Effects of metal ions on physicochemical properties and redox reactivity of phenolates and phenoxyl radicals: Mechanistic insight into hydrogen atom abstraction by phenoxyl radical-metal complexes

Article abstract of DOI:10.1021/ja0036110

Phenolate and phenoxyl radical complexes of a series of alkaline earth metal ions as well as monovalent cations such as Na⁺ and K⁺ have been prepared by using 2,4-di-tert-butyl-6-(1,4,7,10-tetraoxa-13-aza-cyclopentadec-13-ylmethyl) phenol (L1H) and 2,4-di-tert-butyl-6-(1,4,7,10,13-pentaoxa-16-aza-cyclo-octadec-16-ylmethyl) phenol (L2H) to examine the effects of the cations on the structure, physicochemical properties and redox reactivity of the phenolate and phenoxyl radical complexes. Crystal structures of the Mg²⁺- and Ca²⁺-complexes of L^1- as well as the Ca²⁺- and Sr²⁺-complexes of L2^- were determined by X-ray crystallographic analysis, showing that the crown ether rings in the Ca²⁺-complexes are significantly distorted from planarity, whereas those in the Mg²⁺- and Sr²⁺-complexes are fairly flat. The spectral features (UV-vis) as well as the redox potentials of the phenolate complexes are also influenced by the metal ions, depending on the Lewis acidity of the metal ions. The phenoxyl radical complexes are successfully generated in situ by the oxidation of the phenolate complexes with (NH₄)₂[Ce⁴⁺(NO₃)₆] (CAN). They exhibited strong absorption bands around 400 nm together with a broad one around 600-900 nm, the latter of which is also affected by the metal ions. The phenoxyl radical-metal complexes are characterized by resonance Raman, ESI-MS, and ESR spectra, and the metal ion effects on those spectroscopic features are also discussed. Stability and reactivity of the phenoxyl radical-metal complexes are significantly different, depending on the type of metal ions. The disproportionation of the phenoxyl radicals is significantly retarded by the electronic repulsion between the metal cation and a generated organic cation (Ln⁺), leading to stabilization of the radicals. On the other hand, divalent cations decelerate the rate of hydrogen atom abstraction from 10-methyl-9,10-dihydroacridine (AcrH₂) and its 9-substituted derivatives (AcrHR) by the phenoxyl radicals. On the basis of primary kinetic deuterium isotope effects and energetic consideration of the electron-transfer step from AcrH₂ to the phenoxyl radical-metal complexes, we propose that the hydrogen atom abstraction by the phenoxyl radical-alkaline earth metal complexes proceeds via electron transfer followed by proton transfer.

Full text of DOI:10.1021/ja0036110

J. Am. Chem. Soc. 2001, 123, 2165-2175
2165
Effects of Metal Ions on Physicochemical Properties and Redox
Reactivity of Phenolates and Phenoxyl Radicals: Mechanistic Insight
into Hydrogen Atom Abstraction by Phenoxyl Radical-Metal
Complexes
,†
‡
§
,§
Shinobu Itoh,* Hideyuki Kumei, Shigenori Nagatomo, Teizo Kitagawa,* and
Shunichi Fukuzumi*,‡
Contribution from the Department of Chemistry, Graduate School of Science, Osaka City UniVersity,
3
-3-138 Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan, Department of Material and Life Science,
Graduate School of Engineering, Osaka UniVersity, CREST, Japan Science and Technology Corporation,
2
-1 Yamada-oka, Suita, Osaka 565-0871, Japan, and Institute for Molecular Science, Myodaiji,
Okazaki 444-8585, Japan
ReceiVed October 6, 2000
Abstract: Phenolate and phenoxyl radical complexes of a series of alkaline earth metal ions as well as
+
+
monovalent cations such as Na and K have been prepared by using 2,4-di-tert-butyl-6-(1,4,7,10-tetraoxa-
3-aza-cyclopentadec-13-ylmethyl)phenol (L1H) and 2,4-di-tert-butyl-6-(1,4,7,10,13-pentaoxa-16-aza-cyclo-
1
octadec-16-ylmethyl)phenol (L2H) to examine the effects of the cations on the structure, physicochemical
properties and redox reactivity of the phenolate and phenoxyl radical complexes. Crystal structures of the
2
+
2+
-
2+
2+
-
Mg - and Ca -complexes of L1 as well as the Ca - and Sr -complexes of L2 were determined by
2
+
X-ray crystallographic analysis, showing that the crown ether rings in the Ca -complexes are significantly
2
+
2+
distorted from planarity, whereas those in the Mg - and Sr -complexes are fairly flat. The spectral features
UV-vis) as well as the redox potentials of the phenolate complexes are also influenced by the metal ions,
depending on the Lewis acidity of the metal ions. The phenoxyl radical complexes are successfully generated
in situ by the oxidation of the phenolate complexes with (NH4)2[Ce (NO3)6] (CAN). They exhibited strong
absorption bands around 400 nm together with a broad one around 600-900 nm, the latter of which is also
affected by the metal ions. The phenoxyl radical-metal complexes are characterized by resonance Raman,
ESI-MS, and ESR spectra, and the metal ion effects on those spectroscopic features are also discussed. Stability
and reactivity of the phenoxyl radical-metal complexes are significantly different, depending on the type of
metal ions. The disproportionation of the phenoxyl radicals is significantly retarded by the electronic repulsion
(
4
+
+
between the metal cation and a generated organic cation (Ln ), leading to stabilization of the radicals. On the
other hand, divalent cations decelerate the rate of hydrogen atom abstraction from 10-methyl-9,10-dihydroacridine
(AcrH2) and its 9-substituted derivatives (AcrHR) by the phenoxyl radicals. On the basis of primary kinetic
deuterium isotope effects and energetic consideration of the electron-transfer step from AcrH2 to the phenoxyl
radical-metal complexes, we propose that the hydrogen atom abstraction by the phenoxyl radical-alkaline
earth metal complexes proceeds via electron transfer followed by proton transfer.
Introduction
activation of the substrate, which is the initial step of the above
enzymatic reactions. Transition-metal ions bound to a tyrosine
1
Protein radicals are now well recognized to play a crucial
role in several biologically important redox processes. In
particular, organic radicals derived from tyrosine and its
derivatives have been reported to participate in a variety of
enzymatic catalysis such as in class I ribonucleotide reductase
1
radical have also been proven to enhance the radical stability
as well as to control reactivity of the phenoxyl radical species.
One of the most well-documented examples of such systems is
7
galactose oxidase, where a tyrosine radical coordinated to
copper(II) acts as an active species for the oxidation of primary
(RNR), photosystem II, prostaglandin H synthase, galactose
8
alcohols to the corresponding aldehydes. Extensive efforts have
oxidase, cytochrome c oxidase, bovine liver catalase, DNA
photolyase, and so on.1-6 In these cases, the tyrosine radical
so far been made not only to mimic the biochemical reactivity
of galactose oxidase but also to provide valuable insight into
mainly acts as a hydrogen atom acceptor to induce C-H bond
†
Department of Chemistry, Graduate School of Science, Osaka City
(3) Gerfen, G. J.; Bellew, B. F.; Griffin, R. G.; Singel, D. J.; Ekberg, C.
A.; Whittaker, J. W. J. Phys. Chem. 1996, 100, 16739-16748.
(4) MacMillan, F.; Kannt, A.; Behr, J.; Prisner, T.; Michel, H. Biochem-
istry 1999, 38, 9179-9184.
University.
‡
Department of Material and Life Science, Graduate School of Engineer-
ing, Osaka University.
§
Institute for Molecular Science.
(5) Ivancich, A.; Jouve, H. M.; Gaillard, J. J. Am. Chem. Soc. 1996,
118, 12852-12853.
(
1) Stubbe, J.; van der Donk, W. A. Chem. ReV. 1998, 98, 705-762 and
references therein.
2) Babcock, G. T.; Espe, M.; Hoganson, C.; Lydakis-Simantiris, N.;
(6) Aubert, C.; Brettel, K.; Mathis, P.; Eker, A. P. M.; Boussac, A. J.
Am. Chem. Soc. 1999, 121, 8659-8660.
(
McCracken, J.; Shi, W.; Styring, S.; Tommos, C.; Warncke, K. Acta Chem.
Scand. 1997, 51, 533-540.
(7) Ito, N.; Phillips, S. E. V.; Stevens, C.; Ogel, Z. B.; McPherson, M.
J.; Keen, J. N.; Yadav, K. D. S.; Knowles, P. F. Nature 1991, 350, 87-90.
1
0.1021/ja0036110 CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/14/2001

2166 J. Am. Chem. Soc., Vol. 123, No. 10, 2001
Itoh et al.
the general aspects of structures, physicochemical properties,
and functions of phenoxyl radical complexes with a series of
transition-metal ions.9-16
The essential roles of metal ions have also been well
recognized in a variety of enzymatic functions.17-19 A number
of attempts are being made not only to detect but also to
chemically and functionally characterize trace elements or
otherwise inconspicuous metal ions in various enzymes. The
catalytic role of alkaline earth metal ions in enzymatic redox
reactions has particularly merited recent attention, since a
growing number of redox enzymes have been demonstrated to
involve Ca2 at their active sites, where Ca is located near
+
2+
20-24
the redox cofactor.
Although very little is known about the
catalytic roles of Ca2 in such systems, one can assume that
the interaction of the redox cofactor with the cationic species
results in enhancement of the oxidation ability of the enzymes.
In this context, we have recently demonstrated that the oxidation
of primary alkyl alcohols is made possible by coenzyme PQQ
+
(
pyrroloquinolinequinone) of quinoprotein alcohol and glucose
Figure 1. Spectral change observed upon addition of Ca(ClO
4
)
2
2
‚4H O
23,24
2+
-4
dehydrogenases,
system.
of the quinone and stability of the reaction intermediates.
when it coordinates to Ca in the model
It has been shown that Ca enhances electrophilicity
into a CH
3
CN solution of L1H (5.0 × 10 M) in the presence of
-4
25,26
2+
3
Et N (5.0 × 10 M), Inset: plot of the absorbance change at 309 nm
against the concentration of Ca2 added.
+
2
6
However, little is known about the effects of alkaline earth metal
ions on the redox reactivity of organic radicals.
