Heterocyclic donor influences on the binding and activation of CO, NO, and O2 by copper complexes of hybrid triazacyclononane-pyridyl ligands

Article abstract of DOI:10.1016/S0020-1693(99)00292-3

Copper(I) complexes of 1,4-diisopropyl-7-R'-1,4,7-triazacyclononane (R'=2-pyridylmethyl, L(Py); 6-methyl-2-pyridylmethyl, L(6MePy); 5-methyl-2-pyridylmethyl, L(5MePy); 6-phenyl-2-pyridylmethyl, L(6PhPy); 2-quinolylmethyl, L(Quin)) were prepared and characterized by CHN analysis, NMR and FTIR spectroscopy, cyclic voltammetry, and mass spectrometry. An X-ray crystal structure of [L(6PhPy)Cu]SbF₆ was determined and compared to that previously reported for [L(Py)Cu]O₃SCF₃; similar distorted trigonal bipyramidal geometries are adopted with the ligands coordinated in η⁴ fashion. The complexes [LCu]⁺ (L = L(Py) or L(Quin)) form adducts with CO(g) in which the heterocyclic appendage is displaced. With NO(g), [L(Py)Cu]O₃SCF₃ reacts in a disproportionation process to yield N₂O and [L(Py)Cu(ONO)]O₃SCF₃, which was structurally defined by X-ray crystallography. Upon reaction with O₂(g) at -75°C the Cu (I) complexes of L(Py) and L(5MePy) yield trans-1,2-peroxo species [LCuOOCuL]²⁺ as determined by UV-Vis and resonance Raman spectroscopy. In contrast, spectroscopy indicates that low temperature oxygenation of [L(6PhPy)Cu]SbF₆ yields a bis(μ-oxo)dicopper core, postulated to be capped by η³-L(6PhPy) (pyridyl appendage not coordinated). Decomposition of the trans-1,2-peroxo compounds results in hydroxylation of the ligand at the benzylic position of the heterocyclic appendage, but the bis(μ-oxo) complexes decay to give products resulting from N-dealkylation of the heterocycle arm. The different fates of the Cu(I) complexes of L(Py) and L(5MePy) versus that of L(6PhPy) upon oxygenation may be traced to the coordination of the heterocycle; in the former cases, the pyridyl unit remains coordinated, favoring trans-1,2-peroxo generation, whereas pyridyl dissociation facilitated (6PhPy) by the sterically bulky 6-phenyl group on L yields a [η³-L(6PhPy)Cu]⁺ fragment amenable to bis(μ-oxo) core formation. The steric properties of the heterocyclic components of the ligands used in this study thus are important determinants of the reactivity of their Cu(I) complexes with small molecules. (C) 2000 Elsevier Science S.A.

Full text of DOI:10.1016/S0020-1693(99)00292-3

www.elsevier.nl/locate/ica
Inorganica Chimica Acta 297 (2000) 115–128
Heterocyclic donor influences on the binding and activation of
CO, NO, and O2 by copper complexes of hybrid
triazacyclononane–pyridyl ligands
Lisa M. Berreau, Jason A. Halfen, Victor G. Young Jr., William B. Tolman *
Department of Chemistry and the Center for Metals in Biocatalysis, 207 Pleasant St. S.E., Minneapolis, MN 55455, USA
Received 11 March 1999; accepted 11 June 1999
Dedicated to Professor Stephen J. Lippard on the occasion of his 60th birthday.
Abstract
Copper(I) complexes of 1,4-diisopropyl-7-R¦-1,4,7-triazacyclononane (R¦=2-pyridylmethyl, LPy; 6-methyl-2-pyridylmethyl,
L6MePy; 5-methyl-2-pyridylmethyl, L5MePy; 6-phenyl-2-pyridylmethyl, L6PhPy; 2-quinolylmethyl, LQuin) were prepared and character-
ized by CHN analysis, NMR and FTIR spectroscopy, cyclic voltammetry, and mass spectrometry. An X-ray crystal structure of
[L6PhPyCu]SbF6 was determined and compared to that previously reported for [LPyCu]O3SCF3; similar distorted trigonal
bipyramidal geometries are adopted with the ligands coordinated in h4 fashion. The complexes [LCu]+ (L=LPy or LQuin) form
adducts with CO(g) in which the heterocyclic appendage is displaced. With NO(g), [LPyCu]O3SCF3 reacts in a disproportionation
process to yield N2O and [LPyCu(ONO)]O3SCF3, which was structurally defined by X-ray crystallography. Upon reaction with
O2(g) at −75°C the Cu(I) complexes of LPy and L5MePy yield trans-1,2-peroxo species [LCuOOCuL]2+ as determined by UV–Vis
and resonance Raman spectroscopy. In contrast, spectroscopy indicates that low temperature oxygenation of [L6PhPyCu]SbF6
yields a bis(m-oxo)dicopper core, postulated to be capped by h3-L6PhPy (pyridyl appendage not coordinated). Decomposition of the
trans-1,2-peroxo compounds results in hydroxylation of the ligand at the benzylic position of the heterocyclic appendage, but the
bis(m-oxo) complexes decay to give products resulting from N-dealkylation of the heterocycle arm. The different fates of the Cu(I)
complexes of LPy and L5MePy versus that of L6PhPy upon oxygenation may be traced to the coordination of the heterocycle; in the
former cases, the pyridyl unit remains coordinated, favoring trans-1,2-peroxo generation, whereas pyridyl dissociation facilitated
by the sterically bulky 6-phenyl group on L6PhPy yields a [h3-L6PhPyCu]+ fragment amenable to bis(m-oxo) core formation. The
steric properties of the heterocyclic components of the ligands used in this study thus are important determinants of the reactivity
of their Cu(I) complexes with small molecules. © 2000 Elsevier Science S.A. All rights reserved.
Keywords: Dioxygen activation; Vitric oxide; Copper complexes; Macrocyclic complexes
on modeling the structural and functional aspects of the
interactions of CO, NO, and O2 with transition metal
centers in biology is to use multidentate, chelating
N-donor ligands to mimic the common ligation of such
centers by multiple histidine imidazoles [4]. Implemen-
tation of this strategy to model copper active site
chemistry recently has provided significant insights into
the structures of Cu–CO, -O2, and -NO adducts and
pathways of NO and O2 activation [5]. These studies
have revealed that the supporting multidentate N-
donor ligand plays a significant role in controlling the
small molecule binding and activation chemistry, as
1. Introduction
An important objective of bioinorganic chemistry
research is to understand how small molecules such as
carbon monoxide (CO), nitric oxide (NO), and dioxy-
gen (O2) bind to and/or are activated by metalloprotein
active sites during catalytic processes [1], as signaling
agents [2], or as probes of the active site’s structure [3].
A successful strategy in synthetic investigations focused
* Corresponding author. Tel.: +1-612-625 4061; fax: +1-612-624
7029.
E-mail address: tolman@chem.umn.edu (W.B. Tolman)
0020-1693/00/$ - see front matter © 2000 Elsevier Science S.A. All rights reserved.
PII: S0020-1693(99)00292-3

116
L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
Fig. 1. Summary of reactions of O2 and NO with copper(I) complexes supported by the indicated ligands.
isomeric bis(m-oxo)dicopper compounds as well, depend-
ing on the substituent R and other factors [1b]. Other
important reactivity differences are exhibited in reactions
with NO (Fig. 1). Thus, [(TMPA)Cu(CH3CN)]+ yields
N2O and a (m-oxo)dicopper(II) complex [17], TpR,R%Cu-
(CH3CN) binds NO to give isolable adducts TpR,R%Cu-
(NO) that react further to yield N2O and copper(II)–ni-
trite species in a disproportionation process [18], and
[(iPr3TACN)Cu(CH3CN)]+ catalytically oxidizes alco-
hol solvent with concomitant generation of N2O [19].
Factors that influence the course of the chemistry of
copper(I) complexes of these TMPA, TpR,R%, and TACN
(as well as other) ligands include the number of N-donor
atoms (3 for TpR,R% and TACN, 4 for TMPA), the nature
of the N-donors (pyrazolyl for TpR,R%, pyridyl and amine
for TMPA, amine only for TACN), charge of the ligand
(monoanionic for TpR,R%, neutral for TMPA and TACN),
and the size and electronic effects of ligand substituents
that may or may not be oriented similarly with respect to
the substrate binding site(s) on the copper ion. A goal of
current research is to understand how such various,
disparate, and often conflicting influences ‘tune’ small
molecule binding and activation [20]. As part of our
efforts aimed at addressing these issues, we have begun to
investigate the structures and reactivities of copper
complexes of ‘hybrid’ ligands that incorporate features of
two or more of the aforementioned ligand classes. We
report herein the results of studies of copper(I) complexes
supported by one type of such a hybrid ligand comprising
a single heterocyclic donor appended to 1,4-diisopropyl-
1,4,7-triazacyclononane3 (Scheme 1). Our inves-
illustrated by the complementary reactivities of copper(I)
complexes of the tris(pyridylmethyl)amine (TMPA) [6],
tris(pyrazol-1-yl)hydroborate (TpR,R%) [7], and 1,4,7-tri-
azacyclononane (TACN) ligand classes [8], among
others1 (Fig. 1).
For example, reaction of O2 with [(TMPA)-
Cu(CH3CN)]+ at low temperature generates a (trans-1,2-
peroxo)dicopper(II) complex [10] (Fig. 1). A mono-
meric (superoxo)copper intermediate is implicated in this
reaction [11], and by using the derivative ligand
bis(quinolylmethyl)(pyridylmethyl)amine (BQPA) this
Cu–O2 intermediate may be stabilized sufficiently to be
observed via spectroscopy [12]2. In contrast, oxygenation
i
of copper(I) complexes of TpRR% (R=R%=Me, Pr, or
Ph) yields a (m-h2:h2-peroxo)dicopper(II) species [14]
analogous to the oxygenated active sites of hemocyanin,
tyrosinase, and catechol oxidase [1d,15]. Involvement of
a monocopper superoxo intermediate in this oxygenation
is implicated by the isolation and X-ray structural
i
determination of an analog supported by TptBu, Pr that
contains superoxide coordinated in h2 (side-on) fashion
[16]. The complexes [(R3TACN)Cu(CH3CN)]+ also af-
ford (m-h2:h2-peroxo)dicopper species, but can yield
1 Selected examples of related ligand types used in copper-small
molecule chemistry include tris(imidazolyl)phosphine [9a–c],
bis(pyridylmethyl)amine [9d–f], tris(pyrazolylethyl)amine [9g] and
simple diamines [9h–j].
2 An X-ray crystal structure of a monocopper superoxo complex of
a derivative of TMPA with amide arms was reported [13a], but was
shown to be incorrect [13b]. A paper reiterating the previous claims
that this species exists in solution has appeared recently [13c], but in
our view the interpretation of the newly provided supporting evidence
remains ambiguous (cf. NMR data consistent with a small amount of
uncomplexed ligand rather than the putative superoxo compound).
3 Similarly functionalized TACN (and other macrocyclic) ligands
are surveyed in Ref. [21].

