Bridson,
and A. McAuley. Can. J. Chem. 80, 155 (2002).
5. E.J. Gabe, Y. LePage, J.-P. Charland, F.L. Lee, and P.S. White.
J. Appl. Chem. 22, 384 (1989).
References
1
. G.A. Melson (Editor). Coordination chemistry of macrocyclic
compounds. Plenum Press, New York. 1979.
16. Scattering factors from Vol. 4: J.A. Ibers and W.C. Hamilton
(Editors). International tables for X-ray crystallography.
Vol. IV. Kynoch Press, Birmingham, UK. 1974.
17. C.K. Johnson. ORTEP Technical Report ORNL-5138, Oak
Ridge, Tenn. 1976.
18. G.M. Sheldrick. SHELX [computer program]. University of
Göttingen, Göttingen, Germany. 1986 and 1990.
19. B. Bosnich, C.K. Poon, and M.L. Tobe. Inorg. Chem. 4, 1102
(1965).
2
. I. Bertini, I.H.B. Gray, S.J. Lippard, and J.S. Valentine (Edi-
tors). Bioinorganic chemistry. University Science Books, Mill
Valley, California. 1994.
3
4
5
. R.W. Hay. Bio-inorganic chemistry. Ellis Horwood Limited,
Chichester, UK. 1984.
. L. Lindoy. The chemistry of macrocyclic ligand complexes.
University Press, Cambridge, UK. 1989.
. A.I. Ingham, M. Rodopoulos, K.R. Coulter, T. Rodopoulos, S.
Subramanian, and A. McAuley. Coord. Chem. Rev. 233–234,
20. D.G. Fortier and A. McAuley. J. Chem. Soc. Dalton Trans.
101 (1991).
2
55 (2002).
6
. C. Bazzicalupi, A. Bencini, A. Bianchi, M. Cecchi, B.
Escuder, V. Fusi, E. Garcia-España, C. Giorgi, S.V. Luis, C.
Maccagni, V. Marcellino, P. Paoletti, and B. Valtancoli. J. Am.
Chem. Soc. 121, 6807 (1999).
. D.P. Arnold and J. Blok. Coord. Chem. Rev. 248, 299 (2004).
. V. Pavlishchuk, F. Birkelbach, T. Weyhermüller, K. Wieghardt,
and P. Chaudhuri. Inorg. Chem. 41, 4405 (2002).
21. G.W. Bushnell, D.G. Fortier, and A. McAuley. Inorg. Chem.
27, 2626 (1988).
22. R.M. Dyson, S. Kaderli, G.A. Lawrence, M. Maeder, and A.D.
Zuberbühler. Anal. Chim. Acta, 353, 381 (1997).
23. P.M. Brown and G.M. Harris. Inorg. Chem. 7, 1872 (1968).
24. J. Bjerrum and S.E. Rasmussen. Acta Chem. Scand. 6, 1265
(1952).
7
8
9
. D.G. Fortier and A. McAuley. J. Am. Chem. Soc. 112, 2640
25. F. Wagner and E.K. Barefield. Inorg. Chem. 12, 2435 (1973).
26. A. Bencini, A. Bianchi, A. Borselli, M. Ciampolini, P.
Dapporto, E. Garcia-España, M. Micheloni, P. Paoli, J.A.
Ramirez, and B. Valtancoli. Inorg. Chem. 28, 4279 (1989).
(
1990).
1
1
0. D.G. Fortier and A. McAuley. Inorg. Chem. 28, 655 (1989).
1. K.A. Beveridge, A. McAuley, and C. Xu. Inorg. Chem. 30,
2
074 (1991).
1
2. A. McAuley, D.G. Fortier, D.H. Macartney, T.W. Whitcombe,
and C. Xu. J. Chem. Soc. Dalton Trans. 2017 (1994).
©
2005 NRC Canada