2
+
with the divalent metal ions, particularly with Ca . The effects
of metal ions on the oxidation ability of phenoxyl radicals are
also examined by using 10-methyl-9,10-dihydroacridine (AcrH2)
and its 9-substituted derivatives (AcrHR) as substrates, providing
valuable mechanistic insight into the C-H bond activation by
the phenoxyl radical-metal complexes. This study will eventu-
ally stimulate the use of radical-metal complexes also in
nonbiological redox reactions such as metal ion-catalyzed
In this study, we have systematically investigated the effects
of cationic species such as alkaline metal and alkaline earth
metal ions on the structures, physicochemical properties, and
redox reactivity of phenolate and phenoxyl radical species by
using phenol derivatives L1H and L2H containing 1-aza-15-
crown-5-ether and 1-aza-18-crown-6-ether group as the metal
ion binding site, respectively.2
7,28
It is shown that the stability
2
9
electron-transfer reactions of organic radicals.
of phenoxyl radical species is greatly enhanced by the interaction
(
8) Whittaker, M. M.; Ballou, D. P.; Whittaker, J. W. Biochemistry 1998,
7, 8426-8436, and references therein.
9) Itoh, S.; Taki, M.; Fukuzumi, S. Coord. Chem. ReV. 2000, 198, 3-20,
and references therein.
10) Halfen, J. A.; Jazdzewski, B. A.; Mahapatra, S.; Berreau, L. M.;
Wilkinson, E. C.; Que, L., Jr.; Tolman, W. B. J. Am. Chem. Soc. 1997,
3
(
(
1
19, 8217-8227.
11) Wang, Y.; Stack, T. D. P. J. Am. Chem. Soc. 1996, 118, 13097-
3098.
12) Hockertz, J.; Steenken, S.; Wieghardt, K.; Hildebrandt, P. J. Am.
Chem. Soc. 1993, 115, 11222-11230.
13) Sokolowski, A.; Bothe, E.; Bill, E.; Weyherm u¨ ller, T.; Wieghardt,
K. Chem. Commun. 1996, 1671-1672.
14) Adam, B.; Bill, E.; Bothe, E.; Goerdt, B.; Haselhorst, G.; Hilden-
(
1
(
Results and Discussion
Synthesis and Characterization of Phenolate-Metal Com-
plexes. The phenol derivatives containing 1-aza-15-crown-5-
ether (L1H) and 1-aza-18-crown-6-ether group (L2H) have been
prepared by Mannich reaction on 2,4-di-tert-butylphenol with
the corresponding aza crown ether and paraformaldehyde in
refluxing ethanol. Ligands L1H and L2H exhibit an absorption
band at 284 and 283 nm in CH3CN, respectively, which shifted
toward the longer-wavelength region when an alkaline earth
metal ion was added into the solution containing the ligand and
triethylamine as a base. Such a red-shift of the absorption band
is indicative of formation of the phenolate derivatives. A typical
(
(
brand, K.; Sokolowski, A.; Steenken, S.; Weyherm u¨ ller, T.; Wieghardt, K.
Chem. Eur. J. 1997, 3, 308-319.
(15) Sokolowski, A.; M u¨ ller, J.; Weyherm u¨ ller, T.; Schnepf, R.; Hilde-
brandt, P.; Hildenbrand, K.; Bothe, E.; Wieghardt, K. J. Am. Chem. Soc.
997, 119, 8889-8900.
16) Sokolowski, A.; Adam, B.; Weyherm u¨ ller, T.; Kikuchi, A.; Hilden-
brand, K.; Schnepf, R.; Hildebrandt, P.; Bill, E.; Wieghardt, K. Inorg. Chem.
1
(
1
997, 36, 3702-3710.
17) Hughes, M. N. The Inorganic Chemistry of Biological Processes,
nd ed.; John Wiley & Sons: New York, 1984.
18) Kaim, W. Schwederski, B. Bioinorganic Chemistry: Inorganic
Elements in the Chemistry of Life; John Wiley & Sons: New York, 1994.
19) Lippard, S. J.; Berg, J. M. Principles of Bioinorganic Chemistry;
University Science Books: Mill Valley, California, 1994.
20) Limburg, J.; Szalai, V. A.; Brudvig, G. W. J. Chem. Soc., Dalton
Trans. 1999, 1353-1361 and references therein.
21) Einsle, O.; Messerschmidt, A.; Stach, P.; Bourenkov, G. P.; Bartunik,
H. D.; Huber, R.; Kroneck, P. M. H. Nature 1999, 400, 476-480.
22) Riistama, S.; Laakkonen, L.; Wikstr o¨ m, M.; Verkhovsky, M. I.;
Puustinen, A. Biochemistry 1999, 38, 10670-10677.
23) Xia, Z.-X.; Dai, W.-W.; Zhang, Y.-F.; White, S. A.; Boyd, G. D.;
Mathews, F. S. J. Mol. Biol. 1996, 259, 480-501.
24) Ghosh, M.; Anthony, C.; Harlos, K.; Goodwin, M. G.; Blake, C.
Structure 1995, 3, 177-187.
25) Itoh, S.; Kawakami, H.; Fukuzumi, S. J. Am. Chem. Soc. 1997, 119,
(
2
(
2
+
example of the titration of L1H with Ca is shown in Figure
1, where the absorption band at 309 nm due to the Ca -complex
2+
(
(
(27) To distinguish the different states of the phenol moiety of the ligands
-
•
more clearly, the symbols of LnH, Ln , and Ln (n ) 1 or 2) are used to
denote the phenol, phenolate, and phenoxyl radical forms, respectively.
(28) Modulation of redox properties of organic cofactors such as flavin
by noncovalent weak interactions have been extensively studied by Rotello
et al.; Niemz, A.; Rotello, V. M. Acc. Chem. Res. 1999, 32, 44-52 and
references therein.
(29) (a) Fukuzumi, S. Bull. Chem. Soc. Jpn. 1997, 70, 1-28. (b)
Fukuzumi, S. In AdVances in Electron-Transfer Chemistry; Mariano, P. S.,
Ed.; JAI Press: Greenwich, CT, 1992; Vol. 2, pp. 67-175. (c) Fukuzumi,
S.; Itoh, S. In AdVances in Photochemistry; Neckers, D. C., Volman, D.
H., von B u¨ nau, G., Eds.; Wiley: New York, 1998; Vol. 25, pp 107-172.
(d) Fukuzumi, S. In Electron Transfer in Chemistry; Balzani, V., Ed.; Wiley-
VCH: Weinheim, 2000; Vol. 5 in press.
(
(
(
(
(
4
39-440.
(
26) Itoh, S.; Kawakami, H.; Fukuzumi, S. Biochemistry 1998, 37, 6562-
6
571.

H-Atom Abstraction by Phenoxyl Radical-Metal Complexes
Table 1. Spectral Data of the Phenolate Complexes Generated in
J. Am. Chem. Soc., Vol. 123, No. 10, 2001 2167
Table 2. Summary of X-ray Crystallographic Data
the Titration of L1H or L2H with M(ClO ) (M ) Alkaline Earth
4 2
-
a,b
[Ca(L1 )(CH3OH)2]-
3
Metal Ions) in the Presence of Triethylamine in CH CN
-
compound
[Mg(L1 )(CH3OH)]BPh4
BPh4‚2CH3OH
1
λ
max/nm (ꢀ/M- cm-1
)
empirical formula
formula weight
crystal system
space group
a, Å
C50H66NO6BMg
812.19
C53H78NO9BCa
924.09
monoclinic
C2/c (no. 15)
35.486(2)
metal ion
L1-
L2-
2
+
triclinic
Mg
304 (4000)
309 (3940)
310 (4160)
311 (3460)
320 (5520)
-
301 (4400)
306 (4820)
308 (4840)
310 (3560)
2
+
P 1h (no. 2)
13.2719(9)
15.321(1)
12.3308(7)
108.050(3)
94.392(2)
99.524(2)
2329.1(3)
2
Ca
2
+
Sr
2
+
b, Å
17.9443(9)
17.8958(8)
Ba
Na
+
a
c, Å
-
+
a
312 (3100)c
R, deg
â, deg
K
n
+ b
104.803(1)
N( Bu)
4
317 (3740)
317 (4380)
γ, deg
a
V, Å3
Since addition of NaClO
4
or KClO
4
into a CH
3
CN solution of the
11017.3(9)
phenol containing triethylamine resulted in little change of the UV-
Z
8
+
-
+
vis spectrum (see text), the Na -complex of L1 and the K -complex
F(000)
876.00
1.158
23
4000.00
1.114
23
-
Dcalc, g/cm3
T, °C
of L2 were prepared by the reaction of the ligands (L1H and L2H)
and the metal hydrides, NaH and KH, respectively (see Experimental
Section). The spectral data are of the isolated products. b The tetra-
crystal size, mm
µ (Mo KR), cm
diffractometer
00 .. 83 60 × 0.30 × 0.10
Rigaku RAXIS-RAPID
Imaging Plate
Mo KR (0.71069 Å)
55.0
10 .. 63 40 × 0.30 × 0.30
Rigaku RAXIS-RAPID
Imaging Plate
Mo KR (0.71069 Å)
55.0
-
1
n-butylammonium salts of the phenolate were generated in situ by
n
adding tetra-n-butylammonium hydroxide [N( Bu)
4
OH] to the phenol.
c
-
Because of low solubility of [K(L2 )] in CH
taken in CH CN containing 10% CH OH.
3
CN, the spectrum was
3
3
radiation
θmax, deg
2
no. of reflns measd 20277
49660
12570 [I > -10.00σ(I)]
814
0.189; 0.180
0.076
increases with a concomitant decrease in the absorption band
at 284 nm of L1H with clear isosbestic points at 271 and 288
nm. Since the spectral change is completed by the addition of
no. of reflns obsd
no. of variables
9839 [I > -10.00σ(I)]
784
a
b
R ; Rw
0.183; 0.163
0.076
2+
just one equivalent of Ca (see inset of Figure 1), a 1:1-complex
c
R1
-
2+
27
between L1 and Ca is formed effectively. Spectral data
for the titration of the ligands with a series of alkaline earth
metal ions are summarized in Table 1. It should be noted that
-
-
compound
[Ca(L2 )(H2O)]BPh4
[Sr(L2 )(H2O)]BPh4
empirical formula
formula weight
crystal system
space group
a, Å
C51H68NO7BCa
857.99
C51H68NO7BSr
905.53
2
+
addition of NaClO4 instead of divalent cations such as Ca
monoclinic
P21/n (no. 14)
16.7447(6)
18.4217(7)
17.3976(5)
116.619(1)
4797.7(3)
4
monoclonic
P21/n (no. 14)
19.3828(9)
17.9403(8)
30.569(1)
107.3743(8)
10153.8(8)
8
hardly affected the spectrum of the ligand under otherwise the
same experimental conditions. This indicates that association
of the monovalent cation and the phenol derivative is signifi-
cantly small as compared to that with the divalent cations.