L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
117
tigations of the structures, spectroscopic features, and
reactivities with O2, NO, and CO of copper(I) com-
plexes of these ligands have revealed interesting similar-
ities and differences with their TACN- and TMPA-
coordinated counterparts, thus shedding new light
on the general issue of ligand effects on copper
ion-small molecule interactions in chemistry and biol-
ogy. Portions of this work have been communicated
[22,23].
tant mixture was heated at reflux under N2 for 14 h,
after which time it was cooled to room temperature and
then poured into 1 M NaOH (20 ml). The heteroge-
neous mixture was extracted with CHCl3 (3×100 ml)
and the organic extracts dried (Na2SO4) and filtered.
Removal of the solvent under reduced pressure yielded
a yellow–brown oil which was purified by Kugelrohr
distillation (108–116°C, 0.1 torr) to afford the pure
product as a clear, colorless oil (0.304 g, 60%). 1H
NMR (300 MHz, CDCl3): l 7.51 (t, J=7.5 Hz, 1H),
7.41 (d, J=7.5 Hz, 1H), 6.95 (d, J=7.5 Hz, 1H), 3.80
(s, 2H), 2.93 (m, 4H), 2.90 (heptet, J=6.6 Hz, 2H),
2.63 (m, 4H), 2.58 (s, 4H), 2.51 (s, 3H), 0.94 (d, J=6.6
Hz, 12H) ppm; 13C{1H} NMR (75 MHz, CDCl3) d
160.7, 157.2, 136.4, 120.9, 119.7, 64.0, 55.5, 54.8, 52.7,
52.5. 24.5, 18.4 ppm; FTIR (neat, cm−1) 2962, 2823,
2899, 2871, 2802, 1677, 1591, 1580, 1456, 1380, 1359,
1306, 1261, 1169, 1115, 1092, 1034, 999, 973, 882, 851,
788, 769, 709, 568; high resolution electron impact mass
spectrum (HREIMS) calc. for C19H34N4: 318.2783;
found: 318.2784.
2. Experimental
2.1. General procedures
All reagents and solvents were obtained from com-
mercial sources and used as received unless noted other-
wise. Solvents were dried according to standard
procedures and distilled under N2 immediately prior to
use. Air-sensitive reactions were performed either in a
Vacuum Atmospheres inert-atmosphere glovebox under
a N2 atmosphere or by using standard Schlenk and
vacuum-line techniques. Methods used for physical
characterization were as described previously [24]. Pub-
lished procedures were used for the preparation of
1,4-diisopropyl-1,4,7-triazacyclononane (LH) [23], 1,4-
diisopropyl-7-(2-pyridylmethyl)-1,4,7-triazacyclononane
(LPy) [23], 2-chloromethyl-6-methylpyridine hydrochlo-
ride [25], 2-chloromethyl-5-methylpyridine hydrochlo-
ride [26], 2-chloromethyl-6-phenylpyridine [27], and
[Cu(CH3CN)4]SbF6 [28].
2.3. 1,4-Diisopropyl-7-(5-methyl-2-pyridylmethyl)-
1,4,7-triazacyclononane (L5MePy
)
To a solution of LH (0.386 g, 1.81 mmol) in dry
CH3CN (20 ml) were added 5-methyl-2-picolyl chloride
hydrochloride (0.333 g, 1.87 mmol), Na2CO3 90.400 g,
3.77 mmol), and n-Bu4NBr (ꢀ3 mg). The resulting
solution was heated at reflux under N2 for 24 h. Fol-
lowing a work-up procedure identical to that outlined
for L6MePy, a yellow–brown oil was obtained which was
purified by Kugelrohr distillation (102–110°C, 0.1 torr)
to afford the product as a clear, colorless oil (0.347 g,
60%). 1H NMR (300 MHz, CDCl3): l 8.33 (s, 1H), 7.45
(m, 2H), 3.80 (s, 2H), 2.97–2.93 (m, 4H), 2.87 (heptet,
J=6.6 Hz, 2H), 2.66–2.62 (m, 4H), 2.57 (s, 4H), 2.29
(s, 3H), 0.94 (d, J=6.6 Hz, 12H) ppm; 13C{1H} NMR
(75 MHz, CDCl3): l 158.4, 149.1, 136.8, 130.8, 122.5,
63.4, 55.2, 54.8, 52.7, 52.5, 18.4, 18.1 ppm; FTIR (neat,
cm−1) 2966, 2933, 2903, 2821, 1674, 1599, 1569, 1487,
1459, 1380, 1359, 1312, 1263, 1170, 1150, 1114, 1093,
1032, 996, 973, 828, 735, 646, 581, 572, 561, 510, 497;
HREIMS calc. for C19H34N4: 318.2783; found:
318.2784.
2.2. 1,4-Diisopropyl-7-(6-methyl-2-pyridylmethyl)-
1,4,7-triazacyclononane (L6MePy
)
To a solution of LH (0.340 g, 1.60 mmol) in dry
CH3CN (20 ml) were added 2-chloromethyl-6-
methylpyridine hydrochloride (0.284 g, 1.60 mmol),
Na2CO3 (0.393 g), and n-Bu4NBr (ꢀ5 mg). The resul-
2.4. 1,4-Diisopropyl-7-(6-phenyl-2-pyridylmethyl)-
1,4,7-triazacyclononane (L6PhPy
)
To a solution of LH (0.331 g, 1.55 mmol) in dry
CH3CN (15 ml) were added 2-(chloromethyl)-6-
phenylpyridine hydrochloride (0.372 g, 1.55 mmol),
Na2CO3 (0.400 g, 3.77 mmol), and n-Bu4NBr (ꢀ2 mg).
The resulting solution was heated at reflux under N2 for
6h. Following a basic work-up procedure identical to
Scheme 1.