The λmax values of the phenolate-metal complexes are blue-
b, Å
c, Å
â, deg
V, Å3
+
2+
shifted in going from Na to Mg . The stronger Lewis acid
Z
2
+
such as Mg may attract the negative charge of the phenolate
oxygen more strongly than the weaker acids, making the
transition energy of the π-π* of the phenolate ring higher.
F(000)
1848.00
1.188
3888.00
1.198
23
Dcalc, g/cm3
T, °C
23
2+
-
Molecular orbital calculations on the Mg -complex of L1 by
a ZINDO program could reproduce the UV-vis spectrum,
where the calculated value of λmax ) 312 nm is comparable
with the experimental value (λmax ) 304 nm in Table 1). This
absorption band corresponds to the π-π* (HOMO-to-LUMO)
transition.30 The calculated π-π* transition energies of the Na -
crystal size, mm
µ (Mo KR), cm
diffractometer
10 .. 83 10 × 0.30 × 0.10
Rigaku RAXIS-RAPID Rigaku RAXIS-RAPID
0.25 × 0.25 × 0.10
-
1
11.09
Imaging Plate
Mo KR (0.71069 Å)
55.0
Imaging Plate
Mo KR (0.71069 Å)
55.0
21959
5712 [I > 2.00σ(I)]
1108
radiation
2θmax, deg
+
no. of reflns measd 41971
-
no. of reflns obsd
no. of variables
10775 [I > 0.00σ(I)]
complex and the metal-free form of L1 are lower in energy
750
and thereby the calculated λmax values for both species are red-
a
b
R ; Rw
0.162; 0.170
0.065
0.194; 0.248
0.113
shifted (330 and 349 nm, respectively) as compared to that of
c
R1
2
+
-
the Mg -complex of L1 . The result of the MO calculation is
a
2
2)/∑F 2
. b
| - |F
2
- F 2
)2/∑ω(F
2))1/2;
o
R ) ∑(F
o
- F
c
c
o
R
w
) (∑ω(F
||/∑|F |.
o
c
consistent with the experimental observation shown in Table
2
ω ) 1/σ (F
o
). R1 ) ∑||F
o
c
o
3
1
1
.
The 1:1-complexes formed between L1- (and L2 ) and a
-
-
(H2O)]BPh4, and [Sr(L2 )(H2O)]BPh4 are shown in Figure 2a-
series of alkaline earth metal ions were successfully isolated as
d, respectively. As demonstrated by the UV-vis titration, each
complex has one metal ion held in the crown ether cavity, which
is also strongly associated with the phenolate oxygen. Each
complex has additional external ligands such as CH3OH in the
-
2+
-
BPh4 salts except for the Mg -complex of L2 . The structures
2
+
2+
-
of the Mg - and Ca -complexes of L1 as well as the
2+
2+
-
Ca - and Sr -complexes of L2 were determined by X-ray
crystallography. The crystallographic data and selected bond
distances and angles for the X-ray structures are summarized
in Tables 2 and 3. ORTEP drawings of the cationic parts of
2
+
2+
-
2+
Mg - and Ca -complexes of L1 and H2O in the Ca - and
2
+
-
Sr -complexes of L2 , making the coordination number of the
metal center seven or eight.
-
-
-
[
Mg(L1 )(CH3OH)]BPh4, [Ca(L1 )(CH3OH)2]BPh4, [Ca(L2 )-
2
+
-
The Mg -complex ([Mg(L1 )(CH3OH)]BPh4) in Figure 2a
has a nearly perfect pentagonal bipyramidal structure in which
four oxygen atoms [O(2), O(3), O(4), and O(5)] and one tertiary
amine nitrogen atom [N(1)] of the crown ether ring are placed
in the pentagonal plane, and the phenolate and methanol oxygen
atoms [O(1) and O(6)] coordinate to the metal ion in the opposite
30) The ZINDO calculation was carried out on a Mg2+ complex of L1-
(
containing one methanol molecule as an external ligand based on the X-ray
structure of complex shown in Figure 2a.
(
31) The general trend of λmax depending on different metal ions can be
well reproduced by the ZINDO calculations, although the calculated λmax
values are always a little larger than the experimental values.

2168 J. Am. Chem. Soc., Vol. 123, No. 10, 2001
Itoh et al.
Table 3. Selected Bond Lengths (Å) and Angles (deg)a
-
[
3 4
Mg(L1 )(CH OH)]BPh
Mg(1)-O(1)
Mg(1)-O(3)
1.957(2)
2.236(2)
Mg(1)-O(2)
Mg(1)-O(4)
2.158(2)
2.181(2)
Mg(1)-O(5)
Mg(1)-N(1)
2.207(2)
2.290(2)
Mg(1)-O(6)
2.113(2)
O(1)-Mg(1)-O(2)
O(1)-Mg(1)-O(4)
95.97(8) O(1)-Mg(1)--O(3)
91.32(8) O(1)-Mg(1)-O(5)
94.07(7) O(3)-Mg(1)-O(5) 140.42(8) O(3)-Mg(1)-O(6)
101.07(8) O(3)-Mg(1)-N(1) 143.25(8) O(4)-Mg(1)-O(5)
86.41(8)
71.89(8)
O(1)-Mg(1)-O(6) 177.73(9) O(1)-Mg(1)-N(1)
88.50(7) O(4)-Mg(1)-O(6)
141.53(9) O(5)-Mg(1)-O(6)
86.28(9) O(6)-Mg(1)-N(1)
71.41(8)
86.73(9) O(4)-Mg(1)-N(1) 145.26(9)
O(2)-Mg(1)-O(3)
70.42(8) O(2)-Mg(1)-O(4)
77.23(8) O(5)-Mg(1)-N(1)
92.44(8) Mg(1)-O(1)-C(1)
74.08(8)
126.1(2)
O(2)-Mg(1)-O(5) 142.15(9) O(2)-Mg(1)-O(6)
O(2)-Mg(1)-N(1)
72.85(8) O(3)-Mg(1)-O(4)
-
[
Ca(L1 )(CH
3
OH)
2
]BPh
Ca(1)-O(5)
Ca(1)-O(7)
4
3
‚2CH OH
Ca(1)-O(1)
Ca(1)-O(3)
2.315(2)
2.607(2)
Ca(1)-O(2)
Ca(1)-O(4)
2.513(2)
2.554(2)
2.462(2)
2.450(2)
Ca(1)-O(6)
Ca(1)-N(1)
2.444(3)
2.623(3)
O(1)-Ca(1)-O(2)
O(1)-Ca(1)-O(4)
O(1)-Ca(1)-O(6)
O(1)-Ca(1)-N(1)
O(2)-Ca(1)-O(4)
O(2)-Ca(1)-O(6)
O(2)-Ca(1)-N(1)
O(3)-Ca(1)-O(5)
140.22(7)
104.72(7)
79.80(7)
77.49(7)
111.11(7)
124.88(8)
66.12(7)
104.20(7)
O(1)-Ca(1)-O(3)
O(1)-Ca(1)-O(5)
O(1)-Ca(1)-O(7)
O(2)-Ca(1)-O(3)
O(2)-Ca(1)-O(5)
O(2)-Ca(1)-O(7)
O(3)-Ca(1)-O(4)
O(3)-Ca(1)-O(6)
153.41(8)
90.28(7)
85.81(7)
63.42(7)
89.30(7)
74.74(7)
63.37(7)
74.29(7)
O(3)-Ca(1)-O(7)
O(4)-Ca(1)-O(5)
O(4)-Ca(1)-O(7)
O(5)-Ca(1)-O(6)
O(5)-Ca(1)-N(1)
O(6)-Ca(1)-N(1)
Ca(1)-O(1)-C(1)
92.36(8)
65.40(7)
145.10(8)
136.08(8)
66.07(7)
148.24(8)
131.7(2)
O(3)-Ca(1)-N(1)
O(4)-Ca(1)-O(6)
O(4)-Ca(1)-N(1)
O(5)-Ca(1)-O(7)
O(6)-Ca(1)-O(7)
O(7)-Ca(1)-N(1)
128.68(8)
75.94(7)
131.41(8)
149.00(8)
73.35(8)
83.09(9)
-
[
Ca(L2 )(H
2
O)]BPh
4
Ca(1)-O(1)
Ca(1)-O(3)
2.245(2)
2.448(3)
Ca(1)-O(2)
Ca(1)-O(4)
2.506(2)
2.548(2)
Ca(1)-O(5)
Ca(1)-O(7)
2.582(3)
2.356(3)
Ca(1)-O(6)
Ca(1)-N(1)
2.479(3)
2.574(3)
O(1)-Ca(1)-O(2)
O(1)-Ca(1)-O(4)
O(1)-Ca(1)-O(6)
O(1)-Ca(1)-N(1)
O(2)-Ca(1)-O(4)
O(2)-Ca(1)-O(6)
O(2)-Ca(1)-N(1)
O(3)-Ca(1)-O(5)
93.17(8)
86.21(8)
144.75(9)
79.12(9)
129.83(9)
80.75(9)
66.68(9)
81.87(10)
O(1)-Ca(1)-O(3)
O(1)-Ca(1)-O(5)
O(1)-Ca(1)-O(7)
O(2)-Ca(1)-O(3)
O(2)-Ca(1)-O(5)
O(3)-Ca(1)-O(7)
O(3)-Ca(1)-O(4)
O(3)-Ca(1)-O(6)
94.13(9)
149.04(9)
90.30(9)
O(3)-Ca(1)-O(7)
O(4)-Ca(1)-O(5)
O(4)-Ca(1)-O(7)
O(5)-Ca(1)-O(6)
O(5)-Ca(1)-N(1)
O(6)-Ca(1)-N(1)
Ca(1)-O(1)-C(1)
114.95(9)
64.13(9)
80.34(9)
60.79(9)
126.36(9)
66.53(9)
132.3(2)
O(3)-Ca(1)-N(1)
O(4)-Ca(1)-O(6)
O(4)-Ca(1)-N(1)
O(5)-Ca(1)-O(7)
O(6)-Ca(1)-O(7)
O(7)-Ca(1)-N(1)
130.38(9)
124.17(9)
158.88(9)
76.76(9)
79.21(9)
84.60(9)
64.69(9)
112.21(9)
149.78(9)
65.33(9)
113.73(10)
-
[
Sr(L2 )(H
2
O)]BPh
4
molecule 1
molecule 2
molecule 1
molecule 2
Sr(1)-O(1)
Sr(1)-O(2)
Sr(1)-O(3)
Sr(1)-O(4)
2.36(1)
2.63(2)
2.55(2)
2.66(2)
2.38(1)
2.70(1)
2.73(1)
2.69(1)
Sr(1)-O(5)
Sr(1)-O(6)
Sr(1)-O(7)
Sr(1)-N(1)
2.63(1)
2.64(1)
2.59(1)
2.76(2)
2.65(1)
2.68(1)
2.61(1)
2.77(1)
molecule 1
molecule 2
molecule 1
molecule 2
O(1)-Sr(1)-O(2)
O(1)-Sr(1)-O(3)
O(1)-Sr(1)-O(4)
O(1)-Sr(1)-O(5)
O(1)-Sr(1)-O(6)
O(1)-Sr(1)-O(7)
O(1)-Sr(1)-N(1)
O(2)-Sr(1)-O(3)
O(2)-Sr(1)-O(4)
O(2)-Sr(1)-O(5)
O(2)-Sr(1)-O(6)
O(2)-Sr(1)-O(7)
O(2)-Sr(1)-N(1)
O(3)-Sr(1)-O(4)
O(3)-Sr(1)-O(5)
96.1(5)
94.4(5)
92.9(5)
101.8(4)
108.8(4)
177.1(5)
73.8(5)
61.6(6)
120.7(6)
161.9(5)
109.7(5)
83.5(6)
62.6(5)
59.3(6)
119.0(5)
119.8(4)
127.9(4)
96.7(5)
96.2(5)
84.5(5)
157.6(5)
72.3(4)
59.8(4)
121.6(5)
141.8(5)
108.4(5)
68.5(4)
64.2(4)
61.8(5)
109.1(5)
O(3)-Sr(1)-O(6)
O(3)-Sr(1)-O(7)
O(3)-Sr(1)-N(1)
O(4)-Sr(1)-O(5)
O(4)-Sr(1)-O(6)
O(4)-Sr(1)-O(7)
O(4)-Sr(1)-N(1)
O(5)-Sr(1)-O(6)
O(5)-Sr(1)-O(7)
O(5)-Sr(1)-N(1)
O(6)-Sr(1)-O(7)
O(6)-Sr(1)-N(1)
O(7)-Sr(1)-N(1)
Sr(1)-O(1)-C(1)
156.3(5)
82.9(6)
121.0(6)
61.5(5)
122.1(5)
84.8(5)
166.7(5)
61.8(5)
78.7(5)
120.0(5)
74.0(5)
63.8(5)
108.5(5)
139(1)
147.6(5)
74.5(5)
122.6(4)
60.5(5)
119.9(5)
95.0(5)
168.6(5)
59.7(5)
73.2(5)
122.3(5)
73.2(5)
62.9(4)
96.3(5)
140(1)
a
Estimated standard deviations are give in parentheses.