118
L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
that outlined for L6MePy, the crude ligand was isolated
as a yellow–orange oil. This oil was dissolved in
methanol (3 ml) and the resulting pale yellow solution
was added to a methanol/water (4:1 v/v, 5 ml) solution
containing NaClO4 (ꢀ0.4 g). To this solution was
added excess water (5 ml) which led to the deposition of
a viscous tan oil. This oil was then redissolved in
CHCl3 (ꢀ10 ml) and the resulting pale yellow organic
solution was washed with 1 M NaOH (ꢀ10 ml). The
organic phase was removed and the aqueous phase was
extracted a second time with CHCl3 (25 ml). The
combined organic fractions were dried (Na2SO4) and
the solvent was removed to yield the pure product as a
starting material had dissolved (ꢀ10 min). The result-
ing bright yellow solution was then added to Et2O (15
ml) and was cooled to −20°C overnight, which led to
the deposition of the product as a yellow microcrys-
talline solid (408 mg, 68%). 1H NMR (300 MHz,
d6-acetone) l 7.86 (t, J=7.8 Hz, 1H), 7.45 (d, J=7.8
Hz, 1H), 7.31 (d, J=7.8 Hz, 1H), 4.27 (s, 2H), 3.30–
2.78 (m, 14H), 2.78 (s, 3H), 1.28 (d, J=6.6 Hz, 6H),
1.21 (d, J=6.6 Hz, 6H) ppm; 13C{1H} NMR (75 MHz,
d6-acetone) l 158.5, 156.3, 137.6, 123.5, 121.1, 60.3,
57.9, 52.2, 51.5, 50.6, 27.2, 20.0 ppm; FTIR (KBr,
cm−1) 656 (SbF6); UV–Vis (THF) [umax, nm (m, M−1
cm−1)] 364 (3400); FAB-MS (MNBA) m/e (relative
intensity) 381.2 ([M−SbF6]+, 100%). Anal. Calc. for
C19H34N4CuSbF6: C, 37.01; H, 5.56; N, 9.09. Found: C,
36.19; H, 5.38; N, 8.92%.
1
thick, yellow oil (0.243 g, 41%). H NMR (300 MHz,
CDCl3) l 7.98 (d, J=7.5 Hz, 2H), 7.71 (t, J=7.8 Hz,
1H), 7.55 (d, J=8.1 Hz, 2H), 7.46 (t, J=6.9 Hz, 2H),
7.44 (t, J=7.8 Hz, 1H), 3.94 (s, 2H), 2.94–3.02 (m,
4H), 2.91 (heptet, J=6.6 Hz, 2H), 2.64–2.70 (m, 4H),
2.63 (s, 4H), 0.97 (d, J=6.6 Hz, 12H) ppm; 13C{1H}
NMR (75 MHz, CDCl3) l 161.3, 156.4, 139.8, 136.9,
128.7, 126.9, 121.3, 118.4, 64.1, 55.7, 54.8, 52.8, 52.5,
18.4 (14 peaks observed out of 15 expected); FTIR
(neat, cm−1) 3066, 3034, 2963, 2935, 2895, 2857, 2817,
1575, 1454, 1380, 1358, 1309, 1259, 1165, 1112, 1091,
1033, 997, 917, 814, 758, 690, 669, 621, 573, 529;
HREIMS calc. for C24H36N4: 380.2940; found:
380.2941.
2.7. [L5MePyCu]SbF6
To a solution of L5MePy (308 mg, 0.97 mmol) in
CH3CN/THF (2:1, 2 ml total volume) was added solid
[Cu(CH3CN)4]SbF6 (430 mg, 0.93 mmol). The mixture
was stirred at room temperature for ꢀ10 min. The
resulting dull orange solution was added to Et2O (15
ml) and the solution was allowed to stand at ambient
temperature for ꢀ30 min during which time a thick,
dark oil deposited. The remaining bright yellow solu-
tion was decanted away from the dark oil and was
cooled to −20°C for 4 h, which led to the deposition
of the product as a yellow microcrystalline solid (180
2.5. 1,4-Diisopropyl-7-(2-quinolylmethyl)-
1,4,7-triazacyclononane (LQuin
)
1
mg, 29%). H NMR (300 MHz, d6-acetone, 20°C) l
To a solution of LH (0.326 g, 1.53 mmol) in CH3CN
(20 ml) were added 2-chloromethylquinoline hydrochlo-
ride (0.326 g, 1.52 mmol), Na2CO3 (0.400 g, 3.77
mmol), and tetrabutylammonium bromide (ꢀ5 mg).
The resulting solution was heated at reflux for 12 h.
Following a basic work-up procedure identical to that
outlined for L6MePy, the ligand was isolated as a red oil
(0.536 g, 98%) of sufficient purity to be used without
7.77 (d, J=7.8 Hz, 1H), 7.6 (br s, 2H), 3.2 (vbr s, 16H),
2.41 (s, 3H), 1.37 (d, J=5.4 Hz, 6H), 1.24 (d, J=5.3
Hz, 6H) ppm; FTIR (KBr, cm−1) 2977, 2922, 1489,
1462, 1387, 1369, 1144, 1124, 1071, 952, 824, 656
(SbF6); UV–Vis (THF) [umax, nm (m, M−1 cm−1)] 364
(4900); FAB-MS m/e (relative intensity) 381.2 ([M−
SbF6]+, 100%). Anal. Calc. for C19H34N4CuSbF6: C,
37.01; H, 5.56; N, 9.09. Found: C, 36.95; H, 5.53; N,
9.08%.
1
further manipulation. H NMR (300 MHz, CDCl3) l
8.08 (d, J=8.4 Hz, 1H), 8.03 (d, J=8.4 Hz, 1H), 7.78
(d, J=7.8 Hz, 1H), 7.76 (d, J=7.8 Hz, 1H), 7.65 (dt,
J=1.2, 9.3 Hz, 1H), 7.47 (dt, J=0.6, 7.2 Hz, 1H), 3.98
(s, 2H), 2.96 (m, 4H), 2.88 (heptet, J=6.6 Hz, 2H),
2.64 (m, 4H), 2.59 (s, 4H), 0.94 (d, J=6.6 Hz, 12H)
ppm; 13C{1H} NMR (75 MHz, CDCl3) l 162.2, 147.5,
136.0, 129.1, 128.9, 127.5, 127.7, 125.8, 121.4, 64.7,
55.6, 54.8, 52.7, 52.5, 18.4 ppm; HREIMS calc. for
C22H34N4: 354.2783; found: 354.2783.
2.8. [L6PhPyCu]SbF6
To a solution of L6PhPy (373 mg, 0.981 mmol) in
CH3CN/THF (2:1, 2 ml total volume) was added solid
[Cu(CH3CN)4]SbF6 (380 mg, 0.820 mmol). The mixture
was stirred at room temperature for ꢀ10 min and the
resulting bright orange solution was added to Et2O (15
ml). Cooling of the resulting solution overnight at
−20°C led to the deposition of the orange microcrys-
talline product (390 mg, 70%). Crystals suitable for
X-ray diffraction were obtained upon allowing a dilute
solution of the complex in CH3CN/Et2O (ꢀ1:4) to
2.6. [L6MePyCu]SbF6
To a solution of L6MePy (333 mg, 1.05 mmol) in
CH3CN/THF (2:1, 2 ml total volume) was added solid
[Cu(CH3CN)4]SbF6 (430 mg, 0.93 mmol). The mixture
was stirred at room temperature until all of the Cu(I)
1
stand at −20°C for 3–4 days. H NMR (300 MHz,
d6-acetone) l 8.09–8.00 (m, 3H), 7.78 (d, J=7.8 Hz,
1H), 7.62–7.50 (m, 4H), 4.33 (s, 2H), 3.22–3.00 (m,

L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
119
6H), 2.89–2.72 (m, 8H), 1.05 (d, J=6.6 Hz, 6H), 0.97
Table 1
(d, J=6.6 Hz, 6H) ppm; 13C{1H} NMR (75 MHz,
d6-acetone) l 159.3, 157.3, 141.6, 138.2, 129.8, 129.0,
127.4, 123.4, 123.0, 60.4, 57.9, 51.8, 51.5, 50.5, 20.7,
18.6 ppm; FTIR (KBr, cm−1) 659 (SbF6); UV–Vis
(THF) [umax, nm (m, M−1 cm−1)] 388 (2600); FAB-MS
m/z (relative intensity) 443.2 ([M−SbF6]+, 100%).
Anal. Calc. for C24H36N4CuSbF6: C, 42.47; H, 5.35; N,
8.26. Found: C, 42.25; H, 5.21; N, 8.26%.
Spectroscopic properties of copper–dioxygen complexes
a
Complex
UV–Vis umax Resonance Raman b
(nm)
(cm−1
)
(m (M−1 cm−1))
[(LPyCu)2(O2)](CF3SO3)2
550 (10 200),
600 (9700)
822 (771),
530 (506)
584 (557)
[(L6PhPyCu)2(m-O)2](SbF6)2 316 (10 000),
432 (12 000)
550 (11 500),
594 (11 000)
[(L5MePyCu)2(O2)](SbF6)2
823 (772),
530 (502),
476 (457)
2.9. [LQuinCu]SbF6
A solution of LQuin (0.071 g, 0.20 mmol) in EtCN (2
ml) was treated with solid [Cu(CH3CN)]SbF6 (0.090 g,
0.19 mmol) producing a deep red solution. After stir-
ring for 1 h, Et2O (10 ml) was added which resulted in
the precipitation of the product as red microcrystals.
Further purification was accomplished by recrystalliza-
tion from THF/Et2O (0.112 g, 86%). 1H NMR (300
MHz, CD3CN) l 8.41 (d, J=8.4 Hz, 1H), 8.26 (d,
J=8.4 Hz, 1H), 8.02 (d, J=7.8 Hz, 1H), 7.86 (t,
J=7.2 Hz, 1H), 7.68 (t, J=7.2 Hz, 1H), 7.49 (d,
J=8.7 Hz, 1H), 4.33 (s, 2H), 3.20–3.08 (m, 4H),
3.02–2.84 (m, 6H), 2.82–2.60 (m, 4H), 1.21 (d, J=6.6
Hz, 6H), 1.16 (d, J=6.6 Hz, 6H) ppm; 13C{1H} NMR
(75 MHz, CDCl3) l 160.2, 145.2, 137.1, 130.5, 129.3,
128.5, 128.2, 127.4, 121.6, 62.2, 57.7, 53.0, 50.9, 50.5,
19.8 ppm; FTIR (KBr, cm−1) 659 (SbF6); UV–Vis
(CH2Cl2) [umax, nm (m, M−1 cm−1)] 416 (4500); FAB-
MS m/z (relative intensity) 417 ([M−SbF6]+, 100%).
Anal. Calc. for C22H34N4F6CuSb: C, 40.42; H, 5.24; N,
8.57. Found: C, 40.74; H, 5.23; N, 8.72%.
a Measured in either THF:CH3CN (10:1, LPy) or THF (L6PhPy
L5MePy) below −70°C.
,
b The resonance Raman spectra were obtained at 77 K as frozen
THF:CH3CN (10:1, LPy) or THF (L6PhPy, L5MePy) solutions using
either 572 nm (LPy and L5MePy) or 457 nm (L6PhPy) laser excitation;
only 18O sensitive vibrations are given with specific 18O shifts indi-
cated in parentheses.
of CH3CN was added to clarify the solution. The
headspace was then rapidly replaced with NO and the
reaction stirred vigorously under NO with occasional
shaking for 24 h. The headspace was then sampled with
a gas-tight syringe; GC analysis (HP 5180 GC; 6 m
Propak-Q column; flow rate=20 ml min−1; temp.=
30°C) of the headspace gas revealed the presence of
N2O (retention time 4.0 min; yield=63% based on
[LPyCu]+, average of three replicate runs).
2.11. Re6ersible reactions of [LPCu]SbF6 (P=Py or
Quin) with CO(g)
A solution of [LPCu]SbF6 in THF (ca. 0.6 mM) was
purged with CO(g) causing a rapid bleaching of the
yellow–orange color (cf. UV–Vis spectra in Fig. S2).
Analysis of aliquots by FTIR spectroscopy revealed
wCO=2067 (LPy), 2076 (LQuin) cm−1. The colorless so-
lution was then purged with N2 for 5 min, regenerating
the original color and optical spectrum of [LPCu]X
(Fig. S2). This CO/N2 protocol was repeated several
times without appreciable decomposition.
2.10. Reaction of [LPyCu]CF3SO3 with NO(g)
A solution of [LPyCu]CF3SO3 (0.250 g, 0.48 mmol) in
THF/CH3CN (9:1, 10 ml) was exposed to 1 atm NO(g).
After stirring for 30 min, excess NO(g) and the solvent
were removed under vacuum. The blue–green residue
was redissolved in THF, an equal volume of Et2O was
added, and the cloudy solution stored at −20°C
overnight. The resultant blue crystalline mass was col-
lected, washed with Et2O, and dried under vacuum.
Recrystallization from CH2Cl2/Et2O afforded blue crys-
tals of [LPyCu(ONO)]CF3SO3·CH2Cl2 suitable for X-
ray crystallographic analysis (0.247 g, 79%). EPR (9.46
GHz, 77 K, 1:1 CH2Cl2/toluene) gz=2.22, Az=150 G,
gy=2.08; gx=2.00; UV–Vis (THF) [umax, nm (m, M−1
cm−1)] 388 (590), 622 (150); FTIR (KBr, cm−1) 1265
(CF3SO−3 ), 1115 (CF3SO3−), 1028 (CF3SO−3 ), 637
(CF3SO−3 ). Anal. Calc. for C20H34N5O5F3SCl2Cu: C,
37.07; H, 5.29; N, 10.81. Found: C, 37.69; H, 5.26; N,
10.81. Gaseous product analysis was performed as fol-
lows. A Schlenk flask of known volume (23.2 ml) was
charged with a solution of [LPyCu]CF3SO3 (0.050 g,
0.097 mmol) in THF (2.00 ml) under N2. A single drop
2.12. Reactions of [LPCu]SbF6 with O2(g)
Solutions of [LPCu]SbF6 in THF (ꢀ0.5–20 mM)
were cooled to about −75°C and exposed to dry O2
which resulted in immediate color changes as described
in the text. Samples for UV–Vis spectroscopy were
prepared in a quartz UV–Vis cuvette modified for use
at low temperature. Samples for EPR (0.5–2.0 mM)
and resonance Raman (ꢀ20 mM) spectroscopies were
prepared by removing aliquots of the solution with a
pre-cooled (liq. N2) pipette. Data are listed in Table 1.
In experiments aimed at identifying the decomposi-
tion products of the oxygenated species, the solutions

120
L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
(ꢀ0.1 mM) were purged with N2(g) for 10 min after
oxygenation was complete (UV–Vis) and then allowed
to warm to room temperature. The resulting green
solution was treated with aqueous NH4OH (1 ml) and
CHCl3 (5 ml) with vigorous stirring. The organic phase
was then removed, and the blue aqueous phase extracted
with CHCl3 (2×10 ml). The combined organic extracts
were dried (Na2SO4) and evaporated to yield a yellow
(pyridyl or quinolyl) or the type and position (5- or
6-methyl; 6-phenyl) of pyridyl ring substituents. These
new ligands were prepared via treatment of 1,4-diiso-
propyl-1,4,7-triazacyclononane (LH) [22] with the appro-
priate 2-halomethyl-pyridyl or -quinolyl derivative and
were isolated in ꢀ40–60% yield following purification.
All ligands were characterized by NMR spectroscopy
and high resolution mass spectrometry.
residue (mass recovery: L6MePy, 93%; L6PhPy, 88%; LQuin
,
Copper(I) complexes (Fig. 2) of LP [P=Py, 6MePy,
5MePy, 6PhPy, Quin) were prepared by treating the
ligands with [Cu(CH3CN)4]X (X=SbF−6 or CF3SO3−)
in CH3CN or CH3CN/THF (2:1) under an inert atmo-
sphere. The complexes were isolated as yellow to or-
ange–red air-sensitive microcrystalline solids and were
1
88%; L5MePy, 61%) which was analyzed by H NMR
spectroscopy and GC/MS. For P=Py or 5MePy, the
presence of L5MePy and L5MePyO in an approximately 2:1
1
ratio was determined from H NMR spectra. For P=
6MePy, 6PhPy, or Quin, LP and LH were identified in an
approximately 3:2 ratio. The corresponding 2-carbox-
aldehyde coproduct in each of these N-dealkylation
reactions was identified by 1H NMR and GC-MS but was
present only in low yield (B20% of expected amount
based on yield of LH) presumably due to decomposition
under the conditions of the work-up procedure.
1
characterized by CHN analysis, H and 13C NMR and
FTIR spectroscopy, fast atom bombardment mass
spectrometry, and, in several cases, X-ray crystallogra-
phy. Binding of the respective appended heterocyclic
donor to Cu(I) in the complexes was evident from their
UV–Vis spectra, which contained an intense (m 3400–
4500 M−1 cm−1) band with umax$360 (LPy, L6MePy
,
2.13. X-ray crystallography
L5MePy), 388 (L6PhPy), or 420 nm (LQuin) that may be
Crystals of [L6PhPyCu]SbF6 (orange block) and
[LPyCu(ONO)]CF3SO3·CH2Cl2 (blue plate) were at-
tached to a glass fiber with heavy-weight oil and mounted
on a Siemens SMART CCD system for data collection
at 173(2) K. Initial sets of cell constants were calculated
from 70 or 137 strong reflections, respectively, harvested
from three sets of 20 frames. Final cell constants were
calculated from a set of 8192 or 5195 strong reflections
Table 2
X-ray crystallographic data a
[L6PhPyCu]SbF6
[LPyCu(ONO)]
O3SCF3·CH2Cl2
Empirical formula
Formula weight
Crystal system
Space group
C24H36N4F6SbCu C20H34N5O5Cl2F3SCu
679.86
monoclinic
P21/n
18.4464(4)
9.6457(2)
16.0169(2)
90
108.093(1)
90
2708.95(6)
4
646.10
triclinic
(
from the data collection. The space groups P2/n or P1,
respectively, were determined on the basis of systematic
absences and intensity statistics. Successful direct-
methods solutions were calculated which provided most
of the non-hydrogen atoms from the E-map. Several
full-matrix least-squares/difference Fourier cycles were
performed which located the remainder of the non-hy-
drogen atoms. All non-hydrogen atoms were refined with
anisotropic displacement parameters. All hydrogen
atoms were placed in ideal positions and were refined as
riding atoms with relative isotropic displacement
parameters. Calculations were performed using
SHELXTL-PLUS Version 5.0 [29]. Crystal data and refine-
ment parameters are listed in Table 2 and selected bond
distances and angles are shown in Table 3; full reports
are provided as supporting information.
(
P1
,
a (A)
9.8889(2)
12.7742(1)
13.2909(1)
113.534(1)
93.394(1)
109.382(1)
1415.93(3)
2
,
b (A)
,
c (A)
h (°)
i (°)
k (°)
3
,
V (A )
Z
Density (calc.)
1.667
1.520
(g cm−3
)
Temperature (K)
Crystal size (mm)
Diffractometer
173(2)
173(2)
0.18×0.15×0.12 0.50×0.50×0.25
Siemens SMART Siemens SMART
Absorption coefficient 1.844
1.092
(mm−1
)
2q max (°)
Reflections collected
50.04
13275
50.06
7105
Independent reflections 4748
4786
Variable parameters
331
0.0314/0.0658
1.036
372
0.0430/0.1033
1.062
b
3. Results and discussion
R1/wR2
Goodness-of-fit (F2)
Largest difference peak 0.646/−0.533
1.184/−0.757
3.1. Synthesis and spectroscopic characterization of
ligands and copper(I) complexes
−3
,
(e A
)
a
,
Radiation used: Mo Ka (u=0.71073 A).
b R1=S ꢀꢀ Fo ꢀ−ꢀ Fc ꢀꢀ/S ꢀ Foꢀ;
wR2=[S[w(Fo2−Fc2)2]/[w(Fo2)2]]1/2
We prepared a series of ligands (Scheme 1) that differ
with respect to the nature of the heterocyclic donor
where w=1/|2(Fo2)+(aP)2+bP.