axial direction from each other. The bond length between Mg2+
and the phenolate oxygen O(1) (1.957 Å) is significantly shorter
than that with other neutral oxygen and nitrogen atoms (2.11-
oxygen atoms [O(2) and O(3)] and one tertiary amine nitrogen
[N(1)] from the crown ether, alcoholic oxygen [O(6)] of the
external ligand methanol, and the phenolate oxygen [O(1)]. The
calcium ion sitting in the center of the pentagonal plane is
coordinated by the remaining oxygen atoms [O(4) and O(5)]
of the crown ether from the same side of the plane, and the
opposite side of the pentagonal plane is occupied by another
methanol molecule [O(7)] (Figure 2b). The bond length between
the metal ion and the oxygen atoms of the crown ether ring
[O(2)-O(5)] is elongated (2.462-2.607 Å) as compared to the
2
+
+
2
.29 Å), demonstrating the strong interaction between Mg
2
and the phenolate. The crown ether ring is almost flat, and Mg
is located nearly at the center of the pentagonal crown ether
ring.
2+
-
-
The structure of the Ca -complex of L1 ([Ca(L1 )(CH3-
OH)2]BPh4) in Figure 2b is quite different from the structure
2+
-
of the corresponding Mg -complex, [Mg(L1 )(CH3OH)]BPh4
shown in Figure 2a. In this case, coordination number of the
metal center is eight, and the pentagonal plane consists of two
2+
32
reported value for eight-coordinate Ca -complexes (∼2.45 Å).
As a result, the crown ether ring is significantly distorted from

H-Atom Abstraction by Phenoxyl Radical-Metal Complexes
J. Am. Chem. Soc., Vol. 123, No. 10, 2001 2169
Table 4. Redox Potentials (E1/2 vs SCE/V) of the Phenolate
-
-
3
Complexes of L1 and L2 in CH CN Containing 0.1 M
N( Bu) PF
4 6
n
cation
L1-
L2-
Mg2
+
0.59
0.41
0.38
0.33
0.09
-
-
2
+
Ca
0.47
0.36
0.28
-
2
+
Sr
2
+
Ba
Na
+
+
K
0.22
-0.21
n
+
4
-0.14a
a
N( Bu)
a
Determined by SHACV.
2
+
the Ca -complexes described above can be attributed to the
2+
nature of coordination structure of Ca , but to neither the metal
ion charge nor the radius. Such a structural uniqueness of Ca2
may be related to its role as signal transduction elements in
biological systems, although the origin has yet to be clarified.3
Change in the Redox Potentials of Phenolates by the
Complexation with Metal Ions. The alkaline earth metal
complexes of the phenolate derivatives afforded reversible redox
couples in the cyclic voltammetric (CV) measurements in
CH3CN (see Figure S1), and the E1/2 values determined by the
CV measurements are summarized in Table 4. The well-defined
reversibility in the redox couples shown in Figure S1 demon-
strates that the phenoxyl radical-metal complexes are fairly
stable at ambient temperature. On the other hand, reversibility
in CVs of the phenoxyl radical complexes becomes poor upon
going from the divalent metal complexes to the monovalent
alkaline metal complexes and totally irreversible voltammo-
grams were obtained in the case of tetra-n-butylammonium salts
of the phenolate. Thus, the E1/2 values of the tetra-n-butyl-
ammonium salts were determined by the second harmonic ac
+
4-37
-
Figure 2. ORTEP drawing of the cationic part of (a) [Mg(L1 )-
-
-
(
CH
and (d) [Sr(L2 )(H
are omitted for clarity.
3
OH)]BPh
4
-
, (b) [Ca(L1 )(CH
3
OH)]BPh
4 2 4
, (c) [Ca(L2 )(H O)]BPh ,
O)]BPh . The counteranions and hydrogen atoms
2
4
38
voltammetry (SHACV) measurements (Figures S2 and S3).
2
+
planarity. This kind of structural uniqueness of the Ca -
complex is not simply attributed to the difference of the metal
ion radius as discussed below.
-
-
The E1/2 values of the cation complexes of L1 or L2 in
n
+
+
+
2+
Table 4 increase in order: N( Bu)4 < K (Na ) < Ba
<
2+
2+
2+
Sr < Ca < Mg . This order agrees with the Lewis acidity
2+
The unique structure of the Ca -complex is also seen in the
complex of L2 . As in the case of L1 -complex, the metal
center consists of a pentagonal plane with three oxygen atoms
of cationic species derived from the superoxide ion complexes
-
-
39
with cationic species. Thus, the stronger the Lewis acidity of
cationic species, the more positive the E1/2 value, that is, the
stronger the oxidation ability of the complexes. On the other
hand, the effects of the macrocyclic ring size on the redox
[
O(4), O(5), and O(6)], one tertiary amine nitrogen atom [N(1)],
and the phenolate oxygen [O(1)] of the ligand. One side of the
pentagonal plane is also occupied by the remaining oxygen
atoms [O(2) and O(3)] of the crown ether ring, whereas the
opposite side of the pentagonal plane is filled by the external
ligand H2O [O(7)]. In this case as well, the crown ether ring is
bent about 90° along the N(1)-O(4) axis, and the bond length
-
-
potential (L1 vs L2 ) is not so evident.
Characterization of Phenoxyl Radical-Metal Complexes.
-
+
The one-electron oxidation of [Ca(L1 )(CH3OH)2] by (NH4)2-
[
4+
Ce (NO3)6] (CAN) (E1/2 ) 1.0 V vs SCE) in CH3CN resulted
in a spectral change shown in Figure 3. The absorption band at
09 nm due to the phenolate complex immediately decreases
2
+
between Ca and the oxygen atoms of the crown ether ring
3
[
O(2)-O(6)] is elongated (2.448-2.582 Å) as compared to the
together with a concomitant increase in the new absorption
bands at 387, 404, and 682 nm at 25 °C. Similar spectral change
3
2
reported values of eight-coordinated calcium ion (∼2.45 Å).
The external ligand H2O, on the other hand, binds to the metal
(
34) Sigel, H. Calcium and Its Role in Biology, Metal Ions in Biological
Systems; Marcel Dekker: New York, 1984; Vol. 17.
35) Spiro, T. G. Calcium in Biology, Metal Ions in Biology; Wiley-
center fairly strongly (Ca-O(7) ) 2.356 Å).
2+
-
The structure of the Sr -complex of L2 looks rather normal,
even though the ionic radius of Sr (1.12 Å) is larger than
(
2+
Interscience: New York, 1983; Vol. 6.
2
+
33
2+
that of Ca (0.99 Å). The Sr -complex has a distorted
hexagonal bipyramidal structure in which five neutral oxygen
atoms [O(2), O(3), O(4), O(5), and O(6)] and one tertiary amine
nitrogen atom [N(1)] from the crown ether ring occupy the basal
plane, and the phenolate [O(1)] and water oxygen atoms [O(7)]
coordinate to the metal ion center from the opposite axial
positions. In this case, Sr2 is placed at nearly the center of the
macrocyclic ring of the ligand. Thus, the unique structures of
(36) Chazin, W. J. Nat. Struct. Biol. 1995, 2, 707-710.
(
37) Clapham, D. E. Cell 1995, 80, 259-268.
(38) The SHACV method is known to provide a superior approach to
the direct evaluation of the one-electron redox potentials in the presence of
a follow-up chemical reaction: (a) McCord, T. G.; Smith, D. E. Anal. Chem.
1
4
1
969, 41, 1423-1441. (b) Bond, A. M.; Smith, D. E. Anal. Chem. 1974,
6, 1946-1951. (c) Wasielewski, M. R.; Breslow, R. J. Am. Chem. Soc.