L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
121
Table 3
a
,
Selected bond lengths (A) and angles (°) for complexes characterized by X-ray crystallography
b
[LPyCu]O3SCF3
Cu(1)ꢀN(1)
Cu(1)ꢀN(3)
2.167(3)
2.108(3)
Cu(1)ꢀN(2)
Cu(1)ꢀN(5)
2.133(3)
1.941(3)
N(1)ꢀCu(1)ꢀN(2)
N(1)ꢀCu(1)ꢀN(4)
N(3)ꢀCu(1)ꢀN(4)
85.75(11)
86.21(12)
139.67(12)
N(1)ꢀCu(1)ꢀN(3)
N(2)ꢀCu(1)ꢀN(3)
N(2)ꢀCu(1)ꢀN(4)
84.64(11)
87.27(11)
131.05(12)
[L6PhPyCu]SbF6
Cu(1)ꢀN(1)
Cu(1)ꢀN(3)
2.191(3)
2.129(3)
Cu(1)ꢀN(2)
Cu(1)ꢀN(4)
2.070(3)
1.948(3)
N(1)ꢀCu(1)ꢀN(2)
N(1)ꢀCu(1)ꢀN(4)
N(3)ꢀCu(1)ꢀN(4)
84.85(11)
84.60(11)
122.77(11)
N(1)ꢀCu(1)ꢀN(3)
N(2)ꢀCu(1)ꢀN(3)
N(2)ꢀCu(1)ꢀN(4)
84.30(11)
88.28(11)
145.83(11)
[LPyCu(ONO)]O3SCF3 · CH2Cl2
Cu(1)ꢀN(1)
Cu(1)ꢀN(7)
Cu(1)ꢀO(1)
N(1A)ꢀO(2)
2.050(2)
2.292(2)
1.982(1)
1.219(4)
Cu(1)ꢀN(4)
Cu(1)ꢀN(16)
O(1)ꢀN(1A)
2.061(2)
2.001(2)
1.297(4)
N(1)ꢀCu(1)ꢀN(4)
N(1)ꢀCu(1)ꢀN(16)
N(4)ꢀCu(1)ꢀN(7)
N(4)ꢀCu(1)ꢀO(1)
N(7)ꢀCu(1)ꢀO(1)
168.24(9)
82.65(10)
85.49(9)
96.43(10)
107.57(9)
N(1)ꢀCu(1)ꢀN(7)
N(1)ꢀCu(1)ꢀO(1)
N(4)ꢀCu(1)ꢀN(16)
N(7)ꢀCu(1)ꢀN(16)
N(16)ꢀCu(1)ꢀO(1)
84.09(9)
168.4(9)
168.04(10)
96.03(10)
94.42(10)
a Estimated standard deviations indicated in parentheses.
b Selected bond distances and angles for one independent cation (of two) in the asymmetric unit of [LPyCu]O3SCF3; X-ray crystallographic data
for this complex has been previously reported, but metrical information is reproduced here for comparative purposes [22].
spectra at ambient temperature of [L6PhPyCu]SbF6,
attributed to a Cu(I)“heterocycle MLCT transition
[L6MePyCu]SbF6, and [LQuinCu]SbF6 exhibit sharp peaks
consistent with retention of the structures determined
by X-ray crystallography (vide infra) that exhibit h4-co-
ordination of their respective supporting ligands. We
1
[30]. The H NMR spectra of [LPyCu]X (X=O3SCF3−
or SbF−6 ) and [L5MePyCu]SbF6 in CD3CN or acetone-d6
at or above ambient temperature (data acquired up to
i
+55°C) contain sharp peaks for the Pr groups and
4-pyridyl hydrogen, but the remaining resonances are
broad (cf. Fig. S1 in Supplementary material). Sharpen-
ing of the spectrum of [LPyCu]SbF6 occurs as the
temperature is lowered (data acquired to −85°C).
These observations imply the existence of fluxionality in
solution, perhaps involving binding and release of sol-
1
vent, the pyridyl arm, or both. In contrast, H NMR
Fig. 3. Cationic portions of the X-ray crystal structures of (a)
[LPyCu]O3SCF3 [22] and (b) [L6PhPyCu]SbF6 showing non-hydrogen
atoms as 50% ellipsoids.
Fig. 2. Copper(I) complexes prepared in this study.

122
L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
Overall, the coordination geometries in both com-
pounds are similar, the only significant difference be-
tween them being the degree of steric hindrance about
a potential fifth coordination site (cf. space-filling draw-
ings shown in Fig. 4). In [LPyCu]+ this site lies within a
pocket formed by the isopropyl substituents and the
pyridyl donor and appears relatively accessible, whereas
in [L6PhPyCu]+ it is blocked by the phenyl substituent4.
An intermediate degree of shielding of this site in
[LQuinCu]+ is apparent from the preliminary X-ray
structural data and from molecular models. As de-
scribed below, these differences in metal ion accessibility
are of critical importance in small molecule reactivity
studies.
Fig. 4. Space-filling representations of the cationic portions of the
X-ray crystal structures of (a) [LPyCu]O3SCF3 [22] and (b)
[L6PhPyCu]SbF6.
For purposes of comparison to the structures of
[LPyCu]+ and [L6PhPyCu]+, reproductions of the previ-
ously reported X-ray structures of Cu(I) complexes of
related tetradentate tripodal ligands bis(pyridyl-
methyl)(6-carbomethoxy-pyridylmethyl)amine (TMPA%)
[10], tris(quinolylmethyl)amine (TMQA) [12], and tris(6-
phenyl-pyridylmethyl)amine (TPPA) [27] are shown in
Fig. 5. Of the five structures illustrated in Figs. 3–5, that
of [(TMPA%)Cu(CH3CN)]+ is distinguished by the fact
that it adopts a five-coordinate geometry with a bound
surmise that the absence of fluxional behavior for these
complexes is due to the increased steric bulk of their
heterocyclic appendages that makes the metal ion less
accessible to solvent.
3.2. X-ray structures of copper(I) complexes
Drawings of the X-ray structures of the cationic
portions of [LPyCu]O3SCF3 (communicated previously
[22a] but shown here for comparative purposes; only
one of two crystallographically independent but chemi-
cally similar molecules in the asymmetric unit is repro-
duced) and [L6PhPyCu]SbF6 are presented in Fig. 3.
Crystallographic data for the latter structure and se-
lected bond distances and angles for both are listed in
Tables 2 and 3, respectively. A preliminary X-ray crystal
structure of [LQuinCu]SbF6 indicated a topology closely
analogous to those of the compounds coordinated by
LPy and L6PhPy, but problems with pseudosymmetry and
disorder prohibit a full description.
,
acetonitrile (CuꢀNCH CN=1.999(9) A) trans to
a
3
,
weakly-ligated amine donor (CuꢀNamine=2.439(8) A).
In contrast, the remaining four structures comprising
more sterically hindered supporting ligands are qualita-
tively alike insofar as they adopt four-coordinate, ap-
proximately trigonal pyramidal geometries with no
binding of exogenous acetonitrile. Closer inspection,
however, reveals interesting differences between the to-
pologies imposed by LPy and L6PhPy compared to those
of the hindered variants of TMPA. The CuꢀNheterocycle
bond lengths in [LPyCu]SbF6 (1.941(3) A) and
,
[L6PhPyCu]SbF6 (1.948(3) A) are shorter than the corre-
,
In [LPyCu]SbF6 and [L6PhPyCu]SbF6, two amine nitro-
gens and the heterocyclic nitrogen form the basal plane
of a distorted trigonal pyramidal coordination geome-
try, with the third amine nitrogen (N1) occupy ing the
axial position. The intraligand NꢀCuꢀN bond angles
within the basal plane in both structures are significantly
distorted from the idealized value of 120° (LPy,
N(2)ꢀCu(1)ꢀN(3)=87.3(1)° and N(3)ꢀCu(1)ꢀN(4)=
sponding distances in [(TMQA)Cu]+ and [(TPPA)Cu]+
Py
+
,
(avg. 2.01 and 2.12 A). In addition, in [L Cu] and
[L6PhPyCu]+ intraligand NꢀCuꢀN bond angles within
the basal plane are significantly distorted from 120°, but
more regular angles are adopted in [(TMQA)Cu]+
(122.1(2), 120.6(2) and 113.0(2)°) and [(TPPA)Cu]+
(122.0(3), 119.0(3) and 112.2(3)°). This difference may
be attributed to the less constrained nature of the
TMQA and TPPA tripods compared to the LP ligands,
which have three out of the four potential donors
fettered by the TACN macrocycle.
139.7(1)°; L6PhPy
,
N(2)ꢀCu(1)ꢀN(3)=88.3(1)° and
N(2)ꢀCu(1)ꢀN(4)=145.83(11)°), apparently as a result
of the constrained nature of the triazamacrocyclic ring.
Bond angles between the basal ligands and the axial
N-donor N(1) are slightly acute (LPy, avg. 85.4°; L6PhPy
avg. 84.6°), a consequence of a displacement of the
3.3. Electrochemistry
,
copper ion slightly (ꢀ0.15 A) below the basal plane. In
The half-wave potentials for all Cu(I) complexes were
measured by cyclic voltammetry under a N2 atmo-
both structures, the CuꢀNpyridyl bond length (LPy,
6PhPy
,
,
1.948(3) A, respectively) is shorter
1.941(3) A; L
,
than the other CuꢀN bond lengths (2.07–2.21 A),
possibly reflecting increased CuꢀNpyridyl bond order aris-
ing from p-backdonation (Cu(I)d“heterocycle p*).
4 Bonding interactions between the Cu(I) ion and the phenyl ring
carbons in [L6PhPyCu]+ appear to be absent on the basis of Cu-C
,
distances \3 A.

L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
123
Fig. 5. Reproductions of the X-ray crystal structures of previously reported copper(I) complexes of TMPA derivatives.
sphere in CH3CN solution with 0.1 M (Bu4N)(PF6)
(Table 4, reported vs. SCE, Pt electrode). All of the
complexes [LPCu]X (R=Py, 6MePy, 5MePy, 6PhPy,
Quin) exhibited a quasi-reversible, one-electron redox
transformation with peak separations (DEp) of ꢀ90–
110 mV at a scan rate of 0.02 V s−1. Under identical
experimental conditions, the ferrocene/ferrocenium
couple was observed at 0.39 V versus SCE (DEp (0.02 V
basicities of the two heterocyclic donor groups (pKb=
7.94 and 9.12, respectively) [31]. On the basis of the
available solid state structural data that indicate ho-
mologous coordination topologies for [LPyCu]O3SCF3,
[LQuinCu]SbF6, and [L6PhPyCu]SbF6 (vide supra), coordi-
nation geometry effects also would appear to be pre-
cluded as a rationale for the divergent E1/2 values for the
two sets of [LPCu]+ complexes. On the other hand, the
NMR data indicate that the Cu(I) complexes of LPy and
L5MePy (which have low redox potentials) are fluxional in
solution, possibly due to equilibria involving solvent
binding and release. While coordination of CH3CN to
the Cu(I) complexes of these ligands would be expected
to raise rather than lower their redox potentials, solvent
binding to the Cu(II) forms to yield stable five-coordi-
nate geometries might explain their low E1/2 values. Such
solvent coordination, to either the Cu(I) or Cu(II) forms,
would be inhibited sterically in the L6MePy, L6PhPy, and
LQuin cases (as borne out by the NMR spectral results,
vide supra), resulting in higher redox potentials for these
complexes. In addition to this direct ‘blocking’ effect of
the ligand substituents that inhibits solvent coordina-
tion, a less specific shielding of the redox center from
polar solvent by the hydrophobic appendages also may
s
−1)=100 mV).
The data listed in Table 4 indicate that the heterocyclic
appendages have important effects on the Cu(I)/Cu(II)
redox potential. Overall, the E1/2 values for the new
complexes are sufficiently low to favor participation of
the Cu(I) compounds in O2 activation chemistry. The
potentials are lower than that measured previously for
[(iPr3TACN)Cu(CH3CN)]+ [24], which may be
attributed — at least in part — to stabilization of the
Cu(I) state in the latter by coordinated CH3CN, a
‘softer’ donor that favors the cuprous state more than
the heterocyclic appendages on LP. The [LPCu]+
complexes may be divided into two sets on the basis of
their redox behavior; those with E1/2B0 V versus SCE
(R=Py and 5MePy) versus those with E1/2\0 V
(R=Quin, 6MePy, and 6PhPy). Our analysis of how the
nature of the heterocyclic donor group causes the
ꢀ0.2–0.3 V difference in redox potentials between these
two sets of complexes follows those used by Karlin and
co-workers [12], Canary and co-workers [27], and Sorrell
and Jameson [9g] to rationalize data acquired for
complexes of tetradentate tripodal TMPA, BQPA,
TPPA, and related ligands (data listed in Table 4 for
selected cases measured in CH3CN).
The factors which have been shown to influence the
redox potentials of copper complexes include: (1) type of
donor atoms; (2) coordination geometry, and (3) ligand
substituent effects. Differences in the basicity of the
heterocylic donor groups in the LP chelates seem to have
little effect on the position of the Cu(I)/Cu(II) redox
couple. For example, the Cu(I) complex supported by
L6MePy is oxidized at an identical potential to that
supported by LQuin despite significant differences in the
Table 4
Cyclic voltammetry data for Cu(I) complexes in CH3CN vs. SCE a
Complex
E1/2 (V) DEp (mV)
Ref.
[LPyCu]+
−0.07
−0.10
0.12
0.13
0.26
0.36
0.06
−0.24
94
110
104
102
92
this work
this work
this work
this work
this work
[24]
[L5MePyCu]+
[LQuinCu]+
[L6MePyCu]+
[L6PhPyCu]+
[(iPr3TACN)Cu(CH3CN)]+
[(TPPA)Cu(CH3CN)]+
[(TMPA)Cu(CH3CN)]+
110
b
[27]
[12]
c
a For measurements made in this work, 0.1 M (Bu4N)(PF6) was
used with a Pt electrode at 20°C.
b Not reported for this solvent, but in DMF DEp=120 mV.
c Not reported for this solvent, but in DMF DEp=78 mV.

124
L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
cm−1, respectively (Scheme 2 and Fig. S2 in Supple-
mentary material). The loss of the Cu(I)“heterocycle
MLCT feature upon binding of CO to [LPCu]+ attests
to displacement of the heterocyclic arm to yield four-
coordinate adducts [(h3-LP)Cu(CO)]SbF6. Purging of
the colorless solutions of these adducts with N2 for
several minutes resulted in complete restoration of the
original color and charge transfer intensity, indicating
that CO binding and concomitant displacement of the
heterocyclic arm are reversible. An analogous displace-
ment of a pendant pyridyl donor by an exogenous
ligand
was
observed
in
the
reaction
of
[(TMPA)Cu(CH3CN)]+ with PPh3, yielding [(h3-
TMPA)Cu(PPh3)]+ that was structurally characterized
by X-ray crystallography [10]. Whether CO reacts with
[(TMPA)Cu(CH3CN)]+ similarly or simply displaces
CH3CN is not firmly established. The CO adducts
[(R3TACN)Cu(CO)]+ exhibit similar wCO values to
those supported by h3-LP, but due to the lack of a
fourth potential donor do not lose CO and revert to
their respective precursor complex nearly as readily
[32]. In sum, the additional pyridyl or quinolyl donor
group on the 1,4,7-triazacyclononane macrocyclic lig-
and framework in Cu(I) complexes is substituted by
exogenous CO, but CO binding is rendered reversible
by the facility with which the appended donor may
rebind to the metal ion.
Scheme 2.
contribute to the destabilization of the Cu(II) state and
thus to the observation of high potentials for
[L6PhPyCu]+, [LQuinCu]+, and [L6MePyCu]+. Analogous
arguments invoking increased hydrophobicity and/or
solvent ‘blocking’ effects of the quinolyl and phenyl-
3.4.2. Nitric oxide
Exposure of THF/CH3CN (4:1) solutions of
[LPyCu]CF3SO3 to NO at room temperature resulted in
a gradual color change from yellow to blue–green and
the evolution of N2O (yield=65% by GC analysis of
the headspace gas after 24 h). Removal of solvent
followed by recrystallization afforded a single copper-
containing compound, identified as the nitrito complex
[LPyCu(NO2)]CF3SO3 (Scheme 2). The solid state struc-
ture of the complex as determined by X-ray crystallog-
raphy is depicted in Fig. 6; crystallographic data and
selected bond lengths and angles are presented in Ta-
bles 2 and 3, respectively. The central Cu(II) ion is
pyridyl
substituents
of
TMQA
and
TPPA
relative to TMPA and of the methyl groups in the
3,5-dimethylpyrazolyl analog of tris[2-(1-pyrazolyl)-
ethyl]amine) [9g] have been offered to explain the
observed E1/2 values for copper complexes of these
ligands.
3.4. Reacti6ity of copper(I) complexes with small
molecules
Because the principal focus of our studies has been to
characterize ligand structural influences on dioxygen
binding and activation by copper complexes, O2 reac-
tions with the entire set of [LPCu]X compounds were
examined. First, however, we describe the interactions
of selected members of the set with CO and NO that
illustrate key patterns of reactivity arising from specific
features of the LR ligand type.
3.4.1. Carbon monoxide
Treatment of [LPCu]X (P=Quin, X=SbF6; P=Py,
X=O3SCF3) with CO results in bleaching of the color
of the solution and growth of a wCO at 2076 or 2067
Fig. 6. Cationic portion of the X-ray crystal structure of
[LPyCu(ONO)]O3SCF3·CH2Cl2 showing non-hydrogen atoms as 50%
ellipsoids.

L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
125
apparent that small changes in ligand denticity (tri- vs.
tetradentate) and/or ligand type (aromatic vs. aliphatic
N-donor) can dramatically alter the pathway of a Cu(I)/
NO reaction.
[LPyCu]++3NO“N2O+[LPyCu(ONO)]2+
(1)
3.4.3. Dioxygen
The course of low temperature oxygenations of the
copper(I) complexes depends on the nature of the hetero-
cyclic appendage (Scheme 3). Oxygenation of cold (ca.
−75°C) yellow solutions of [LPyCu]O3SCF3 or
[L5MePyCu]SbF6 in THF results in an immediate color
change to deep purple. UV–Vis and resonance Raman
spectroscopic data obtained for these solutions are sim-
ilar to those of the fully characterized trans-1,2-peroxo
complex [(TMPA)2Cu2(O2)]2+ [10] (Table 1). Thus, the
purple species exhibit diagnostic CT absorption bands
and resonance enhanced (uex=572 nm) 18O-sensitive
features in the Raman spectrum attibutable to peroxide
OꢀO and CuꢀO vibrations (Fig. S3 in Supplementary
material). The purple solutions are EPR silent, indicative
of strong antiferromagnetic coupling between the Cu(II)
ions. On the basis of the combined data, we formulate
the oxygenation products as trans-1,2-peroxodicopper
complexes [(LPCu)2(O2)](X)2 (P=Py, X=O3SCF3; P=
5MePy, X=SbF6) analogous to the TMPA-supported
species.
Instead of generating a purple solution upon low
temperature oxygenation, [L6PhPyCu]SbF6 in THF reacts
with O2 to produce intense yellow solutions at concentra-
tions suitable for UV–Vis spectroscopic analysis (ꢀ1
mM) and orange–brown solutions at higher concentra-
tions (ꢀ10 mM). Spectroscopic data (Table 1) indicate
that the oxygenated product in this case is a bis(m-
oxo)dicopper species [(h3-L6PhPyCu)2(m-O)2](SbF6)2
(Scheme 3). Particularly diagnostic spectral features
similar to those identified previously for the bis(m-
oxo)dicopper unit (e.g. for the complex of Bn3TACN as
shown in Table 1) are the intense CT absorptions,
resonance enhanced (uex=457 nm) 18O-sensitive [Cu2(m-
O)2]2+ core vibrations in the Raman spectrum, and EPR
silence. Further characterization of [(L6PhPyCu)2(m-
O)2](SbF6)2 by techniques such as mass spectrometry or
manometry was precluded due to the rapid rate of
decomposition of the complex at about −75°C (vide
infra). Due to the similarities of the spectral properties
to simple R3TACN derivatives, we hypothesize that
formation of the bis(m-oxo)dicopper complex supported
by the L6PhPy ligand is accompanied by loss of coordina-
tion of the pyridyl group. Thus, the different fates of the
Cu(I) complexes of LPy and L5MePy versus that of L6PhPy
upon oxygenation are directly related to the coordination
of the heterocyclic appendage. In the LPy and L5MePy
cases the pyridyl unit remains coordinated, favoring
Scheme 3.
bound by all donors of the LPy ligand and a single O atom
of the nitrite ion in a square pyramidal geometry (~=
0.003) with an amine N atom (N7) occupying the axial
,
coordination site (Cu1ꢀN7=2.292(2) A). Nitrite ion
,
bond alternation (D(NꢀO)=0.078 A) and a long
,
Cu1···O2 distance (2.779 A) offer convincing evidence of
weak or non-existent Cu1ꢀO2 bonding. The CuꢀOnitrite
,
bond length (1.982(2) A) is only slightly longer than the
Cu(I)ꢀOnitrite bond length observed in the dicopper(I)ꢀni-
trite complex {[(iPr3TACN)Cu]2(m-NO2)}PF6 (1.968(2)
,
,
A) [24] and in [(TMPA)Cu(NO2)]PF6 (1.938(2) A)
[33] wherein the nitrite O atom occupies an axial position
in the trigonal-bipyramidal complex (~=0.74).
These CuꢀOnitrite bond lengths are uniformly shorter than
those observed in the (h2-O,O)-nitrite complexes
i
,
[(Ph3P)2Cu(NO2)] (2.191(4) A) [34], {[( Pr3TACN)Cu]2-
,
(m-NO2)}[B(Arf)4]2 (2.047(9) and 2.13(1) A) [24], and
TpMe2Cu(NO2) [18b].
The observed products of the reaction of NO with
[LPyCu]+ are indicative of a NO reductive disproportion-
ation reaction (Eq. (1)) similar to that observed with
tris(pyrazolyl)hydroborate-copper(I) complexes [18]. In
the reactions of NO with these latter complexes mononu-
clear copperꢀnitrosyl complexes were identified. How-
ever, no spectroscopically identifiable intermediates were
observed upon treatment of solutions [LPyCu]CF3SO3
with NO either at 25 or −78°C. The NO reactivity we
observed differs significantly from that seen with com-
plexes of TMPA [17] or R3TACN [19] (Fig. 1). It is thus