976, 98, 4222-4229. (d) Arnett, E. M.; Amarnath, K.; Harvey, N. G.;
+
Cheng, J.-P. J. Am. Chem. Soc. 1990, 112, 344-355. (e) Fukuzumi, S.;
Fujita, M.; Maruta, J.; Chanon, M. J. Chem. Soc., Perkin Trans. 2 1994,
1
597-1602. A well-defined symmetrical SHACV trace was obtained for
-
(
32) Katz, A. K.; Glusker, J. P.; Beebe, S. A.; Bock, C. W. J. Am. Chem.
Soc. 1996, 118, 5752-5763.
33) Handbook of Chemistry and Physics, 61st ed.; CRC Press: Boca
Raton, FL, 1981.
the one-electron oxidation of the tetra-n-butylammonium salts of L1 and
L2 in CH3CN as shown in Figure S1 and S2, in which the intersection
-
(
with the dc potential axis corresponds to the redox potential (E1/2).
(39) Fukuzmi, S.; Ohkubo, K. Chem. Eur. J. 2000, 6, 4532-4535.

2170 J. Am. Chem. Soc., Vol. 123, No. 10, 2001
Itoh et al.
Table 5. UV-Vis Data (λMax/nm and ε/M- cm-1) of the Phenoxyl Radical Complexes
1
cation
L1•
L2•
2
+
Mg
393 (2130)
387 (2140)
385
408 (2630)
404 (2580)
402
738 (180)
682 (230)
672
388 (1370)
388 (2660)
380
406 (1600)
400 (3210)
398
404
-
407 (2040)
404 (1880)
640 (170)
694 (300)
668
626
-
656 (130)
645 (170)
2
+
Ca
2
+ a
Sr
2
+ a
Ba
Na
387
406
627
-
-
+
388 (1180)
-
386 (2700)
406 (1270)
-
406 (2460)
635 (180)
-
635 (220)
+
K
389 (2200)
384 (2260)
n
+
N( Bu)
4
a
Precipitation of Sr(NO
3
)
2
and Ba(NO
3
)
2
caused by the CAN oxidation prevented us to determine the accurate value of ε.
•
+
Figure 4. Resonance Raman spectrum of [Ca(L1 )(NO
3
)] (solid line)
-
3
(1.5 × 10 M) in CH
3
CN at -40 °C and that of the same compound
decomposed at 25 °C (dotted line).
•
+
Figure 3. UV-vis spectrum of [Ca(L1 )(NO
3
)] (solid line) generated
Table 6. Resonance Raman Data (cm-1) of the Phenoxyl Radical
-
-4
by the oxidation of [Ca(L1 )(CH
3
OH)]BPh
CN at 25 °C.
4
(dotted line) (5.0 × 10
-
4
3
Complexes in CH CN at -40 °C
M) by CAN (5.0 × 10 M) in CH
3
L1•
L2•
was observed in the oxidation of other phenolate complexes by
CAN under the same experimental conditions. The λmax and ε
values of the oxidation products are listed in Table 5. All of
the spectra exhibit a similar shape having strong absorption
bands around 400 nm together with a very broad one in the
near-infrared (NIR) region (600-900 nm), which is typical for
cation
Mg2
ν
7a
′
ν
8a
′
ν
7a
′
8a
ν ′
nda
1587
1586
nd
+
1517
1525
1524
1506
1506
-
1590
1586
1584
nd
nd
-
1507
1523
1519
1509
-
2
+
Ca
Sr
2
+
2
+
Ba
Na
+
-
nd
9-16,40
phenoxyl radical species.
The ZINDO calculation on
+
K
1505
•
27
phenoxyl radical L1 suggested that the 400 nm band is mainly
associated with a π-π* transition (HOMO-1-to-LUMO+1);
both molecular orbitals exist on the phenyl ring of the phenoxyl
radical species. It is also suggested that the broad band in the
NIR region is due to the HOMO-1-to-LUMO transition, where
the LUMO orbital exists mainly on the C-O bond of the
phenoxyl radical species. Thus, the 400 nm bands are hardly
affected by the complex formation, whereas the 600-900 nm
bands are significantly altered, depending on the type of metal
ions (Table 5). It is interesting to note that the NIR band shifts
toward the lower-energy direction with an increase in the Lewis
acidity of the metal ion when the E1/2 value is shifted to the
positive direction (Table 4), although some exceptions are seen
a
nd: Not detected.
1
1
505-1525 cm- together with a small one at around 1590 cm-1
as listed in Table 6. A typical example of the RR spectrum is
shown in Figure 4, and spectra of others are presented as
Supporting Information (Figures S4-S12). These bands are
vibrational modes characteristic of phenoxyl radicals and are
assigned to the modes ν7a′ and ν8a′ which predominantly consist
of the C-O stretching and the Cortho-Cmeta stretching, respec-
4
1
tively. The higher energy of the C-O stretching of the
phenoxyl radical (ν7a′) as compared to that of a phenolate (ν7a
1250 cm ) clearly indicates that the C-O bond of the
phenoxyl radical species has a partial double bond character,
-
1 10
≈
•
in the case of L2 . This shows a sharp contrast to the case of
2
+
and this character is most pronounced in the Ca -complexes
the phenolate complexes, where the λmax of the phenolate shifts
toward the higher-energy direction with an increase in the Lewis
acidity of the metal ion (Table 1). The stronger interaction
between the phenoxyl radical oxygen and the metal ion may
result in a decrease in the C-O centered LUMO level, thereby
leading to a decrease in the HOMO-1-to-LUMO transition
energy.
(
Table 6).
The electrospray ionization mass spectrum (ESI-MS) shown
in Figure S13 exhibits only a set of prominent peaks with a
mass and an isotope distribution pattern being consistent with
those of the Ca -complexes of the phenoxyl radical species
L1 and L2 with one nitrate ion as a counteranion coming from
2
+
•
•
•
+
•
+
CAN: [Ca(L1 )(NO3)] and [Ca(L2 )(NO3)] . Oxidation of the
phenolate complexes of other metal ions also gave similar
The resonance Raman spectra of the phenoxyl radical-metal
complexes excited at 406.7 nm display one prominent peak at
(
40) Johnston, L. J.; Mathivanan, N.; Negri, F.; Siebrand, W. Can. J.
(41) Schnepf, R.; Sokolowski, A.; M u¨ ller, J.; Bachler, V.; Wieghardt,
K.; Hildebrandt, P. J. Am. Chem. Soc. 1998, 120, 2352-2364.
Chem. 1993, 71, 1655-1662.

H-Atom Abstraction by Phenoxyl Radical-Metal Complexes
J. Am. Chem. Soc., Vol. 123, No. 10, 2001 2171
Table 7. ESR Data of the Phenoxyl Radical Complexes in CH
at -40 °C
3
CN
Scheme 1
L1•
L2•
cation
g
a
H
/G
g
H
a /G
2
+
Mg
2.0039
2.0043
2.0045
2.0045
2.0046
-
6.42
6.77
7.49
6.06
6.42
-
2.0046
2.0042
2.0043
2.0042
-
6.06
6.77
6.42
6.30
-
2
+
Ca
2
+
Sr
2
+
Ba
Na
+
+
K
2.0045
2.0046
5.35
n
+
4.17a
N( Bu)
4
2.0048
5.88
a
The a
H
value was determined by computer simulation using ESRaII
Table 8. Second-order Rate Constants kdec (M- s-1) for the
1
version 1.01 (Calleo Scientific Publisher) on a Macintosh personal
computer.
Decomposition of Phenoxyl Radical Complexes in CH
3
CN at 25 °C
cation
L1•
L2•
results, confirming the formation of the corresponding phenoxyl
radical-metal complexes.
Mg2
+
108
78.6
134
643
844
-
15.1
51.9
-
2
+
Ca
The phenoxyl radical-metal complexes exhibit a doublet ESR
signal at g ) 2.0039∼2.0046 in CH3CN at -40 °C. The ESR
spectra of the phenoxyl radical species are presented as
Supporting Information (Figures S14-S25), and their ESR
parameters are summarized in Table 7. The hyperfine splitting
into a doublet is due to a spin-nucleus coupling with one of the
benzylic methylene protons. Coordination of the phenoxyl
oxygen to the metal center will inhibit the free rotation around
the C(2)-C(2′: benzylic carbon) bond, fixing the steric relation
between the benzylic protons and the aromatic ring. As a result,
one of the C-H bond is nearly perpendicular to the aromatic
plane but another C-H bond is nearly horizontal to the phenoxyl
2
+
Sr
+
Na
n
+
N( Bu)4
962
The decrease in the absorption band due to the phenoxyl radical
complex obeys second-order kinetics, indicating that the decay
occurs via a bimolecular disproportionation reaction (see Figure
S26). The second-order rate constants (kdec) for the decomposi-
tion of the phenoxyl radical complexes were determined from
the second-order plots (the inset of Figure S26), and the results
are given in Table 8. The mass spectral analysis on the final
reaction mixture indicated that N-dealkylation products (3,5-
di-tert-butyl-2-hydroxy-benzaldehyde and the aza-crown ether)
were formed together with the original phenol derivative
radical ring. Thus, only one methylene proton exhibits the
hyperfine coupling.4
2,43
Thus, the differences in the hyperfine
coupling constants aH among the phenoxyl radical-metal
complexes are largely attributed to the differences in the
molecular geometry of the phenoxyl radical complexes as seen
in the phenolate complexes discussed above (Figure 2). It should
be noted that the g values of the phenoxyl radical complexes
2
7
(
LnH). Thus, the disproportionation reaction between the two
phenoxyl radical-metal complexes yields the original phenolate
-
+
complex [M(Ln )] and the organic cation complex [M(Ln )],
the latter of which further decomposes into 3,5-di-tert-butyl-
2-hydroxy-benzaldehyde and the corresponding crown ether by
hydrolysis as shown in Scheme 1. The hydrolysis of organic
n
+
2+
decrease upon going from N( Bu)4 to Mg . The deviation of
g values of the phenoxyl radical complexes from the free spin
value (ge ) 2.0023) can be expressed in terms of the spin-
orbit coupling constant of the oxygen atom (ê) and the energy
gap (δ) between the singly occupied orbital (SOMO) and the
nonbonding orbital as given by eq 1.44 A constant (C) is a value
+
cation Ln into the aldehyde derivative and the crown ether
+
will afford one proton (H ) which may bind to the initially
formed phenolate complex [M(Ln )] to provide the phenol
derivative (LnH).
-
The kdec values in Table 8 are regarded as measure of the
kinetic stability of the phenoxyl radicals, which vary signifi-
cantly depending on the type of cationic species. The kinetic
determined by the coefficients of the SOMO and nonbonding
orbitals of the phenoxyl radical complexes.44 The stronger
interaction of metal ions with the nonbonding orbital with an
increase in the Lewis acidity of metal ion, the larger the energy
2
+
stabilities of the Ca -complexes are most high among the
4
6
phenoxyl radicals examined (Table 8). Since the dispropor-
tionation reaction between the two phenoxyl radicals involves
both the oxidation and reduction, the phenoxyl radical species
must act as both the oxidant and reductant. The stronger Lewis
gap (δ), and thereby the smaller is the g value as observed
experimentally.45
g ) g + Cê/δ
(1)
e
•
acid would accelerate the reduction of phenoxyl radical (Ln )
-
Kinetic Stability of the Phenoxyl Radical-Metal Com-
plexes. The phenoxyl radical-metal complexes generated by
the CAN oxidation gradually decompose at room temperature.
to the phenolate (Ln ) as indicated by the more positive value
of E1/2 which corresponds to the reduction potential of Ln•
(Table 4), but at the same time it would decelerate the oxidation
process due to the repulsive interaction between the oxidized
(
42) Itoh, S.; Taki, M.; Kumei, H.; Takayama, S.; Nagatomo, S.;
Kitagawa, T.; Sakurada, N.; Arakawa, R.; Fukuzumi, S. Inorg. Chem. 2000,
9, 3708-3711.
43) It has been reported that hyperfine coupling constant (hfc) of
+
product (Ln ) and the metal cations. As a result of these
3
contradictory factors, the decomposition (disproportionation)
reaction is slowest in the case of a mild Lewis acid such as
(
phenoxyl radicals can be estimated by the angle-dependent McConnell-
2+
Ca . The kinetic stability of the phenoxyl radical-metal
2
type relationship: aC-H ) FC1B cos θ, where aC-H is the hfc of the
complexes is enhanced in the L2-ligand system as compared
methylene proton, FC1 is the spin density at the C1 position, B is a constant
equal to 162 Hz, and θ is the dihedral angle defined in Figure 8 of Babcock’s
paper: Babcock, G. T.; El-Deeb, M. K.; Sandusky, P. O.; Whittaker, M.
M.; Whittaker, J. W. J. Am. Chem. Soc. 1992, 114, 3727-3734.
2
+
2+
to the L1-ligand system (e.g., compare kdec of Ca - and Sr -
•
•
complexes of L1 and L2 ), although the stability of the metal-
free radicals is nearly the same. The higher kinetic stability of
(
44) Wertz, J. E.; Bolton, J. R. Electron Spin Resonance Elementary
Theory and Practical Applications; McGraw-Hill: New York, 1972.
45) Fukuzumi, S.; Okamoto, T. J. Am. Chem. Soc. 1993, 115, 11600-
1601.
(
(46) Precipitation of Ba(NO3)2 caused by the CAN oxidation prevented
•
2+
1
the kinetic studies of the decomposition process of [Ba(Ln )] .

2172 J. Am. Chem. Soc., Vol. 123, No. 10, 2001
Itoh et al.
Scheme 2
-
3
Figure 5. Spectral change for the oxidation of AcrH
2
(5.0 × 10 M)
CN at -40 °C. Interval:
6 s. Inset: first-order plot based on the absorption change at 404 nm.
Ph)) and M(LnH) and a subsequent electron transfer from the
•
+
-4
by [Ca(L1 )(NO
1
3
)] (5.0 × 10 M) in CH
3
•
•
+
resulting AcrR to another M(Ln ), generating AcrR and
-
M(Ln ) (Scheme 2). A relatively large primary kinetic isotope
H
D
(M-1 s-1) for the
2
effects k 2/k 2 ) 4.9-12.5 were obtained when AcrD2 was
employed instead of AcrH2, confirming that the initial hydrogen
atom transfer is the rate-determining step. Thus, the k2 values
Table 9. Second-order Rate Constants k
Oxidation of AcrH , AcrD , and AcrHR (R ) Me, Et, and Ph) by
CN at -40 °C
2
2
•
M(L1 ) in CH
AcrH
15
3
•
(kH
/kD
)a
can be regarded as a measure of H -abstraction ability of the
M
2
AcrD
2
2
2
AcrHMe AcrHEt AcrHPh
phenoxyl radical-metal complexes.
2
+
Mg
1.2 (12.5)
0.86 (8.4)
0.48 (7.1)
2
+
There are two possibilities in the mechanisms of hydrogen-
transfer reactions of NADH analogues (AcrH2) to an organic
radical oxidant, that is, a one-step hydrogen transfer or electron
Ca
7.2
3.4
193
0.26 0.11 0.048
2
+
Sr
+
Na
40
(4.9)
2
9a,47
transfer followed by proton transfer.
A one-step (direct)
a
Primary kinetic deuterium isotope effect.
hydrogen transfer from AcrH2 to the phenoxyl radicals would
be retarded by the binding of metal ions to the phenoxyl radicals
to which the hydrogen atom is to be bound. The reaction rate
for the tetra-n-butylammonium salt was too fast to be determined
•
L2 -complexes can be attributed to the larger macrocyclic ring
of L2 , that may protect the phenoxyl radical center more
effectively than the smaller crown ether ring in L1 . Such a
steric effect of the crown ether ring is not expected in the metal-
free radicals, since there is no attractive interaction between
the phenoxyl radical and the crown ether moiety.
C-H Bond Activation. Reactivity of the phenoxyl radical-
metal complexes toward C-H bond activation has been
examined in the hydrogen-transfer reaction from 10-methyl-
•
•
+
accurately, and the Na -complex shows significantly higher
reactivity than the alkaline earth metal complexes (Table 9). In
the case of alkaline earth metal complexes, however, the k2 value
2
+
2+
2+
increases in order: Sr < Ca < Mg with increasing the
Lewis acidity of metal ions, when the binding of metal ions to
the phenoxyl radical becomes stronger. In addition, the k 2/k 2
value increases with increasing the binding strength of metal
ions to the phenoxyl radical in order: Na < Sr < Ca
Mg . This is completely opposite from what is expected for a
one-step hydrogen transfer reaction in which the primary kinetic
H
D
9
(
,10-dihydroacridine (AcrH2) and its 9-substituted derivatives
+
2+
2+
AcrHR, R ) Me, Et, and Ph). The reactions were performed
<
2
+
at -40 °C, where the self-decomposition (disproportionation)
of the phenoxyl radical complexes was negligibly slow. The
•
+
H
D
spectral change for the reaction of [Ca(L1 )(NO3)] with AcrH2
is shown in Figure 5, in which the characteristic absorption
bands due to the phenoxyl radical complex (387, 404, and 682
nm) decreases in the course of the reaction. The final spectrum
of the product is identical to that of the authentic sample of
deuterium isotope effects (k 2/k 2) should decrease with increas-
ing the binding strength of metal ions to the phenoxyl radical
when the driving force of hydrogen transfer decreases.48 Thus,
an alternative mechanism, that is, electron transfer followed by
proton transfer shown in Scheme 3 should be considered to
account for the reactivity of the alkaline earth metal complexes
of phenoxyl radicals and the primary kinetic deuterium isotope
effects in Table 9.
+
AcrH (396, 416, and 440 nm). From the absorption intensity
at 440 nm (ꢀ ) 2150 M-1 cm ) is estimated the yield of AcrH
-1
+
as 76%.
In contrast to the self-decomposition of the phenoxyl radical-
metal complexes (vide supra), the oxidation of AcrH2 obeys
first-order kinetics as shown in the inset of Figure 5, where the
first-order rate constant (kobs) was determined from the slope
of the first-order plot. The rate was first-order with respect to
the AcrH2 concentration, and the second-order rate constant (k2)
was obtained from the slope of a linear plot of kobs versus
Judging from the one-electron oxidation potential of AcrH2
0
49
(E ox vs SCE ) 0.81 V) which is higher than the one-electron
reduction potentials of the phenoxyl radical-metal complexes
(equivalent to the one-electron oxidation potentials of the
phenolate complexes in Table 4), the free energy change of
electron transfer from AcrH2 to the phenoxyl radical complexes
(47) Fukuzumi, S.; Tokuda, Y.; Chiba, Y.; Greci, L.; Carloni, P.;
[AcrH2]. The k2 values for the oxidation of AcrH2 as well as
Damiani, E. J. Chem. Soc., Chem. Commun. 1993, 1575-1577.
(48) (a) Pryor, W. A.; Kneipp, K. G. J. Am. Chem. Soc. 1971, 93, 5584-
5585. (b) Ishikawa, M.; Fukuzumi, S. J. Chem. Soc., Faraday Trans. 1990,
9
,9-dideuterated derivative (AcrD2) and 9-substituted derivatives
•
(AcrHR) by the series of L1 complexes are listed in Table 9.
8
6, 3531-3536.
49) Fukuzumi, S.; Tokuda, Y.; Kitano, T.; Okamoto, T.; Otera, J. J.
Am. Chem. Soc. 1993, 115, 8960-8968.
The overall reaction consists of a formal hydrogen atom transfer
(
•
•
from AcrHR to M(Ln ) to afford AcrR (R ) H, D, Me, Et, or

H-Atom Abstraction by Phenoxyl Radical-Metal Complexes
J. Am. Chem. Soc., Vol. 123, No. 10, 2001 2173
Scheme 3
0
0
0
is positive [∆G et (in eV) ) e(E ox - E red) > 0, e is the
elementary charge], and thereby the electron-transfer step is
endergonic. In such a case, the overall rate of hydrogen transfer
(k2) which consists of electron-transfer and proton-transfer steps
would be slower than the initial electron-transfer rate (ket). The
0
maximum ket value is evaluated from the ∆G et value by eq 2,
11
-1 -1
where Z is the frequency factor taken as 1 × 10
M
s , and
kB is the Boltzmann constant.
0
et
k ) Z exp(-∆G /k T)
(2)
et
B
0
Figure 6 shows a plot of log k2 versus ∆G et in reference to
0
a linear correlation log ket versus ∆G et, calculated by eq 2
+
•
(
denoted by ET line). The k2 value of the Na -L1 complex is
significantly above the ET line, whereas those of the alkaline
earth metal-L1 complexes are much lower than the ET line.