126
L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
trans-1,2-peroxo generation, whereas pyridyl dissocia-
tion facilitated by the sterically bulky 6-phenyl group
on L6PhPy yields a [h3-L6PhPyCu]+ fragment amenable to
bis(m-oxo) core formation. Note that the steric features
associated with the [h3-L6PhPyCu]+ fragment are very
similar to [LiPr2BnCu]+, which also yields a bis(m-
infra), perhaps due to less steric protection by the
6-methyl versus the 6-phenyl group. Finally, treatment
of a THF solution of [LQuinCu]SbF6 with dioxygen at
about −75°C resulted in a bleaching of the orange
color of the Cu(I) complex over the course of ꢀ1–2
min with formation of a pale green solution as the final
product. This bleaching corresponds to the loss of the
416 nm MLCT band of [LQuinCu]SbF6. Careful moni-
toring of the oxygenation process (mild bubbling) at
about −75°C, scanning every ꢀ10 s, did not reveal
the formation of any spectroscopically identifiable cop-
perꢀdioxygen species.
Both types of dioxygen adducts described above, the
trans-1,2-peroxo and bis(m-oxo)dicopper complexes, de-
compose upon warming, but via different pathways.
We reported previously that warming of solutions of
[(LPyCu)2(O2)](O3SCF3)2 followed by extraction of the
copper ions with NH4OH yielded a mixture of LPy and
a new modified ligand LPyO in a 4:1 ratio (ꢀ75% total
recovery) (Scheme 4) [22]. We also have found that the
analogous compound [(L5MePyCu)2(O2)](SbF6)2 exhibits
similar reactivity upon decomposition, yielding L5MePy
and L5MePyO in a ꢀ2:1 ratio (ꢀ61% total recovery).
In contrast to the ligand oxygenation resulting from
decay of the trans-1,2-peroxo complexes, decompos-
oxo)dicopper core (rather than
a
(m-h2:h2-per-
oxo)dicopper isomer) upon oxygenation [32].
The results of the oxygenations of THF solutions of
[L6MePyCu]SbF6 and [LQuinCu]SbF6 are less clear than
those described above. In the case of [L6MePyCu]SbF6,
treatment with O2 at −75°C caused an increase in the
intensity of the yellow color of the solution followed by
bleaching to a pale green solution. UV–Vis spectro-
scopic monitoring of the oxygenation process (ꢀ5 min,
mild O2 bubble) revealed the formation of a new fea-
ture at 430 nm that grew in and then rapidly decayed.
During this process, the MLCT band of the Cu(I)
complex (364 nm) gradually diminished, but the spectra
obtained were always indicative of the presence of both
the Cu(I) complex and the initial oxygenated product.
We speculate that a bis(m-oxo)dicopper core akin to
that seen for the L6PhPy case forms, again due to pyridyl
arm loss facilitated by the steric hindrance of the 6-sub-
stituent, but in this instance it is more reactive (vide
Scheme 4.

L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
127
ition of the bis(m-oxo)dicopper complex [(L6PhPyCu)2(m-
O)2](SbF6)2 results in oxidative N-dealkylation of the
heterocyclic appendage. Thus, warming of solutions of
[(L6PhPyCu)2(m-O)2](SbF6)2 followed by removal of the
copper ions from the green solution by treatment with
NH4OH resulted in the isolation of a 3:2 mixture of
L6PhPy and LH and a trace amount of 6-phenyl-2-
pyridinecarboxaldehyde (ꢀ87% total recovery). This
type of N-dealkylation reactivity was shown previously
[35] to be the predominant pathway of decomposition of
bis(m-oxo)dicopper complexes. However, the observa-
tion of N-dealkylation at the pseudobenzylic position of
the ligand contrasts with the isopropyl substituent cleav-
age seen upon decomposition of {[(iPr2BnTACN)-
Cu]2(m-O)2}(ClO4)2. The differences observed in the re-
gioselectivity of these N-dealkylation reactions may
indicate that additional factors associated with the
6PhPy substituent may influence the observed oxidative
chemistry.
oxo complexes, but when the 6-substituents are present
the heterocycle is ejected and bis(m-oxo)dicopper com-
plex formation is observed (L6PhPy) or implicated (LQuin
and L6MePy). The trans-1,2-peroxo and bis(m-oxo) spe-
cies decompose differently to yield ligands that are
either oxygenated or N-dealkylated, respectively. These
results imply different pathways for the decay of the
structurally divergent species, but also may be explained
by invoking similar initial steps for the oxygenated
complexes (e.g. attack at ligand methylene CꢀH bond to
yield an ꢀOH group) but a different fate for the subse-
quent intermediate (ketone formation vs. N-dealkyla-
tion) due to inherent differences in the reactivity of the
ligands.
When compared to the ligands portrayed in Fig. 1,
the hybrid ligands described herein display behavior in
their copper(I)-small molecule chemistry that is interme-
diate between that seen when R3TACN and TMPA are
used. This behavior for the LP ligands can be traced to
their ability to adopt h4 or h3 coordination geometries,
which is, in turn, subject to the steric influences of the
heterocyclic ring substituents. Thus, our studies show
that ligand substituent size and shape are significant
determinants of the reactivity of Cu(I) compounds with
CO, NO, and O2 and provide further support for the
notion that such steric effects are of general importance
in metallobiochemistry and catalysis.
Although copperꢀdioxygen species could not be con-
clusively identified in low-temperature oxygenation re-
actions of either [L6MePyCu]SbF6 or [LQuinCu]SbF6,
similar ligand product ratios to those observed for the
decomposition of [(L6PhPyCu)2(m-O)2](SbF6)2 (ꢀ3:2
LP:LH, \85% total recovery; P=6MePy, Quin) were
obtained. Therefore, we propose that the L6PhPy, L6MePy
,
and LQuin systems all proceed through bis(m-
oxo)dicopper intermediate complexes which decompose
via a CꢀH activation process at the pyridyl- or quinolyl-
methyl position of the heterocyclic appendage. Our
inablity to observe and/or characterize the intermediate
bis(m-oxo)dicopper complexes in the L6MePy and LQuin
systems suggests that the thermal stability of this family
of complexes is highly dependent upon the nature of the
heterocyclic appendage, an effect that at present we do
not understand.
5. Supplementary material
VT 1H NMR data for [LPyCu]SbF6 (Fig. S1), UV–Vis
data for the reversible binding of CO to [LPyCu]SbF6
(Fig. S2), resonance Raman spectra for the trans-peroxo
complexes supported by LPy and L5MePy (Fig. S3), and
full reports of X-ray crystallographic data (22 pages).
4. Summary and conclusions
Acknowledgements
A series of Cu(I) complexes of hybrid ligands com-
prising 1,4-diisopropyl-1,4,7-triazacyclononane ap-
pended to variably substituted pyridyl or quinolyl
donors have been characterized and their reactions with
CO, NO, and O2 explored. With CO, displacement of
the heterocyclic arm occurs reversibly to yield adducts
with the ligand coordinated only through its macro-
cyclic N donor atoms in h3 fashion. Disproportionation
of NO is promoted by the Cu(I) complex of LPy,
yielding N2O and a Cu(II)ꢀNO2− complex in which the
nitrite is h1-O-bound and LPy is coordinated h4. The
reactions with O2 vary according to the specific hetero-
cyclic appendage, with steric effects of the heterocyclic
substituents being of paramount importance. Thus, the
pyridyl unit remains bound to the metal in the less
hindered cases (LPy and L5MePy) to yield trans-1,2-per-
We thank the National Institutes of Health
(GM47365 to W.B.T. and a postdoctoral fellowship to
L.M.B.), the National Science Foundation (National
Young Investigator Award to W.B.T.), and the Alfred
P. Sloan and Camille & Henry Dreyfus Foundations for
financial support. We also thank Dr Carole Toia for
assistance.
References
[1] Selected lead references: (a) A.L. Feig, S.J. Lippard, Chem. Rev.
94 (1994) 759. (b) B.J. Wallar, J.D. Lipscomb, Chem. Rev. 96
(1996) 2625. (c) L. Que Jr., R.Y.N. Ho, Chem. Rev. 96 (1996)
2607. (d) E.I. Solomon, U.M. Sundaram, T.E. Machonkin,
Chem. Rev. 96 (1996) 2563. (g) J.P. Klinman, Chem. Rev. 96
(1996) 2541.