The k2 value of the Na -L1 complex, which is much larger
than the corresponding ket value, indicates that the hydrogen
transfer proceeds via a direct one-step hydrogen transfer rather
than via electron transfer. In such a case, the primary kinetic
deuterium isotope effect is ascribed to the direct transfer of
•
+
•
Figure 6. Plot of the rate constant of hydrogen transfer from AcrHR
•
(
R ) H, Me, Et, and Ph) to the phenoxyl radical complexes M(L1 )
+
•
H
2
(log k ) vs the free energy of electron transfer from AcrHR to the
hydrogen atom from AcrH2 to the Na -L1 complex. The k 2/
k 2 value (4.9) agrees with the value (4.8) reported for the
0
phenoxyl radical complexes (∆G et). The solid line shows the depen-
D
0
dence of the calculated rate constant of electron transfer (ket) on ∆G et
hydrogen atom transfer from AcrH2 to hydrogen peroxyl
based on eq 2, see text.
radical.50 On the other hand, the k2 values of alkaline earth
•
metal-L1 complexes, which are much smaller than the
corresponding maximum ket values, indicate that the hydrogen
transfer proceeds via electron transfer followed by proton
transfer as shown in Scheme 3 rather than a direct one-step
hydrogen transfer. In such a case, the primary kinetic deuterium
•
+
isotope effect is ascribed to proton transfer from AcrH2 to
-
the alkaline earth metal-L1 complexes. The large primary
kinetic isotope effects (7.1-12.5) observed for alkaline earth
•
metal-L1 complexes (Table 9) are consistent with those
•+
reported for proton transfer from AcrH2 to semiquinone radical
4
8b
anions (7.2-10.4).
The diminished reactivity of AcrH2 toward the hydrogen-
2+
•
transfer reaction with the Ca -L1 complex when AcrH2 is
replaced by AcrHR (R ) Me, Et, and Ph) shown in Table 9 is
also consistent with the electron-transfer-proton-transfer mech-
anism in Scheme 3, since such diminished reactivity of AcrH2
is also observed for a hydride transfer reaction from AcrHR to
tetracyanoethylene (TCNE), which has been shown to proceed
via electron transfer from AcrHR to TCNE followed by proton
transfer from AcrHR•+ to TCNE . The k2 values of hydrogen-
•- 51
Figure 7. Comparison of the rate constant (log k
2
•
) of hydrogen transfer
2+
•
2+
transfer reactions from AcrHR to the Ca -L1 complex are
compared with the kh values (kh: the rate constant of hydride
transfer from AcrHR to TCNE)5 in the logarithmic plots in
Figure 7, where the k2 values are in parallel with the kh values.
Such a parallel relationship confirms the validity of the electron-
transfer-proton-transfer mechanism shown in Scheme 3.
The coupling of electron transfer and proton transfer is a well-
established alternative for the formal hydrogen atom transfer
from AcrHR (R ) H, Me, Et, and Ph) to Ca (L1 ) and the rate constant
h
(log k ) of hydride transfer from AcrHR to TCNE (the data were taken
1
from ref 51).
is, electron transfer followed by proton transfer or vice versa,
or a concerted process in one quantum mechanical tunneling
53
event. The present study demonstrated for the first time clearly
that the hydrogen atom abstraction by the phenoxyl radical-
alkaline earth metal complexes proceeds via electron transfer
followed by rate-determining proton transfer, providing valuable
insight into the C-H bond activation mechanism by tyrosine
radicals in biological systems.
4
9
in many organic reactions. Such a proton-coupled electron
transfer has been suggested to play an important role in
biological electron transfer.52 It is often difficult to determine
whether the proton-coupled electron transfer is consecutive, that
(52) (a) Okamura, M. Y.; Feher, G. Annu. ReV. Biochem. 1992, 61, 861.
(
b) Malmstr o¨ m, B. G. Acc. Chem. Res. 1993, 26, 332-338. (c) Ferguson-
(
50) Fukuzumi, S.; Ishikawa, M.; Tanaka, T. J. Chem. Soc., Perkin Trans.
1989, 1037-1045.
51) Fukuzumi, S.; Ohkubo, K.; Tokuda, Y.; Suenobu, T. J. Am. Chem.
Soc. 2000, 122, 4286-4294.
Miller, S.; Babcock, G. T. Chem. ReV. 1996, 96, 2889-2907. (d) Hoganson,
C. W.; Babcock, G. T. Science 1997, 277, 1953-1956. (e) Cukier, R. I.;
Nocera, D. G. Annu. ReV. Phys. Chem. 1998, 49, 337-369.
2
(
(53) Cukier, R. I. J. Phys. Chem. A 1999, 103, 5989-5995.

2174 J. Am. Chem. Soc., Vol. 123, No. 10, 2001
Itoh et al.
Experimental Section
yield (348.8 mg): IR (neat) 3000 (hydrogen bonding phenolic OH),
-1
1
1
361, 1241, and 1129 cm (O-H, C-O); H NMR (400 MHz, CDCl
δ 1.29 (s, 9H), 1.44 (s, 9H), 2.86 (t, J ) 5.9 Hz, 4H), 3.63-3.61 (m,
H), 3.67 (t, J ) 5.9 Hz, 4H), 3.69 (s, 4H), 3.72-3.70 (m, 4H), 3.80
s, 2H), 6.84 (d, J ) 2.2 Hz, 1H), 7.21 (d, J ) 2.2 Hz, 1H); HRMS
m/z 437.3152, calcd for C25 437.3141; UV-vis (CH
3
)
General. Reagents and solvents used in this study were commercial
products of the highest available purity and were further purified by
4
(
5
4
the standard methods. 10-Methyl-9,10-dihydroacridine (AcrH
its 9,9-dideuterated derivative (AcrD ) were prepared by the reported
procedures, and 9-substituted derivatives of AcrH (AcrHR, R ) Me,
2
) and
2
H43NO
5
3
CN) λmax
49
2
284 nm (ꢀ ) 2250 M cm-1).
,4-Di-tert-butyl-6-(1,4,7,10,13-pentaoxa-16-aza-cyclooctadec-16-
ylmethyl)-phenol (L2H). This compound was prepared using 1-aza-
8-crown-6 in place of 1-aza-15-crown-5 by a similar procedure for
the synthesis of L1H in 83% yield: IR (neat) 3000 (hydrogen bonding
-1
)
5
1
Et, and Ph) were obtained from the previous study. IR spectra were
recorded on a Shimadzu FTIR-8200PC and UV-vis spectra on a
Hewlett-Packard 8453 photodiode array spectrophotometer. H NMR
spectra were recorded on a JEOL FT-NMR GX-400 spectrometer. ESR
spectra were recorded on a JEOL JES-ME-2X spectrometer. The g
values were determined using a Mn marker as a reference. Mass
spectra were recorded on a JEOL JNX-DX303 HF mass spectrometer
or a Shimadzu GCMS-QP2000 gas chromatograph mass spectrometer.
ESI-MS (electrospray ionization mass spectra) measurements were
performed on a JEOL JMS-700T Tandem Mstation or a PE SCIEX
API 150EX. Molecular orbital calculations were performed by using a
CAChe program (Version 3.2). The UV-visible electronic transitions
are calculated with ZINDO using INDO/1 parameters after optimizing
geometry with Augmented MM3 and MOPAC PM3. The cyclic
voltammetry measurements were performed on a BAS 100 W or a ALS
2
1
1
-
1
1
phenolic OH), 1360, 1241, and 1123 cm (O-H, C-O); H NMR
400 MHz, CDCl ) δ 1.27 (s, 9H), 1.41 (s, 9H), 2.84 (t, J ) 5.5 Hz,
2
+
(
4
3
H), 3.61-372 (m, 20H), 3.80 (s, 2H), 6.82 (d, J ) 2.2 Hz, 1H), 7.19
(
4
d, J ) 2.2 Hz, 1H); HRMS m/z 481.3409, calcd for C27
H
47NO
6
-
1
-1
81.3403; UV-vis (CH
3
CN) λmax ) 283 nm (ꢀ ) 2360 M cm ).
-
[Na(L1 )]. NaH in oil (70 wt %) (0.4 mmol, 13.7 mg) was washed
with dry n-hexane three times to remove the oily material before the
reaction. Then, a dry THF solution (5 mL) of L1H (0.4 mmol, 175
mg) was added to NaH slowly. The mixture was stirred for several
minutes at room temperature. After removal of insoluble material by
filtration, addition of n-hexane (30 mL) to the filtrate gave colorless
6
00 electrochemical analyzer in deaerated CH
NBu PF as supporting electrolyte. The Pt and Au working electrodes
BAS) were polished with BAS polishing alumina suspension and rinsed
with acetone before use. The counter electrode was a platinum wire.
The measured potentials were recorded with respect to an Ag/AgNO
0.01 M) reference electrode. The E1/2 values (vs Ag/AgNO ) are
3
CN containing 0.10 M
micro crystal in 55% yield: UV-vis (CH CN) λmax ) 320 nm (ꢀ )
3
4
6
-
1
-1
-
+
5
[
520 M cm ); FAB-MS m/z 460 (Na(L1 ) + H ). Anal. Calcd for
(
-
Na(L1 )], C25
H
42NO
5
Na: C, 65.33; H, 9.21; N, 3.05. Found: C, 65.07;
H, 9.16; N, 3.05.
3
-
[
K(L2 )]. This complex was prepared in a grove box ([O
2
] < 1
(
3
5
5
ppm, [H
2
O] < 1 ppm) as follows. KH in oil (35 wt %) (0.4 mmol,
converted to those vs SCE by adding 0.29 V. All electrochemical
measurements were carried out at 25 °C under an atmospheric pressure
of nitrogen.
4
5.8 mg) was washed with dry n-hexane three times to remove the
oily material before the reaction. Then, a dry THF solution (5 mL) of
L1H (0.4 mmol, 192.6 mg) was added to KH slowly. The mixture
was stirred for several minutes at room temperature. After removal of
the solvent, the resulting material was washed with dry n-hexane to
+
Resonance Raman Measurement. The 406.7 nm line of a Kr laser
(model 2060 Spectra Physics) was used as the exciting source. Visible
resonance Raman scattering was detected with a liquid nitrogen cooled
CCD detector (model LN/CCD-1340 × 400PB, Princeton Instruments)
attached to a 1 m single polychlomator (model MC-100DG, Ritsu Oyo
Kogaku). The slit width and slit height were set to be 200 µm and 20
mm, respectively. Wavenumber width per channel is 0.35 cm , which
determines the wavenumber resolution in this measurement. The laser
power used was 8.9 mW at the sample point. All measurements were
carried out at -40 °C with a spinning cell (1000 rpm). Raman shifts
were calibrated with indene, and accuracy of the peak positions of the
afford white solid in 25% yield: UV-vis (CH CN) λmax ) 312 nm (ꢀ
3
-
1
-1
-
+
)
3100 M cm ); FAB-MS m/z 520 (K(L2 ) + H ). Anal. Calcd
-
for for [K(L2 )]‚2H
2
O, C27
H50NO
7
K: C, 58.34; H, 9.07; N, 2.52.
-
1
Found: C, 58.95; H, 8.57; N, 2.48.
-
[
Mg(L1 )(CH
3
OH)]BPh
4
. To a mixture of L1H (0.2 mmol, 87.5
mg) and triethylamine (0.2 mmol, 28.7 µL) in methanol (5 mL) was
added Mg(ClO (0.2 mmol, 44.6 mg) in methanol (5 mL) dropwise
with stirring at ambient temperature. Addition of NaBPh (0.4 mmol,
4 2
)
4
-
1
136.9 mg) in methanol (5 mL) to the mixture gave colorless micro
Raman bands was (1 cm
.
crystals, which were recrystallized from methanol to give pure material
X-ray Structure Determination. Data of X-ray diffraction were
collected by Rigaku RAXIS-RAPID imaging plate two-dimensional
area detector using graphite-monochromated MoKR radiation (λ )
of the Mg2 -complex in 61% yield: UV-vis (CH
+
CN) λmax ) 304
3
-
1
-1
- +
nm (ꢀ ) 4000 M cm ); FAB-MS m/z 460 (Mg(L1 ) ). Anal. Calcd
-
for [Mg(L1 )(CH
3
OH)]BPh
4 6
, C50H66NO BMg: C, 73.93; H, 8.19; N,
0
.71070 Å) to 2θ max of 55.0 °. The intensity measurement was
1
.72. Found: C, 73.86; H, 8.29; N, 1.70.
undertaken by sealing them in a glass capillary tube containing the
mother liquid. All of the crystallographic calculations were performed
by using teXsan software package of the Molecular Structure Corpora-
tion [teXsan: Crystal Structure Analysis Package, Molecular Structure
Corp. (1985 and 1999)]. The crystal structure was solved by the direct
methods and refined by the full-matrix least squares. All non-hydrogen
atoms and hydrogen atoms were refined anisotropically and isotropi-
cally. The summary of the fundamental crystal data and experimental
parameters for structure determinations is given in Table 2. The
experimental details including data collection, data reduction, and
structure solution and refinement as well as the atomic coordinates and
-
[Ca(L1 )(CH
3
OH)
2
]BPh
4
. This complex was prepared in a similar
-
manner for the synthesis of [Mg(L1 )(CH
(ClO
UV-vis (CH
m/z 476 (Ca(L1 ) ). Anal. Calcd for [Ca(L1 )(CH
NO BCa: C, 71.22; H, 8.20; N, 1.63. Found: C, 71.36; H, 7.91; N,
1.82.
[Sr(L1 )(CH OH) ]BPh . This complex was prepared in a similar
3
OH)]BPh
4
by using Ca-
(91% yield):
)
4 2
2
‚4H O (0.2 mmol, 62.2 mg) instead of Mg(ClO
4
)
2
-1
-
1
3
CN) λmax ) 309 nm (ꢀ ) 3940 M cm ); FAB-MS
-
+
-
3 2 4 70
OH) ]BPh , C51H -
7
-
3
2
4
-
3
4
4 2
manner for the synthesis of [Mg(L1 )(CH OH)]BPh by using Sr(ClO )
(0.2 mmol, 57.3 mg) instead of Mg(ClO ) (84% yield): UV-vis
4
2
-
1
-1
Biso/Beq, anisotropic displacement parameters, and intramolecular bond
(CH CN) λ
) 310 nm (ꢀ ) 4160 M cm ); FAB-MS m/z 524
3
max
-
+
-
distances and angles have been deposited in the Supporting Information
(Sr(L1 ) ). Anal. Calcd for [Sr(L1 )(CH OH) ]BPh , C H NO BSr:
3
2
4
51 70
7
(
S23-S26).
C, 67.49; H, 7.77; N, 1.54. Found: C, 66.98; H, 7.44; N, 1.68.
Syntheses. 2,4-Di-tert-butyl-6-(1,4,7,10-tetraoxa-13-aza-cyclopen-
-
[
Ba(L1 )(CH
3
OH)
2
]BPh
4
. This complex was prepared in a similar
tadec-13-ylmethyl)-phenol (L1H). A mixture of 1-aza-15-crown-5 (1
mmol, 219.3 mg), paraformaldehyde (2 mmol, 60.0 mg), and 2,4-di-
tert-buthylphenol (2 mmol, 412.7 mg) in ethanol (10 mL) was refluxed
for 1 day. Removal of the solvent gave an oily material from which
-
manner for the synthesis of [Mg(L1 )(CH
ClO
UV-vis (CH
3
OH)]BPh
4
by using Ba-
(88% yield):
(
)
4 2
2
‚3H O (0.2 mmol, 78.1 mg) instead of Mg(ClO
4
)
2
-1
-1
3
CN) λmax ) 311 nm (ꢀ ) 3460 M cm ); FAB-MS
- +
-
m/z 574 (Ba(L1 ) ). Anal. Calcd for [Ba(L1 )(CH
NO BBa: C, 63.99; H, 7.37; N, 1.46. Found: C, 63.76; H, 7.07; N,
.67.
Ca(L2 )(H
3 2 4 70
OH) ]BPh , C51H -
2 3
L1H was isolated by column chromatography (SiO , CHCl ) in 80%
7
1
(
54) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals; Butterworth-Heinemann: Oxford, 1988.
55) Mann, K.; Barnes, K. K. Electrochemical Reactions in Nonaqueous
Systems; Marcel Dekker Inc.: New York, 1990.
-
[
2
O)]BPh
4
. This complex was prepared in a similar
-
(
manner for the synthesis of [Ca(L1 )(CH
vis (CH CN) λmax ) 306 nm (ꢀ ) 4820 M cm ); FAB-MS m/z 520
3
OH)]BPh
4
70% yield: UV-
-
1
-1
3

H-Atom Abstraction by Phenoxyl Radical-Metal Complexes
J. Am. Chem. Soc., Vol. 123, No. 10, 2001 2175
-
+
-
(
7
Ca(L2 ) ). Anal. Calcd for [Ca(L2 )(H
2
O)]BPh
4
, C51
H
68NO
7
BCa: C,
Acknowledgment. This work was partially supported by
Grants-in-Aid for Scientific Research Priority Area (Nos.
1228205, 11228206) and Grants-in-Aid for Scientific Research
(Nos. 11440197 and 12874082) from the Ministry of Education,
Science, Culture and Sports, Japan.We also thank Professor
Yasushi Kai and co-workers, especially Dr. Nobuko Kanehisa,
of Osaka University for their assistance in the X-ray measure-
ments and Professor Hiroshi Tsukube of Osaka City University
for his helpful discussion.
1.39; H, 7.99; N, 1.63. Found: C, 71.24; H, 7.89; N, 1.69.
Sr(L2 )(H
similar manner for the synthesis of [Sr(L1 )(CH
yield: UV-vis (CH
MS m/z 568 (Sr(L2 ) ). Anal. Calcd for [Sr(L2 )(H
51.5
H70NO7.5BSr: C, 67.12; H, 7.66; N, 1.52. Found: C, 66.87; H,
.30; N, 1.24.
-
[
2
O)]BPh
4
‚0.5CH
3
OH. This complex was prepared in a
1
-
3
OH)
2
]BPh
4
60%
CN) λmax ) 308 nm (ꢀ ) 4840 M cm-1); FAB-
-
1
3
-
+
-
2
O)]BPh
4 3
‚0.5CH OH,
C
7
-
[
Ba(L2 )(H
2
O)](BPh
4
)‚0.5CH
3
OH. This complex was prepared
-
in a similar manner for the synthesis of [Ba(L1 )(CH
3
OH)
2
]BPh
4
-
1
-1
8
9% yield: UV-vis (CH CN) λmax ) 310 nm (ꢀ ) 3560 M cm );
3
-
+
-
FAB-MS m/z 618 (Ba(L2 ) ). Anal. Calcd for [Ba(L2 )(H
2
O)]BPh
70NO7.5BBa: C, 63.69; H, 7.26; N, 1.44. Found: C,
3.54; H, 6.99; N, 1.54.
Kinetics. The phenoxyl radical species of the alkaline and alkaline
4
‚
0
6
3
.5CH OH, C51.5H
Supporting Information Available: The cyclic voltammo-
-
gram of [Ca(L1 )(CH3OH)]BPh4 (Figure S1), the SHACVs of
-
-
tetra-n-butylammonium salts of L1 and L2 (Figures S2 and
earth metal complexes were generated in situ by adding an equimolar
4
+
-4
S3), resonance Raman spectra of the phenoxyl radical species
amount of (NH
CH
4
)
2
[Ce (NO
3
)
6
] (CAN, 5.0 × 10 M) into a deaerated
-
4
•
+
•
CN solution of the corresponding phenolate complex (5.0 × 10
(Figures S4-S12), ESI-MS of [Ca(L1 )(NO3)] and [Ca(L2 )-
3
+
M) in a UV cell (1 cm path length, sealed tightly with a silicon rubber
(NO )] (Figure S13), ESR spectra (Figures S14-S25) of the
3
cap) at 25 °C. The rate constants for the self-decomposition reaction
phenoxyl radical complexes, spectral change for the decomposi-
(k
dec) of the phenoxyl radical complexes were determined by following
a decrease in the absorption band due to the phenoxyl radical.
The oxidation of AcrH and AcrHR by the phenoxyl radical species
was initiated by adding the substrate into the CH CN solution, when
•
+
tion of [Ca(L1 )(NO3)] (Figure S26), and details about the
X-ray structure determination including crystallographic data
S27-S30) (PDF). An X-ray crystallographic file (CIF). This
2
(
3
the absorption of the phenoxyl radical complexes reached a maximum
in a few minutes after the addition of CAN into the phenolate complex
solution at -40 °C. The rate constants (kobs) for the redox reactions
were also determined by following a decrease in the absorption due to
the phenoxylradical species.
material is available free of charge via the Internet at
http://pubs.acs.org.
JA0036110