128
L.M. Berreau et al. / Inorganica Chimica Acta 297 (2000) 115–128
[2] S. Moncada, J. Stamler, S. Gross, E.A. Higgs, (Eds.), The Biology
J.A. Halfen, V.G. Young Jr., W.B. Tolman, Inorg. Chem. 35
(1996) 6339. (c) M. Harata, K. Jitsukawa, H. Masuda, H. Einaga,
Bull. Chem. Soc. Jpn 71 (1998) 637.
of Nitric Oxide, Part 5, Portland Press, London, 1996.
[3] See for example: (a) V.I. Chen, A.M. Orville, M.R. Harpel, C.A.
Frolik, K.K. Surerus, E. Mu¨nck, J.D. Libscomb, J. Biol. Chem.
264 (1989) 21677. (b) B.J. Reedy, N.J. Blackburn, J. Am. Chem.
Soc. 116 (1994) 1924.
[14] N. Kitajima, K. Fujisawa, C. Fujimoto, Y. Moro-oka, S. Hashi-
moto, T. Kitagawa, K. Toriumi, K. Tatsumi, A. Nakamura, J.
Am. Chem. Soc. 114 (1992) 1277.
[4] K.D. Karlin, Science 261 (1993) 701.
[15] (a) K.A. Magnus, B. Hazes, H. Ton-That, C. Bonaventura, J.
Bonaventura, W.G.J. Hol, Proteins 19 (1994) 302. (b) K.A.
Magnus, H. Ton-That, J.E. Carpenter, Chem. Rev. 94 (1994) 727.
(c) E.I. Solomon, F. Tuczek, D.E. Root, C.A. Brown, Chem. Rev.
94 (1994) 827. (d) C. Eiken, F. Zippel, K. Bu¨ldt-Karentzpoulos,
B. Krebs, FEBS Lett. 436 (1998) 293.
[5] Reviews: (a) K.D. Karlin, S. Kaderli, A.D. Zu¨berbu¨hler, Acc.
Chem. Res. (1997) 139. (b) W.B. Tolman, Acc. Chem. Res. 30
(1997) 227. (c) S. Fox, K.D. Karlin, in: J.S. Valentine, C.S. Foote,
A. Greenberg, J.F. Liebman (Eds.), Active Oxygen in Biochem-
istry, Blackie Academic & Professional, Chapman & Hall, Glas-
gow, Scotland, 1995. (d) N. Kitajima, Y. Moro-oka, Chem. Rev.
94 (1994) 737. (e) K.D. Karlin, Z. Tyekla´r, A.D. Zuberbu¨hler, in:
J. Reedijk (Ed.), Bioinorganic Catalysis, Marcel Dekker, New
York, 1993. (f) T.N. Sorrell, Tetrahedron 40 (1989) 3. (g) K.D.
Karlin, Y. Gultneh, Prog. Inorg. Chem. 35 (1987) 219. (h) W.B.
Tolman, in: H.H. Thorp (Ed.), Synthetic Modeling of the Inter-
actions of Nitrogen Oxides with Copper Proteins, vol. 7, Ameri-
can Chemical Society, Washington, 1995.
[6] (a) K.D. Karlin, J.C. Hayes, J. Shi, J.P. Hutchinson, J. Zubieta,
Inorg. Chem. 21 (1982) 4106. (b) J. Zubieta, K.D. Karlin, J.C.
Hayes, in: K.D. Karlin, J. Zubieta (Eds.), Copper Coordination
Chemistry: Biochemical and Inorganic Perspectives, Adenine
Press, New York, 1983. (c) G. Anderegg, F. Wenk, Helv. Chim.
Acta 50 (1967) 2330.
[16] K. Fujisawa, M. Tanaka, Y. Moro-oka, N. Kitajima, J. Am.
Chem. Soc. 116 (1994) 12079.
[17] P.P. Paul, K.D. Karlin, J. Am. Chem. Soc. 113 (1991) 6331.
[18] (a) C.E. Ruggiero, S.M. Carrier, W.E. Antholine, J.W. Whittaker,
C.J. Cramer, W.B. Tolman, J. Am. Chem. Soc. 115 (1993) 11285.
(b) C.E. Ruggiero, S.M. Carrier, W.B. Tolman, Angew. Chem.,
Int. Ed. Engl. 33 (1993) 895. (c) J.L. Schneider, S.M. Carrier, C.E.
Ruggiero, V.G. Young Jr., W.B. Tolman, J. Am. Chem. Soc. 120
(1998) 11408.
[19] S. Mahapatra, J.A. Halfen, W.B. Tolman, J. Chem. Soc., Chem.
Commun. (1994) 1625.
[20] A pertinent example: M.S. Nasir, R.R. Jacobson, J. Zubieta, K.D.
Karlin, Inorg. Chim. Acta 203 (1993) 5.
[21] K.P. Wainwright, Coord. Chem. Rev. 166 (1997) 35.
[22] (a) J.A. Halfen, V.G. Young Jr., W.B. Tolman, J. Am. Chem. Soc.
118 (1996) 10920. (b) J.A. Halfen, Ph.D. Thesis, University of
Minnesota, 1996.
[7] (a) S. Trofimenko, Chem. Rev. 93 (1993) 943. (b) G. Parkin, Adv.
Inorg. Chem. 42 (1995) 291. (c) N. Kitajima, W.B. Tolman, Prog.
Inorg. Chem. 43 (1995) 419.
[8] K. Wieghardt, P. Chaudhury, Prog. Inorg. Chem. 35 (1988) 329.
[9] (a) T.N. Sorrell, W.E. Allen, P.S. White, Inorg. Chem. 34 (1995)
952. (b) W.E. Allen, T.N. Sorrell, Inorg. Chem. 36 (1997) 1732.
(c) W.E. Lynch, J.D.M. Kurtz, S. Wang, R.A. Scott, J. Am.
Chem. Soc. 116 (1994) 11030. (d) I. Sanyal, M. Mahroof-Tahir,
M.S. Nasir, P. Ghosh, B.I. Cohen, Y. Gultneh, R.W. Cruse, A.
Farooq, K.D. Karlin, S. Liu, J. Zubieta, Inorg. Chem. 31 (1992)
4322. (e) S. Itoh, H. Nakao, L.M. Berreau, T. Kondo, M.
Komatsu, S. Fukuzumi, J. Am. Chem. Soc. 120 (1998) 2890. (f)
H.V. Obias, Y. Lin, N.N. Murthy, E. Pidcock, E.I. Solomon, M.
Ralle, N.J. Blackburn, Y.M. Neubold, A.D. Zuberbu¨hler, K.D.
Karlin, J. Am. Chem. Soc. 120 (1998) 12960. (g) T.N. Sorrell,
D.L. Jameson, Inorg. Chem. 21 (1982) 1014. (h) V. Mahadevan,
Z. Hou, A.P. Cole, D.E. Root, T.K. Lal, E.I. Solomon, T.D.P.
Stack, J. Am. Chem. Soc. 119 (1997) 11996. (i) A.P. Cole, D.E.
Root, P. Mukherjee, E.I. Solomon, T.D.P. Stack, Science 273
(1996) 1848. (j) J.S. Thompson, J. Am. Chem. Soc. 106 (1984)
8308.
[23] J.A. Halfen, W.B. Tolman, Inorg. Synth. 34 (1998) 75.
[24] J.A. Halfen, S. Mahapatra, E.C. Wilkinson, A.J. Gengenbach,
V.G. Young Jr., L. Que Jr., W.B. Tolman, J. Am. Chem. Soc. 118
(1996) 763.
[25] A. Furstner, A. Ernst, Tetrahedron 51 (1995) 773.
[26] A. Nanthakumar, S. Fox, N.N. Murthy, K.D. Karlin, J. Am.
Chem. Soc. 119 (1997) 3898.
[27] C. Chuang, K. Lim, Q. Chen, J. Zubieta, J.W. Canary, Inorg.
Chem. 34 (1995) 2562.
[28] G.J. Kubas, Inorg. Synth. 19 (1979) 90.
[29] SHELXTL-PLUS Version 5.0, Siemens Industrial Automation,
Madison, WI.
[30] M. Melnik, L. Macaskova, C.E. Holloway, Coord. Chem. Rev.
126 (1993) 71.
[31] D.D. Perrin, Dissociation Constants of Organic Bases in Aqueous
Solutions, Supplement 1972, Butterworth, London, 1972.
[32] S. Mahapatra, J.A. Halfen, E.C. Wilkinson, G. Pan, X. Wang,
V.G. Young Jr., C.J. Cramer, L. Que Jr., W.B. Tolman, J. Am.
Chem. Soc. 118 (1996) 11555.
[10] Z. Tyekla´r, R.R. Jacobson, N. Wei, N.N. Murthy, J. Zubieta,
K.D. Karlin, J. Am. Chem. Soc. 115 (1993) 2677.
[33] N. Komeda, H. Nagao, Y. Kushi, G. Adachi, M. Suzuki, A.
Uehara, K. Tanaka, Bull. Chem. Soc. Jpn. 68 (1995) 581.
[34] J.A. Halfen, W.B. Tolman, Acta Crystallogr., Sect. C 51 (1995)
215.
[11] K.D. Karlin, N. Wei, B. Jung, S. Kaderli, A.D. Zuberbu¨hler, J.
Am. Chem. Soc. 113 (1991) 5868.
[12] N. Wei, N.N. Murthy, Q. Chen, J. Zubieta, K.D. Karlin, Inorg.
Chem. 33 (1994) 1953.
[35] S. Mahapatra, J.A. Halfen, W.B. Tolman, J. Am. Chem. Soc. 118
(1996) 11575.
[13] (a) M. Harata, K. Jitsukawa, H. Masuda, H. Einaga, J. Am.
Chem. Soc. 116 (1994) 10817. (b) L.M. Berreau, S. Mahapatra,
.